Poster Session 4

T. Kon¹, N. Shoji-Harada¹, H. Tokuno², Y. Hamanaka¹, A. Ebata¹, M. Iida¹, M. Sato¹, M. Yanagaki¹, A. Yamazaki¹, A. Sakamoto¹, M. Makita¹, O. Hinano¹, T. Abe², M. Miyashita¹; ¹Department of Breast and Endocrine Surgical Oncology, Graduate School of Medicine, Tohoku University, Sendai, JAPAN, ²Department of Clinical Biology and Hormonal Regulation, Graduate School of Medicine, Tohoku University, Sendai, JAPAN.

Background:In breast cancer, changes in the fecal metabolites and gut microbiota have been shown to influence drug sensitivity and resistance; however, differences in microbial composition between subtypes and their impact on cancer risk remain poorly understood. Here, we compared the fecal metabolites and gut microbiota between healthy controls(HCs) and breast cancer(BC) patients with subtype-specific analyses. Methods:43 HCs and 29 BC patients feces were collected. Samples from BC patients were collected before treatment initiation, when pre-treatment samples were unavailable, samples were collected during treatment. Metabolomic analysis was performed with LC-TOFMS/MS. 16S metagenomic analysis was performed and the composition was evaluated using ALDEx2. Statistical analysis was conducted using the Wilcoxon-Mann-Whitney test, and p values were adjusted for multiple comparisons using the Benjamini-Hochberg method.Results:The median age was 71 years (range: 36-95) for HCs and 57 years (range: 32-87) for BC patients. The subtypes included 18 cases of hormone receptor (HR)-positive/human epidermal growth factor receptor 2 (HER2)-negative, 6 cases of HER2-positive, and 5 cases of triple-negative breast cancer (TNBC). Concerning metabolomic analysis, the PCA plot indicated that the metabolomic profile was different between the HCs and BC patients. AMP, asparagine, diethanolamine, N-Acetylglucosamine 1-phosphate, NAD⁺, nicotinamide, riboflavin, terephthalate and UMP were increased in the BC patients. ROC analysis for these metabolites yielded an AUC over 0.79 in discriminating BC patients from HCs. In gut microbial analysis, β diversity assessed showed no significant difference between HCs and BC patients, while α diversity measured by the Chao1 was reduced in BC patients (q< 0.001). Similarly, the HR+/HER2- group showed no significant difference in β diversity, but α diversity assessed by Chao1 was reduced in the HR+/HER2- group (p< 0.05). Because the number of HER2-positive and TNBC was small, we couldn’t come to the conclusion about the difference in these two groups. ALDEx2 revealed that Clostridiales and Bacteroidales at the order level were changed when comparing HCs and BC patients (effect size > 0.4, p < 0.01, adjusted p = 0.27, 0.19). Furthermore, comparison between the HR+/HER2- and TNBC subtypes showed several changes in Clostridiales (effect size > 0.8, p < 0.05, adjusted p = 0.86). Even within the same species, the direction of change varied among subtypes.Conclusion:This study provides novel insights into the metabolome in breast cancer, highlighting as potential biomarkers. Microbial diversity was reduced in BC patients and the gut microbiota was different between breast cancer subtype.

A. Han¹, H. Choi¹, J. Ahn², H. Ahn², J. Shin², J. Song², Y. Park²; ¹Surgery, YUWMC, Wonju, KOREA, REPUBLIC OF, ²Internal medicine, SMC, Seoul, KOREA, REPUBLIC OF.

Background) Pathologic complete response (pCR) has been widely used as a surrogate marker for relevant outcomes in breast cancer patients undergoing neoadjuvant treatment. However, the prognostic value of pCR varies by subtype, with triple-negative breast cancer (TNBC) showing one of the strongest associations.Notably, some patients achieving pCR still experience recurrence, suggesting that pCR and survival outcomes such as event-free survival (EFS) may represent distinct biological endpoints from different processes.This study investigated whether the tumor-aorta ratio based on Hounsfield units (TAR) and the neutrophil-to-lymphocyte ratio (NLR) function as distinct predictive biomarkers for pCR and EFS in TNBC patients treated with the KEYNOTE-522 regimen. Materials and Methods) Patients with TNBC who received neoadjuvant systemic therapy according to the KEYNOTE-522 regimen were included. Clinicopathologic data, including TAR and NLR values, were collected and analyzed. Group comparisons were performed using independent t-tests or Mann-Whitney U tests, as appropriate. Statistical analyses were conducted using SPSS version 29, with a p-value <0.05 considered statistically significant. Results) Initial TAR was significantly associated with pCR (p = 0.022), supporting its role as a predictive biomarker for immediate tumor response. In contrast, EFS was more strongly associated with post-treatment NLR, a systemic immune marker (p < 0.001).TAR decreased consistently after treatment, regardless of pCR status or the occurrence of subsequent events, suggesting that TAR may reflect structural tumor regression.In contrast, NLR exhibited divergent post-treatment patterns. Among patients without events, NLR increased from 1.99 ± 0.85 to 2.50 ± 2.88. However, in patients who experienced events, NLR decreased from 2.26 ± 1.05 to 1.21 ± 0.45. Further analysis revealed that neutrophil counts changed in opposite directions depending on event occurrence, while lymphocyte counts decreased across both groups. Conclusion)This study suggests that pCR and EFS may require distinct predictive biomarkers, with TAR and NLR reflecting different biological processes: localized tumor response versus systemic immune activity. Importantly, in this uniform cohort of TNBC patients treated with an identical neoadjuvant regimen, an increase in NLR—driven primarily by increased neutrophil counts—was associated with better prognosis. These findings raise the possibility that neutrophil-mediated immune responses may contribute to improved clinical outcomes, underscoring the potential relevance of systemic immune dynamics in the context of outcome.

A. Korlimarla¹, V. Sasimouli², L. Krishna², S. Shree³, H. P.S⁴, B. Sivaraman⁵, S. Appachu⁶, S. Srihari⁷, G. Guhan⁸, P. Advani⁹, S. B.S²; ¹Molecular Oncology, Sri Shankara Cancer Hospital and Research Centre, bengaluru, INDIA, ²Surgical Oncology, Sri Shankara Cancer Hospital and Research Centre, bangalore, INDIA, ³Surgical Oncology, Sri Shankara Cancer Hospital and Research Centre, Bangalore, INDIA, ⁴Molecular Oncology, Sri Shankara Cancer Hospital and Research Centre, bangalore, INDIA, ⁵Molecular Oncology, Sri Shankara Cancer Hospital and Research Centre, BAngalore, INDIA, ⁶Medical Oncology, Sri Shankara Cancer Hospital and Research Centre, bangalore, INDIA, ⁷Pathology, Sri Shankara Cancer Hospital and Research Centre, bangalore, INDIA, ⁸Molecular Oncology, Sri Shankara Cancer Hospital and Research Centre, Bangalore, INDIA, ⁹Hematology and Oncology, Mayo Clinic, florida, FL.

Background: Breast cancer remains the leading cause of cancer-related mortality in women globally, with metastatic disease accounting for >90% of deaths. Current staging and histopathological features inadequately predict individual recurrence risk. Approximately 20-30% of early-stage patients develop distant metastases despite optimal therapy, while others are overtreated. Circulating microRNAs (MIRNAs) represent a paradigm shift toward liquid biopsy, offering real-time tumor assessment through minimally invasive blood sampling. These stable molecules reflect dynamic tumor behavior and metastatic potential, making them ideal for personalized surveillance strategies.Methods:. This prospective proof-of-concept study included 46 breast cancer patients from Sri Shankara Cancer Hospital, India: 37 with metastatic disease (plasma collected at progression, 2024) and 9 disease-free controls (≥3-year follow-up). MIRNA isolation used MIRVANA method with samples processed within 2 hours, stored at -80°C. Expression levels of MIR-21, MIR-30d, and MIR-425 were quantified by TaqMan qRT-PCR, normalized to MIR-16 control. Logistic regression analysis of the transcript levels yielded a probability of predictive accuracy ranging from 0-1. Cut-off of high expression was chosen based on the best sensitivity and specificity by ROC curve analysis. Clinicopathological parameters (CPP) were correlated with MIR expression(Mann-Whitney U test). A combined model of MIR and CPP was built using logistic regression and ROC curve analysis.Results: Patient demographics(in Table) Difference in median age, subtypes and nodal status are interesting to note. MIR-21 showed elevated expression in metastatic cases, particularly hormone receptor-positive tumors. MIR-30d demonstrated significant correlation with grade 3 and node-positive status (p=0.038 for N0 vs N1, p=0.006 for N0 vs N3), with high expression associated with poor progression-free survival (p=0.0057, HR=3.24). MIR-425 exhibited significant differential expression between metastatic and non-metastatic groups with strong correlation to advanced nodal stages (p=0.0018). Univariate analysis with all clinicopathological parametres with DFS revealed LN status as significant predictor (AUC=0.68). Similarly combined MIR score of MIR-425 and MIR-30d achieved 72% predictive accuracy of DFS (AUC=0.716). Interestingly, multivariate analysis of MIR score and LN with DFS, enhanced the AUC to 0.81.Conclusions: This proof-of-concept study suggests circulating MIR-30d and MIR-425 may serve as potential biomarkers for breast cancer recurrence risk stratification. The integrated signature shows promise for enhancing surveillance strategies. Assay also standardized methods for using plasma to measure circulating MIRNA levels. Validation in larger multicenter cohorts is underway for clinical assay development.

A. B. Duffy¹, M. Cristofanilli², C. Reduzzi², W. V. Williams³, G. Del Priore³, S. Chumsri⁴, C. Tang⁵, T. Iwase⁶, N. T. Ueno⁶, D. L. Adams¹; ¹Research, Creatv Microtech Inc., Monmouth Junction, NJ, ²Medicine, Weill Cornell Medicine, New York, NY, ³Management, BriaCell Therapeutics Corp., Philadelphia, PA, ⁴Oncology, Mayo Clinic Cancer Center, Jacksonville, FL, ⁵Management, Creatv Microtech Inc., Potomac, MD, ⁶Medicine, University of Hawai’i Cancer Center, Honolulu, HI.

Background: Currently, the gold standard in cancer diagnosis, prognosis, and treatment is assessing cancer cell mitosis in tumor tissue biopsies, which identifies patients with more aggressive cancer subtypes that may require more tailored treatment regimens. Circulating tumor cells (CTCs), cancer cells that detach from primary tumors, enter the bloodstream, and seed metastatic sites, are a well-known non-invasive blood biomarker which can also be used to aid in stratifying metastatic breast cancer (mBC) patients with highly aggressive subtypes. Initial pilot studies have identified a particular subtype of CTCs which are undergoing mitosis that appears to correlate with very poor survival outcomes compared to non-mitotic CTCs alone. In a multi-institutional prospective study, we isolated CTCs from the blood of n=167 mBC patients to categorize mitotic CTCs and evaluate their association with progression-free survival (PFS) and overall survival (OS) over a 3 year period. Methods: Peripheral blood samples (7.5ml) were collected from n=167 mBC patients progressing on systemic therapies and prior to new systemic therapy induction (i.e. chemotherapy n=102, hormone therapy n=13, immunotherapy n=63, targeted therapy n=62). Samples were processed using CellSieve microfilters to isolate CTCs via size exclusion and fluorescently stained for CD45, Cytokeratin, and DAPI. CTCs were imaged, enumerated, and subtyped based on the presence of ≥1 mitotic event, using previously established visual indicators. PFS and OS were assessed over 3 years by censored univariate and multivariate analyses. Results: CTCs were detected in ~47% (n=79/167) of patients, while mitotic CTCs were identified in a subset of this group, ~48% (n=38/79). The presence of any CTC was associated with poor outcomes (PFS HR=0.49, 95%CI=0.3-0.7, p=0.0002 and OS: HR=0.42, 95%CI=0.3-0.7, p=0.0002). However, the presence of mitotic CTCs correlated with significantly poorer outcomes compared to non-mitotic CTCs (PFS: HR=0.3, 95%CI=0.2-0.5, p<0.0001 and OS: HR=0.3, 95%CI=0.2-0.6, p=0.0002). Further, patients with no CTCs had a median PFS (mPFS) of 6 months and median OS (mOS) of 22 months, patients with non-mitotic CTCs had a mPFS of 4 months and mOS of 14.5 months; patients with mitotic CTCs had a mPFS of 2.4 months and mOS of 4.4 months. Notably, 95% of patients with a mitotic CTC progressed within 6 months, regardless of systemic therapy type. Interestingly, there were 5 patients who survived longer than 1 year with a mitotic CTC all of which were receiving specific targeted therapies. Conclusion: In this study, the detection of CTCs indicated poor outcomes in patients. Additional cytological analysis of CTCs further identified CTCs undergoing mitosis which correlated to highly aggressive disease with the poorest survival rates. These results indicate that while the presence of CTCs is a prognostic indicator for survival, subtyping CTCs and identifying mitosis may correlate to therapy response. This study highlights the need for larger, more comprehensive studies to expand upon and validate these findings.

N. Lopetegui-Lia¹, A. Davenport¹, J. Gill², A. M. Roy¹, S. Myers³, D. Agnese³, E. Burke³, T. Y. Andraos⁴, D. M. Schimming¹, S. Hulett¹, H. LeFebvre¹, M. Gatti-Mays¹, D. Stover¹; ¹Medical Oncology, The Ohio State University, Columbus, OH, ²Internal Medicine, Tennova North Knoxville Medical Center, Powell, TN, ³Surgical Oncology, The Ohio State University, Columbus, OH, ⁴Radiation Oncology, The Ohio State University, Columbus, OH.

Introduction: In the United States, breast cancer (BC) is the most common cancer and leading cause of cancer-related death among young women. Yet approximately only 17% of young adult (YA) patients (age < 40 years) carry a known cancer susceptibility mutation, and the full spectrum of somatic genomic alterations unique to this population remains poorly defined. YA also face worse outcomes than those with average-onset (≥40 years) BC, even after adjusting for tumor characteristics. How genomic features contribute to these disparities and influence outcomes is not well understood. This study aims to characterize clinical and genomic alterations in YA with BC at our institution.Methods: Patients consented to the Total Cancer Care protocol (as part of the Oncology Research Information Exchange Network) at the Ohio State University, whose tumors underwent RNA sequencing and tumor/germline whole exome DNA sequencing were included for analysis. YA were considered < 40 years, and average-onset were considered ≥40 years. We focused on clinical characteristics, as well as mutations, copy number alterations, and mRNA expression changes in genes relevant to BC. Descriptive statistics were used to characterize the frequency and co-occurrence of alterations. The Chi-square test was used to assess associations between categorical variables, and the Wilcoxon rank-sum test was applied to compare continuous variables between groups. Kaplan-Meier survival analyses and log-rank tests were conducted to explore associations between genomic alterations and overall survival (OS). Results: A total of 825 patients were included in the analysis: 98.2% (n=813) were women, 1.4% (n=12) were men, and 0.4% (n=3) had unreported sex. There were 10% (n=83) of patients aged <40 years, and 90% (n=742) were aged ≥40 years. The median age at diagnosis in the <40 cohort was 36.0 years, compared to 56.3 years in the ≥40 cohort. In both age cohorts, most patients were White: 89.3% (n=75) in the <40 cohort and 90.4% (n=672) in the ≥40 cohort. Black patients comprised 10.7% (n=9) and 7% (n=53) in <40 and ≥ 40 cohorts, respectively (p=0.974). Higher pathologic lymph node (LN) involvement was more common in younger patients: 20.3% (n=17) of <40 cohort had pN2 or pN3 disease versus (vs) 15.3% (n=114) in the ≥40 cohort (p=0.066). The median tumor mutational burden (TMB) was significantly lower in the <40 cohort: 5.5 vs 9.7 in the ≥40 cohort (p = 0.001). Genes significantly enriched in the <40 cohort (compared to ≥40 years) were: NBPF12 22.5% (n=16) vs 10.8% (n=73), p =0.007; CDK17 15.3% (n=11) vs 4.5% (n=30); ZNF280B 11.1% (n=8) vs 3.85% (n=26). Genes enriched in the ≥40 cohort were: CDH1 20.1% (n=136) vs 4.2% (n=3); p = 0.003; PIK3CA 39.1% (n=264) vs 18.1% (n=13), p = 0.0003. Looking at specific PIK3CA alterations, H1047 mutations were significantly more frequent in the ≥40 cohort. Lastly, the probability of OS was significantly worse in patients aged <40 compared to those ≥40 years (p= 0.02). Conclusion: While this analysis corroborates existing evidence that, compared to patients ≥40 years diagnosed with BC, YA patients exhibited higher rates of advanced LN involvement, lower TMB, and significantly worse OS, novel findings included differences in genomic alterations by age. Specifically, differences in the frequency of targetable alterations such as PIK3CA highlight biologically and clinically meaningful differences in BC by age, underscoring the need for age-specific therapeutic strategies and further investigation.

M. Gonzalez Rodriguez¹, K. Amilliano², S. Pernas³, P. Sánchez⁴, I. Blancas⁵, E. Gal-Yam⁶, M. Lopez-Flores⁷, E. Lopez-Miranda⁸, T. Curiel⁹, C. Rodriguez¹⁰, A. Pous¹¹, E. Galve-Calvo¹², E. Sanfeliu¹³, M. Pujol¹⁴, C. Polo¹⁵, C. Guardia¹⁶, L. Paré¹⁷, M. Marín-Aguilera¹⁷, A. Herrera¹⁸, A. Sebastian¹⁹, A. Prat¹, O. Martínez-Sáez¹, T. Pascual¹; ¹Medical Oncology, Hospital Clínic Barcelona, Barcelona, SPAIN, ²Medical Oncology, University Hospital Sant Joan de Reus, Tarragona, SPAIN, ³Medical Oncology, Catalan Institute of Oncology (ICO)-L’Hospitalet de Llobregat, Barcelona, SPAIN, ⁴Medical Oncology, Hospital Clínico Universitario Virgen de la Arrixaca, Murcia, SPAIN, ⁵Medical Oncology, San Cecilio University Hospital, Granada, SPAIN, ⁶Medical Oncology, Jusidman Cancer Center, Sheba Medical Center, Ramat Gan, ISRAEL, ⁷Medical Oncology, University Hospital of León, León, SPAIN, ⁸Medical Oncology, Hospital Universitario Ramon y Cajal, Madrid, SPAIN, ⁹Medical Oncology, University Hospital of Santiago de Compostela (SERGAS), Santiago de Compostela, SPAIN, ¹⁰Medical Oncology, Hospital Universitario de Salamanca-IBSAL, Salamanca, SPAIN, ¹¹Medical Oncology, Catalan Institute of Oncology (ICO)-Badalona, Germans Trias I Pujol University Hospital, Barcelona, SPAIN, ¹²Medical Oncology, Basurto University Hospital, Bilbao, SPAIN, ¹³Pathology Department, Hospital Clínic Barcelona, Barcelona, SPAIN, ¹⁴Medical Oncology, Translational Genomics and Targeted Therapies in Solid Tumors, August Pi i Sunyer Biomedical Research Institute (IDIBAPS), Barcelona, SPAIN, ¹⁵Operational department, SOLTI Cancer Research Group, Barcelona, SPAIN, ¹⁶Scientific department, SOLTI Cancer Research Group, Barcelona, SPAIN, ¹⁷Scientific department, Reveal Genomics, Barcelona, SPAIN, ¹⁸Marketing department, Reveal Genomics, Barcelona, SPAIN, ¹⁹Operational department, Reveal Genomics, Barcelona, SPAIN.

Background: HER2-positive breast cancer (BC) accounts for approximately 20% of all breast tumors. Despite major advances in HER2-targeted therapies, many patients with early-stage HER2-positive BC remain overtreated, leading to unnecessary toxicity and increased healthcare costs. HER2DX is a genomic diagnostic tool that stratifies patients by risk of recurrence and likelihood of achieving pathological complete response (pCR) after neoadjuvant therapy. The DEFINITIVE trial (www.thedefinitivetrial.eu) evaluates the clinical utility of HER2DX to guide treatment de-escalation or escalation in early-stage HER2-positive BC. We report preliminary feasibility and performance metrics from the first 80 screened patients. Methods: DEFINITIVE (NCT06446882) is an international, multicenter, randomized, open-label, two-arm clinical trial designed to evaluate the impact of HER2DX-guided treatment versus standard of care on health-related quality of life, safety, and efficacy outcomes. Eligible patients include premenopausal or postmenopausal women and men with stage II-IIIA HER2-positive BC suitable for neoadjuvant therapy.This interim analysis includes all patients screened from study initiation on October 30, 2024, through June 19, 2025. All patients underwent centralized screening with HER2DX using formalin-fixed paraffin-embedded (FFPE) tumor samples shipped to a reference laboratory. Operational endpoints included the number of samples received, HER2DX testing success rate, and turnaround time (TAT) from sample receipt to result delivery. All data were collected prospectively as part of the trial workflow. Results:As of the data cutoff, 80 patients had been screened across 11 sites in 2 countries. A total of 80 samples were received by the central laboratory, and HER2DX testing was successfully completed in 78 cases, yielding a success rate of 97.5%. Test failures were due to insufficient tumor content or low RNA purity. The median TAT was 8 working days (interquartile range [IQR] 7-10; range 3-16), confirming the feasibility of rapid centralized testing.The median time from tumor biopsy to sample reception at the central lab was 29.5 days (IQR 22-42), reflecting real-world variability in local processing and logistics.Among the 78 patients with available HER2DX results, 42.3% were classified as having low pCR likelihood, 30.8% as intermediate, and 26.9% as high. Based on HER2DX risk score, 47.4% were considered low-risk and 52.6% high-risk. ERBB2 mRNA expression levels also reflected biological heterogeneity, with 11.6% classified as low, 25.6% as intermediate, and 62.8% as high expression. Conclusions: Centralized HER2DX-based screening in the DEFINITIVE trial has proven operationally feasible on an international scale. High testing success rates, acceptable TAT, and biologically informative distributions support the integration of genomic classifiers into real-time treatment decision-making in prospective clinical trials. Funding:This project has received funding from the European Union’s Horizon Europe research and innovation programme under grant agreement No. 101136953.

P. P. Advani¹, J. Harvey¹, J. Polley¹, M. Weidner¹, A. Arnold¹, D. Haley¹, R. Bollam¹, R. Rao¹, S. Chumsri¹, K. Hugghins¹, M. Acampora¹, C. Vivelo², H. Hernan³, L. Hrycyniak⁴, J. Zweng⁵, H. Gierman⁶, S. Bhattacharya¹, D. Mukhopadhyay¹; ¹Hematology and Oncology, Mayo Clinic, Jacksonville, FL, ²Medical Affairs, Elephas Biosciences, Madison, WI, ³Clinical team, Mayo Clinic, Madison, WI, ⁴R and D, Mayo Clinic, Jacksonville, WI, ⁵Hematology and Oncology, Elephas Biosciences, Madison, WI, ⁶Chief Scientific Officer, Elephas Biosciences, Madison, WI.

Background: Immune checkpoint inhibition (ICI) in combination with chemotherapy is associated with improved outcomes in a subset of patients with triple negative breast cancer (TNBC). However, biomarkers for predicting patient response, such as PD-L1 expression, have limited predictive efficiency. Voabil et al. has shown that ex vivo cytokine profiling of live tumor tissue resections treated with αPD-1 can predict patient response to ICI (Voabil et al., Nat Med. 2021). However, this method requires large amounts of tissue to address intra-specimen tumor heterogeneity, which is not feasible with standard of care (SOC) diagnostic core needle biopsies (CNBs) for breast cancer, where tissue is scarce. We show for the first time that ex vivo cytokine profiling is possible in live tumor fragments (LTFs) generated from CNBs in breast tissue.Methods: We present a cytokine profiling platform to assess response to ICI treatment in LTFs generated from CNBs. CNBs were collected from patients who were identified as potential candidates for SOC ICI (by treating physician) and enrolled in one of two prospective observational clinical trials (NCT05520099 and NCT06349642). CNBs were cut by an automated, proprietary cutting instrument to generate LTFs ofapproximately 300 µm in thickness. To address tissue heterogeneity and mitigate tissue scarcity in CNBs, we used a sequential treatment strategy in which control and ICI treatment were applied to LTFs in the same well, in a phased approach. This strategy provides a solution for the variability observed in cross-well comparisons. LTFs were encapsulated in hydrogel and treated with IgG control followed by αPD-1 (BioXCell). Cytokine production rates were measured longitudinally using multiplex assays assessing>45 cytokines, including 15 cytokines assayed by Voabil et al., and the fold changes of cytokine production rates in response to ICI over IgG control were calculated. Results: Using non-breast tumor tissue, we validated that our platform can measure changes in cytokine production in LTFs following sequential treatment, and that these measurements are comparable to those observed in a cross-well setting. In a small cohort of TNBC specimens, we show changes in cytokine production rates in response to ex vivo sequential control and ICI treatment. Ten cytokines, including IFN-γ and CXCL10, shown to be predictive of patient response to ICI in Voabil et al. were found upregulated in response to αPD-1 in TNBC specimens. Future determination of correlation with clinical response data (pathological complete response for early stage and RECIST 1.1 for metastatic patients) gathered from NCT05520099 and NCT06349642 will be used to validate platform responders and non-responders. To date, 11 patients with confirmed TNBC diagnoses have enrolled in these clinical studies. Conclusion: We report, for the first time to our knowledge, the ability to detect immune checkpoint inhibitor responses in live TNBC tissue obtained via CNBs. This proof-of-concept lays the groundwork for integrating functional tissue responses with clinical correlation data, with important implications for guiding current ICI therapies and advancing this platform for use in next-generation immuno-oncology treatments for TNBC.

M. Bakre¹, T. Durgekar¹, S. BA¹, P. Shrivastava¹, G. Basaran², G. Ozge², T. Korkmaz³; ¹R and D, OncoStem Diagnostics Pvt Ltd, Bangalore, INDIA, ²Medical Oncology, Acibadem University Scool of Medicine, Istanbul, TURKEY, ³Medical Oncology, Acibadem Maslak Hospital, Istanbul, TURKEY.

Objective: ~70% of HR+/HER2- early breast cancer have low risk of breast cancer recurrence. Hence, prognostication in these patients to assess chemotherapy benefit is a huge value add. Online tools such as Nottingham prognostic index (NPI), PREDICT etc are often used as they are quick and free. However use of genomic prognostic tests like Oncotype DX (ODX), Mammaprint (MP) etc is increasing. CanAssist Breast (CAB) is a proteomics-based prognostic test that uses an AI-based algorithm to segregate EBC patients as ‘low or high’ risk for recurrence. CAB is validated in global studies and clinically used on ~10,000 patients to date in the Indian subcontinent, UAE, Turkey, Iran, and Saudi Arabia. Comparative analysis of prognostic tests is important to evaluate the relative performance of different prognostic tests in a population and assess how a new test performs in relation to established, validated ones. We have compared CAB with NPI, PREDICT, ODX, and MP, and here we showcase the results of those comparative studies.Methods: A patient cohort of 1474 from Europe, India and US was used to compare CAB with NPI and PREDICT. NPI risk groups were categorized into three prognostic groups: good (GPG-NPI index ≤ 3.4), moderate (MPG 3.41-5.4), and poor (PPG > 5.4). Patients with chemotherapy benefit of 15.5) categories. Agreement between CAB and NPI/PREDICT risk groups were assessed by kappa coefficient. Retrospective comparison of risk stratification by CAB with ODX and MP was done with 109 (US and India) and 43 (EU) patients, and prospectively with a total of 116 Turkish patients- 58 patients in each group. Accuracy/ negative predictive value was calculated using MedCalc. Concordance of CAB with ODX or MP was calculated using the overall percentage agreement.Results: Risk proportions generated by all tests were: CAB low:high 74:26; NPI good:moderate:poor prognostic group- 38:55:7; PREDICT low:high 63:37. 65% of NPI-MPG patients were called low risk by CAB. From PREDICT high risk patients, CAB segregated 51% as low risk, thus preventing over-treatment in these patients. Overall, there was a fair agreement between CAB and NPI [κ=0.31(0.278-0.346)] / PREDICT [κ=0.398 (0.35-0.446)], with a concordance of 97% / 88% between CAB and NPI/PREDICT low risk categories. In cohorts with mostly T1N0 patients, NPI and PREDICT segregated more as low risk compared to CAB, suggesting that T1N0 patients with aggressive biology are missed by online tools but not by CAB. Comparison of CAB with ODX retrospectively (n=109) and prospectively (n=58) showed similar low:high risk proportions as 83:17 and 79:21 for CAB; 90:10 and 83:17 for ODX. An overall concordance of 75% and 65%, and low risk concordance of 82% and 77% was observed between both the tests in retrospective and prospective studies. Retrospective (n=43) and prospective (n=58) comparison of CAB with MP showed similar low:high risk proportions of 65:35 and 66:34 for CAB; 56:46 and 52:48 for MP. Overall concordance between the two tests was 75% in the retrospective study and 62% in the prospective study, while low risk concordance was 83% and 77%, respectively.Conclusion: CAB provided unbiased risk stratification across cohorts of various geographies with minimal impact by clinical parameters, whether compared with online tools and tumor based prognostic tests. CAB is useful for all EBC patients and specifically in NPI-MPG and PREDICT high risk patients for making accurate decisions on chemotherapy use. CAB shows good concordance with ODX and MP in low risk categories. This, coupled with high accuracy comparable with ODX, shows that CAB is an excellent, cost-effective, and quick alternative.

S. Shivhare¹, C. Sanchez¹, R. Larson¹, L. Dobrolecki², M. Lewis², J. Chen³, M. Sebastian⁴, A. Evdokimova⁵, D. Goncharova⁵, A. Alexander⁶, A. Nasrazadani⁶, R. Layman⁶, B. Lim⁶, V. Valero⁶, A. Lucci³, The MDACC Inflammatory Breast Cancer Team, W. Woodward¹; ¹Department of Breast Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX, ²Lester and Sue Smith Breast Center, Baylor College of Medicine, Houston, TX, ³Department of Breast Surgical Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX, ⁴Department of Veterinary Medicine & Surgery and Department of Translational Molecular Pathology, The University of Texas MD Anderson Cancer Center, Houston, TX, ⁵BostonGene RnD, BostonGene Corporation, Waltham, MA, ⁶Department of Breast Medical Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX.

Purpose: Inflammatory breast cancer (IBC) is a rare yet exceptionally aggressive subtype that accounts for a disproportionate number of breast cancer deaths. A significant barrier to effective treatment is its ability to invade the lymphatic system and evade immune suppression. Emerging research highlights the potential role of C-C chemokine receptor type 7 (CCR7) in driving tumor progression and modulating immune responses within IBC. Baseline changes in CCR7 gene expression have not been reported in IBC. We investigated CCR7 gene amplification, as well as protein expression and spatial distribution in patient tumors, patient-derived xenografts (PDXs), and IBC cell lines. Methods: PDX characterization of non-IBC models for CCR7 was performed using data from the BCM PDX portal (PDX Insights), and sections from selected PDX were stained for multiplex immunofluorescence (mIF) and imaged with a Nikon Ni-E microscope. Cell lines were subjected to immunoblotting for CCR7, and conditioned media were analyzed by ELISA for CCR7 ligands CCL19 and CCL21. Clinical analysis was performed retrospectively, based on an IRB-approved protocol using genomic data from 18 hormone receptor-positive (HR+) HER2-negative (HER2-) IBC and 24 HR-negative (HR-) HER2- IBC cases which were compared to 150 and 104 subtype-matched non-IBC cases, respectively. Results: CCR7 gene expression did not differ between IBC and non-IBC cases independent of subtype. Four IBC patients in each subtype had gain or amplification of CCR7, while six patients in each subtype had deletion. mRNA expression did not correlate with copy number alterations (CNAs). In triple-negative and HER2+ IBC cell lines, CCR7 protein expression was high. CCR7 ligands were not detected in conditioned media of IBC cell lines. Among non-IBC PDX models characterized from PDX Insights, 9/53 demonstrated gain or amplification of CCR7, while 25/53 demonstrated loss. CCR7 gene expression was not concordant with CNAs; we observed 11% CCR7 gene expression in HR+ HER2- IBC patients and 16% in HR- HER2- IBC patients. mIF analysis of PDX sections from BCM-3104, BCM-3807, and BCM-15029, representing CCR7 amplified/high expression (+/+), non-amplified/high expression (-/+), and non-amplified/low expression (-/-), respectively, revealed significant membranous expression only in the +/+ PDX context with concurrent high expression of podoplanin and ligand CCL21. CCR7 protein and gene expression were roughly correlated, with -/+ expression level being intermediate between -/- and +/+. H&E staining showed tumor necrosis and an increase in inflammatory infiltrate relative to CCR7 gene expression. Conclusion: Patient data suggest that CCR7 CNAs are common in both inflammatory and non-inflammatory breast cancer phenotypes within HR+ HER2- and triple-negative breast cancer subtypes and are not concordant with gene expression overall. In PDX tissues, membranous CCR7 was predominantly found in the amplified PDX context. Additional work is needed to distinguish the impact of membranous versus cytoplasmic CCR7 expression. CCL21 showed greater expression than CCL19, highlighting potential ligand-specific functionality of the CCR7 axis. These characterizations provide important baseline data for further consideration of CCR7 as a target for aggressive breast cancer.

A. A. Davis¹, L. S. Welch², J. Saha³, L. Bucheit², M. D. Lipsyc-Sharf⁴, A. Bardia⁴, C. X. Ma¹, M. Cristofanilli⁵, E. L. Podany¹; ¹Department of Medicine, Washington University in St. Louis, St. Louis, MO, ²Medical Affairs, Guardant Health, Palo Alto, CA, ³Outcomes and Evidence, Guardant Health, Palo Alto, CA, ⁴Department of Medicine, Division of Hematology/Oncology, UCLA Health, Santa Monica, CA, ⁵Department of Medicine, Division of Hematology/Oncology, Weill Cornell Medicine, New York, NY.

Background: Black patients (pts) are more likely to be diagnosed with mBC and have higher mortality rates than White pts, in part due to social determinants of health, higher rates of de novo mBC, and possible genomic differences by race and ethnicity. Most variant classification efforts have focused on White populations, creating uncertainty about prevalence and significance of VUS in Black pts. It is also unclear if mutation frequency differences exist in genes with known targets, such as PIK3CA and ESR1. We examined a real-world (rw) database for racial differences in VUS, ESR1 and PIK3CA detected by circulating tumor DNA (ctDNA), matched therapy use, and outcomes by race for ESR1 and PIK3CA. Methods: Pts with mBC and ctDNA (Guardant360) results post-mBC diagnosis were identified from Guardant INFORM (n=46,286), which includes de-identified records with clinical genomic data, claims, and race when available. For Black and White pts (n=34,149), overall VUS rates were analyzed by race, as defined by OncoKB annotation. ESR1 and PIK3CA codons of interest [ESR1 ligand binding domain (LBDmut); PIK3CA helical or kinase domains (HKmut)], targeted therapy matching (elacestrant and alpelisib), and outcomes were evaluated by race. Outcomes were assessed using rw time to treatment discontinuation (rwTTD), time to next treatment (rwTTNT), and overall survival (rwOS) in months, stratified by race. Chi-squared tests compared proportions; log-rank tests compared time-to-event outcomes. Results: There were 12.5% Black (5,791/46,286) and 61.2% White (28,358/46,286) pts included. Mean VUS frequency was similar between groups (2.40 Black vs 2.47 White, p>0.05). While White pts with ESR1 mutations more frequently had ESR1 LBDmut vs Black pts, there was no difference in matched therapy use with elacestrant by race for ESR1 LBDmut (Table 1). Black pts with ESR1 LBDmut and 2L elacestrant had significantly shorter rwTTD (3.5 vs 4.2, p=0.024) and shorter numeric rwTTNT (5.9 vs 8.9, p=0.32). White pts also had significantly more PIK3CA mutations, yet among PIK3CA HKmut, there was no difference in matched therapy use with alpelisib by race (Table 1). Black pts with PIK3CA HKmut treated with 2L alpelisib compared with white pts had numerically longer rwTTD (4.4 vs 3.7, p=0.48), rwTTNT (8.7 vs 6.6, p=0.33), and rwOS (33.9 vs 25.9, p=0.38); none reached statistical significance. Conclusion:In this large, rw analysis of a diverse patient population, differences in VUS rate, ESR1/PIK3CA matched therapy use, and outcomes were not observed between Black/White pts, suggesting the need for additional exploration of race-based differences on mBC outcomes. This study underscores the need for continued inclusion of diverse populations in genomic discovery and prospective studies to clarify known outcome disparities.

Y. Inoue¹, T. Osako², M. Akiya², T. Chiba², Y. Haruyama¹, U. Nakadaira¹, J. Masuda³, Y. Kimura¹, A. Iesato¹, M. Nishimura³, Y. Ozaki³, T. Maeda¹, N. Uehiro¹, N. Yamashita¹, T. Kobayashi³, T. Sakai¹, T. Takano³, T. Ueno¹; ¹Breast Surgical Oncology, Cancer Institute Hospital of JFCR, Tokyo, JAPAN, ²Division of Pathology, Cancer Institute of JFCR, Tokyo, JAPAN, ³Breast Medical Oncology, Cancer Institute Hospital of JFCR, Tokyo, JAPAN.

Background: We are conducting an ongoing project using archival tissue samples from primary breast cancer cases treated with surgery alone at our institution, without any adjuvant systemic therapy. Tissue microarrays (TMAs) have been constructed, and various protein expressions are being re-evaluated using contemporary immunohistochemical techniques. This treatment-naïve cohort provides a unique opportunity to assess the intrinsic prognostic significance of biomarker expression, free from the confounding effects of systemic therapy.Androgen receptor (AR) is a steroid hormone receptor expressed in a subset of breast cancers and has attracted growing interest as a potential prognostic and therapeutic marker. Forkhead-box protein A1 (FOXA1) is a pioneer transcription factor known to act as a super-enhancer of the estrogen receptor (ER), playing a crucial role in mammary gland development. It is well established as a favorable prognostic marker in ER-positive, HER2-negative breast cancer. In recent years, however, FOXA1 expression has also been reported in other subtypes, including ER-negative tumors, although its clinical significance in these contexts remains uncertain. Whether FOXA1 functions primarily as a true prognostic marker or rather as a surrogate for endocrine responsiveness remains an open question. In this study, we report the results of immunohistochemical analyses for AR and FOXA1 in a treatment-naïve historical cohort, aiming to clarify their intrinsic prognostic relevance across breast cancer subtypes. Methods: We retrospectively analyzed primary breast cancer cases operated on at our institution between 1979 and 1982, from which TMAs were constructed by sampling three representative areas of the invasive tumor component. All pathological features, including immunohistochemical markers, were re-evaluated using current standardized protocols. Clinical outcomes were obtained from institutional records. Results: Among 935 eligible cases, 758 were available for AR and FOXA1 immunohistochemical evaluation. Subtype distribution based on ER and HER2 status was as follows: ER+/HER2− (n = 538, 71.3%), ER+/HER2+ (n = 47, 6.2%), ER−/HER2+ (n = 74, 9.8%), and ER−/HER2− (n = 96, 12.7%), comparable to current subtype frequencies.Using a 10% cutoff for positivity, AR expression was observed in 69.1% of ER+/HER2−, 57.5% of ER+/HER2+, 44.6% of ER−/HER2+, and 16.7% of ER−/HER2− tumors. In survival analysis, strong AR expression was significantly associated with improved distant recurrence-free survival (DRFS) (p = 0.0239).FOXA1 expression, using the same cutoff, was detected in 96.4% of ER+/HER2−, 100% of ER+/HER2+, 92.9% of ER−/HER2+, and 23.2% of ER−/HER2− tumors. Notably, FOXA1 was expressed in a subset of ER-negative tumors. FOXA1 expression was not significantly associated with DRFS, including in ER-positive tumors. In the ER−/HER2− subtype, although no significant difference in DRFS was observed based on Kaplan-Meier analysis, among patients who experienced recurrence, those with FOXA1-positive tumors had a significantly longer disease-free interval (DFI) (p = 0.0005). Conclusion: In this historical cohort of breast cancer patients untreated with adjuvant therapy, strong AR expression was significantly associated with improved DRFS. FOXA1 was expressed across all intrinsic subtypes and, contrary to previous reports, did not demonstrate prognostic significance in ER-positive tumors in the absence of systemic therapy. These findings suggest that FOXA1 may be more reflective of endocrine responsiveness rather than intrinsic tumor aggressiveness. The observation that FOXA1-positive tumors had a longer DFI suggests a potential role in late recurrence, particularly in ER-negative subtypes, and warrants further investigation.

D. Trapani¹, S. Deshmukh², S. Wu², J. Xiu², N. Lin³, G. Curigliano¹, P. Spanheimer⁴, S. Gandhi⁵, S. Chumsri⁶, S. Graff⁷, M. Lustberg⁸, G. Sledge Jr², S. Tolaney³, J. Leone⁹; ¹oncology and hematology, University of Milan, milan, ITALY, ²Caris Life Sciences, Caris Life Sciences, Phoenix, AZ, ³breast oncology, Dana farber cancer institute, Boston, MA, ⁴medical oncology, University of North Carolina at Chapel Hill, Chapel Hill, Chapel Hill, NC, ⁵medical oncology, Winship Cancer Institute of Emory University, milan, GA, ⁶medical oncology, Mayo Clinic, Jacksonville, FL, ⁷medical oncology, Lifespan Cancer Institute, Legorreta Cancer Center, Brown University, Providence, RI, ⁸medical oncology, Yale School of Medicine, New Haven, CT, ⁹breast oncology, Dana farber cancer institute, milan, MA.

Background: Antibody-drug conjugates (ADCs) are increasingly used in the treatment of breast cancer due to their superior efficacy compared to conventional chemotherapy. The management of male breast cancer is largely extrapolated from clinical data in females and evidence regarding the efficacy of ADCs in men remains limited. This study aimed to compare survival outcomes between male and female breast cancer patients treated with ADCs across the spectrum of human epidermal growth factor receptor 2 (HER2) expression. Methods: We included total of 2,954 (Male: 34, Female: 2,920 breast cancer samples) from patients who underwent treatment with ADCs. We excluded patients with triple-negative breast cancer status due to the rarity of this subtype in males and the worse prognosis of triple-negative breast cancer, so that the cohorts would be more homogeneous without overrepresentation of triple-negative breast cancer in females. We categorized patients by HER2 status as: HER2-null (if there was 0% stain on HER2 IHC), HER2-ultra low (if HER2 IHC was 0 but with stain 1-10%), HER2-low (if HER2 IHC was 1+ or 2+ with a negative chromogenic in situ hybridization [CISH] assay), and HER2-positive (if HER2 IHC was 3+ or 2+ with a positive CISH assay). Real-world median overall survival was derived from insurance claims and calculated from the start of first ADC with either trastuzumab deruxtecan (T-DXd) or trastuzumab emtansine (T-DM1) or sacituzumab govitecan (SG) to last contact using Kaplan-Meier. Statistical significance was assessed using chi-square and Mann-Whitney U. Results: The distribution of sex by HER2 expression group was: HER2-null: 8 males vs 612 females; HER2-ultra low 6 males vs 467 females; HER2-low 11 males vs 903 females; HER2-positive 9 males vs 938 females. Median survival from treatment initiation with either T-DXd or SG in the HER2-null cohort was 21.31 months in males vs 15.72 months in females (HR 0.82, 95% CI 0.33 – 2.01, p=0.67); in the HER2-ultra low cohort was 17.73 months in males vs 14.47 months in females (HR 0.84, 95% CI 0.26 – 2.63, p=0.76); in the HER2-low cohort was 11.08 months in males vs 18.62 months in females (HR 1.53, 95% CI 0.63 – 3.7; p=0.34). In the HER2-positive cohort, 5/9 males (55.5%) and 544/938 females (58%) were treated with T-DXd. Median survival from treatment initiation with either T-DXd or T-DM1 in the HER2-positive cohort was 36.55 months in males vs 46.52 months in females (HR 1.77, 95% CI 0.84 – 3.74, p=0.12). Conclusions: In this study, survival appears to be comparable between males and females with breast cancer across the spectrum of HER2 expression when both are treated with ADCs, although the survival in the HER2-positive cohort appears numerically lower in males, consistent with data on HER2-positive MaBC. An important limitation is the small sample size of the male breast cancer cohorts, and the lack of a categorization of HER2 in HR+ vs HR-, for the small MaBC sample size. However, to our knowledge this is the first analysis evaluating outcomes of men treated with ADCs across the spectrum of HER2 expression.

J. Hundal¹, R. Plagens², A. Elliott³, G. Sledge³, C. Travaline⁴, S. Graff⁵, M. Lustberg⁶, M. Vasekar⁷, J. Leone⁸; ¹Taussig Cancer Institute, Cleveland Clinic, Cleveland, OH, ²Taussig Cancer Institute, Caris Life Science, Irving, TX, ³Caris, Caris Life Sciences, Irving, TX, ⁴Hershey Cancer Institute, Penn State, Hershey, PA, ⁵Legorreta Cancer Center, Brown University, Providence, RI, ⁶Yale Cancer Center, Yale University, New Haven, CT, ⁷Hershey Cnacer Institue, Penn State University, Hershey, OH, ⁸Dana-Farber Cancer Institute, Dana-Farber Cancer Institute, Harvard Medical School,, Boston, MA.

Background: Metabolic dysregulation, including hyperglycemia (HG) and type 2 diabetes mellitus (T2DM), is increasingly prevalent among breast cancer (BC) patients (pts) and may influence treatment outcomes. In hormone receptor-positive (HR+) BC, metabolic comorbidities could affect tumor biology and response to endocrine therapy. This study integrates real-world clinical outcomes with multi-omic tumor profiling to evaluate the potential impact of HG/T2DM on endocrine therapy outcomes and tumor molecular signatures in pts with HER2-/HR+ BC. Methods: This retrospective analysis utilized BC samples profiled by Caris Life Sciences (Phoenix, AZ), including next-generation sequencing (592-gene panel or whole exome for DNA and whole transcriptome for RNA) and immunohistochemistry (IHC). Tumors were classified as HER2- (≤1+ intensity or ≤10% cells stained by IHC, or 2+/>10% and CISH-negative) and HR+ (ER or PR ≥1+/≥1%). Pts with no prior systemic therapy who initiated first-line aromatase inhibitor and/or fulvestrant (AI/Fulv) within 100 days post-biopsy were identified using linked claims data. ICD-10 codes were used to define HG and T2DM status, and medication exposure was assessed for insulin, metformin, and GLP-1 receptor agonists (Ins/Met/GLP-1RA). Clinical endpoints included overall survival (OS; time from first AI/Fulv to death or last contact) and time on treatment (ToT; from first to last AI/Fulv), as assessed via Cox proportional hazards models and log-rank tests. Among pts without reported exposure to Ins/Met/GLP1-RA, tumor molecular profiles were compared between pts with HG/T2DM and metabolically healthy controls (no HG/T2DM) using Mann-Whitney U and chi-square tests. Results: A total of 2,074 pts with HER2-/HR+ BC treated with first-line AI/Fulv were included in the analysis, comprising three cohorts: 1,119 pts without HG/T2DM and no exposure to Ins/Met/GLP-1RA; 398 pts with HG/T2DM and no exposure to these agents; and 557 pts with HG/T2DM who had received Ins/Met/GLP-1RAs. Median OS was 49, 48, and 52 months, respectively (p = 0.71), and median ToT was 17, 21, and 19 months, respectively (p = 0.37). Subgroup analyses stratified by CDK4/6 inhibitor exposure in first- or later-line therapy revealed consistent patterns, with no statistically significant differences in survival across metabolic strata. Molecular analyses identified notable distinctions in tumors from pts with HG/T2DM. Pathogenic mutations in PIK3CA (50.4% vs. 42.9%; p = 0.01), CASP8 (1.5% vs. 0.1%; p < 0.001), ACVR1B (2.1% vs. 0.2%; p = 0.02), and TSC2 (0.9% vs. 0.1%; p = 0.02) were more prevalent in the HG/T2DM group. Deletion events in CYLD (2.8% vs. 0.5%; p < 0.01), B2M (0.7% vs. 0.0%; p = 0.02), CDKN2C (1.1% vs. 0.1%; p = 0.02), RAD54L (1.1% vs. 0.1%; p = 0.02), and FANCA (2.8% vs. 0.9%; p = 0.03) were similarly enriched. Amplifications of HSP90AB1 and SS18 were detected only in HG/T2DM tumors (0.7% and 1.1%, respectively; p = 0.02 for both), while MDM2 amplification was more frequent in controls (3.1% vs. 0.7%; p = 0.03). Conclusion: Metabolic dysfunction was not associated with significant differences in clinical outcomes among pts receiving endocrine therapy. However, tumors from pts with HG/T2DM exhibited distinct molecular alterations, including higher frequencies of PIK3CA mutations, MDM2 and HSP90AB1 amplifications, and deletions in CYLD, B2M, and CDKN2C, suggesting enhanced proliferative signaling and reduced immune activation. These findings may contribute to endocrine resistance and underscore the need for biomarker-driven, metabolically informed treatment strategies. Prospective studies are warranted to validate these observations and inform future therapeutic approaches.

B. To¹, V. Prasath¹, D. Veney¹, C. Rolfo¹, V. Adalsteinsson², N. Lin³, S. M. Tolaney³, H. A. Parsons³, R. Wesolowski¹, D. Quiroga¹, D. Stover¹; ¹Medical Oncology, The Ohio State University Comprehensive Cancer Center, Columbus, OH, ²Cancer Diagnostics, Broad Institute of MIT and Harvard, Cambridge, MA, ³Medical Oncology, Dana-Farber Cancer Institute, Boston, MA.

Background: Metastatic triple-negative breast cancer (mTNBC) is a poor prognosis subset ofbreast cancer characterized by shorter survival and higher circulating tumor DNA (ctDNA) ‘tumorfraction’ (TF) relative to other breast cancer subtypes. TF is the proportion of ctDNA originatingfrom tumors cells, and early changes in ctDNA TF may correlate with therapy response,potentially offering a rapid response biomarker for early phase studies. Furthermore, there is littleknown regarding TF dynamics in the hours immediately after initiation of systemic therapy –including whether there is a ‘surge’ in ctDNA.Methods: Banked plasma samples were identified from four completed phase Ib/II clinical trialsenrolling patients with mTNBC: heat shock protein 90 (HSP90) inhibitor onalespib with paclitaxel(NCT02474173), MEK inhibitor trametinib alone and in combination with AKT inhibitorGSK2141795/uprosertib (NCT01964924), multi-kinase inhibitor cabozantinib monotherapy(NCT02260531), and JAK1/2 inhibitor ruxolitinib monotherapy(NCT01562873). For theonalsepib+paclitaxel study, plasma was collected pre-infusion, end-of-infusion (EOI), then0.5/1/2/4/6/8/24 hours post-infusion for 1) onalespib alone (day -7); 2) paclitaxel alone (day 1);and 3) onalespib + paclitaxel (day 8). For the remaining studies, plasma was collected pre-treatment on cycle 1 day 1 (C1D1) and prior to cycle 2 day 1 (C2D1). Plasma samples underwentshallow whole genome sequencing (sWGS) and TF determination. The primary objectives wereto evaluate change in TF and association with best overall response and progression-freesurvival.Results: To evaluate acute ctDNA TF change in the onalespib+paclitaxel trial, ctDNA TF wassuccessful in 313/317 (98.7%) of available plasma samples from 14 patients. There wassignificant decline in TF from pre-infusion to 6 hours for paclitaxel alone (Wilcoxon signed rankp=0.03) but no significant change for onalespib alone or onalespib + paclitaxel from pre-infusion to 6- or 24-hours. There was significant decline in TF from pre-infusion day -7 to cycle 1, day 9(median TF 16% to 6.5%, Wilcoxon signed rank p=0.004). No significant ‘surge’ was identifiedand despite significant decline in TF over the first full cycle of therapy. To evaluate change in TFacross multiple Phase Ib/II clinical trials of targeted therapies in mTNBC, we identified 57 totalpatients across studies (onalsepib+paclitaxel trial n=8; trametinib trial n=22; cabozantinib trialn=19; ruxolitinib trial n=8) with available plasma sample prior to C1D1 (T1) and follow-up plasmasample 14-28 days after therapy initiation (T2). Change in ctDNA from T1:T2 was significantlyassociated with best overall response (ANOVA p=0.01). Clearance of ctDNA at T2 (TF=0) wasassociated with significantly longer PFS (median PFS 10.3 months clearance vs. 2.3mo noclearance; log-rank p=0.013). Similarly, TF that was stable or declined T1:T2 was associated withsignificantly improved PFS relative to TF increase (log-rank p=0.003).Conclusion: There was no significant ‘surge’ in ctDNA TF within minutes to 24-hours of infusionof a targeted therapy, cytotoxic chemotherapy, or both in combination. Stability or decline ofctDNA TF 14-28 days after initiation of targeted therapy trial was associated with objectiveresponse and prolonged PFS. This supports further research on ctDNA dynamics immediatelyafter systemic therapy initiation, particularly in the context of early phase trials for mTNBC.

B. Conte¹, S. D’Alfonso¹, M. Mellai², D. Cora², F. Favero², V. Martini², N. Negrini¹, I. Taglialatela¹, T. Landolfo¹, B. Ruffilli³, S. Nardin⁴, F. D’Avanzo⁴, V. Rossi⁴, C. Branni⁴, M. Graeser⁵, N. Harbeck⁵, A. Gennari¹; ¹Medicina Traslazionale, Università del Piemonte Orientale, Novara, ITALY, ²UPO-CAAD, Università del Piemonte Orientale, Novara, ITALY, ³Medical Oncology, Università del Piemonte Orientale, Novara, ITALY, ⁴Medical Oncology, Ospedale Maggiore della Carità, Novara, ITALY, ⁵Oncology, West German Study Group, Munich, GERMANY.

Transcriptomic profiling of circulating T lymphocytes predicts response to neoadjuvanttherapy in patients with early breast cancer.Benedetta Conte¹, Sandra D’Alfonso², Marta Mellai², Davide Cora’¹, Francesco Favero¹, Veronica Martini³,Letizia Negrini¹, Ida Taglialatela¹, Tommaso Lupo Landolfo¹, Beatrice Ruffilli¹, Giulia Isingrini¹, FrancescaD’Avanzo³, Valentina Rossi³, Simone Nardin³, Carmen Branni³, Monika Graeser⁴, Nadia Harbeck⁴,Alessandra Gennari¹.1. Oncology, Department of Translational Medicine, Università del Piemonte Orientale, Novara, Italy.2. Department of Health Sciences, Università del Piemonte Orientale, Novara, Italy.3. Division of Oncology, University Maggiore Hospital della Carità, Novara, Italy.4. West German Study Group, Moenchengladbach, GermanyBackground. In early breast cancer (eBC), anti-tumor immunity is a key determinant of response toneoadjuvant therapy (NAT). Immune profiling currently relies on static tissue biomarkers such astumor-infiltrating lymphocytes (TILs). However, the anti-tumor immune response is a dynamicprocess undergoing substantial changes during NAT. Compelling findings show that on-treatmentTILs are a stronger predictor of pCR than baseline TILs during NAT. Circulating immune cells mayoffer a non-invasive means to monitor systemic immune response during NAT. Here we present thepreliminary data for the development of a blood-based immune assay that analyzes transcriptomicdata from circulating immune cells.Methods. Baseline circulating T lymphocytes were collected from 95 eBC patients treated at ourinstitution between 2021–2023. The cohort included 39 HER2+ cases treated with NAT plus single(n=12) or dual HER2 blockade (n=27), 36 triple-negative cases treated with NAT alone (n=19) orNAT plus pembrolizumab (n=17), and 20 HR+/HER2- cases treated with anthracyclines-basedchemotherapy. Bulk RNA was extracted for gene expression analysis. 33 immune signatures werecalculated with known prognostic role in tumor tissue. Significant analysis of microarrays (SAM)was applied to identify differences in T cell polarization between the pCR and RD group.Results. 47 patients (49%) achieved pCR. The table below shows the genomic featuresdownregulated and upregulated in pCR. At SAM analysis, T cells of responders showed lowerexpression of genes linked to Tregs (FOXP3, CCR4), monocyte recruitment (CD163, IL1R2A), andsystemic inflammation (FGFR1, IL6R). A tissue-based gene signature associated to neutrophilactivation was upregulated in non-responders, along with chemokines linked to neutrophilrecruitment (CXCL8, CXCL2, CXCL1). Similarly, a tissue-based signature associated to IL6-JAK-STAT activation pathway was downregulated in T cells of responders.Conclusions. The transcriptomic profiling of T cells carries predictive information in eBC andreplicates the prognostic information captured by tissue-based signatures. This data supports thedevelopment of blood-based immune assays in eBC GENOMIC FEATURESlog2 FOLD CHANGEFALSE DISCOVERY RATECOX6CP1-0.542.09COL18A1-0.542.09CXCL8-0.502.09Neutrophil_Chemotaxis_NatureMed_20188-0.482.09S100A8-0.462.09S100A9-0.432.09FOXP3-0.422.09CCL4L2-0.402.09CXCL1-0.382.09CD163-0.382.09ITGA6-0.382.09CXCL2-0.372.09IGF1R-0.372.09BCL2-0.362.09CD14-0.362.09IL2RA-0.362.09CXCL3-0.332.09FGFR1-0.332.09COX7CP2-0.332.09PIM2-0.332.09CCDC154-0.322.09CCR2-0.312.09CD27-0.312.09IL6R-0.312.09IL6_JAK_STAT3_Pathway_GSEASig-0.292.09

R. C. Chidebe¹, O. Akapo¹, S. C. Ipiankama¹, O. Uzodinma², S. Bornengo³, H. Harnik⁴, I. I. Okoye⁵; ¹Patients Care and Research, Project PINK BLUE – Health & Psychological Trust Centre, Abuja, NIGERIA, ²From Testing To Targeted Treatments, The Synergist, Dijon, FRANCE, ³From Testing To Targeted Treatments, The Synergist, Belguim, BELGIUM, ⁴From Testing To Targeted Treatments, The Synergist, Brussels, BELGIUM, ⁵UNN Centre for Clinical Trials, University of Nigeria Teaching Hospital, Enugu State, NIGERIA.

Background Globally, precision medicine (PM) is revolutionizing the treatment of different cancers, including metastatic breast cancer (mBC). However, patient education on PM is largely unknown in Nigeria. The study explored the perspectives of cancer patients on the PM materials and their access to targeted therapies. Methods Participants were (N=147) breast cancer (BC) patients recruited from two cancer support groups in Nigeria. Using a patient engagement approach, From the Testing to Targeted Treatments co-created patient educational materials (PEMs) on “biomarker testing for cancer treatment,” “targeted therapy for cancer,” and “questions to ask your healthcare provider.” The PEMs were adapted and reviewed at Project PINK BLUE. Using a mixed-methods design, 76 of the participants completed an online survey on the usefulness of PEMS, and 71 participated in three focus group discussions on access to targeted therapies. Results The majority of the participants, 71.1% reported an increase in awareness of PM, 80.3% found the materials helpful for making informed decisions, and 90% rated the materials highly for clarity. Themes were: “financial barriers to targeted therapies,” “poor awareness of targeted therapies,” and “lack of access to PEMs.” The qualitative findings showed that some of the cancer patients were taking targeted therapies but were unaware of their treatment. A participant said, “We feel the cost is too high, and even when we have the money, it’s still not affordable.” Conclusion Our finding underscores the urgent need to accelerate access to PM for patients in low-resource settings.

H. Yi¹, K. Lee¹, C. Park¹, D. Lee¹, S. Cho¹, J. Koh², H. Ryu², S. Im¹; ¹Department of Internal Medicine, Seoul National University Hospital, Seoul, KOREA, REPUBLIC OF, ²Department of Pathology, Seoul National University Hospital, Seoul, KOREA, REPUBLIC OF.

Background: Bone is the most common site of metastasis in breast cancer, and patients developing skeletal-related events show poor prognosis. Bone metastasis is a complex process involving interactions between many different molecules, which include receptors and proteins related to chemokine signaling, cell adhesion and migration, or bone metabolism regulation. We aimed to demonstrate the correlation between bone metastasis and expression of these molecules in breast cancer. Methods: Clinicopathologic parameters were collected from 1,440 stage I-III breast cancer patients (1,436 female and 44 male, median age 50 years, range 25-90) who underwent curative surgery at Seoul National University Hospital from October 2000 to December 2013. Immunohistochemistry (IHC) for CXCR4, CXCR6, MGP, osteocalcin, periostin, PTH1R, PTHLH and RANKL were performed on tissue microarray (TMA) constructed from the resected breast cancer tissue. The expression levels of each marker were assessed as H-scores, and we applied cut-off values of 0, 100, 200, and 300 to dichotomize into positive or negative. Results: There were 966 patients (67.1%) with hormone receptor (HR)-positive and HER2-negative breast cancer, 225 patients (15.6%) with HER2-positive breast cancer, and 249 patients (17.3%) with triple negative breast cancer (TNBC). As of February 2024, at median follow-up duration of 138 months, 196 patients (13.6%) had experienced recurrence. This included 65 patients with local recurrence and 131 patients with distant metastasis (49 patients had bone-only metastasis and 82 patients had visceral metastasis as well). There were 186 patients (12.9%) who died from any cause. CXCR6 H-scores were higher in HR-positive breast cancer (mean H-score 33.90) compared to HER2-positive breast cancer (mean H-score 15.58, p=0.0006) or TNBC (mean H-score 16.78, p=0.0056). CXCR6 positivity was significantly associated with longer overall survival (OS) (5-year survival rate (5YSR) 95.8% vs. 95.0%, p=0.049 for cutoff 0). In HR-positive breast cancer, CXCR6 positivity was also correlated with longer bone-specific relapse-free survival (RFS) (5YSR 98.0% vs. 96.8%, p=0.024 for cutoff 0). In contrast, CXCR6 positivity was correlated with shorter bone-specific RFS in HER2-positive breast cancer and TNBC across various cutoff values. RANKL positivity was correlated with shorter OS (5YSR 95.6% vs. 93.9%, p=0.049 for cutoff 200) and RFS (5YSR 92.8% vs. 88.5%, p=0.026 for cutoff 200). In contrast, bone-specific RFS prolonged as RANKL positivity increased (5YSR 97.4% vs. 96.8%, p=0.044 for cutoff 100). The positive correlation to bone-specific RFS was maintained in RANKL-positive, HR-positive breast cancer (5YSR 98.2% vs. 96.3%, p=0.018 for cutoff 100). However, in TNBC, the RANKL score was negatively correlated with RFS across various cutoff values. Periostin was significantly less expressed in TNBC (mean H-score 23.79) compared to other subtypes (mean H-score 48.37, p<0.001 for HR-positive breast cancer and mean H-score 49.14, p<0.001 for HER2-positive breast cancer). Periostin positivity was related to longer OS (5YSR 97.7% vs. 95.6%, p=0.03 for cutoff 0) and bone-specific RFS (5YSR 99.5% vs. 96.7%, p=0.014 for cutoff 100) in HR-positive breast cancer, but not in other subtypes. Conclusion: Expression levels of CXCR6, RANKL, and periostin were correlated with prognosis and bone-specific survival in breast cancer. Interestingly, the clinical impact of each marker differed with respect to each breast cancer subtypes, possibly suggesting a different underlying mechanism of bone metastasis in HR-positive breast cancer compared to the others. Further investigations into the differences between subtypes and establishing proper cutoff values are warranted.

Y. Cheng¹, J. Bradley¹, K. Morris¹, I. Ivanova¹, M. Barnett¹, A. Séguret¹, O. Eve¹, A. Zyoud¹, G. Benitez¹, A. Robinson¹, A. Turnbull², J. Dixon², R. Hockett¹, H. Chuang¹, R. Kuo¹; ¹Wobble Genomics, Wobble Genomics Ltd, Edinburgh, UNITED KINGDOM, ²Edinburgh Cancer Research Centre, MRC Institute of Genetics and Cancer, The University of Edinburgh, Edinburgh, UNITED KINGDOM.

Background: Accurate quantification of tumor HER2 expression is essential for guiding breast cancer therapy selection, as HER2 status directly determines treatment decisions and eligibility for targeted therapies. Current HER2 testing methods, primarily based on IHC and ISH, often lack the precision to confidently distinguish certain HER2 statuses, particularly in borderline cases, which can lead to under- or overtreatment. Developing a more accurate and reliable HER2 quantification method will enable safer and more effective treatment choices, ensuring patients receive the most appropriate therapy while minimizing unnecessary exposure to side effects. Ultimately, this will lead to better clinical outcomes and improved quality of care for breast cancer patients. Methods: We have developed a novel platform that leverages long-read sequencing to capture full-length mRNA and generate isoform-level expression profiles. By integrating tumor and blood transcriptome data from breast cancer patients, we characterized the expression of HER2 and 728 other cancer-related genes (based on COSMIC Cancer Gene Census and established ADC targets) to identify patterns that can be used to distinguish patients with different HER2 statuses. Patients: Twelve tumor biopsies and matched blood samples were collected from nine breast cancer patients with HER2 IHC scores of 0 or 1+ (ongoing study; numbers reported as of abstract submission). In addition, blood transcriptome profiles were analyzed from a separate cohort of 95 control patients with benign breast condition or no cancer on imaging. Results: In the 12 tumor samples and 729 genes analyzed, 29,804 non-reference transcript isoforms which had not been previously reported in the reference human genome annotations were identified in multiple patients. Of these, 16,533 (55.5%) were not found in any of 95 control samples, representing novel isoforms potentially unique to breast cancer tumors. For 551 genes, we identified non-reference isoforms shared between tumor and matched blood samples, including in HER2. Comparative analysis of tumor samples with different HER2 statuses (IHC 1+ vs. IHC 0) revealed significant differences in HER2 expression aligned with the IHC scores. The IHC 1+ group exhibited not only higher expression levels (median 79.5 vs. 37.5 parts per million; p = 0.015), but also greater transcript isoform diversity (median 38 vs. 9 transcripts per sample; p = 0.015) at the HER2 locus. On average, 23.3% of HER2 isoforms identified in HER2 IHC 1+ patients contained novel combinations of known splice donor/acceptor sites, as opposed to 11.5% in HER2 IHC 0 patients. Notably, some of these alternative splicing patterns were also detected in matched blood samples, offering valuable insight that could facilitate non-invasive diagnosis of patient HER2 status from blood. Conclusions: Our novel approach utilizing full-length RNA sequencing enables more comprehensive characterization and precise quantification of tumor-associated expression of HER2 and other cancer-related genes, providing new insights into transcript diversity and expression patterns that are critical for improving diagnostic accuracy, refining treatment selection, and ultimately enhancing patient outcomes.

S. Lakis¹, E. Tsolaki¹, N. Korfiatis¹, A. Goussia¹, H. P. Kourea¹, A. Batistatou¹, M. Bobos¹, K. Papadopoulou², A. Charchanti¹, M. Bai¹, O. Tzaida¹, K. Petraki¹, P. Arapantoni¹, T. Koletsa¹, D. Pectasides¹, A. Koutras¹, D. Kalapanida¹, F. Dimitrakopoulos¹, E. Aravantinou-Fatorou¹, N. Spathas¹, A. Psyrri¹, H. Gogas¹, F. Zagouri¹, G. Fountzilas¹; ¹Data Office, Hellenic Cooperative Oncology Group (HeCOG), Athens, GREECE, ²Laboratory of Molecular Oncology, Hellenic Foundation for Cancer Research, Thessaloniki, GREECE.

Background: HER2-low breast cancers are a mixture of luminal and triple negative tumors. Precise molecular characterization can help refine treatment strategies with novel antibody-drug conjugates and improve outcomes in this patient subgroup. Methods: We used central ER/PgR (HR) and HER2/ERBB2 IHC/FISH assessment to define HER2-low (IHC 1+ or 2+/FISH non-amplified), HER2-zero (IHC 0) and Luminal A/ Luminal B/ TNBC subtypes across 7 prospective adjuvant studies in which eBC patients were treated with dose dense sequential chemotherapy. Retrospective analysis of clinico-pathological data and IHC markers from tissue microarray cores was performed on N=2,751 cases, whereas DNA next generation sequencing (NGS) data were available in a subset of N=1,120 tumors. Heterogeneity was investigated by comparing HER2-low concordant (HER2-low_c) cases, when all cores were scored consistently as HER2-low (>0) and HER2-low discordant (HER2-low_d) cases, when at least one core was scored “0”. Results: First, HER2-low tumors were more likely to be HR + (p<0.001). Second, within HR + disease only, the proportion of Luminal B was higher in HER2-low (51%) than HER2-zero (41.6%), whereas Luminal A showed the opposite pattern (p< 0.001); PIK3CA mutations on the other hand were more common in HER2-zero (33.6% vs. 24.7%, p=0.0036). Third, a statistically significant progressive increase in ERBB2 average copies was observed from HER2-zero to HER2-low_d and HER2-low_c irrespective of HR status (p<0.001). Finally, HER2 low was not associated with survival in either HR + or HR – disease. Conclusion: Our data suggests that HR + HER2-low and HER2-zero eBC differ with respect to the status of the estrogen receptor pathway. When accounting for inter-core variability, differences in HER2 IHC levels correlate with the underlying ERBB2 copy number; however, standard HER2 IHC scoring thresholds did not influence prognosis in this cohort.

M. A. Bravo, D. C. Sotelo-Rodríguez, J. Caicedo, S. Cervera, A. Ruiz-Patiño, W. A. Mantilla, S. X. Franco; Breast Cancer, Fundación Centro de Tratamiento e Investigación sobre Cáncer(CTIC), Bogota, COLOMBIA.

Abstract San Antonio Breast Cancer Symposium Title: Genomic Characterization and Genetic Testing Gaps in Early-Onset Breast Cancer in Colombia Background:Breast cancer (BC) in women under 50 years represents a biologically aggressive disease with distinct genomic features and higher hereditary burden. Latin American data on this subgroup remain limited. We aimed to describe the clinical and molecular characteristics of young BC patients in a comprehensive cancer center in Colombia, and to evaluate the yield of genetic testing. Methods:A retrospective cohort of 113 women <50 years diagnosed with BC at the Luis Carlos Sarmiento Angulo Cancer Treatment and Research Center (CTIC) between 2019 and 2024 was analyzed. Demographic, clinical, histopathologic, imaging, and molecular data were extracted. Tumor subtypes were defined based on ER, PR, and HER2 status. Germline sequencing was performed in 95% of cases (n=108). Chi-squared and t-tests were used to assess associations between mutation status, age, and clinical variables. Results:Among 108 women diagnosed with breast cancer at ≤50 years of age who underwent comprehensive germline genetic testing, 76 patients (70.4%) had complete immunohistochemical data, enabling subtype classification. The most common subtype was Luminal (n=36, 47.4%), followed by triple-negative (n=18, 23.7%) and HER2-enriched tumors (n=8, 10.5%). The luminal group showed a mutation frequency of 13.9% (5/36), while HER2-enriched tumors exhibited a rate of 12.5% (1/8). BRCA1 was the most frequent germline mutation in triple-negative tumors, identified in 18.2% (2/11) of cases. In contrast, Luminal and HER2-enriched subtypes showed a predominance of non-canonical mutations, including genes such as MUTYH, POLE, and APC, detected in 40.9% (9/22) and 14.3% (2/14) of cases, respectively. When considering all germline findings, 47 out of 108 patients (43.5%) harbored at least one germline alteration. The most frequently altered genes across the cohort were BRCA1 (7.4%), PALB2 (5.6%), BRCA2 (3.7%), and CHEK2 (1.9%). Additionally, 24.1% had variants in non-canonical genes, including both pathogenic and uncertain significance variants. Patients carrying germline mutations were diagnosed at a significantly younger age (mean 34.9 ± 6.1 years) compared to non-carriers (mean 41.8 ± 5.4 years; p < 0.001), emphasizing the contribution of hereditary predisposition to early-onset breast cancer and supporting the implementation of broad germline testing in all young patients, irrespective of tumor subtype. Conclusions:Germline pathogenic variants are frequent in early-onset breast cancer, particularly in triple-negative and Luminal subtypes. BRCA1 predominates in triple-negative cases, while PALB2 and CHEK2 are found in hormone receptor–positive tumors. Additionally, 24.1% of patients carried alterations in non-canonical genes, reflecting marked genomic heterogeneity. Mutation carriers were diagnosed at a significantly younger age, supporting early and comprehensive genetic testing in all young patients to guide personalized care and familial risk assessment.

M. Yasukawa¹, S. Sato², N. Saito³, S. Saji⁴, N. Masuda⁵, T. Yamanaka¹, S. Fujiwara¹, K. Goda¹, N. Mori⁶, T. Takano⁷, K. Watanabe⁸, Y. Naito⁹, H. Tada¹⁰, M. Kitada¹¹, T. Iwasa¹², T. Morimoto¹³, A. Takahashi¹, M. Isoda¹, T. Yamashita¹, D. Hoshino³; ¹Department of Breast Surgery and Oncology, Kanagawa Cancer Center, Kanagawa, JAPAN, ²Morphological Information Analysis Laboratory, Kanagawa Cancer Center Research Institute, Kanagawa, JAPAN, ³Cancer Biology Division, Kanagawa Cancer Center Research Institute, Kanagawa, JAPAN, ⁴Department of Medical Oncology, Fukushima Medical University, Fukushima, JAPAN, ⁵Department of Breast Surgery, Graduate School of Medicine, Kyoto University, Kyoto, JAPAN, ⁶Department of Breast Surgery, Seirei Hamamatsu General Hospital, Shizuoka, JAPAN, ⁷Department of Breast Medical Oncology, The Cancer Institute Hospital of JFCR, Tokyo, JAPAN, ⁸Department of Breast Oncology, National Hospital Organization Hokkaido Cancer Center, Hokkaido, JAPAN, ⁹Department of Breast and Medical Oncolog, National Cancer Center Hospital East, Chiba, JAPAN, ¹⁰Department of Surgery, Division of Breast and Endocrine Surgery, Tohoku University Hospital, Miyagi, JAPAN, ¹¹Department of Breast Disease Center, Asahikawa Medical University Hospital, Hokkaido, JAPAN, ¹²Department of Medical Oncology, Kindai University Hospital, Osaka, JAPAN, ¹³Department of Breast Surgery, Yao Municipal Hospital, Osaka, JAPAN.

Background: Malignant tumor cells grow invasively and form distant metastases after moving through multiple tissue barriers. Invasion requires cell locomotion, accompanied by the degradation of the extracellular matrix (ECM) by matrix metalloproteinases (MMPs). MT1-MMP (Membrane Type 1 MMP) is an integral membrane protease that degrades a variety of protein components within the extracellular milieu. The substrates of MT1-MMP include a variety of components of the ECM membrane proteins, such as type 1 collagen. Type 1 Collagen is a physical varier not only affecting cancer cell invasion but also drug accessibility to cancer cells. It is reported that MT1-MMP regulates small compound drug resistances via activation of integrin signaling through type 1 collagen cleavage and higher expression of MT1-MMP was considered a marker of increased malignancy and poorer prognosis. However, it remains unclear whether larger drugs, such as antibodies, can be affected by MT1-MMP-related collagen remodeling. Here, we conducted a retrospective analysis of MT1-MMP on treatment effectiveness and survival using the JBCRG-M06/EMERALD clinical trial. The clinical trial is a Phase III study comparing two first-line chemotherapeutic treatments for HER2-positive local advanced or metastatic breast cancer: trastuzumab, pertuzumab, and taxane therapy versus trastuzumab, pertuzumab, and eribulin mesylate therapy. Method: In the EMERALD study, 231 specimens were collected from 231 patients who consented to biomarker research at the time of enrollment. Among these 182 specimens were collected from primary tumors, and 49 from metastatic sites. The tissue samples were immunostained with anti-MT1-MMP antibody and classified into low (score1), moderate (score2), and high (score3) expression based on staining proportion and intensity. We investigated the correlation between MT1-MMP expression and the therapeutic effects of trastuzumab and pertuzumab combination with eribulin mesylate or taxane, as well as Progression-Free Survival (PFS). Results: Of the 231 patients in this study, 123 were assigned to the eribulin group and 108 to the taxane group, with MT1-MMP scores 1/2/3 of 28/52/43 in the eribulin group and 26/43/39 in the taxane group, respectively. The median PFS in the eribulin group was 14.7 months and 13.4 months in the taxane group, and it was similar to the results for all patients in the EMERALD study. There was no difference between the eribulin and taxane groups. Whereas the median PFS for MT1-MMP scores 1/2/3 was 10.5/13.2/18.4 months, and it tended to be longer with higher MT1-MMP scores (p=0.0306). Although higher MT1-MMP scores tended to improve response rates overall, there was no significant difference between the two treatment groups. In this study of patients with HER2-positive local advanced or metastatic breast cancer patients, higher expression of MT1-MMP may be with better PFS and response rates. Conclusion: The MT1-MMP score is unable to distinguish the combined trastuzumab and pertuzumab with eribulin mesylate or trastuzumab and pertuzumab with taxane therapy. However, it may be a prognostic factor for HER2-positive local advanced or metastatic breast cancer patients treated with the trastuzumab and pertuzumab combination and eribulin mesylate or taxane.

S. Ding¹, J. Wu², D. Liu¹, Y. Fang¹, W. Wang¹, L. Zhu¹; ¹Department of Breast and Thyroid Surgery, Shanghai General Hospital,Shanghai Jiao Tong University School of Medicine, Shanghai, CHINA, ²Comprehensive Breast Health Center, Ruijin Hospital,Shanghai Jiao Tong University School of Medicine, Shanghai, CHINA.

BackgroundTriple-negative breast cancer (TNBC) had great tumor heterogeneity and lacks effective therapeutic targets. The addition of immunotherapy to neoadjuvant chemotherapy (NAC) has dramatically improved pathologic complete response rates (pCR) in TNBC, but some patients may be overtreated as they could achieve pCR with NAC only. Thus, identifying the molecular features associated with efficacy of NAC and finding robust biomarkers to select patients who are sensitive to chemotherapy was crucial.Patients and MethodsRNA sequencing analysis was performed on paired tumor samples collected before and after neoadjuvant treatment with nab-paclitaxel and carboplatin from 44 TNBC participants enrolled in the prospective NeoPATH clinical trial (ClinicalTrial.gov identifier: NCT0390780). Transcriptomic features were compared in patients with different treatment responses. Changes in molecular biology after treatment were illustrated in matched pre- treatment and post-treatment samples as well. Univariate regression and lasso regression were utilized to construct the predictive signature for NAC. ResultsThe results showed that patients achieving pCR had upregulated immune-related signatures suggesting more active immune response and decreased metabolic signatures indicating less energy production. Increases in both anti-tumor immune cell infiltration and immune molecular expression were observed after neoadjuvant chemotherapy, inferring the activation of immune microenvironment. Multiple immune-related signatures were found to be positively associated with treatment response, while metabolic signatures were negatively associated with the response. Based on these findings, the immune-metabolic predictive signatures were further built and validated to predict the treatment response of neoadjuvant nab-paclitaxel and carboplatin for TNBC (AUC, 0.719 in the train set, 0.787 and 0.717 in the validation sets).ConclusionsOverall, our study delineates distinct molecular profiles between pCR and non-pCR TNBC patients following neoadjuvant nab-paclitaxel/carboplatin. The developed immune-metabolic signature effectively identifies treatment-sensitive tumors, while observed transcriptomic changes post-therapy may inform future targeted drug development.

J. Tsuchida, M. Nagahashi, M. Komatsu, S. Urano, M. Kuroiwa, S. Yoshida, G. Sugimoto, Y. Togashi, A. Mitsuyoshi, H. Kanaoka, A. Hattori, T. Higuchi, A. Nishimukai, K. Murase, M. Shimoda, Y. Miyoshi; Division of Breast and Endocrine Surgery, Department of Surgery, Hyogo Medical University, Nishinomiya, Hyogo, JAPAN.

Background: The tumor immune microenvironment affects treatment response and outcomes in patients with metastatic breast cancer (MBC). Cytokines, such as tumor necrosis factor-α (TNF-α) and interleukin (IL)-6, play an important role in tumor progression and immune regulation, and they are associated with outcomes in patients with MBC. Although the significance of each cytokine as a prognostic factor has been reported, there is a paucity of reports on the relationship between the integrated indicator of multiple cytokines and outcomes. In this study, we established an integrated indicator of multiple cytokines and examined the relationship with patient outcomes. In addition, since inflammatory cytokines are thought to affect cancer progression and metastasis by acting on T cells, myeloid-derived suppressor cells (MDSCs), and regulatory T cells (Tregs), we also examined the relationship between the integrated cytokine indicator and immune cell profile. Materials and Methods: This retrospective study included 43 patients with MBC treated with chemotherapy. Baseline values for cytokine levels, including TNF-α, soluble interleukin-2 receptor (sIL-2R), IL-6, IL-8, transforming growth factor (TGF)-β, and thymidine kinase 1 (TK1), were measured in peripheral blood on the day before the first treatment with chemotherapy. Each cytokine was classified into low and high groups using cutoff values defined by ROC curves for overall survival (OS), and the progression-free survival (PFS) and OS between low and high groups were compared by Kaplan-Meier plots. CD4+ and CD8+ lymphocytes, MDSCs, and Tregs levels were determined in the blood by flow cytometry analysis. Results: Of the patients, 46.5% (n = 20) were treated with eribulin, 9.3% (n = 4) with trastuzumab deruxtecan, 9.3% (n = 4) with paclitaxel and bevacizumab, 7.0% (n = 3) with trastuzumab emtansine, and 27.9% (n = 12) with other chemotherapeutic agents. Patients treated as first-line were 55.8% (n = 24), and those treated as second-line or later were 44.2% (n = 19). Patients with high levels of TNF-α, sIL-2R, IL-6, and TK1 showed a shorter OS compared to those with low levels (P = 0.01, 0.01, 0.01, and <0.01, respectively). IL-8 and TGF-β were not associated with OS (P = 0.07 and 0.06, respectively). Patients with high levels of sIL-2R, TGF-β, and TK1 showed a shorter PFS compared to those with low levels (P = 0.02, 0.01, and 0.04, respectively). TNF-α, IL-6, IL-8 and TK1 levels were not associated with PFS (P = 0.11, 0.10, 0.07 and 0.04, respectively). A “cytokine score” was developed based on the number of elevated cytokines among TNF-α, sIL-2R, IL-6, and TK1. Patients with 0-1 elevated cytokine were classified as low score, and those with 2-4 as high score. The high cytokine score group had shorter OS and PFS compared to the low score group (P < 0.01 and P = 0.03, respectively). Multivariable analyses revealed that cytokine score were an independent prognostic factor for both OS (P < 0.01) and PFS (P = 0.02). Furthermore, patients with high cytokine scores had lower proportions of CD4+ and CD8+ lymphocytes (P = 0.01 and 0.04, respectively), and higher levels of MDSCs (P < 0.01), with no difference in Tregs (P = 0.30). Conclusion: Our results show that the levels of TNF-α, sIL-2R, IL-6, and TK1 were associated with outcomes in patients with MBC treated with chemotherapy. The cytokine score, based on the number of elevated cytokines, is an important prognostic indicator for both OS and PFS. Furthermore, since a high cytokine score was associated with higher levels of MDSCs, and lower levels of CD4+ and CD8+ lymphocytes, the cytokine score might be associated with patient outcomes through immune reaction. Patients with high cytokine scores have worse outcomes, which may reflect a poor tumor microenvironment.

W. Wen¹, P. Pu¹, H. Xu¹, T. Shang¹, L. Cong¹, G. Qiao², Z. Jia¹, Y. Liu¹, Y. Huang¹, R. Zhou¹, Y. Li², J. Liu¹; ¹Department of Breast Surgical Oncology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, CHINA, ²Department of Breast Surgery, The Affiliated Yantai Yuhuangding Hospital of Qingdao University, Yantai, CHINA.

Background: Nipple discharge is a common clinical manifestation of breast disease, yet current diagnostic tools such as ultrasound, mammography, MRI, ductography, and cytology have limited accuracy in early cancer detection. Copy number variation (CNV) profiling of DNA from body fluids has shown promise in early diagnosis of cancers like bladder and cervical, but its application to nipple discharge remains unexplored. This study aimed to establish a noninvasive, CNV-based diagnostic approach that reflects primary breast tumor biology with high specificity and sensitivity. Methods: We analyzed CNV data from 1,086 tumor and 942 normal samples in the TCGA-BRCA cohort using GISTIC2.0, identifying recurrent CNV regions with high frequency and statistical significance (q50%, biological interpretability (oncogenes in amplifications, tumor suppressors in deletions), consistency with gene expression changes, and prognostic relevance. A final panel of 20 CNV features (12 amplifications, 8 deletions) was established and validated across multiple breast cancer cohorts and their matched normal tissues, when available. Diagnostic performance was further tested using nipple discharge DNA an in-house cohort of patients with benign and malignant breast disease. Additionally, single-cell RNA sequencing was performed on tumor tissues from five breast cancer patients in the cohort. Using inferCNV, the CNV features detected in nipple discharge were traced back to distinct epithelial subclusters using inferCNV, and the functional characteristics of the dominant CNV-contributing subpopulations were further analyzed. Results: In the TCGA-BRCA dataset, only 8 of 118 adjacent normal tissues were panel-positive, yielding a specificity of 93.22%. External validation demonstrated consistent diagnostic performance: GSE87048 (sensitivity 77.0% in tumor tissue, specificity 99.0% in matched peripheral blood mononuclear cells), GSE118527 (sensitivity 92.8%, specificity 100%), METABRIC (sensitivity 95.7%), and GSE31427 (Stage I-II; sensitivity 66.9%). In the MSK-CHORD PanCancer cohort, the panel showed the highest detection rate in breast cancer (78.3%) compared to colorectal, prostate, lung, and pancreatic cancers, supporting tissue specificity. In nipple discharge samples (n = 39 malignant, n = 18 benign), the panel achieved a sensitivity of 71.79% and a specificity of 94.44%, highlighting its potential for accurate, noninvasive early detection. In five patients from the in-house cohort, single-cell RNA sequencing of tumor tissues revealed that CNV features detected in nipple discharge originated from specific epithelial subclusters. These CNV-contributing subpopulations exhibited enrichment in pathways related to cell junction assembly and tissue morphogenesis. Conclusion: This study provides the first evidence that CNV profiling of nipple discharge DNA is a feasible and tumor-specific approach for early breast cancer detection. The selected CNV panel demonstrates robust diagnostic performance across independent datasets and early-stage tumors, supporting its clinical utility as a novel, minimally invasive screening tool.

Y. Liu, M. Parrella, S. Schwaegerle, S. Berry, C. Betts, M. Landers; Advanced Biopharma Solutions, Akoya Biosciences, Marlborough, MA.

Background Antibody-drug conjugates (ADCs) have emerged as a promising therapeutic class for solid tumors, especially in breast cancer (BC), with a focus on agents targeting HER2 and Trop2. Currently, two ADC drugs targeting HER2: Trastuzumab emtansine (T-DM1) and fam-trastuzumab deruxtecan-nxki (T-DXd), and two targeting Trop2: Sacituzumab govitecan (SG) and datopotomab deruxtecan-dlnk (Dato-DXd) have gained approval by the FDA in treating metastatic BC, with additional ADCs in development. Despite the advances, challenges remain in patient selection including identifying the HER2 “ultra-low” patient subgroup, characterizing HER2/Trop2 co-expression, and ensuring reliable evaluation of biomarkers on the cell membrane surface. Here, we present the fully validated Opal™ multiplex-immunofluorescence (mIF) BC-ADC panel, leveraging the IVD-grade Akoya PhenoImager HT imaging system. The BC-ADC panel provides simultaneous assessment of BC standard-of-care biomarkers with Ki67 and hormone receptor (HR), ADC-specific targets HER2 and TROP2, together with a proprietary membrane marker cocktail facilitating specific biomarker membrane evaluation. Methods A proprietary membrane cocktail, comprised of multiple membrane-specific biomarkers, was developed to specifically label and visualize cell membranes. The membrane cocktail performance supporting accurate cell definition and subcellular localization was assessed against ground truth boundaries created by board-certified pathologists. Biomarkers TROP2, HER2, Ki67, ER, PR and the membrane cocktail, plus DAPI were tested for staining order effects and spectral unmixing, then developed into an single slide Opal™ 6-color multiplex immunofluorescence panel. HER2-positive/HER2-negative breast cancer tissue was utilized for assay development and verification. Board-certified pathologists evaluated panel validation according to the pre-set acceptance criteria including: absence of spectral crosstalk or umbrella effect by antibody drop-test; assay robustness by inter/intra-day triplicate precision studies, and biomarker sensitivity and specificity confirmed by equivalency between “gold-standard” IHC and mIF. HER2-negative breast cancer samples utilized for testing contained samples spanning HercepTest score 0+ (null or ultra-low), score 1+ (low), and score 2+ (low). Assay HER2 signal sensitivity was evaluated by pathologists to differentiate HER2-negative scores qualitatively. Results The membrane cocktail consistently delineates cell membrane in both tumor and stromal compartments. The finalized mIF assay met the predefined validation criteria as assessed by board-certified pathologists. Antibody drop-testing revealed no umbrella effect or spectral crosstalk was observed between channels. The precision study demonstrated high reproducibility in inter-day comparison and intra-day comparison. Strong equivalency of each biomarker was observed between IHC and the mIF panel. Pathologist review confirmed the high sensitivity in differentiating HER2 expression levels across HER2-negative cases: null/ultra-low/low. Conclusion The validated 5-plex/6-color BC-ADC mIF panel demonstrated robust analytical performance: high sensitivity and specificity across all biomarkers, especially in HER2-negative samples, stratifying HER2 low and ultra-low subgroups. The mIF panel showed reliable isolation of markers and robust assay performance. Results also suggested that the custom membrane cocktail can enhance the subcellular evaluation of biomarkers. The study will be supplemented by an independent patient verification cohort to support forthcoming clinical-utility studies.

L. Naghi, K. G. Roth, S. Lindsey, A. Seuylemezian, J. Bonner, L. Kruper, S. B. Gruber, J. Mortimer, K. McDonnell; Medical Oncology, City of Hope, Duarte, CA.

Purpose: To characterize the prevalence of germline pathogenic or likely pathogenic (PLP) variants in Fanconi Anemia (FA) pathway genes among patients with invasive breast cancer (IBC) and evaluate opportunities for clinical actionability through germline and somatic analysis. Background: Disruption of the FA DNA repair pathway can impair homologous recombination (HR). While high-penetrance PLP variants in BRCA1, BRCA2, and PALB2 are established breast cancer susceptibility genes and contribute to HR deficiency (HRD), the clinical consequences of other FA germline PLP variants remain unclear. Using the updated cohort from the City of Hope INSPIRE Study, we assessed the prevalence and functional impact of germline PLP variants in 17 FA pathway genes. HRD, mutational signatures, and loss of heterozygosity were used to evaluate effects on HR integrity. These findings may have implications for HRD-related precision therapeutic intervention and clinical trial eligibility. Methods: This retrospective study included consented patients with histologically confirmed IBC enrolled in the INSPIRE Study as of January 14, 2025. Germline sequencing of 155 cancer predisposition genes was performed; patients with multiple FA variants were classified according to the FA gene of the highest penetrance. A cohort subset underwent paired tumor-normal whole exome sequencing (>400x tumor, >180x normal), including HRD scoring (scarHRD), LOH determination, and SBS3 assessment (SigProfiler). Clinical data were obtained from cancer registries and medical records. Clinical trial eligibility criteria were evaluated using ClinicalTrials.gov. Results: Of 8,242 patients with IBC, 7,982 underwent germline sequencing; 592 (7.4%) had a PLP variant in an FA gene, including 197 (33.3%) in non-high-penetrance (non-HP) genes. Among 1,833 patients with paired tumor-normal sequencing, 145 (7.9%) had an FA PLP variant, of which 53 (2.89%) were in non-HP genes. HRD scores were ≥42 in 45.3% of non-HP carriers (Table 1). 24 of 53 patients (45.3%) with a non-HP FA germline PLP variant had HRD-positive (≥42) tumors. HRD-positivity is more likely among patients with germline PLP variants in non-HP FA pathway genes compared to those without a variant in any FA gene (RR 1.51, 95% CI 1.12 to 2.05; p < 0.0077). Stage data were available for 1,548 patients; 128 (8.3%) were stage IV at diagnosis, including 7/128 (5.5%) with non-HP FA gene PLP variants. Conclusion: Germline PLPs in non-HP FA genes are present in a meaningful subset of patients with IBC and may be functionally relevant. These findings support the value of expanded germline testing and trial eligibility criteria that encompass the broader FA pathway and suggest that non-HP FA gene variants may serve as biomarkers for identifying patients likely to benefit from HRD-targeted therapies.

F. Klemm¹, F. Mohammad², A. C. Tuck¹, X. Xing¹, G. Morales², A. Willig¹, E. Smith¹, M. Basgall¹, A. Francois¹, L. Canetti³, E. Jorge³, Z. Xu¹, F. Penault Llorca⁴, M. Ferreira³; ¹Data Science, SOPHiA GENETICS, Rolle, SWITZERLAND, ²Data Science, SOPHiA GENETICS, Boston, MA, ³Pathology Unit, OncoGèn Auvergne, Centre Jean Perrin, Clermont-Ferrand, FRANCE, ⁴Department of Pathology and Biopathology, Centre Jean Perrin, UMR INSERM 1240, University Clermont Auvergne, Clermont-Ferrand, FRANCE.

Background: Liquid biopsy (LBx) enables non-invasive molecular profiling and monitoring of cancer but has been impeded by operational, computational and financial barriers. While centralized LBx testing is widely used in clinical trials and specialized centers, locally or regionally performed decentralized testing has the potential to broaden access to molecular profiling and reduce turnaround times. We evaluated the analytical performance and real-world utility of the decentralized MSK-ACCESS® powered with SOPHiA DDM™ LBx assay in patients with breast cancer. Methods: Analytical characteristics were assessed by comparing results from the decentralized LBx assay against a reference orthogonal method in two cohorts: breast cancer (n=17) and prostate cancer patients (n=55). Metrics included the mean molecular coverage, number of variants per sample and mean variant allele frequency (VAF) for single nucleotide variants (SNVs) and insertions/deletions (indels). Similar analytical performance across both cohorts was used to support assay validity. Subsequently, mutational profiles were characterized in the breast cancer cohort, with a focus on clinically relevant alterations in ESR1 and PIK3CA. These findings were extended to a larger real-world dataset from the SOPHiA GENETICS user base. Results: The MSK-ACCESS® powered with SOPHiA DDM™ assay demonstrated comparable performance across both breast and prostate cancer cohorts, with similar mean duplex molecular coverage (n_breast=1791, n_prostate=1638), median somatic variant counts (n_breast=3, n_prostate=2.5), and median somatic VAFs (VAF_breast=0.6%, VAF_prostate=0.6%). In breast cancer samples, recurrent ESR1 mutations (e.g. E380Q, Y537S and D538G) and hotspot PIK3CA mutations (e.g. H1047R) mediating resistance were observed down to 0.06% VAF. These variants were recapitulated in the larger real-world breast cancer cohort (n=219; ESR1_mut VAF_min=0.11%, PIK3CA_mut VAF_min=0.14%) confirming the solution’s reproducibility and robustness in routine decentralized use. Conclusions: The decentralized LBx assay showed consistent analytical performance across tumor types and reliably detected clinically relevant variants in breast cancer. These findings demonstrate the feasibility and clinical utility of the decentralized MSK-ACCESS® powered with SOPHiA DDM™ testing, supporting its broader implementation in precision oncology.

T. Hanamura¹, K. Yokoyama¹, S. Takahashi¹, H. Kiyohara¹, S. Nakagawa¹, R. Ishida¹, M. Terao¹, T. Okamura¹, N. Kumaki¹, A. Katayama², S. Kurozumi³, N. Niikura¹; ¹Department of Breast Oncology, Tokai University School of Medicine, Isehara, JAPAN, ²Department of Pathology, Shizuoka Cancer Center, Shizuoka, JAPAN, ³Department of Breast Surgery, International University of Health and Welfare, Narita, JAPAN.

Purpose: An immunosuppressive tumor microenvironment (TME) contributes to therapeutic resistance in breast cancer. Immune checkpoint inhibitors (ICIs) combined with chemotherapy have emerged as a promising strategy to overcome this barrier. Understanding immune regulation within the TME is essential for developing effective biomarkers and therapies. Our previous work suggested an immunomodulatory role for ZAG (AZGP1), with both clinical and in vitro findings consistently indicating its immunosuppressive activity, potentially contributing to therapeutic resistance. This study evaluates the relationship between ZAG expression and response to neoadjuvant chemotherapy (NAC) in triple-negative breast cancer (TNBC). Methods: To determine the cellular source of AZGP1, we reanalyzed a public TNBC single-cell RNA-seq dataset using Seurat v5.2.1. UMAP clustering was used to annotate cell types. Feature plots and violin plots were employed to visualize cell type-specific AZGP1 expression. Gene set enrichment analysis (GSEA) was performed using bulk transcriptomic data from four NAC-treated TNBC cohorts to explore pathways associated with pathological complete response (pCR) and AZGP1 expression. ROC analysis was used to assess the predictive value of AZGP1 for pCR and to determine optimal cut-off values. For validation, we analyzed NAC-treated TNBC cases at our institution. Pre-treatment core needle biopsy (CNB) samples were evaluated for tumor-infiltrating lymphocytes (TILs) based on International TILs Working Group criteria. ZAG expression was assessed by immunohistochemistry. The ZAG score was defined as the percentage of tumor cells showing moderate-to-strong staining. ROC analysis was also performed to evaluate the ZAG score’s predictive value for pCR (ypT0/isN0, ypT0N0), overall ypN0, and ypN0 in cN1 cases. Results: Single-cell RNA-seq analysis confirmed that AZGP1 expression was restricted to tumor epithelial cells. GSEA revealed that pCR was associated with enrichment of immune-related gene sets such as ALLOGRAFT_REJECTION and INTERFERON_GAMMA_RESPONSE, while AZGP1 expression negatively correlated with these pathways. In four public TNBC cohorts, AZGP1 showed modest predictive value for pCR (AUC 0.557-0.599; not significant), but high-AZGP1 groups had significantly lower pCR rates in three cohorts (p < 0.05). In our institutional cohort of 51 cases (94.1% stage II-III, mean age 57.2), 86.3% received anthracycline-taxane regimens. Pathological CR (ypT0/isN0) was achieved in 31.4% of cases. ZAG scores were significantly higher in tumors with low TILs (p < 0.05). High ZAG expression was associated with poorer response to NAC, particularly in terms of nodal clearance in cN1 cases (AUC = 0.748, p < 0.05). The ZAG-high group showed significantly lower rates of pCR (ypT0N0), overall ypN0, and ypN0 in cN1 cases (all p < 0.05). Conclusion: ZAG (AZGP1), shown to be tumor-derived via single-cell RNA-seq, acts as a paracrine immune modulator in TNBC, supporting its validity as a tumor-intrinsic marker in bulk transcriptomic analyses. High AZGP1 expression correlated with suppressed immune gene programs and reduced chemotherapy sensitivity. Our data confirmed that elevated ZAG expression predicts poor NAC response and reduced nodal clearance, particularly in cN1 cases. ZAG represents a promising tumor-intrinsic biomarker and potential immunotherapeutic target in TNBC.

M. O. Gbadamosi, I. Lopes de Lima, K. H. Streeks, E. Molchan, M. S. Makarem, K. L. Coleman; Department of Pharmacotherapy and Translational Research, University of Florida, Gainesville, FL.

Background: PD-1 immune checkpoint inhibitors (ICIs) have reshaped cancer treatment paradigms. While PD-1 ICIs displays monotherapy efficacy in some tumor types, others, such as breast cancer (BC), require a synergistic combination of ICIs and chemotherapy for clinical benefit. Beyond cytotoxicity, chemotherapy synergize with ICIs via chemoimmunomodulation (CIM), which comprise the immunomodulatory signals elicited by chemotherapy. While many studies have examined molecular features influencing chemoresistance, few have focused on the molecular features that influence CIM, despite the overlap between CIM and the hallmarks of response to PD-1 ICIs. This paucity may be driven by inadequate means for phenotypically delineating CIM and presents a challenge to selection of the most synergistic agents for PD-1 ICIs for a given patient. Methods: We leveraged transcriptomic data from 20 pre-/post-treatment specimens obtained from BC patients (N = 2 reps/sample) treated with neoadjuvant chemotherapy, including anthracycline, cyclophosphamide, taxanes, and platinating agent, to identify CIM induction profiles. CIM induction was assessed using Δ log₂(TPM+1) values for 377 genes obtained from gene sets and literature related to autophagy, lysosomes, cellular stress, and immune function. We developed a novel iterative k-means-based unsupervised clustering approach to identify discrete CIM induction states using CIM gene induction values. Overrepresentation analysis was used to identify transcriptomic programs induced in each CIM induction states. Differential gene expression (DEG) analysis and inferential statistics were used on matched pre-treatment specimens to evaluate baseline differences between states. The abundances of immune cell types were deconvoluted using CIBERSORT. Results: Our iterative unsupervised approached identified two discrete states: one with high (H-CIM; N = 10) and one with low CIM (L-CIM; N = 10) induction. Only one replicate from patients was separately clustered. Stability, robusticity, and the binary nature of the clusters were confirmed using standard cluster evaluation metrics. There were no significant differences in clinical characteristics or treatments between states. The H-CIM state had induction of antigen presentation, upregulation of T-cell populations, and phagocytosis transcriptomic programs while the L-CIM group was primarily characterized by the induction of MYC-governed reactive oxygen species (ROS) detoxification and proteostasis programs (FDR-adjusted P (adjP) < 0.05). Baseline DEG analysis identified upregulated immunosuppressive features such as S100A9 (Fold Change (FC) = 8.9; adjP = 0.04) and HP (FC = 21.2; adjP = 0.008) in the L-CIM group which was supported by a significantly higher baseline amount of M2 macrophages (P < 0.001). Proteostasis safeguards SEC61G (FC = 3.3; adjP = 0.04) and SEC31B (FC =2.6; adjP = 0.002) were also upregulated in the L-CIM state suggesting a predisposition for protection against cellular stress. Interestingly, there was no significant difference in baseline MYC expression between the L-CIM and H-CIM group (P = 0.34). Conclusion: Our unsupervised approach enabled classification of distinct CIM induction states using paired pre- and post-treatment transcriptomic analysis, allowing interrogation of CIM in a manner not captured via traditional chemoresistance studies alone. We uncover underappreciated molecular programs, like proteostasis and ROS detoxification, that may underlie blunted CIM induction in BC. Future work will focus on characterizing baseline molecular features that prompt drug-specific CIM induction states, with the goal of informing and guiding the personalization of chemoimmunotherapy in breast cancer.

R. Mukhtar¹, P. Norwood², A. Borowsky³, S. Alkhafaji⁴, K. Giridhar⁵, C. Vaklavas⁶, A. Elias⁷, M. Wei⁶, M. Goetz⁵, N. Onishi⁸, O. Olopade⁹, L. Van ‘t Veer¹⁰, L. Huppert¹¹, W. F. Symmans¹², L. Brown Swigart¹⁰, C. Wu¹, C. Yau¹, D. Yee¹³, M. Magbanua¹¹, E. Price¹⁴, N. Hylton¹⁴, L. Esserman¹, J. Chien¹¹; ¹Surgery, University of California, San Francisco, San Francisco, CA, ²Statistics, Quantum Leap Healthcare Collaborative, San Francisco, CA, ³Pathology, University of California, San Francisco, San Francisco, CA, ⁴Labaratory Medicine, University of California, San Francisco, San Francisco, CA, ⁵Oncology, Mayo Clinic, Rochester, MN, ⁶Medical Oncology, Huntsman Cancer Institute, Salt Lake City, UT, ⁷Medicine, University of Colorado Anschutz, Aurora, CO, ⁸Radiology & Biomedical Imaging, University of California, San Francisco, San Francisco, CA, ⁹Medicine, University of Chicago, Chicago, IL, ¹⁰Laboratory Medicine, University of California, San Francisco, San Francisco, CA, ¹¹Medicine, University of California, San Francisco, San Francisco, CA, ¹²Anatomical Pathology, MD Anderson, Houston, TX, ¹³Medicine, University of Minnesota, Minneapolis, MN, ¹⁴Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA.

Background Invasive lobular carcinoma (ILC) differs from invasive ductal carcinoma (IDC) in its estrogen receptor (ER) signaling pathways, response to systemic therapy, and imaging features. We evaluated differences in early change in the proliferation marker Ki67 and features on magnetic resonance imaging (MRI) after novel endocrine based neoadjuvant therapy in lobular versus non-lobular cases on the I-SPY2 EOP Trial. Methods EOP is a prospective, randomized NET trial enriched for patients with ILC. Patients with stage II-III ER positive HER2 negative breast cancer with low molecular risk (Mammaprint low or Mammaprint Hi1 plus high Sensitivity to Endocrine therapy Index) were randomized to one of several endocrine therapy-based arms. Treatment arms include novel oral SERDS with or without CDK4/6 inhibitor, aromatase inhibitors, and SERMs with or without ovarian function suppression in pre-menopausal patients. Patients undergo repeat core needle biopsy with central Ki67 assessment after 3 weeks of treatment and serial breast MRI. Here, we compare endocrine therapy responsiveness as measured by change in Ki67 at 3 weeks and functional tumor volume (FTV) on pre-surgical MRI in lobular versus non-lobular cases. Results We evaluated 253 patients of whom 148 (58.5%) had non-lobular cancer (93.9% IDC), and 105 (41.5%) had lobular cancer (17.1% mixed ILC/IDC). Overall, mean age was 53.3 years, 63.8% of tumors were grade 2, and 86.5% were Mammaprint (MP) Low with the remaining being Hi1, with higher rates of MP Low in the lobular versus non-lobular cohorts (92.4% versus 83.4%, p=0.0358). At baseline, tumors in the lobular cohort had significantly lower mean Ki67 than those in the non-lobular cohort (11.2% [SD 12.8] versus 16.4% [SD 14.4%], p=0.0038). Consistent with this, the proportion with baseline tumor Ki67 <10% was significantly higher in the lobular cohort than the ductal cohort (60.4% versus 39.3%, p=0.001). At 3-week biopsy, mean Ki67 remained lower in the lobular versus non-lobular group (3.4% versus 5.4%, respectively, p=0.0415). The proportion of patients with Ki67 <10% at 3 weeks did not differ significantly between lobular and non-lobular cases (88.6% versus 79.7%, respectively, p=0.08). Among the 40 lobular cases and 88 non-lobular cases with baseline Ki67 ≥10%, week 3 Ki67 decreased to <10% in 90% of lobular cases and 72.7% of non-lobular cases (p=0.07). On MRI, mean FTV at baseline was similar among lobular and non-lobular cases (21 cc³ versus 16.3 cc³). On pre-surgical MRI, patients in the lobular cohort had significantly higher mean residual FTV than those in the non-lobular cohort (5.7 cc³ versus 3.8 cc³, p=0.034). Interestingly, when splitting the lobular cohort into mixed lobular and pure lobular, mean residual MRI FTV was 7.5 cc³ in mixed cases and 3.8 cc³ in pure lobular cases, reflecting that FTV decreased the least in mixed lobular cases (decrease of 51.1% in mixed cases, 69.7% in pure lobular cases, and 73.9% in non-lobular cases, p=0.0255). Conclusions These results highlight differences in lobular versus non-lobular breast cancer. Lobular cases had lower baseline Ki67, and similar rates of decrease in Ki67 after 3 weeks of NET than non-lobular cases. Despite this, mean FTV after completion of 6 months of NET was higher in lobular tumors. These findings may reflect differences in early versus late indicators of response to NET, or differential response patterns by histology. Alternatively, standard response measures such as Ki67 and FTV may need lobular-specific modifications. Ongoing work includes central assessment of Ki67 from surgical specimens, analysis of changes in gene expression signatures, and evaluation of surgical outcomes such as breast conservation rates by histologic subtype.

R. Upadhyay¹, A. Alexander², Q. Ye³, K. Chen³, C. Yam², G. J. Hogan⁴, S. Zhang⁴, M. L. LaBella⁴, S. Ratzel⁴, The MDACC Inflammatory Breast Cancer Team, A. Lucci⁵, W. A. Woodward⁶, B. Lim²; ¹Hematology/Oncology Fellowship Program, MD Anderson Cancer Center, Houston, TX, ²Department of Breast Medical Oncology, MD Anderson Cancer Center, Houston, TX, ³Department of Bioinformatics and Computational Biology, MD Anderson Cancer Center, Houston, TX, ⁴Research & Development, Myriad Genetics, Inc., Salt Lake City, UT, ⁵Department of Breast Surgical Oncology, MD Anderson Cancer Center, Houston, TX, ⁶Department of Breast Radiation Oncology, MD Anderson Cancer Center, Houston, TX.

Background: Inflammatory breast cancer (IBC) is a rare, aggressive subtype characterized by rapid progression and a high risk of relapse despite curative-intent therapy. Outcomes are particularly poor for patients with residual disease following neoadjuvant treatment. Circulating tumor DNA (ctDNA)-based assays offer a noninvasive means of detecting minimal residual disease (MRD) and may enable real-time, risk-adapted decisions regarding adjuvant therapy. However, feasibility and performance data for personalized ctDNA monitoring in IBC remain limited. Methods: We prospectively enrolled adults with hormone receptor (HR)-positive (≥10% ER or PR) stage III IBC who did not achieve a pathologic complete response (pCR) to neoadjuvant chemotherapy to a single-arm phase II trial (NCT02971748). Eligible patients had no prior immunotherapy or active autoimmune disease and received pembrolizumab 200 mg IV every three weeks plus physician-chosen endocrine therapy for up to 24 months, starting during or after radiation. The primary endpoint was 2-year event-free survival (EFS); secondary endpoints included safety, toxicity, and overall survival (OS). As an exploratory analysis, the Myriad Precise ultrasensitive MRD assay was used to perform whole-genome sequencing of paired tumor/normal DNA and to subsequently design 1,000-variant capture panels. Plasma was collected postoperatively, serially during therapy, and at end-of-treatment (EOT) for ctDNA analysis. Quantitative tumor fraction (TF) and quality control (QC) metrics were tracked longitudinally. Results: Of 23 enrolled patients, 19 (83%) provided sufficient tumor or normal tissue for personalized panel design. A total of 153 out of 154 plasma samples (99%) passed all quality control (QC) metrics. ctDNA was detected in 42 of 154 samples (27%), with 23 (96%) ctDNA-positive calls below 0.01% tumor fraction (TF), underscoring the need for ultrasensitive detection. One patient who cleared ctDNA detection early remains recurrence-free (early molecular clearance). Two patients who were negative at baseline became MRD positive 10 and 15 months prior to clinical progression. For all patients with disease progression <24 months, we observed persistent and rising ctDNA levels ahead of radiographic relapse with lead times up to 18 months. In contrast, all patients who were MRD negative at the post-operative baseline and throughout treatment remained progression-free at 24 months. Correlative analyses of baseline MRD status and end-of-treatment (EOT) conversion patterns with EFS and overall survival (OS) are ongoing and will be presented. Conclusions: Personalized ctDNA MRD testing is feasible and highly sensitive in the postoperative HR-positive IBC setting, with >99% QC pass rate and detection capability below 0.001% TF. Preliminary data show that ctDNA kinetics correlate with clinical outcomes and ctDNA detection often precedes radiographic relapse by months. These findings support incorporating MRD monitoring into adjuvant trials for real-time therapeutic adaptation in high-risk breast cancer. Acknowledgments: The participants included in the current project were recruited through a Merck supported investigator-initiated trial (NCT02971748). The WGS and ctDNA analyses were performed by the Myriad and MDACC teams through a Myriad and MDACC strategic alliance. The Morgan Welch IBC Research and Clinic Program is funded by Texas State.

I. URDANIBIA¹, A. VELASCO¹, G. RODRIGUEZ², N. TUSET², A. GASOL², S. MORALES²; ¹Medical Oncology, IRB LLEIDA, LLEIDA, SPAIN, ²Medical Oncology, HospItal arnau de vilanova de lleida, LLEIDA, SPAIN.

Background: Tumor-educated platelets (TEPs) store cancer-related mRNAs and tolerate routine manipulation, providing a stable and minimally invasive window into disease biology. We wondered whether serial TEPs-RNA profiling reflects therapy-induced changes and predicts response in Luminal (HR+/HER2) breast cancer.Methods: In a prospective luminal cohort (n = 12), platelet-rich plasma was collected at diagnosis (T0), post-neoadjuvant therapy (T1) and post-surgery (TF). Fifty-five immune, proliferative and endocrine genes were quantified by RT-qPCR. Data were analysed with PLS-DA, limma differential expression, Spearman correlation with tumour shrinkage, and logistic regression for nodal status.Results: PLS-DA cleanly separated T0, T1 and TF, indicating stepwise transcriptomic re-programming. Twenty genes differed between T0 vs T1 and T0 vs TF (FDR < 0.05); five differed between T1 vs TF. Higher post-neoadjuvant IKBKG and VDR correlated with greater tumour reduction (r ≈ -0.6, p < 0.01). A two-gene model (CACNA2D1 + IKBKG) classified axillary involvement with 85% accuracy. A baseline five-gene signature (ADRB2, IL13, CACNA2D1, MET, ERBB2) was identified that distinguishes patients who achieved pathological complete response (pCR), showing an AUC of 0.87 and clear segregation in hierarchical clustering. Enrichment analysis revealed pathways related to cell cycle control, DNA repair, and p53 regulation, highlighting biological processes associated with treatment response.Conclusions: Sequential TEPs-RNA profiling reflects molecular effects of therapy and surgery and yields baseline biomarkers of complete response in luminal breast cancer. Platelet RNA warrants prospective validation as an accessible complement to tissue-based assays for personalized treatment.

C. Wei, L. Mengbo, Z. Hui; Department of Breast Surgery, Shengli Clinical Medical College of Fujian Medical University, Fuzhou, CHINA.

Background: Dephosphorylation regulates key signaling pathways in breast cancer, influencing tumor progression, metastasis, and treatment resistance. However, the role of dephosphorylation-related genes (DRGs) in breast cancer (BRCA) and their impact on patient outcomes has not been fully explored. This research sought to explore the prognostic relevance of NRGs in BRCA through the integration of bulk RNA sequencing (bulk RNA-seq) and single-cell RNA sequencing (scRNA-seq) data.Methods: The bulk RNA-seq and scRNA-seq data for BRCA, along with DRGs, were sourced from public databases. Utilizing differential expression analysis, Cox regression analysis, and a machine learning algorithm, this study identified prognostic genes. Leveraging prognostic genes, a risk model was crafted, further dichotomizing BRCA patients into two distinct risk groups. Moreover, enriched pathways and the immune microenvironment of the different risk groups were explored. Subsequent analyses involved examining the expression of prognostic genes at the single-cell level, identifying key cells, and conducting pseudo-time analysis.Results: NOS1, CTNNA2, ACTN2, and KRT5 were identified as prognostic genes in this study. The constructed risk model demonstrated robust predictive ability, with area under the curve (AUC) values surpassing 0.60 at 1-, 2-, and 3-year survival. Subsequently, the functional pathways were related to cellular functions and immune response (such as “apoptosis” and “chemokine signaling pathway”). Moreover, there were six types of immune infiltrating cells that showed significant differences between two risk groups. Additionally, scRNA-seq analysis identified eight cell types, with smooth muscle cells as the key cells, and the expression of KRT5 showed dynamic changes during the different stages of smooth muscle cells differentiation.Conclusion: NOS1, CTNNA2, ACTN2, and KRT5 were identified as prognostic genes for BRCA, highlighting the crucial prognostic role of NRGs in BRCA, which could be particularly important for identifying targeted therapeutic strategies.

L. P. Genovez, H. J. Kim, S. M. Sanches, M. G. Cesca; Clinical Oncology, A. C. Camargo Cancer Center, São Paulo, BRAZIL.

Background: Trastuzumab deruxtecan (T-DXd), an antibody-drug conjugate targeting HER2, has shown significant clinical benefit in patients with HER2-low metastatic breast cancer. The DESTINY-Breast04 trial led to its approval in this setting, particularly among hormone receptor-positive (HR+) patients. Given its topoisomerase I inhibitor payload and DNA-damaging mechanism, tumors with homologous recombination deficiency, such as those harboring BRCA1/2 mutations, may be particularly sensitive to T-DXd. While prior studies have explored this hypothesis, the predictive role of BRCA mutations in the efficacy of T-DXd in HR+/HER2-low disease remains unclear and has not been systematically quantified. Objectives: To evaluate whether BRCA1/2 mutations are associated with improved efficacy of trastuzumab deruxtecan in patients with hormone receptor-positive, HER2-low metastatic breast cancer, by comparing objective response rate (ORR) and progression-free survival (PFS) between BRCA-mutated and BRCA wild-type populations. Methods: A systematic review and meta-analysis were conducted on June 23, 2025 according to PRISMA guidelines. Literature searches were performed in PubMed, Embase, Cochrane Library, and ASCO and ESMO websites databases. Eligible studies reported T-DXd outcomes stratified by BRCA mutation status in HR+/HER2-low metastatic breast cancer. Two reviewers independently performed study selection and data extraction. The primary endpoints were objective response rate (ORR) and progression-free survival (PFS). Analyses were performed using R software (v.4.2.2) with random-effects models. The protocol for this systematic review and meta-analysis was registered in the PROSPERO international prospective register of systematic reviews (ID: CRD420251089678). Results: A total of 3,194 records were retrieved, and after deduplication, 2,461 unique records were screened. Three studies were included: two randomized trials (DESTINY-Breast04 and DESTINY-Breast06) and one large real-world cohort (Tarantino et al., 2025). Patients with BRCA1/2 mutations demonstrated a significantly higher ORR compared to BRCA wild-type counterparts, with a pooled odds ratio of 4.48 (95% CI 1.66-12.09; I² = 0%). For PFS, the pooled hazard ratio favored BRCA-mutated patients (HR 0.56, 95% CI 0.15-2.04), although with substantial heterogeneity (I² = 90.3%). Conclusions: This is the first meta-analysis to evaluate the predictive impact of BRCA1/2 mutations on the efficacy of T-DXd specifically in HR+/HER2-low metastatic breast cancer. BRCA mutations appear to be associated with significantly higher response rates and a potential progression-free survival advantage. These results highlight the need to consider BRCA status as a predictive biomarker in future clinical trials of T-DXd and support its role in guiding therapeutic decisions in HER2-low disease.

J. Dea¹, C. Chavany¹, R. P. Valle¹, R. Hernandez-Gonzalez², M. Jendoubi¹; ¹Biomarker Assay Development, Milagen, Inc., Emeryville, CA, ²Departamento de Investigacion Experimental y Bioterio, Instituto Nacional de Ciencias Medicas y Nutricion, Mexico City, MEXICO.

Background: Breast cancer (BC) is the most frequent cancer in women from western countries. 30% of BC patients are likely to develop recurrence and risk remains indefinitely. Monitoring recurrence and evaluating response to therapy are important aspects of clinical decision making in the treatment of BC. According to clinical guidelines, BC follow-up is based on clinical examination and bilateral mammogram only. Serum tumor markers such as cancer antigen 15-3 (CA15-3) and other imaging techniques are used by many physicians, but without clear recommendations from clinical societies. We used an in-house developed modular proteomic array platform, the Matrix Protein Array to identify a protein biomarker (BF-09 antigen) that was differentially expressed in BC tissues versus normal counterparts, present in early and late stages and measurable in serum. The development of an ELISA assay to establish its diagnostic performance in discriminating BC patients from healthy individuals and benign cases was previously described ^1-2. The present study is designed to assess the utility of our assay for monitoring BC recurrence and therapy response. Methods: The concentration of BF-09 antigen was measured by ELISA assay in 379 retrospective serial serum samples from 85 BC patients and changes in biomarker concentration were correlated to change in disease status as determined by standard of care. BC patient samples were obtained from a biobank at the IRCCS San Raffaele Pisana Research Center (Italy). Patients were followed for up to 10 years (median 23.1 months, range 2.9-203 months) and were examined periodically with imaging to document the remission or progression of the disease. 67% of patients were diagnosed with early-stage BC (I-III) while 33% had metastatic breast cancer (stage IV). Treatments received by patients were neoadjuvant therapy (NAT, 16.5%), adjuvant therapy (AT, 18.8%), metastatic breast cancer therapy (MBCT, 35.3%), or multiple therapies (AT± NAT and/or MBCT, 29.4%). A significant increase or decrease in BF-09 concentration was defined as a change equal or superior to 2.5 times the %CV of the assay as compared to a previous point and interpreted as progression or no progression, respectively. Sensitivity (SE), specificity (SP), negative (NPV) and positive predictive values (PPV) as well as overall concordance of the assay with disease status were calculated. Results: Total concordance of BF-09 serum level with disease status was not affected by BC subtypes (luminal A/B, HER2 positive or triple negative) or stage (Chi-square, p≥0.05). Overall performance of the assay in detecting progressive disease was 81.8% sensitivity at 92.6% specificity with 68.7% PPV and 96.2% NPV. The BF-09 assay could identify BC recurrence both in early BC patients following curative surgery/treatment and in metastatic BC patients during therapy follow up, in some cases years before patients started suffering from clinical symptoms. We were also able to show the utility of BF-09 antigen assay in monitoring therapy response in the neoadjuvant setting prior to surgery. Conclusion: Our findings show that the BF-09 biomarker assay quantitatively reflects therapy response both in the (neo)adjuvant therapy and metastatic therapy settings and detects cancer relapses in a timely manner. This appears to be the first time a protein biomarker has shown high accuracy and applicability for these clinical indications and this could offers new opportunities for BC patient management. Further testing is needed to confirm these results in a prospective trial population. ¹ Chavany et al. New breast cancer marker BF-09 is overexpressed in tumor extracts and secreted in serum, Biochemistry and Biophysics Reports, Volume 43, 2025,102097. ² Chavany et.al. A Serum Biomarker for the Early Detection of Breast cancer. Cancer Res (2024) 84 (9_Supplement): PO5-07-08.

V. Carreno, A. H. Chang, I. G. Fernandez, R. Qamar, M. T. West, K. B. Wisinski, M. N. Sharifi; Department of Medicine, University of Wisconsin, Madison, Madison, WI.

Background: Breast cancer is a heterogeneous disease with multiple transcriptionally defined subtypes including luminal A and B hormone receptor (HR) positive subtypes, the HER2 amplified subtype, and the basal subtype most commonly associated with clinically triple negative breast cancer (TNBC). These transcriptional subtypes have differing disease biology, clinical prognosis, and treatment vulnerabilities, and studies utilizing paired tissue biopsies have demonstrated that subtype switching/evolution can occur from primary tumor to metastatic recurrence, as well as in metastatic disease over time on treatment. However, understanding the role of subtype switching in treatment response and resistance has been limited by the challenges of obtaining serial tissue biopsies from patients. Liquid biopsies, including circulating tumor DNA (ctDNA) and circulating tumor cells (CTCs), are a source of tumor-derived material shed into the blood of patients with cancer. Because liquid biopsies require only a peripheral blood draw, they are uniquely suited for longitudinal sampling to evaluate for subtype evolution in metastatic breast cancer. Here, we report the development of a targeted gene expression assay for genes associated with estrogen and HER2 signaling as well as luminal and basal identity in breast cancer CTCs, to address this need. Methods: 15mL peripheral blood was collected in EDTA vacutainer tubes from patients with metastatic breast cancer. CTCs were isolated with automated mTAE technology integrating live cell CTC capture using anti-EpCAM-conjugated paramagnetic particles with RNA extraction. CTC RNA underwent pre-amplification followed by qRT-PCR for genes related to epithelial cell identity, estrogen signaling, HER2 and other growth factors, basal differentiation, as well as controls for immune cell content and housekeeping genes. Expression values were calculated as 33-Ct values. Epithelial/CTC content score was calculated by summing the expression of the epithelial cell identity genes (EpCAM, KRT18, KRT19). Samples with epithelial content score > 10 were classified as having high CTC content. Results: CTCs were detected in 34/51 (66.6%) of HR+ samples, 3/3 (100%) of TNBC samples, and 2/3 (66.6%) of HER2+ samples. 27/51 (52.9%) of HR+ samples, 2/3 (66.6%) of TNBC samples, and 1/3 (33.3%) of HER2+ samples were classified as having high CTC content. In patients with progressing disease (24 samples from 21 patients), 20/24 (83.3%) had CTCs detected and 17/24 (70.8%) had high CTC content. In patients with stable disease (19 samples from 15 patients) 14/19 (73.7%) samples had CTCs detected and 8/19 (42.1%) samples had high CTC content. Hierarchical clustering of high CTC content samples by expression of the estrogen signaling, HER2/growth factor, and basal genes demonstrated clusters consistent with differing transcriptional subtypes. Among HR+ HER2 non-amplified samples (n=22), higher estrogen signaling gene expression was associated with shorter progression free survival (median 2.2 versus 6.9 months, log rank p<0.05) Conclusions: In a pilot cohort of 37 patients with metastatic breast cancer, we demonstrate the feasibility of detecting gene expression relevant to breast cancer transcriptional subtypes via liquid biopsy circulating tumor cell gene expression profiling and relatively high sensitivity even in patients with stable disease, with CTCs detected in 73% of samples including 42% with high CTC content. Interestingly, in HR+ breast cancer CTCs in this cohort, higher estrogen-regulated gene expression was associated with shorter progression free survival, suggesting that this assay could also be utilized to monitor response to anti-estrogen therapy in this patient population.

Z. Qu¹, S. Pei², K. Li³, F. YI²; ¹Department of Biostatistics, Epidemiology, and Informatics, Perelman School of Medicine, University of Pennsylvania, PHILADELPHIA, PA, ²Department of Breast Surgical Oncology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, CHINA, ³Department of Clinical Laboratory, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, CHINA.

Background: Cancer-related fatigue (CRF) is a debilitating and underrecognized symptom in breast cancer, often preceding treatment and co-occurring with psychological and physical complaints. However, its pathobiological basis remains poorly understood. We leveraged the China PERSEVERE study (NCT07010939), a national prospective cohort of early breast cancer patients, to investigate baseline symptom clusters and identify molecular correlates of CRF using multi-omics profiling. Methods: PERSEVERE is an ongoing multi-center cohort study designed to evaluate treatment-related toxicities and quality of life in women with stage I-III breast cancer in China. Baseline assessments, all conducted prior to surgery or neoadjuvant treatment, include detailed sociodemographic and clinical data, biospecimen collection (plasma, PBMCs, and tumor tissue), and validated patient-reported outcome (PRO) measures (EORTC QLQ-C30/BR23, FA12, IPAQ, HADS, PSS-10, PSQI, SSRS). This cross-sectional analysis included 168 newly diagnosed breast cancer patients enrolled to date, all of whom had complete baseline PROs and biospecimen collection. Clinically significant CRF was defined as a QLQ-C30 fatigue score ≥40 or any FA12 fatigue domain score ≥40. Symptom interrelationships were analyzed using Spearman correlation and hierarchical clustering. Multivariable logistic regression identified predictors of baseline CRF. A nested case-control subgroup of 14 participants (7 with high fatigue scores and 7 matched controls based on age and clinical stage) underwent untargeted baseline serum proteomic and metabolomic profiling. Results: At baseline, 39/168 participants (23.2%) met the criteria for clinically significant CRF. CRF scores showed moderate correlation with poor sleep quality (PSQI; r=0.37, p<0.001), anxiety (HADS-A; r=0.32, p<0.05), and perceived stress (PSS-10; r=0.25, p=0.01), forming a distinct symptom cluster. Patients with CRF also exhibited significantly higher systolic blood pressure, breast-specific symptom burden (BR23), and elevated IL-6 levels (all p<0.05). In adjusted models, BR23 symptom score and waist-to-hip ratio (WHR) remained independent predictors of CRF. In the multi-omics subgroup, 52 differentially expressed proteins were identified, enriched in cytokine signaling (e.g., IL-6 receptor activity) and mitochondrial pathways (e.g., oxidative phosphorylation). Metabolomic analysis revealed 76 altered metabolites, implicating disrupted kynurenine metabolism, glutathione balance, and neuroactive signaling. Integrated correlation analysis linked CRF severity with altered inflammatory markers (CRP, IL6R) and mitochondrial dysfunction (e.g., antimycin A1, tiopronin) across both omics layers. Conclusions: CRF is a core feature of a symptom cluster involving sleep and psychological distress, even before systemic therapy. This integrated clinical and molecular analysis from the PERSEVERE cohort highlights inflammation and mitochondrial dysfunction as plausible contributors to baseline fatigue. These findings inform early identification of high-risk individuals and suggest future intervention targets. Longitudinal follow-up is underway to validate predictive trajectories and biological correlates of CRF over time.

L. Pontolillo¹, L. Bucheit², L. Tung³, E. Nicolò¹, M. Serafini¹, N. Bayou¹, B. Pasto¹, L. Gerratana⁴, E. Andreopolou¹, E. Bria⁵, C. Reduzzi¹, M. Cristofanilli¹; ¹Department of Medicine, Division of Hematology-Oncology, Weill Cornell Medicine, New York, NY, ²Medical Affairs, Guardant Health, Palo Alto, CA, ³Bioinformatics, Guardant Health, Palo Alto, CA, ⁴Department of Medicine, University of Udine, Udine, ITALY, ⁵Translational Medicine and Surgery, Catholic University of Sacread Heart, Rome, ITALY.

Background: Hormone receptor (HR) and HER2 status, typically assessed on tissue biopsies, are essential for breast cancer (BC) management. As multiple therapeutic options are available for receptors positive metastatic BC (mBC), ongoing subtype assessment may be recommended to monitor changes under the selective pressure and/or evolution of disease. Cell-free DNA (cfDNA) showed reliability for biomarker evaluation and high concordance with tissue-based methods and may be an opportunity for non-invasive, real -time subtype assessment. This study aimed to evaluate agreement between cfDNA methylation-based molecular breast subtyping (MBS) and tissue status in a clinical cohort who underwent cfDNA testing as part of routine clinical care. Methods: Blood samples drawn from mBC patients with HR+ disease on the most recent tissue biopsy and tested with Guardant360 at Weill Cornell Medicine from Nov 2024-June 2025 were queried. A novel cfDNA MBS signature was applied to test results through bioinformatics and/or residual sample testing. MBS confidence scores for HR+/HER2+ where tumor fraction (TF) >0.5% were applied. In this research setting, MBS predicted overall subtype was generated to compare it to the most recent tissue biopsy available. Clinical and genomic factors were annotated for correlation; Fisher’s exact and t-tests were used for statistical analysis.Results: 73 samples were screened of which 42 (57.5%) were evaluable for MBS (TF>0.5%). Of those not evaluable, 16/31 (51.6%) had maximum variant allele frequency <1% and two had no circulating tumor DNA detected. Notably, 24/31 (77%) non-evaluable samples had >1 genomic alteration annotated as clonal hematopoiesis. Two had undetermined results with TF>0.5% but confidence below MBS thresholds, thus 40 samples were included in agreement analysis. A total of 31/40 (77.5%) samples had initial MBS agreement with tissue of whom 27 (87.1%) were HR+/HER2+. The median confidence score from MBS for HR+ prediction was 80% (range: 30%->90%). Of 9/40 (22.5%) samples with initial MBS disagreement with tissue, one (11.1%) differed in HR status (MBS HR-/tissue HR+) and 8 (88.9%) differed in HER2 status (3 MBS HER2+/tissue HER2-; 5 MBS HER2-/tissue HER2+). For the HR case, concomitant genomic analysis did not identify any alterations typically associated with HR+ (ESR1, PIK3CA, PTEN, AKT1). When exploring clinical factors, among samples that had MBS/tissue agreement vs. disagreement, there was no significant difference in: average time from tissue to ctDNA testing (521 days vs. 494 days), number of metastatic sites (2.84 vs. 2.78), number of prior therapies (2.78 vs. 2.35) or prior exposure to CDK4/6 inhibitors (64.2% vs. 77.8%); all p>0.05. Although TF was higher in cases with agreement (mean: 23.4% vs. 20.5%; median: 8.06% vs. 5.26%) this difference was not significant (p>0.05). Upon clinical review of cases with disagreement,1 had MBS HR+/HER2- yet had HR+/HER2- on previous tissue biopsies, concordant with the MBS subtype. A second case had MBS HR+/HER2+ yet HR+/HER2- on the most recent tissue biopsy done in 2023; this case had previous HR+/HER2+ biopsies in 2017 (primary diagnosis) and 2020 (metastatic relapse) and received anti-HER2 therapies, suggesting MBS may capture heterogeneity even after therapeutic intervention. Thus, adjusted agreement of MBS with additional clinical information was 82.5% (33/40).Conclusions: In this mBC cohort, cfDNA MBS showed reliable agreement to tissue biopsy and replicated previously reported accuracy. Subtype switching, heterogeneity and/or lower levels of MBS confidence may contribute to disagreement with tissue. MBS may be a feasible non-invasive option for continuous subtype assessment for HR+ mBC.

L. Foffano¹, A. Davis², C. Reduzzi³, E. Podany², A. Medford⁴, K. Clifton², M. Velimirovic⁵, B. Pastò¹, L. Pontolillo³, R. Occhiogrosso Abelman⁶, C. Gianni⁷, S. Tapiavala², E. Molteni¹, M. Lypsic-Sharf⁸, E. Nicolò³, E. Andreopoulou³, W. Gradishar⁹, F. Puglisi¹, C. Ma², A. Bardia⁸, L. Gerratana¹, M. Cristofanilli³; ¹Department of Medicine, University of Udine, Udine, ITALY, ²School of Medicine, Washington University, St Louis, MO, ³Department of Medical Oncology, Weill Cornell Medicine, New York, NY, ⁴Oncology, Massachusetts General Hospital, Boston, MA, ⁵Taussig Cancer Institute, Cleveland Clinic, Cleveland, OH, ⁶Department of Oncology, Massachusetts General Hospital, Boston, MA, ⁷Department of Oncology, IRST Dino Amadori, Meldola, ITALY, ⁸Jonsson Comprehensive Cancer Center, University of California, Los Angeles, CA, ⁹Feinberg School of Medicine, Northwestern University, Chicago, IL, USA, IL.

Background: Hormone receptor–positive/HER2-positive metastatic breast cancer (HER2+/HR+ MBC) represents a biologically distinct subtype, characterized by extensive crosstalk between estrogen receptor (ER) and HER2 signaling pathways. The treatment landscape for HER2+ MBC is rapidly evolving, with recent data from the DESTINY-Breast09 trial supporting consideration of T-DXd in the first line and the PATINA trial demonstrating that adding the CDK4/6 inhibitor palbociclib to maintenance therapy significantly prolonged progression-free survival in patients with HR+/HER2+ disease. The aim of this analysis was to compare genomic profiles and outcomes across HR+/HER2–, HER2+/HR+, and HER2+/HR– subtypes through ctDNA assessment. Methods: This retrospective study analyzed a multi-institutional cohort of 1318 patients (pts) with MBC and ctDNA testing with the Guardant360 NGS panel within a large academic consortium (PMAC). HR and HER2 status were defined based on the most recent tissue biopsy. Associations between single nucleotide variants (SNVs), copy number variations (CNVs) and breast cancer subtypes were tested by multinomial logistic regression (MLR) in terms of Relative Risk Ratio (RRR). The impact of prognosis was evaluated through Cox regression for overall survival (OS), defined from time of baseline ctDNA collection. Results: Among 1318 pts with MBC, 1104 (83.8%) had HR+/HER2–, 127 (9.6%) HER2+/HR+, and 87 (6.6%) HER2+/HR- disease. Multivariable MLR, designed with HR+/HER2- as the reference subtype, investigated clinical variables and genomic alterations across subtypes. Clinically, the presence of nodal (RRR 1.82, p < 0.001) and soft tissue (RRR 2.60, p < 0.001) metastases were more common for HER2+/HR+ but not for HER2+/HR-, while central nervous system (CNS) metastases were associated with both HER2+/HR+ (RRR 3.41, p < 0.001) and HER2+/HR- (RRR 2.73, p = 0.006). Considering single genes alterations, at multivariable analysis, PIK3CA SNVs were less common (RRR 0.51, p = 0.010) and ERBB2 CNVs more common (RRR 41.17, p < 0.001) for HER2+/HR+, while ERBB2 CNVs were also associated with HER2+/HR- (RRR 85.64, p < 0.001). Considering pathway alterations, HER2+/HR+ had fewer PI3K SNVs (RRR 0.47, p = 0.003) and ER SNVs (RRR 0.52, p = 0.022), while in HER2+/HR- there were fewer ER SNVs (RRR 0.05, p = 0.004) and more MYC CNVs (RRR 5.10, p = 0.029). In OS analysis, both HER2+/HR+ (HR 0.55, p = 0.002) and HER2+/HR- (HR = 0.61, p = 0.027) had a significantly favorable prognostic impact compared to HR+/HER2-. Within the HR+/HER2- cohort, several pathway alterations were associated with worse prognosis, including ER SNVs (HR 1.33, p = 0.007), P53 SNVs (HR 1.46, p < 0.001), PI3K CNVs (HR 1.54, p = 0.028) and MYC CNVs (HR 1.84, p = 0.001). In HER2+/HR+, a significant prognostic impact emerged for PI3K SNVs (HR 2.66, p = 0.016), P53 SNVs (HR 2.46, p = 0.036) and RTK CNVs (HR 3.82, p = 0.004), while in HER2+/HR- it was observed only for PI3K CNVs (HR 79.65, p = 0.003). Conclusions: Our ctDNA-based analysis confirms that HER2+/HR+ MBC is a clinically and molecularly distinct subtype. Compared with HR+/HER2– disease, HER2+/HR+ MBC is characterized by higher ERBB2 amplification and a lower frequency of ER SNVs, suggesting reduced dependence on estrogen signaling despite hormone receptor positivity, and the presence of distinct mechanisms of endocrine resistance. Prognostically, both HER2+/HR+ and HER2+/HR– subtypes had significantly more favorable overall survival, suggesting an intrinsic advantage linked to the success of therapeutics aimed at targeting HER2. These findings therefore highlight the potential value of comprehensive baseline and longitudinal mutational profiling to guide initial and maintenance therapy selection and support personalized treatment approaches in HR+/HER2+ disease.

S. Li, Z. Ningning, S. Qing, W. Guanwen, Q. Fanli, W. Long, Q. Yang, Z. Xiaohua; Breast Cancer Center, Chongqing University Cancer Hospital, Chongqing, CHINA.

Background and Purpose: Although immune checkpoint inhibitors (ICIs) combined with chemotherapy have been established as the standard treatment regimen for triple-negative breast cancer (TNBC), the proportion of patients deriving significant clinical benefit remains limited. Identifying novel therapeutic targets and developing synergistic strategies to enhance the efficacy and broaden the applicability of ICIs represent critical areas of ongoing research. Methods: A series of comprehensive experiments, including 4T1 orthotopic breast cancer mouse models, tumor-T cell co-culture systems, transcriptome sequencing, flow cytometry, immunoblotting, and qPCR, were performed to systematically elucidate the anti-tumor immune mechanisms of the drug combination. Results: The combination of CDK4/6 and EZH2 inhibitors significantly inhibited 4T1 orthotopic tumor growth and promoted CD8⁺ T cell infiltration. Transcriptomic analysis identified differential gene enrichment in pathways related to cell adhesion and migration. qPCR demonstrated a notable upregulation of chemokine mRNAs (e.g., CXCL9, CXCL10, CXCL11) in the combination treatment group, while in vitro migration assays confirmed enhanced recruitment of CD8+ T cells. Mechanistically, the combination therapy activated the MAVS-TBK1 axis via dsRNA signaling (but not dsDNA) signaling, leading to STAT1 phosphorylation and increased IFN-β secretion, which subsequently upregulated chemokine expression (no such effects were observed with H151, a STING pathway inhibitor). Furthermore, the combination therapy synergistically enhanced the efficacy of anti-PD-1 monoclonal antibodies. Conclusion: This study demonstrates that the combination of CDK4/6 and EZH2 inhibitors induces chemokine expression and facilitates CD8⁺ T cell infiltration via the MAVS-TBK1-STAT1 axis, thereby enhancing anti-tumor immunity and providing a novel synergistic approach for TNBC immunotherapy.

W. Chen¹, X. Qin¹, X. Xu², X. Lang³, G. Xie⁴, W. Liang¹; ¹Department of Breast Surgery, Changzhou No.2 People’s Hospital, the Third Affiliated Hospital of Nanjing Medical University, Changzhou, CHINA, ²Department of Thyroid and Breast Surgery, Northern Jiangsu People’s Hospital, Yangzhou, CHINA, ³Department of General Surgery, Sunshine Union Hospital, Weifang, CHINA, ⁴Department of Breast Surgery, The People’s Hospital of Chuxiong City, Chuxiong, CHINA.

Background: Accumulating evidence has demonstrated that epithelial-mesenchymal transition (EMT) plays a critical role in breast cancer (BRCA) initiation, invasion, metastasis, and prognosis. BRCA still has a high mortality rate owing to recurrence and metastasis. As a core component of the biological signal regulatory network, EMT urgently requires accurate biomarkers to evaluate the prognosis of BRCA. Methods: We obtained 1223 BRCA samples from The Cancer Genome Atlas and 1184 EMT-related genes from the dbEMT2.0 public database. Differentially expressed genes were filtered, and univariate Cox and LASSO-Cox regression analyses were performed to select prognosis-related gene signatures. A novel risk score model was developed and validated using external cohorts. Moreover, mutual confirmation between KEGG/GO and gene set enrichment analyses was performed to reveal potential molecular mechanisms. Finally, a nomogram including the risk score model and several clinical parameters was established to help individualize patients’ survival predictions.Results: A total of 381 differentially expressed EMT-associated genes (EAGs) were identified and 15 EAGs relevant to survival prognosis were used to calculate the risk score formula. Patients were divided into high- and low-risk groups based on the specific risk score model. Samples in the low-risk group had a higher overall survival. One UCSC cohort and three Gene Expression Omnibus cohorts verified the reliability of the model. A nomogram was established to predict survival, including the risk score, age, pathological N stage, and pathological M stage. Ultimately, the results of gene set enrichment analysis revealed that the potential molecular mechanism was mainly concentrated in the extracellular region.Conclusions: The 15-gene signature model based on EMT is significant for predicting overall survival and revealing possible molecular mechanisms as potential targets and tools for pharmacologic and regenerative medicine biomarkers.

W. Chen¹, X. Qin¹, X. Xu², X. Lang³, G. Xie⁴, W. Liang¹; ¹Department of Breast Surgery, Changzhou No.2 People’s Hospital, the Third Affiliated Hospital of Nanjing Medical University, Changzhou, CHINA, ²Department of Thyroid and Breast Surgery, Northern Jiangsu People’s Hospital, Yangzhou, CHINA, ³Department of General Surgery, Sunshine Union Hospital, Weifang, CHINA, ⁴Department of Breast Surgery, The People’s Hospital of Chuxiong City, Chuxiong, CHINA.

Background: Accumulating evidence has revealed that epithelial-mesenchymal transition (EMT) plays a crucial role in tumor progression and the immune microenvironment, which further results in a high rate of recurrence and metastasis. The EMT immune signaling pathway provides a great perspective for designing personalized therapies. Methods: In this study, 1223 RNA-seq samples were obtained from the TCGA-BRCA dataset. A total of 381 EMT-related differentially expressed genes were analyzed and combined with clinical parameters, and the matrix was divided into training and testing cohorts at a ratio of 7:3. The training cohort was used to develop an EMT signature, including GKN2, FZD2, NDRG2, SCUBE2, ALX4, CCL19, SDC1, EZR, CPEB1, and HRG genes, and the accuracy of this signature was validated by testing and GSE158309 cohorts. Results: A risk score model and clinical parameters were used to establish a nomogram for predicting prognosis. The C-index (0.719), calibration curves, and model comparison with four previous studies demonstrated the reliability of the EMT signature, the biological phenotypes of which were tested for functional enrichment, immune infiltration, and tumor mutation. Additionally, patients’ responses to immunotherapy and chemotherapy were assessed. Our results showed that the low-risk group had higher immune infiltration, tumor mutational burden, microsatellite instability levels, immune checkpoint inhibitor expression, tumor immune dysfunction and exclusion scores, and immunophenoscores, which could predict patient sensitivity to immunotherapy. Moreover, low-risk patients exhibit better sensitivity to chemotherapy. Conclusion: This novel EMT signature offers excellent potential for predicting the prognosis, tumor immune heterogeneity, and therapeutic responses in breast cancer.

Y. Lin, J. Wu, Y. Li, F. Fu, C. Wang; Department of Breast Surgery, Fujian Medical University Union Hospital, Fuzhou, CHINA.

Background: Triple-negative breast cancer (TNBC) is more aggressive and has higher metastasis rate compared with other subtypes of breast cancer. RAN binding protein 2 (RANBP2) is a less-explored SUMO E3 ligase and has been shown to play key roles in various cellular cycle processes. However, its biological function and molecular mechanism in TNBC remain largely unclear. Methods: RANBP2 expression was evaluated in TNBC patients’ s tumor tissues with public database and immunohistochemical staining methods. Correlation between RANBP2 expression and patients’ prognosis was also analyzed. The effects of RNABP2 on TNBC cell proliferation ability were detected by CCK-8, colony formation, apoptosis and cell cycle analysis. Tumor formation assay was constructed to examine the effects of RANBP2 on TNBC growth in vivo. Proteomic profiling, Western blot, protein stability tests and rescue experiments were further applied to evaluate the detailed mechanism. Results: The expression of RANBP2 was indicated to be upregulated in TNBC tumor tissues and high expression of RANBP2 was correlated with poor prognosis of TNBC patients. Functional studies demonstrated that RANBP2 overexpression promoted TNBC cell proliferation both in vitro and in vivo, whereas RANBP2 knockdown inhibited proliferation, induced apoptosis, and caused G1-S phase cell cycle arrest. Baculoviral IAP Repeat Containing 5 (BIRC5) was identified as a downstream effector of RANBP2. Mechanistically, RANBP2 post-translationally regulated BIRC5 by facilitating its degradation via the proteasomal pathway. Functional rescue experiments further demonstrated that BIRC5 overexpression could partially reverse the anti-proliferative effects induced by RANBP2 knockdown in TNBC cells. Conclusion: RANBP2 serves as a critical regulator of TNBC development and progression. Targeting RANBP2/BIRC5 axis may provide new insights into therapeutic strategies and research directions for TNBC treatment.

G. Gastelum Martinez, M. Rezaei, P. Miller, C. Pine, Y. Anbalagan, N. Bishopric, B. Hudson, M. Lippman; Department of Oncology, Lombardi Comprehensive Cancer Center Georgetown University, Washington, DC.

Resistance to antiestrogen therapy is a major challenge to the treatment of estrogen receptor–positive (ER⁺) metastatic breast cancer (MBC). While endocrine therapies such as tamoxifen, aromatase inhibitors, and fulvestrant are initially effective, most patients ultimately develop resistance, leading to progressive metastatic disease. Psychosocial stress, widely reported in breast cancer patients, is a clinically relevant factor that worsens outcomes. Elevated chronic levels of norepinephrine (NE), a hallmark of stress physiology, correlate with poor prognosis and increased metastatic burden in breast cancer. We and others have shown the clear negative effect of chronic adrenergic stress in driving immunomodulatory changes that favor tumorigenesis. However, the direct, tumor-intrinsic effects of adrenergic stimulation remain underexplored and not well characterized.Here, we explore a previously unrecognized mechanism by which adrenergic stress drives ER⁺ breast cancer cells toward a therapy-resistant and metastatic phenotype. We first confirmed that human ER⁺ breast tumor, and culture murine and human ER⁺ breast cancer cells, express α2A-, β1-, and β2-adrenergic receptors at both transcript and protein levels using integrated transcriptomic analysis and validated through immunocytochemistry. Functionally, NE promotes proliferation in both estrogen-rich and estrogen-deprived conditions and cooperates with 17β-estradiol (E2) to significantly enhance clonogenicity, anchorage-independent sphere formation, CD44 expression, and migration—features consistent with cellular plasticity and partial cancer stem-like phenotypes. Crucially, we show that NE partially rescues ER⁺ breast cancer cells from growth inhibition by fulvestrant and tamoxifen, suggesting adrenergic signaling acts as a bypass mechanism that subverts endocrine blockade. We also observe that long-term exposure to NE plus E2 confers persistent cellular changes, as these cells maintain enhanced clonogenic potential even after withdrawal of stimuli. Pharmacologic interrogation using selective adrenergic antagonists abrogates these NE-mediated phenotypes in ER+ breast cancer cells and identifies α2A-adrenergic signaling as a key driver of endocrine resistance. Our findings establish that chronic adrenergic stimulation provides selective pressure on ER⁺ breast cancer cells and drives therapy-resistance, and transition of cells into stem-like states through tumor-intrinsic mechanisms independent of immune suppression. This work increases the potential role of stress in ER⁺ breast cancer and supports a new therapeutic paradigm—targeting adrenergic signaling not just to modulate the immune microenvironment, but to directly disrupt tumor plasticity and restore endocrine sensitivity. These insights provide strong rationale for incorporating adrenergic inhibitors into treatment strategies and open the door for potential clinical trials leveraging stress-targeted therapies to overcome therapeutic resistance for patients with endocrine-resistant MBC.

Y. Wu¹, K. Wang², Z. Li³; ¹Department of Internal Medicine, SUNY Downstate Health Sciences University, Brooklyn, NY, ²Clinical and Translational Research Center, Shanghai First Maternity and Infant Hospital, Tongji University School of Medicine, Shanghai, CHINA, ³Department of Breast Surgery, Shanghai First Maternity and Infant Hospital, Tongji University School of Medicine, Shanghai, CHINA.

Background: Tryptophan (Trp), an essential amino acid, is metabolized through the kynurenine, indole, and serotonin pathways. Imbalances in Trp metabolism have been implicated in cancer progression. However, clinical trials targeting indoleamine-2,3-dioxygenase (IDO), a key enzyme in Trp catabolism, have yielded unsatisfactory results, prompting the exploration of alternative therapeutic targets within the Trp metabolic network. Emerging evidence discovers that several Trp-derived metabolites can bind to and activate the aryl hydrocarbon receptor (AhR), a ligand-dependent transcription factor that is essential in immune response and tumorigenesis. Although AhR is widely expressed in various breast cancer subtypes, its precise role in breast cancer development remains unclear, likely due to disease complexity and the diversity of AhR ligands. Thus, investigating Trp metabolism in relation to AhR activation may provide novel insights into the role of Trp-activated AhR pathway in breast cancer progression. Methods: Clinical data and serum samples were collected from patients with breast cancer (n = 13) and age-matched individuals with breast fibroadenoma (n = 9). Eighteen Trp metabolites were quantified using Ultra-High Performance Liquid Chromatography (UHPLC). Metabolite levels were normalized using Z-score transformation and visualized via heatmap. Hierarchial clustering was performed to assess sample grouping. AhR mRNA expression levels were compared between primary breast cancer tissues (n = 1111) and benign tissues (n = 113) by retrieving RNA-sequencing data from The Cancer Genome Atlas (TCGA) Breast Invasive Carcinoma cohort. Results: Differential levels of Trp metabolites were identified between breast cancer and fibroadenoma groups, with Trp, cinnavalininate, 5-hydroxyindole-3-acetic acid and picolinic acid showing the most significant differences. Heatmap visualization revealed distinct Trp metabolite profiles across all samples. Hierarchical clustering showed partial segregation between the breast cancer and fibroadenoma groups. Analysis of TCGA data demonstrated significantly reduced AhR mRNA expression in breast cancer tissues compared to benign tissues. Conclusion: Trp metabolism is disrupted in breast cancer. The partial clustering of breast cancer and fibroadenoma samples reflects both the biological heterogeneity of breast cancer and the overlapping metabolic features between the two groups. The observed alterations in Trp metabolism, along with reduced AhR expression in breast cancer compared to benign tissues, suggest that impaired Trp-activated AhR signaling may play a critical role in breast cancer pathogenesis. These findings support further investigation into the Trp-activated AhR pathway as a potential therapeutic target in breast cancer management.

E. Fountzilas¹, K. Papadopoulou¹, N. Korfiatis¹, A. Goussia¹, A. Christopoulou¹, E. Moirogiorgou¹, N. Tsoukalas¹, G. Petrakis¹, C. Gouedard², F. Zagouri¹, D. Stefanou¹, F. Kyriakou³, E. Vorrias³, S. Karageorgopoulou¹, F. Dimitrakopoulos¹, A. Vernadou¹, S. Murray², G. Fountzilas¹; ¹Data Office, Hellenic Cooperative Oncology Group (HeCOG), Athens, GREECE, ²Molecular Oncology, BioPath Innovations SA, Athens, GREECE, ³Medical Oncology, German Oncology Center, Limassol, CYPRUS.

Background: Robust clinico-genomic predictors of survival in dnMBC remain ill defined. We employed next generation sequencing (NGS) to explore the mutational spectrum of a Greek dnMBC cohort alongside key clinicopathological factors and evaluated their prognostic impact for overall survival (OS). Patients and Methods: Formalin-fixed, paraffin-embedded tumor tissue blocks were collected from the HeCOG Biobank for 106 patients with dnMBC alongside peripheral blood samples for 91 cases. NGS was performed on Ion Torrent platforms with AmpliSeq DNA panels targeting genes recurrently altered in BC. Variants classified as pathogenic or likely pathogenic (PVs) according to Clinvar and ACMG/AMP criteria were then processed for analysis. Cox proportional-hazards regression models were applied to assess prognostic relevance of both genomic and clinicopathological variables. Results: The total study population consisted of 106 women with a median age of 63 years (31-86); 67 patients (64%) were postmenopausal Overall, 62.5% presented with a PS of 0 and 74% with NST histology. Regarding tumor profiling, 82% were hormone receptor (HR) positive, 28.8% HER2+, 50% grade 3. Family history of cancer was positive in 17.3% of patients. Notably, 42% of patients underwent some type of breast surgery. First-line treatment included a combination of aromatase inhibitor (AI) and CDK4/6 inhibitor in 36 (34%) patients, chemotherapy (CT) alone in 27 (23.6%), combination of CT and other targeted agents in 30 (28.3%), hormonal therapy only in 7 (6.6%), while for 6 patients (5.6%) treatment data were missing. Moreover, 11 patients (10.3%) were diagnosed with locoregional (LR) disease only, 45 (42.5%) with distant and 42 (39%) with LR and distant metastases. Regarding IHC subtypes tested locally, 70 patients (66%) were classified as Luminal A/B, 31 (29.2%) HER2+ and 5 (4.7%) as TNBC. At least one PV was present in 69 of 106 tumors (65.1 %), with most carrying PVs in PIK3CA (40 cases; 37.7 %) and TP53 (29 cases; 27.4 %). Recurrent, albeit less frequent, PVs were also observed in ERBB2, GATA3, PTEN, CDH1 and AKT1 (each present in ≥ 2 tumors). Meanwhile, germline sequencing revealed single PVs in two patients, one in TP53 and the other in PTEN. In univariate analyses, none of the standard clinicopathological parameters (age, menopausal status, performance status, histology, T-stage, grade, ER/HER2 status) or individual PVs reached statistical significance for OS. No significant difference was found between Luminal and HER2+ subgroups. Conclusion: The mutational landscape of our Greek dnMBC series reveals a high prevalence of actionable PIK3CA, TP53 and ERBB2 alterations. These findings emphasize further the biological heterogeneity of dnMBC disease and the need for the identification of novel integrative biomarkers beyond standard clinico-genomic parameters to improve prognosis and guide therapeutic decisions. The study was registered on ClinicalTrials.gov under the identifier NCT05758948

A. Abukhdeir¹, K. Cagin², J. Borgia², M. Cobleigh¹; ¹Department of Medicine, Rush University, Chicago, IL, ²Department of Anatomy and Cell Biology, Rush University, Chicago, IL.

Background: Approximately 6% of triple-negative breast cancers (TNBCs) carry mutations in the PIK3R1 gene (PIK3R1^MUT). We previously showed that PIK3R1 knock-out causes increased phosphorylation of MEK and enhanced sensitivity to MEK inhibitors, trametinib and binimetinib. Alpelisib targets p110-α in the PI3K/AKT/mTOR pathway. We hypothesized that alpelisib might also inhibit growth of PIK3R1^MUT TNBC. This study tested binimetinib, alpelisib, and their combination in PDX models with PIK3R1^MUT. Methods: Four TNBC PDX models were studied (Table 1). All models had wild-type PIK3CA. Control mice had wild-type PIK3R1, while experimental mice had PIK3R1^MUT. Animals were treated with vehicle control, binimetinib (10 mg/kg BID), alpelisib (35 mg/kg QD), or their combination for at least 28 days. Tumor volumes were recorded biweekly. The genetic backgrounds of each model are listed. All mice had co-mutations in PI3K or MAPK pathway genes but lacked co-mutations in both pathways. Tumors were analyzed using mass spectrometry to elucidate downstream protein and phosphoprotein expression. Results: Endpoint analyses showed no significant tumor growth reduction with single-agent binimetinib or alpelisib in Models A and B. Binimetinib alone resulted in a slightly significant effect in Model C (p<0.01). However, a highly significant effect of both single agents was observed in Model D (p<0.001). Combination therapy with binimetinib and alpelisib significantly reduced tumor growth in all models but was most effective in models C and D (p<0.001, p<0.001, p<0.0001, and p<0.0001 in Models A-D, respectively). The mass spectrometry results will be presented at the meeting. Conclusions: Mutations in PIK3R1 sensitize TNBCs to combined MEK and PI3K pathway inhibition. Co-mutations in PIK3R1 wt PDX models may also sensitize TNBC to MEK and PI3K inhibitors. These findings suggest that combining inhibitors of MEK and PIK3CA could inhibit growth of TNBCs. Disclaimer: AA is employed by AstraZeneca but contributed to this publication in his own capacity. The views expressed are his own and do not necessarily represent the views of AstraZeneca.

M. R. Lloyd¹, E. L. Podany², Z. Bagheri³, E. Rizk¹, A. Dedeoglu⁴, A. Putur⁴, C. Ping², J. Hesse², A. Golden⁵, S. Addison², J. J. Tao⁶, N. Vidula⁴, R. O. Abelman⁴, K. K. Clifton², I. Sanidas⁴, A. Varkaris⁴, D. Juric⁴, N. Vasan⁷, C. X. Ma², G. M. Wulf¹, L. W. Ellisen⁴, S. A. Wander⁴; ¹Department of Medicine, Division of Hematology / Oncology, Beth Israel Deaconess Medical Center, Boston, MA, ²Department of Medicine, Division of Hematology / Oncology, Washington University in St. Louis, St. Louis, MO, ³Department of Graduate Education, Harvard Medical School, Boston, MA, ⁴Department of Medicine, Division of Hematology / Oncology, Massachusetts General Hospital Cancer Center, Boston, MA, ⁵Department of Medicine, Division of General Medicine and Geriatrics, Washington University in St. Louis, St. Louis, MO, ⁶Department of Medicine, Division of Hematology / Oncology, Columbia University Irving Medical Center, New York, NY, ⁷Perlmutter Cancer Center, NYU Langone Health, New York, NY.

Introduction: HR+/HER2- MBC remains incurable, with therapeutic resistance limiting treatment durability. Capivasertib, the first FDA-approved AKT inhibitor for MBC with PIK3CA/AKT1/PTEN alterations, is now widely utilized. However, predictors of response remain poorly understood. This study examined baseline genomic alterations and prior therapy exposures for associations with progression-free survival (PFS) in patients treated with capivasertib. Methods: This retrospective cohort included patients from three U.S. centers treated between November 2023 (capivasertib approval) and June 2025 (data cutoff). Follow up continued until progression, toxicity-related discontinuation, death, or data cutoff. Eligible patients had HR+/HER2- MBC treated with capivasertib plus fulvestrant and baseline next-generation sequencing, as defined by testing within 6 months of therapy start. Two cohorts were analyzed: 1) patients with baseline circulating tumor DNA (ctDNA) sequencing only, and 2) patients with baseline ctDNA or tissue sequencing. Patients were identified via pharmacy claims; clinical and sequencing data were extracted and de-identified. The primary variable examined was the presence of baseline oncogenic genomic alterations; variants of uncertain significance and genes altered in <3 patients were excluded. Other exposures included prior lines of metastatic therapy and prior PI3K inhibitor (PI3Ki) use. The primary outcome was PFS, analyzed using the Kaplan-Meier method and Cox regression. Results: In cohort 1, we identified 45 patients treated with capivasertib and fulvestrant with baseline ctDNA-only sequencing. For this group, median age was 67, 96% were postmenopausal, and two-thirds had visceral and/or brain metastases. Median number of prior lines of therapy was 2; 96% had received a CDK4/6 inhibitor, 60% fulvestrant, 51% chemotherapy, and 22% prior PI3Ki. Median PFS in cohort 1 was 253 days (95% CI 147-392); 21 patients remained on therapy at data cutoff (approximately half had >6 months follow up). Baseline PIK3CA mutations (n=30) trended toward longer PFS (305 vs 194 days; HR 0.58, 95% CI 0.24-1.47); PTEN alterations (n=6) showed no PFS difference (HR 1.09, 95% CI 0.12-4.53), and AKT1 mutations (n=2) were too few to analyze. Pathogenic CCND1 alterations (n=4; 3 amplifications, 1 truncating mutation) were associated with shorter PFS (101 vs 305 days; HR 4.91, 95% CI 1.22-15.40), as were truncating ARID1A mutations (n=4; HR 6.35, 95% CI 1.19-23.45). ESR1 mutations (n=12), all co-occurring with a PI3K pathway alteration (PIK3CA, PTEN, or AKT1), trended toward worse PFS (147 vs 356 days; HR 1.86, 95% CI 0.66-4.82). PFS was numerically longer in patients with ≤1 prior line of therapy vs 2-3 prior lines (305 vs 202 days; HR 1.17, 95% CI 0.36-3.53) or ≥4 prior lines (305 vs 151 days; HR 1.44, 95% CI 0.51-4.06). Prior PI3Ki use trended toward shorter PFS (147 vs 305 days; HR 1.67, 95% CI 0.64-4.07). In cohort 2, 53 patients had baseline ctDNA or tissue sequencing; demographic and clinical features were consistent with cohort 1. FGFR1 alterations, too infrequent to analyze in the ctDNA-only group, trended toward shorter PFS (n=3; 88 vs 253 days; HR 2.33, 95% CI 0.58-6.92). Other clinical-genomic associations remained similar. Conclusions: In this study of patients with HR+/HER2- MBC treated with capivasertib and fulvestrant, alterations in CCND1 and ARID1A in ctDNA were associated with poorer PFS, while ESR1 and PI3K pathway co-mutations trended toward worse outcomes. Patients treated in the second-line or earlier, and those without prior PI3Ki exposure, may derive greater benefit. These findings are exploratory and warrant validation but suggest specific biomarkers that may correlate with risk for early progression on an AKT inhibitor.

T. Hatano¹, T. Tanei¹, S. Seno², Y. Sota¹, N. Masunaga¹, C. Mishima¹, M. Tsukabe¹, T. Yoshinami¹, K. Shimazu¹; ¹Breast and Endocrine Surgery, Osaka University, Suita-city, JAPAN, ²Graduate School of Information Science and Technology, Osaka University, Suita-city, JAPAN.

Background: Breast cancer tumors often have intratumoral heterogeneity (ITH), which has been associated with therapeutic resistance. Tumors with high ITH show human epidermal growth factor receptor 2 (HER2) heterogeneity, impacting the effectiveness of HER2-targeted therapies. Our recent study identified HER2 ITH as an independent prognostic factor for poor outcomes in HER2-positive breast cancer. In this study, we investigated the association between HER2 ITH and resistance to Anti-HER2 Neoadjuvant Chemotherapy (NAC). Methods: The study included 97 patients of primary HER2-positive breast cancer treated with Anti-HER2 NAC. Breast tumor samples were obtained from vacuumassisted breast biopsy before NAC. HER2 gene amplification was assessed by fluorescence in situ hybridization (FISH), and histograms of HER2 gene copy numbers were generated. Using Gaussian mixture model, the histogram data were analyzed and classified into two groups, high HER2 heterogeneity (HH) and low HER2 heterogeneity (LH). The relationship between HER2 ITH and treatment response was evaluated with the pathological complete response (pCR) rate. Results: Of 97 patients,18 (18.6%) were classified into the HH group, and 79 (81.4%) into the LH group. The pCR rate in the HH group was significantly lower at 28% (5/18) compared to 65% (51/79) in the LH group (P = 0.01). Multivariate analysis of pathological parameters revealed that HER2 ITH was the significant predictor of pCR rate (P = 0.02), following estrogen receptor status. Conclusions: The assessment of HER2 ITH may be valuable in predicting therapeutic outcomes in HER2-positive breast cancer. Our novel approach of HER2 ITH method using FISH histograms could potentially serve as a useful tool in predicting resistance to anti-HER2 NAC.

l. quan¹, R. Huiteng², y. jian¹, y. peng¹; ¹Department of Medical Oncology, cancer hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, CHINA, ²Department of Data Science, School of Data Science, The Chinese University of Hong Kong, Shenzhen, shenzhen, CHINA.

IntroductionWhile breast cancer (BC) remains the most prevalent malignancy among women worldwide, the dynamics of the intratumoral microbiome during tumor initiation, progression, and treatment have been largely overlooked. Prior studies are predominantly cross-sectional and limited by indirect microbial inference from RNA-seq data.Method This study presents the first comprehensive longitudinal analysis of intratumoral microbiota across breast tissue samples using high-depth 16S rRNA sequencing (11W tags). Samples included: 165 benign nodules (82 non-transforming, 83 that later progressed to cancer with matched malignant tissues); 180 primary BC tissues and 165 benign controls; and 101 neoadjuvant therapy (NAT) specimens (15 pCR, 86 non-pCR, with paired pre/post-treatment samples). Microbial diversity, differential taxa, functional predictions, network structures, and correlations with clinical characteristics were systematically analyzed. In addition, two predictive models were developed to malignancy transformation and NAT response.ResultsCompared to NNT group, NDT group exhibited lower microbial diversity, with a relative enrichment of genera such as Bacteroides, Weissella, Paenibacillus, and Methyloversatilis, whereas the NNT group showed higher abundances of Halomonas, Dietzia, Nesterenkonia, and Aeromicrobium (P<0.05). Functional analysis revealed that amino acid and vitamin metabolism-related pathways were significantly more active in the NNT group, while lipid metabolism pathways were enriched in the NDT group (P<0.05). During the malignant transformation of nodules, the abundances of Aeromicrobium, Dietzia, Halomonas, and Nesterenkonia markedly increased, whereas Paenibacillus and Faecalibacterium significantly decreased (P<0.05). Second, in comparisons between BC tissues and benign lesions, genera such as Methyloversatilis, Weissella, Caulobacter, and Bosea were enriched in tumor tissues, while Delftia, Dietzia, Halomonas, Aeromicrobium, Nesterenkonia, and Acinetobacter were notably reduced (P<0.05). Furthermore, with advancing tumor stages, the relative abundances of Dietzia, Aeromicrobium, Delftia, Nesterenkonia, Halomonas, and Nitriliruptor gradually declined, suggesting that these taxa may be associated with the early-stage tumor microenvironment (P<0.05). Finally, following NAT, microbial diversity significantly increased, and the microbial community composition underwent a notable reconstruction. After treatment, alpha diversity rose, with genera such as Aeromicrobium, Halomonas, Dietzia, Nesterenkonia, and Nitriliruptor showing significant increases in abundance (P<0.05). Functional prediction analysis indicated enhanced activities in pathways related to amino acid metabolism, lipid metabolism, and terpene metabolism, while pathways involved in glycan metabolism were diminished (P<0.05). Importantly, we developed two predictive models with high clinical relevance: one stratifying malignancy risk in nodules (AUC=1.0), and another predicting NAT response (AUC=0.82). ConclusionOur findings highlight the clinical relevance of microbial signatures in early diagnosis, malignancy prediction, and response to therapy, offering novel insights for precision oncology and microbiome-based interventions in breast cancer.

E. Ciruelos¹, T. Pascual², B. Adamo³, M. Alva Bianchi⁴, I. Blancas⁵, L. Carnerero⁶, R. Villanueva-Vázquez⁷, B. Fullana⁷, M. Melé⁸, J. Salvador Bofill⁹, J. M. Cejalvo¹⁰, A. Ferrando-Díez¹¹, Y. Izarzugaza¹², M. Borrell¹³, X. González Farré¹⁴, G. Villacampa¹⁵, M. Paes Dias¹⁶, F. M. Juan¹⁶, M. Oliveira¹³; ¹Medical Oncology, University Hospital 12 De Octubre / Instituto de Investigacion Sanitaria Hospital 12 de Octubre (imas12) / SOLTI Cancer Research Group, Madrid, SPAIN, ²Medical Oncology, Institute of Cancer and Blood Diseases, Hospital Clinic of Barcelona / Translational Genomics and Targeted Therapies in Solid Tumors group, August Pi i Sunyer Biomedical Research Institute (IDIBAPS) / SOLTI Cancer Research Group, Barcelona, SPAIN, ³Medical Oncology, Institute of Cancer and Blood Diseases, Hospital Clinic of Barcelona / Translational Genomics and Targeted Therapies in Solid Tumors group, August Pi i Sunyer Biomedical Research Institute (IDIBAPS), Barcelona, SPAIN, ⁴Medical Oncology, University Hospital 12 De Octubre, Madrid, SPAIN, ⁵Medical Oncology, Hospital Universitario Clínico San Cecilio / Medicine Department, Granada University / Instituto de Investigación Biosanitaria de Granada (ibs Granada), Granada, SPAIN, ⁶Medical Oncology, Hospital Universitario Clínico San Cecilio / Instituto de Investigación Biosanitaria de Granada (ibs Granada), Granada, SPAIN, ⁷Medical Oncology, Institut Català d’Oncologia (ICO Hospitalet), L’Hospitalet de Llobregat, SPAIN, ⁸Medical Oncology, Hospital Universitari Sant Joan de Reus, Reus, SPAIN, ⁹Medical Oncology, Hospital Universitario Virgen del Rocio, Instituto de BioMedicina de Sevilla (IBIS), Sevilla, SPAIN, ¹⁰Medical Oncology, Hospital Clínico Universitario de Valencia (INCLIVA) / SOLTI Cancer Research Group, Valencia, SPAIN, ¹¹Medical Oncology, Institut Català d’Oncologia Badalona (ICO Badalona), Badalona, SPAIN, ¹²Medical Oncology, Hospital Universitario Fundación Jiménez Dí, Madrid, SPAIN, ¹³Medical Oncology, Vall d’Hebron University, Hospital and Vall d’Hebron Institute of Oncology (VHIO) / SOLTI Cancer Research Group, Barcelona, SPAIN, ¹⁴Medical Oncology, IOR – Instituto Oncologico Dr. Rosell, Hospital General de Catalunya / SOLTI Cancer Research Group, Barcelona, SPAIN, ¹⁵Clinical Research, SOLTI Cancer Research Group / Statistics Unit, Vall d’Hebron Institute of Oncology (VHIO), Barcelona, SPAIN, ¹⁶Clinical Research, SOLTI Cancer Research Group, Barcelona, SPAIN.

Background: Sacituzumab Govitecan (SG) is an antibody-drug conjugate (ADC) containing a TROP2-directed antibody bound to a topoisomerase I inhibitor payload. SG is approved for the treatment of patients (pts) with metastatic triple-negative breast cancer who have received at least two prior systemic therapies, including one for advanced disease, and pts with hormone receptor-positive (HR+)/HER2-negative metastatic breast cancer (mBC) after endocrine therapy and ≥2 systemic treatments in the advanced setting. SG efficacy does not seem to correlate to TROP2 protein expression, and no additional efficacy biomarkers have been identified. Methods: SOLTI ACROSS-TROP2 (NCT06236269) is a phase II, open-label, single-arm trial investigating biomarkers of efficacy and mechanisms of resistance to SG in 100 pts with HR+/HER2- mBC who progressed during or after treatment with CDK4/6 inhibitors and received up to one prior chemotherapy or ADC regimen for metastatic disease. Pts received SG at 10 mg/kg via IV infusion on Days 1 and 8 of each 21-day cycle. Mandatory tumor biopsies were obtained at baseline and after 2-3 weeks of treatment (C2D1). The primary objective is to evaluate the change in CelTIL score (a composite score of tumor cellularity and tumor infiltrating lymphocytes) between baseline and C2D1 tumor samples (Nuciforo et al, Ann Oncol 2019). Key secondary endpoints include progression-free survival (PFS), overall response rate (ORR), and safety. Here we present the planned interim analysis results from the first 50 enrolled pts. Results: Between April and December 2024, 50 pts were enrolled. Median follow-up was 8.6 months (m). Median age was 58 (range 30-79), all pts were female, most were postmenopausal (78%), had no visceral metastases (64%), and received SG as second line for mBC after chemotherapy or ADC (70%). At data cut-off (May-25), 18/50 (36%) remained on treatment and 43/50 (86%) were evaluable for the primary endpoint. The mean increase in CelTIL score from baseline to C2D1 was 4.33 points (95% CI 0.99-7.65), indicating a significant overall change, with a greater and significant increase in patients that achieved radiological response (6.33 vs 3.13 points). PFS events occurred in 28/50 pts (maturity 56%). Median PFS was 6m (4.2-NR) both overall and among pts treated in second line for mBC. ORR was 37% and 38%, respectively. Grade ≥3 treatment-related adverse events (TRAEs) occurred in 48.0% of pts. Most common TRAEs were fatigue (58%), anemia (58%), neutropenia (54%), nausea (46%) and diarrhea (38%). Treatment discontinuation rate due to TRAEs was 8%. Conclusions: In this interim analysis of SOLTI ACROSS-TROP2, one cycle of SG led to an increase in CelTIL score, particularly among patients who achieved a radiological response. Consistent efficacy was observed in terms of PFS and ORR. The final results from the full cohort of 100 patients, along with a comprehensive translational analysis of paired pre- and on-treatment biopsies, are expected to provide further insights into biomarkers of response and mechanisms of resistance to SG in HR+/HER2- mBC.

V. D. Bertoni¹, B. Cavalcante², G. Rocha¹, C. Gurgel²; ¹Human Pathology Pos Doc Program, Gonçalo Moniz Institute, Oswaldo Cruz Foundation (IGM-FIOCRUZ/BA, Brazil), Salvador, BRAZIL, ²LPBM, Gonçalo Moniz Institute, Oswaldo Cruz Foundation (IGM-FIOCRUZ/BA, Brazil), Salvador, BRAZIL.

Identifying Prognostic Epithelial-Mesenchymal Transition and Tumor Stem Cell Markers in Breast Cancer SubtypesDybal, Vanessa^1,2,3 & Cavalcante, Bruno^1,3 & Rocha, Gisele^1,3,4 & Gurgel, Clarissa^{1, 3,4}1.Gonçalo Moniz Institute, Oswaldo Cruz Foundation (IGM-FIOCRUZ/BA, Brazil) ;2. Oncology Multidisciplinary Assistence (AMO, DASA, BA, Brazil)3.Department of Pathology and Forensic Medicine, School of Medicine, Federal University of Bahia, Brazil4.D ́Or Institute for Research and Education (IDOR), BA,Brazil Introduction: Breast cancer (BC) remains the most prevalent malignancy among women, excluding skin tumors. Despite significant therapeutic advancements, a high relapse rate, affecting approximately 20% of all cases, persists. To improve clinical outcomes and develop novel therapeutic strategies, a deeper understanding of mechanisms driving tumor resistance, such as epithelial-mesenchymal transition (EMT) and tumor stem cell (TSC) biology, is crucial.Objective: This study aimed to identify immunohistochemically targetable markers related to EMT and TSC in different breast cancer subtypes with prognostic significance.Methods: We analyzed a public The Cancer Genomic Atlas (TCGA) breast cancer patient database encompassing clinical, histopathological, transcriptomic, and prognostic data. Bioinformatics tools were employed to identify differentially expressed genes (DEGs) impacting survival, analyze associated biological processes, construct co-expression networks, and investigate interactions with EMT and TSC-related pathways. Patients were categorized by their clinical immunohistochemical subtypes: triple-negative (TNBC), HER2-positive, and hormone receptor-positive (HR+) tumors. Protein-protein interaction (PPI) network analysis was integrated to identify the most relevant proteins within each subtype that correlated with survival.Results: Analysis of 989 BC patients from the Firehose Legacy TCGA database revealed DEGs and co-expression gene communities correlated with breast cancer clinical subtypes. Within these communities, DEGs related to EMT and TSC were identified, and those with a significant impact on survival were selected. PPI analysis pinpointed proteins within each community exhibiting a high degree of interaction and clinical endpoint impact. Following integration with genomic findings, the following potential protein markers were identified for further investigation in paraffin-embedded patient material: CCNB1 and CDCA8 (TNBC); COMP, P4HA3, and ECM2 (HER2-positive tumors); PPARG, ADIPOQ, LIPE, LEP, FABP4, DGAT2; CAV1,KLF4, FOS, IGF1, CLDN5, FOB, ALDH1A1, EGR1,PGM5, AKAP12, CAV2 (HR+ tumors). Conclusion: We identified subtype-specific prognostic protein markers associated with EMT and TSC in breast cancer. These findings are currently undergoing validation through immunohistochemical analysis of paraffin-embedded material from Brazilian patients, Brazilian patients RNA analysis, and in commercial breast cancer tissue microarray.

Y. Gao¹, M. Zaidi², A. Naqvi², F. Azim², S. Satta³, C. Brewer³, C. Palsuledesai³, E. Kalashnikova³, K. Sun⁴, P. Niravath⁴, H. Mai⁴, J. McKenzie³, A. Rodriguez³, M. Liu³, A. Puri⁴; ¹Department of Hematology/Oncology, Houston Methodist Neal Cancer Center, Houston, TX, ²Department of Internal Medicine, Houston Methodist Neal Cancer Center, Houston, TX, ³Oncology, Natera, Inc., Austin, TX, ⁴Breast Medical Oncology, Houston Methodist Neal Cancer Center, Houston, TX.

Background: Recent studies have demonstrated the utility of longitudinal circulating tumor DNA (ctDNA) testing for monitoring treatment response and predicting long-term outcomes in patients (pts) with breast cancer (BC). In this study, we evaluated ctDNA detection rates post-surgery in pts with HER2+ BC in the adjuvant treatment and surveillance settings. Methods: This retrospective study included 59 patients with stage I-IV HER2+ BC. Commercial CTDNA testing was performed postoperatively at the physician’s discretion. A clinically validated, tumor-informed ctDNA assay (Signatera^TM, Natera, Inc.) was used to detect and quantify ctDNA in blood samples collected after primary resection and serially during surveillance. ctDNA status and dynamics were correlated with clinicopathological features and surgical outcomes. Results: The cohort included 59 patients with HER2+ BC; the majority of pts had clinical T1-3 stage disease at diagnosis (T1, 25%, 15/59; T2, 20%, 12/59; and T3, 24%, 14/59), the remaining pts were T4 (7%; 4/59) or had unknown stage (24%; 14/59). Lymph node involvement was reported in 46% of pts (27/59; stage N1, 22/59; N2, 3/59; N3, 2/59); 47% had grade 3 (28/59) and 46% had grade 2 (27/59) disease. Sixty-four percent of cases (38/59) were estrogen receptor (ER)-positive; 36% (21/59) were ER-negative. Neoadjuvant treatment was received by 71% (42/59) of pts, this information was unavailable for 4% (2/59); the rest of the pts (25%; 15/59) did not receive neoadjuvant treatment. Among pts receiving neoadjuvant therapy (n=42), residual cancer burden (RCB) class was available for 27 (RCB0: 22%, 6/27; RCB1: 19%, 5/27; RCB2: 40%, 11/27; RCB3: 19%, 5/27).Among pts with longitudinal ctDNA results and complete clinical information available (n=42), ctDNA-positivity any time after surgery was observed in 14% (6/42) with a median time to ctDNA-positivity of 34 months (range: 14-792) from surgery. Of these six ctDNA-positive patients, one experienced recurrence 13 months after the first positive ctDNA test, one experienced ctDNA clearance after adjuvant therapy and remained event-free for 4.6 mos, and four remained recurrence-free with a median follow-up since ctDNA-positivity of 6.7 months and ongoing monitoring. Among patients with a favorable response to neoadjuvant treatment (RCB 0/1; n=11), no ctDNA-positivity was observed after surgery. Among pts with residual disease at surgery (RCB 2/3; n=10), ctDNA-positivity was observed in 20% (2/10). Among pts with persistent ctDNA-negativity during surveillance (n=30, median follow-up of 15.4 months, range: 5.9-50.0 months), no clinical recurrences were reported. Conclusions: In addition to clinicopathological features and surgical outcomes, longitudinal ctDNA monitoring helps identify pts with HER2+ BC who may be at high risk of recurrence. Further research is needed to assess the true utility of a ctDNA-guided treatment approach in this patient population.

X. Zeng, Z. Sun, Q. Hu, M. Tang, C. Liu, Z. Liu, T. Yu; Department of Breast Surgery, Jiangxi Cancer Hospital & Institute, Nanchang, CHINA.

The INAVO 120 clinical trial confirmed the efficacy of first-line palbociclib combined with inavolisib in hormone receptor-positive (HR+) advanced breast cancer patients harboring PIK3CA mutations who developed resistance to adjuvant endocrine therapy. However, the comparative effectiveness of alternative CDK4/6 inhibitors (e.g., abemaciclib, ribociclib) in the first line setting, or the potential of crossline therapy (using abemaciclib or ribociclib combined with inavolisib) following palbociclib resistance in PIK3CA-mutant patients remains unclear. Our prior research identified the G protein-coupled estrogen receptor (GPER) as a pivotal initiator of endocrine resistance in HR+ advanced breast cancer, further driving activation of the PI3K/AKT/mTOR (PAM) signaling pathway. We infer that GPER signaling modulates endocrine therapy sensitivity in PIK3CA-mutant HR+ advanced breast cancer and could serve as a novel biomarker to guide personalized clinical treatment decisions. To demonstrate this hypothesis, we identified MCF-7 (E545K-mutant) and T47D (H1047R-mutant) breast cancer cell lines harboring PIK3CA mutations, and developed tamoxifen-resistant (MCF-7/TAM-R, T47D/TAM-R) and palbociclib-resistant (MCF-7/Palb-R, T47D/Palb-R) models to recapitulate first- and second-line endocrine resistance in clinic. Interestingly, GPER expression was significantly upregulated in all above resistant cell lines and promoting ABCG2 efflux pump expression via the GPER/PI3K/AKT pathway, as validated by quantitative PCR and Western blot. In GPER^high tamoxifen-resistant cells, combinations of inavolisib with abemaciclib or ribociclib exhibited superior antiproliferative effects compared to palbociclib-based regimens, which assessed by MTT assays, flow cytometry for apoptosis, and cell cycle analysis. Knockdown of GPER/ABCG2 axis using siRNA normalized the efficacy of palbociclib, abemaciclib, and ribociclib in combination with inavolisib, underscoring GPER’s critical role in diverse CDK4/6 inhibitors sensitivity. Notably, in palbociclib-resistant cells, combinations of inavolisib with abemaciclib or ribociclib significantly outperformed inavolisib monotherapy. Furthermore, GPER/ABCG2 knockdown not only partially reversed palbociclib resistance, but also enhanced sensitivity to abemaciclib- and ribociclib-based regimens, demonstrating reduced tumor growth and improved therapeutic response upon GPER signaling inhibition. Therefore, assessing GPER signaling activation in tumor biopsies may enable precise identification of PIK3CA-mutant HR+ advanced breast cancer patients likely to benefit from specific endocrine and targeted therapy combinations. Integrating PI3K, CDK4/6, and GPER-targeted therapies represents a promising strategy to overcome endocrine resistance in the future, offering a path toward personalized treatment paradigms with improved clinical outcomes.

X. Li, Y. Ni, S. Zhang, C. Yang, G. Li, Y. Zhang, X. Chen, X. Yang, J. Su, B. Li; Innovation Center, Burning Rock Biotech, Shanghai, CHINA.

Breast Cancer is a major global health challenge and remains one of the most common cancers among women. Although traditional tissue biopsies provide valuable diagnostic insights, their invasive nature makes them unsuitable for long-term monitoring. Liquid biopsy, which analyzes circulating tumor DNA (ctDNA) in blood samples, offers a minimally invasive alternative with growing potential in breast cancer care. Recent studies have demonstrated that detecting specific mutations, such as ESR1, in ctDNA can guide treatment decisions, significantly reducing disease progression and mortality. Furthermore, ctDNA monitoring can accurately predict long-term outcomes, identifying patients at low risk of relapse. These findings highlight the importance of liquid biopsy in guiding personalized treatment strategies and improving patient outcomes, even in challenging low-tumor-burden scenarios. To establish an ultra-deep targeted sequencing for low-ctDNA detection, we collected blood samples from non-cancer controls and breast cancer patients. An advanced UMI-tagging protocol was applied to process 20 – 60ng of plasma cfDNA and 100ng of WBC genomic DNA, enabling targeted enrichment of 168 cancer-related genes through designed panels. Library sequencing was conducted on the NovaSeq6000 platform, producing a raw depth of 35,000X, and a mean unique molecular depth of 4000X. A multi-step data processing pipeline was implemented to eliminate technical noise and normalize biological heterogeneity. First, reads with identical UMIs and endpoints position were used to collapse and merge to generate high-accuracy consensus sequences, effectively remove random errors introduced during PCR amplification and sequencing. Building on this, a machine-learning-driven noise model was developed to systematically identify and filter recurrent artifacts, including DNA damage patterns and sequencer-specific error signatures. Additionally, paired WBC-sequencing was applied into the workflow to distinguish somatic mutations from non-cancer biological aberrations, such as clonal hematopoiesis-derived variants, ensuring specificity in detecting tumor-associated alterations. To evaluate analytical performance, ground-truth reference materials were utilized, with subsequent confirmation through a dilution series across different tumor allele fraction (TAF) levels. The limit of detection (LoD) was rigorously assessed across different variant classes at defined input masses. For panel-wide SNVs, the median LoD is 0.25% with a 20ng input, and this threshold could be further reduced to 0.15% with a 50ng input. For panel-wide indels, the median LoD improved from 0.25% at a 20ng input to 0.14% at 50ng input. Additionally, the limit of blank (LoB) was defined from 51 cancer-free donors with paired WBC controls, resulting in an exceptionally low error rate of 4.88*10^-7 for both the SNVs and indels across the entire gene panel. This panel covers whole coding regions of key breast cancer-associated genes, such as PTEN, ESR1, PIK3CA, BRCA1, BRCA2 and PALB2. To further validate precision and accuracy, a combined evaluation was performed on 44 clinical samples derived from breast cancer patients with an expanded panel. This analysis showed a positive percent agreement (PPA) of 83.0% for panel-wide SNVs, 85.7% for panel-wide indels, along with a negative percent agreement (NPA) of 99.9%. In this study, we demonstrated that the accuracy and consistency of the advanced ultra-deep mutation sequencing technologies. Combined with machine-learning tools, this approach offers a practical, non-invasive diagnostic tool for mutation profiling, applicable even when tumor levels are low.

P. Chapani¹, N. Pryma¹, N. LeVasseur¹, T. Nghiem¹, M. Marra², J. Laskin¹, Z. Mitri¹; ¹Medical Oncology, BC Cancer Vancouver, Vancouver, BC, CANADA, ²Michael Smith Genome Sciences Centre, BC Cancer Vancouver, Vancouver, BC, CANADA.

Background. Young women breast cancer (YWBC), defined as diagnosed at age <40, represent 5% of all breast cancer cases, yet remains the leading cause of cancer related death in this age group. YWBC is more aggressive and associated with worse outcomes compared to breast cancer diagnosed in women age >40 (non-YWBC). Despite this, YWBC is underrepresented and understudied. In this study, we characterize clinical and pathologic features, as well as outcomes of YWBC enrolled in the Personalized Oncogenomics (POG) program at BC Cancer. Methods. Patients with metastatic breast cancer (MBC) enrolled in POG, a personalized oncology platform aimed at characterizing the molecular landscape of advanced cancers, were included and categorized into cohorts as YWBC or non-YWBC. Data collected included patient and tumor characteristics at initial diagnosis and metastatic recurrence, germline mutations, treatments received, and survival outcomes. Clinical tumor subtypes were classified as hormone receptor positive HER2 non-amplified, HER2-amplified, and triple negative breast cancer (TNBC). As all patients enrolled in the POG program undergo whole-genome and transcriptome analysis (WGTA), we evaluated for potential targetable alterations. Results. Patient Characteristics. Between 2012 and 2025, 249 MBC patients were enrolled, with 60 (24%) YWBC and 189 (76%) non-YWBC. Median age at breast cancer diagnosis was 35 (IQR 33 – 38) for YWBC and 52 (IQR 45 – 59) for non-YWBC. In this study population, 7.6% of patients had germline mutations in BRCA, PALB2, or TP53, with no difference between the cohorts. Clinico-Pathologic Features. Compared to non-YWBC, YWBC were more likely to have stage 2-3 disease (76.7% vs 58.7%, p = 0.07), node positive disease (68.3% vs 55%, p = 0.09) and more likely to receive chemotherapy (75% vs 63.5%, p=0.3). De-novo MBC was observed in 20% of YWBC, and 24.9% of non-YWBC. The distribution of tumor subtype at initial diagnosis was similar between the 2 cohorts (p =0.46). Notably, 13% of patients had a tumor subtype switch from initial diagnosis to MBC, partly driven by an increase in TNBC in YWBC, rising from 23.3% to 35%. Treatment and Survival Outcomes. Patients in both cohorts were enrolled to POG after a median of 2 prior lines of therapy for MBC. Compared to stage 1 at initial diagnosis, stages 2 and 3 were associated with shorter time to MBC development (p=0.06 and p=0.01, respectively). There was no significant difference in median overall survival from initial breast cancer diagnosis (7.97 vs 7.53 years, log-rank test p=0.86) or from MBC diagnosis (3.13 vs 2.85 years, log-rank test p=0.47) between YWBC and non-YWBC, respectively. Compared to other subtypes, TNBC was associated with a 79% (p <0.001) and 97% (p<0.001) increase in rate of death from metastatic and initial diagnosis respectively. Additionally, there was a non-significant trend towards increase in rate of death from metastatic diagnosis in patients who had tumor subtype switch and in those with 3 or more metastatic sites. Molecular Profiling. The most common mutations identified were in the PI3K/AKT/PTEN pathway in 50%, ESR1 mutations in 18%, and somatic BRCA mutations in 16% of metastatic samples. Notably, a composite score from WGTA identified 12% of tumors with potential sensitivity to immune checkpoint inhibitor therapy. Compared to non-YWBC, there was a non-significant trend for YWBC tumors to harbor more somatic BRCA mutations, and less alterations in the PI3K/AKT/PTEN pathway.Conclusion. In our cohort, YWBC were characterized by a higher stage at initial breast cancer diagnosis and enriched for TNBC tumors at time of metastatic recurrence. Efforts to characterize matched primary tumors aim to elucidate molecular factors driving metastatic progression in YWBC, and to identify potential therapeutic targets for this high-risk group are ongoing.

L. Jackson, G. A. Pestano; R&D, Biodesix Inc., Louisville, CO.

Background: Estrogen receptor 1 (ESR1) mutations are a critical mechanism of acquired resistance to aromatase inhibitor (AI) therapy in hormone receptor-positive (HR+) breast cancer, occurring in approximately 20-40% of patients during treatment. The detection of ESR1 mutations in circulating tumor DNA (ctDNA) through liquid biopsy represents a new approach to real-time monitoring of treatment resistance and therapeutic decision-making. Recent interim results from the landmark SERENA-6 Phase III trial demonstrated the potential clinical utility of ctDNA-guided treatment switching, where median progression-free survival was 16.0 months in the camizestrant group versus 9.2 months in the aromatase-inhibitor group when treatment was switched based on ESR1 mutation detection. This study represented the first global Phase 3 trial to demonstrate the potential clinical utility of using ctDNA for treatment guidance and breast cancer management.Methods: We have developed a droplet digital PCR (ddPCR) test for the detection of ESR1 variants in plasma-derived cfDNA. The ESR1 ddPLEX kitted assay, in combination with a supplementary 4-plex Expert Design assay, targets the 11 most prevalent ESR1 variants (97% coverage of ESR1 mutations in breast cancer). Analytical sensitivity studies were performed using synthetic DNA controls diluted into wild-type background DNA at varying input concentrations. Serial dilutions were tested to establish the limit of detection (LOD) across different DNA input amounts, with variant allele frequencies (VAF) ranging from 0.025% to 0.2%. Results: Our studies demonstrated robust analytic performance of the ESR1 ddPCR assay across multiple input concentrations. With 30 ng cfDNA input, we achieved analytical sensitivity ranging from 0.025-0.5% VAF, enabling detection of low-level ESR1 mutations present in early resistance development. When DNA input was reduced to 10 ng, reflecting real-world clinical sample limitations, the assay maintained excellent sensitivity of 0.025-0.1% VAF. The assay also demonstrated high specificity with no false positive calls in wild-type controls.Conclusions: We have successfully developed a highly sensitive and specific ddPCR test for the detection of multiple ESR1 variants in plasma. The assay performed as designed and meets the analytical requirements required for further clinical development. The achieved sensitivity of 0.025-0.1% with as little as 10 ng of cfDNA positions this ddPCR assay as a potentially valuable tool for liquid biopsy monitoring applications in breast cancer. Clinical validation studies may further support the initial utility of ctDNA guided treatment selection demonstrated in SERENA-6. The availability of this test could enable detection of emerging resistance mutations from AI therapy in HR+ breast cancer in advance of traditional monitoring methods, e.g. imaging.

S. Rosenberg, T. Escobar Gil, J. Nemunaitis; Internal Medicine, University of New Mexico, Albuquerque, NM.

Introduction: Aromatase inhibitors (AIs) are standard therapy for hormone receptor-positive breast cancer in postmenopausal women but are frequently associated with musculoskeletal symptoms (AIMSS) that impair quality of life and adherence. We summarize current evidence on biologic and imaging markers associated with AIMSS, to both clarify its underlying mechanisms and identify potential predictors, with the goal of guiding future strategies to improve AI adherence. Methods: We conducted a scoping review of original studies published between 2010 and 2025 that examined biologic and imaging markers for aromatase inhibitor related musculoskeletal symptoms in breast cancer patients, according to the PRISMA guidelines. Databases include PubMed, EMBASE, COCHRANE, JAMA, and Medline. Eligible studies reported on at least one of the following: breast cancer, aromatase inhibitors, arthralgia, biomarker, biological marker, inflammatory marker, imaging, radiology. Data were extracted in duplicate and synthesized narratively. Risk of bias was assessed using the ROBINS-I tool. Results: Thirteen studies met inclusion criteria. Participants were mainly postmenopausal women aged 50-67 with early-stage hormone receptor-positive breast cancer. Most studies evaluated markers after initiation of AI therapy, though a few included baseline assessments for comparison. Biomarkers included inflammatory markers (CRP, IL6, MCP-1), metabolic and lipidomic profiles (oxylipins, 8-HETE), hormonal markers (estradiol, sex binding protein), and genetic polymorphisms (CYP19A1, ESR1, HSD17B2). One study identified BMI as a clinical risk factor, and suggested IGF-1 as a potential biomarker, though it was not directly measured. Elevated inflammatory cytokines and lower estradiol levels were associated with AIMSS in several studies. Genetic findings were inconsistent, though newer studies identified novel variants like HSD17B2. Imaging assessments varied, with some studies performing baseline imaging and others focusing on early and serial imaging after AI initiation. Imaging methods included ultrasound, MRI, and shear wave elastography (SWE), with SWE showing promise as a quantifiable tool to detect early tendon stiffness. Discussion: Our review summarizes several markers that could be associated with in AIMSS, including inflammatory and hormonal pathways, genetic variants, and early imaging changes. However, these studies were limited by small sample sizes, heterogeneous methods, and have issues with reproducibility. While some markers show promise for predicting symptoms, the current evidence mainly advances understanding of AIMSS mechanisms. Future studies should integrate these markers to develop predictive tools, with the goal of improving AIMSS and AI adherence.

M. Pizon, D. Schott, K. Pachmann; Research, Laboratory Pachmann, Bayreuth, GERMANY.

Background:Breast cancer remains a leading cause of cancer-related morbidity and mortality worldwide. Emerging evidence suggests that a subpopulation of circulating tumor cells—circulating cancer stem cells (cCSCs)—plays a pivotal role in metastasis, therapy resistance, and disease relapse. However, functional assessment of these cells in a minimally invasive manner poses a significant clinical challenge. Tumorsphere assays offer a unique in vitro platform to enrich and evaluate the self-renewal capacity of cCSCs, potentially reflecting the biological aggressiveness of the tumor. Establishing a reliable method for cultivating tumorspheres from peripheral blood may provide a novel biomarker for monitoring tumor progression and facilitating patient stratification.Materials and Methods:A total of 98 breast cancer patients, representing stages I to IV, were enrolled in this study. Peripheral blood samples were collected from each patient to detect cCSCs using the stemtrac® method, followed by a standardized tumorsphere-forming assay.Results:Tumorsphere count showed a significant correlation with disease stage and metastatic status. Patients with stage IV disease exhibited a higher median tumorsphere count compared to those with stage I (30 vs. 10; p < 0.002). Metastatic patients had more tumorspheres than non-metastatic patients (30 vs. 15; p < 0.005), with the highest counts observed in cases with multiple metastases (median 61; p < 0.001). Additionally, higher tumorsphere counts were associated with features of tumor aggressiveness, including HER2/neu positivity (35 vs. 10; p < 0.05), increased Ki-67 index, and higher histological gradeConclusion:These findings underscore the clinical relevance of tumorsphere assays as a minimally invasive method for monitoring tumor biology. The number of tumorspheres derived from peripheral blood may serve as a surrogate biomarker for disease progression, metastatic potential, and tumor aggressiveness, offering promising implications for real-time patient stratification and personalized treatment planning in breast cancer.

N. Lopetegui-Lia¹, A. Davenport¹, J. Gill², A. M. Roy¹, S. Myers³, D. Agnese³, E. Burke³, T. Y. Andraos⁴, D. M. Schimming¹, S. Hulett¹, H. LeFebvre¹, M. Gatti-Mays¹, D. Stover¹; ¹Medical Oncology, The Ohio State University, Columbus, OH, ²Internal Medicine, Tennova North Knoxville Medical Center, Powell, TN, ³Surgical Oncology, The Ohio State University, Columbus, OH, ⁴Radiation Oncology, The Ohio State University, Columbus, OH.

BackgroundThe rising incidence of Breast cancer (BC) in young women requires further genomic characterization, as only a fraction carry a germline mutation. Given the growing interest in therapies targeting the phosphoinositide 3-kinase/AKT (PI3K/AKT), mammalian target of rapamycin (mTOR), and phosphatase and tensin homolog (PTEN) pathways, central to BC pathogenesis, we evaluated whether PI3K/AKT/mTOR pathway alterations and gene expression differed by age at BC diagnosis.Methods: In this analysis of patients’ tumor/germline whole exome DNA and RNA sequencing data from the Ohio State University Total Cancer Care registry, 71 profiled patients were young adults (YA) aged < 40 years, and 675 were non-YA aged ≥40 years. Descriptive statistics characterized the frequency and co-occurrence of BC-specific alterations and mRNA expression profiles by age groups. ResultsThe prevalence of these BC-relevant mutations are reported in Table1. PIK3CA mutation occurred more frequently in non-YA patients (16.9% YA versus (vs) 39% non-YA; p<0.0002). The frequency of driver mutations were similar between cohorts (73% in YA vs 82% in non-YA). Most hotspot mutations were enriched in non-YA group, particularly H1047R (15.11% n=102 non-YA vs 4.23%, n=3 YA; p= 0.01). Others include: E545K (6.52%, n=44 vs 2.82%, n=2, p=0.302), E542K (4.59%, n=31 vs 1.41%, n=1; p=0.352); H1047L (1.93%, n=13 vs none, p=0.625) and N345K (1.78%, n=12 vs 1.41% n=1; p=1.00). There were no age-based differences in AKT1 mutation (5.6% YA vs 10.8% non-YA patients; p=0.220). Hotspot alterations E17K, E17G, and L52R were numerically enriched, albeit nonsignificant, in non-YA. There were no significant differences in the presence of mTOR by age (16.9% YA and 21.5% non-YA; p=0.450). Most alterations in both groups were variants of unknown significance (17/17 in YA, 157/159 in non-YA). YA and non-YA cohorts had similar frequency of PTEN mutations (12.7% YA vs 10.5% non-YA; p=0.550). The proportion of PTEN driver mutations also did not differ (82% YA and 93% non-YA). Conclusion PIK3CA mutations were more prevalent among non-YAs highlighting potential age-associated oncogenic drivers. Although AKT1, mTOR, and PTEN gene mutations did not differ between age groups, in general, oncogenic/hotspot mutations occurred more frequently in patients ≥40 years. As management of BC evolves to incorporate PI3K/AKT/mTOR inhibitors, age-specific somatic alterations may uncover therapeutic vulnerabilities and inform customized treatment. Table:

J. Wu¹, M. Liu¹, W. Li¹, N. Zhang¹, Y. Liu¹, Y. Zhao², X. Zeng¹; ¹Department of Breast Center, Chongqing Key Laboratory for Intelligent Oncology in Breast Cancer, Chongqing University Cancer Hospital, Chongqing, CHINA, ²Chongqing Key Laboratory of Translational Research for Cancer Metastasis and Individualized Treatmen, Chongqing University Cancer Hospital, Chongqing, CHINA.

Background Existing studies have primarily focused on the clinical efficacy of Apatinib, a selective VEGFR2 inhibitor, leaving a significant gap in understanding the biological mechanisms underlying its resistance. In this study, we identify 3-hydroxy-3-methylglutaryl-CoA synthase 2 (HMGCS2), a rate-limiting enzyme in ketogenesis, as a novel biomarker associated with Apatinib resistance in HER2-negative breast cancer . Furthermore, we investigate its biological function and molecular mechanism.Methods Metabolomics analysis and immunohistochemical verification were performed on Apatinib-treated clinical samples to identify genes associated with Apatinib resistance. The biological role of these genes in HER2-negative breast cancer was validated through in vitro cell experiments and in vivo animal studies. The specific molecular mechanisms involved were elucidated using proteomics analysis , along with techniques such as co-immunoprecipitation and western blotting.Results HMGCS2 was found to be overexpressed in Apatinib-refractory tumors and was associated with resistance to Apatinib. Compared to normal mammary epithelial cells, HMGCS2 expression levels were significantly elevated in breast cancer cells. Knockdown of HMGCS2 via shRNA effectively suppressed tumor proliferation and invasion in HER2-negative breast cancer cell lines and reduced tumor burden in xenograft mouse models. Conversely, overexpression of HMGCS2 markedly enhanced malignant phenotypes in vitro. Mechanistic investigations revealed that RAB23-dependent β-catenin/TCF activation served as the primary driver of carcinogenesis.Conclusion Our study revealed a critical role of HMGCS2 in mediating Apatinib resistance in HER2-negative breast cancer. Mechanistically, HMGCS2 promoted malignant progression in HER2-negative breast cancer by activating the β-catenin/TCF transcriptional axis through RAB23-mediated signaling, highlighting HMGCS2 as a potential therapeutic target for precision oncology.

S. Jung¹, J. Yisun², S. Yang¹, Y. Sun Young², Y. Ji Young¹, L. Eun-Shin¹, O. Harim³, S. Jongmin³, K. Sangmin², L. Jeong Eon²; ¹Department of Surgery, Korea University Hospital, Seoul, KOREA, REPUBLIC OF, ²Department of Breast Cancer Center, Samsung Medical Center, Seoul, KOREA, REPUBLIC OF, ³Department of Pathology, Korea University Hospital, Seoul, KOREA, REPUBLIC OF.

Kinesin family member 11 (KIF11), also known as Eg5, is a motor protein critical for mitotic spindle formation and proper cell division. While KIF11 has been implicated in promoting cell proliferation and invasion in various cancer types, its regulatory mechanisms in breast cancer remain poorly understood. In this study, we investigated the regulation of KIF11 expression and assessed the therapeutic potential of its specific inhibitor, ispinesib, in breast cancer models. Analysis of the GSE4922 dataset revealed that estrogen receptor-positive (ER+) breast cancer patients with high KIF11 expression exhibited significantly poorer survival compared to those with low expression. Moreover, KIF11 levels were elevated in lymph node-positive (LN+) patients and those with advanced-stage disease, suggesting its potential role in tumor progression. We further found that KIF11 expression was positively correlated with MAPK1 expression and negatively correlated with STAT3 expression. Pharmacological inhibition of MEK1/2 using binimetinib significantly reduced KIF11 expression, whereas overexpression of constitutively active MEK1 (CA-MEK1) increased its expression. KIF11 knockdown (KD) also induced G2/M phase arrest and inhibited cell growth in MCF7 breast cancer cells. Finally, we evaluated the pharmacological efficacy of ispinesib in regulating the cell cycle and cell growth. As expected, ispinesib induced G2/M phase arrest and completely inhibited cell growth in ER+ breast cancer cells. These findings support KIF11 as a prognostic marker and potential therapeutic target in ER+ breast cancer, regulated at least in part by the MAPK signaling pathway.

S. Lee¹, S. Lee¹, I. Lee¹, S. Park², J. Kim³, N. Son⁴; ¹Department of surgery, International St. mary’s hospital, Incheon, KOREA, REPUBLIC OF, ²Division of Breast surgery, Department of surgery, Yonsei University College of Medicine, Seoul, KOREA, REPUBLIC OF, ³Yonsei University College of Medicine, Yongin, International St. mary’s hospital, Yongin, KOREA, REPUBLIC OF, ⁴Department of Statistics, Keimyung University, Daegu, KOREA, REPUBLIC OF.

PurposeBRCA mutations significantly increase the lifetime risk of breast cancer. While they are strongly associated with triple-negative breast cancer, they are also found in hormone receptor-positive, HER2-negative subtypes, in which multigene assays are commonly used for risk stratification and treatment decision-making. However, the relationship between BRCA status and multigene assay outcomes remains unclear, particularly in Asian populations. This study investigates the association between clinical indications for BRCA testing and multigene assay results in Korean breast cancer patients. Methods This multicenter study included 2,542 patients with hormone receptor-positive, HER2-negative breast cancer who underwent multigene testing between 2013 and 2024. BRCA testing was performed in patients aged under 40 years, those with bilateral breast cancer, or those with a personal or family history of breast or ovarian cancer within third-degree relatives. Multi-gene assays included Oncotype DX, MammaPrint, EndoPredict, and OncoFREE, and each result was classified into high-risk or low-risk groups based on each test-specific criteria. Statistical analyses were performed using chi-square tests and logistic regression in SPSS. Results Patients were categorized as BRCA-tested group (n=798) or non-tested group (n = 1,596). In the BRCA-tested group, 19.5% were classified as high risk and 80.5% as low risk by multigene assay. pTstage, pNstage, histologic grade, progesterone receptor status, and Ki-67 level was independently associated with high risk and multigene assay result. BRCA mutation was identified in 44 patients (7 BRCA1 mutation, 37 BRCA2 mutation). BRCA mutation was independently associated with high-risk classification on multigene assay (OR 3.554, 95% CI 1.767-7.151, p < 0.001). In subgroup analysis of BRCA mutation group, among 30 patients with a family history, 20 (66.7%) were classified as high risk. Multivariate analysis also confirmed that familial history of breast or ovarian cancer was independently associated with high-risk in BRCA-mutation patients. (OR 9.752, 95% CI 1.449-65.651, p <0.05). An association between BRCA testing indications and high-risk multigene assay results was not statistically significant when only one indication was present. Conversely, the presence of two or more indications demonstrated a potential correlation with high-risk classification. (OR 5.282, p <0.061) Conclusion In Korean patients with hormone receptor-positive, HER2-negative breast cancer, BRCA mutation was significantly associated with high-risk classification on multigene assay. In BRCA-mutated patients, the presence of a family history further increased the likelihood of high-risk classification. Additionally multiple BRCA testing indications predicted high-risk classification in BRCA mutation patients, but this trend was not observed in those without a detected mutation.

D. Schott, M. Pizon, A. Riess, K. Pachmann; Research, Laboratory Pachmann, Bayreuth, GERMANY.

Background:The most critical and life-threatening feature of malignant tumors is their ability to metastasize to distant organs, ultimately leading to patient mortality. Most metastases become clinically detectable only months or even years after the surgical removal of the primary tumor, indicating that they originate from tumor cells disseminated into the circulation and persisting in the body as minimal residual disease (MRD).It is widely assumed that the majority of circulating tumor cells (CTCs) rapidly disappear from the bloodstream following surgery, with a reported half-life of only a few hours. However, some of these cells clearly survive, either by recirculating or by entering protective niches within the body. Using the standardized maintrac® method, we demonstrated that substantial numbers of viable tumor cells can still be detected in peripheral blood several days after surgical intervention in patients with both lung and breast cancer. These findings strongly suggest that a fraction of CTCs is indeed capable of escaping rapid elimination.This raises an important question: why do most conventional CTC detection methods fail to identify these cells, both before and after surgery? How do tumor cells released into the bloodstream manage to remain undetected?Methods:Using fluorochrome- labeled anti-EpCAM antibody cells were dyed with the maintrac® approach right after blood draw and after resting the samples for up to 24 hours at room temperature on the shelf. At all times the presence of platelets was monitored using red fluorescing anti CD36 or anti-CD41/CD61 antibody.Results: Immediately after blood draw, no EpCAM-positive (green-fluorescing) CTCs were detectable. However, clusters of red-fluorescing platelets were observed, which, under transmitted light, were shown to be adherent to the circulating tumor cells. After 24 hours of resting at room temperature, a substantial number of viable EpCAM-positive tumor cells became detectable, with platelets visibly detached. Time-lapse imaging over 24 hours revealed progressive detachment of platelets from tumor cells, accompanied by increased anti-EpCAM staining intensity. These findings indicate that platelets form a cloak around circulating tumor cells in circulation, physically obstructing antibody binding and thereby preventing detection. Only upon platelet detachment could the tumor cells be reliably identified.Conclusion:Circulating tumor cells may evade detection by being cloaked in platelets. This platelet-mediated masking may not only hinder diagnostic detection but could also play a functional role in immune evasion and metastasis formation. Platelets expressing immune checkpoint molecules like PD-L1 may further suppress anti-tumor immune responses. These findings underscore the importance of considering platelet interactions in liquid biopsy techniques and highlight a potential target to improve cancer diagnostics and therapy.

H. Choi¹, B. Hassan¹, R. O’Connor², F. Michael¹, D. Mankoff¹, R. Zhou¹; ¹Radiology, University of Pennsylvania, Philadelphia, PA, ²Pathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, PA.

OBJECTIVES: Glutaminolysis is a metabolic pathway deployed by aggressive cancers including triple-negative breast cancer (TNBC) to support their growth and progression. Immune checkpoint inhibitors (ICI) are FDA approved treatment modalities for TNBC; however, treatment responses are modest and variable. To overcome the heterogeneity in response, there is a critical need to understand the unique characteristics underlying immune activity responding to ICI. Our prior study has shown that TNBC tumors are enriched with myeloid-derived suppressor cells (MDSCs), which were significantly reduced by a pan-glutamine metabolic inhibitor JHU083, a pool compound of DRP-104/ Sirpiglenastat being evaluated in clinical trials. JHU083/DRP-104 is the prodrug form of DON (6-Diazo-5-oxo-L-norleucine). Here we examined the impact of glutamine antagonism in cancer versus immune cells traversing the tumor microenvironment. PET imaging is an important tool to identify patients most suited to the therapeutic promise of ICI. Leveraging the recent development of a radio-labeled antibody against CD69, a well characterized early-activation cell surface marker, we tested whether this immune-PET marker can detect the enhanced activation of T cells in the TNBC tumor immune microenvironment. METHODS: Murine (4T1, EO771) and human (HCC1806) TNBC and ER+ (MCF7) cell lines obtained from ATCC, T cells and MDSCs isolated from the spleen, were treated with JHU083 (0.1µM) to assess ROS and lipid ROS levels using Dihydroethidium and C11BODIPY assays. The impact of MDSC on T cell activation was examined by co-culturing T cells with MDSCs for one day in culture media containing 0.05 µg/ml PMA and 0.1 µg/ml ionomycin and either 0.1 µM DON or without DON. For in vivo imaging studies, 4T1 cells were inoculated into the mammary fat pad of BALB/c mice. When tumors grew to approx. 100 mm³, mice were randomized and enrolled in control (no treatment), ICI (treated with PD-1 at 250 and CTLA4 at 100 μg/mouse twice/week i.p.) or ICI plus JHU083 (1.82 mg/kg, i.p. every other day from day 0) group. After one week treatment, each mouse received a 100 ~ 300 µCi of [⁸⁹Zr]-DFO-H1.2F3 (anti-CD69 mAb) via tail vein injection and PET scan was performed at 48 and 72 hours post-injection. After PET imaging, FACS analysis was performed to profile immune cells in the tumor. RESULTS: 0.1 µM DON does not induce cytotoxicity in TNBC cells, MDSCs or T cells, however it markedly increases cellular reactive oxygen species and lipid peroxidation in cancer cells and MDSCs but not in the activated or naïve T cells. CD69 PET signals (%ID/g) in 4T1 tumors are significantly higher after ICI (13%, p<0.05) or JHU083 plus ICI treatment (15%, p<0.05) compared to controls (9%). FACS analysis of tumors after PET imaging reveals elevated CD69⁺ cell population in treated tumors compared to controls. Meanwhile the CD45⁺CD11b⁺ cell fraction in ICI treated tumors (76%) remains similar to controls (84%) but decreased substantially after JHU083 plus ICI treatment (19%). In co-cultures, the presence of MDSCs reduced the fraction of CD69⁺ and granzyme B⁺ CD8⁺ T cells, confirming MDSC’s immune suppressive activity. However, adding 0.1 µM DON in co-culture reversed the functional competence of T cells. DISCUSSIONS & CONCLUSION: By selectively disrupting metabolism in cancer cells and MDSCs while sparing T cells that provide anti-tumor control, glutamine antagonism has immediate translational relevance for adoptive immunotherapies. Although ICI therapy removes PD-1 and CTLA4 mediated inhibition on T cells, it fails to inhibit tumor growth likely due to predominant presence of MDSCs which suppress CD8⁺ T cells as demonstrated by co-culture data. Combining JHU083 with ICI reduces MDSC infiltration to the tumor, leading to enhanced anti-tumor immunity detected by [⁸⁹Zr]CD69 PET.

K. V. Giridhar¹, D. Zahrieh², M. Hanna³, A. Singh⁴, E. Al-Hattab⁵, P. Schumacher², J. Carroll¹, J. Hoppenworth¹, T. Haddad¹, A. Tevaarwerk¹, B. Rudder⁴, G. M. Choong¹, S. Basu⁵, R. A. Leon-Ferre¹, T. O’Brien⁴, A. Lewis³, F. de Snoo⁶, L. Stork-Sloots⁶, F. K. Kuhr⁷, S. Lazaro⁸, R. Prendergast⁸, D. S. Spinner⁹, M. P. Goetz¹; ¹Medical Oncology, Mayo Clinic, Rochester, MN, ²Quantitative Health Sciences, Mayo Clinic, Rochester, MN, ³Medical Oncology, Mayo Clinic Health System, Albert Lea, MN, ⁴Medical Oncology, Mayo Clinic Health System, Mankato, MN, ⁵Medical Oncology, Mayo Clinic Health System, Eau Claire, WI, ⁶Medical Affairs, Medex15, Amsterdam, NETHERLANDS, ⁷Medical Affairs, Menarini Silicon Biosystems, Huntingdon Valley, PA, ⁸Clinical Laboratory, Menarini Silicon Biosystems, Huntingdon Valley, PA, ⁹Market Access & Reimbursement, Menarini Silicon Biosystems, Huntingdon Valley, PA.

Introduction Serial assessments of circulating tumor cells (CTCs) in metastatic breast cancer (MBC) provide independent prognostic significance for progression free survival and overall survival. A low CTC burden (<5 CTCs / 7.5mL blood) at baseline is favorable prognostically, and reduction of CTCs during treatment indicates response to treatment (Janni, CCR 2025). Despite increasing evidence supporting its clinical validity, real-time serial CTC enumeration has not been utilized broadly in clinical practice in the U.S. The ACT-MBC study is a prospective observational study to evaluate the impact of serial CTC testing on treatment decisions and response assessment (NCT05662345). Provider (physician or advanced practitioner) perception of CTC testing is measured with questionnaires at baseline (PQ1), and for each pt during the first restaging visit (PQ2), and at the end of the study (PQ3). We previously reported PQ1 results (Giridhar, SABCS 2024). Here, we report on results of PQ2. Methods ACT-MBC prospectively enrolled patients (pts) at Mayo Clinic Rochester and Mayo Clinic Health Systems sites (Mankato, Albert Lea and Eau Claire). Pts with MBC were enrolled at time of progressive disease during any line of therapy (TNBC) or ≥2^nd line metastatic therapy for ER+/HER2- disease. Pts were tested for CTCs using the FDA-cleared CellSearch® assay which included enumeration and biomarker testing on CTCs for ER, HER2, and PD-L1. The treating clinician determined the regimen and timing of restaging scans. CTC testing was performed once per cycle (every 3-4 weeks) with results returned to the treating provider within 7 days. We descriptively summarized PQ2 responses from providers who had at least one on-treatment CTC result available by the first restaging visit. We report the proportion (90% CI) of providers who agreed or strongly agreed that assessing CTCs enhanced their ability to determine response to therapy and that the CTC results helped guide their decision to continue or change therapy (co-primary endpoints) while accounting for within provider correlation. Results As of June 25, 2025, 15 providers enrolled a median of 2 pts each (range: 1 to 15 pts) and completed PQ2 surveys on 40 pts. 7 providers (47%) practiced in the community setting and 4 (27%) were nurse practitioners. Most pts were female (98%), White (98%), and had ER+/HER2- MBC (80%). Baseline (C1) CTC results were elevated (≥5 CTCs) in 24 (60%) pts (CTC range 0-609). At C2 (n=39), CTC values remained undetected in 7 pts, decreased in 21 pts, increased in 8 pts, and remained unchanged in 3 pts. At PQ2 (n=40), based on clinician interpretation of radiographic findings, results showed 2.5% of pts with complete response, 20% with partial response, 25% with stable disease, 17.5% with mixed response, and 35% with progressive disease. Clinicians reported that CTC results were concordant with imaging results in 34 (85%) responses, discordant with imaging in 7.5% of responses, and unclear in the remaining 7.5% of responses. Most clinicians (82.5%) agreed or strongly agreed that assessing CTCs enhanced their ability to determine response to therapy (90% CI: 68.3%, 96.7%). Further, 70% of providers agreed or strongly agreed that the CTC results helped guide their decision to continue or change therapy (90% CI: 50.8%, 89.2%). 32.5% of the time, CTCs results influenced the timing of the first restaging scans. In 11/25 pts without progressing disease, CTC results influenced the timing of future scans. In the 7/8 pts with a mixed response at first restaging scans with decreasing CTCs, 4 remained on the same systemic therapy. Conclusions Oncologists in academic and community practice responded that serial CTC assessments enhance ability to determine response of therapy, guide decisions to continue or change treatments, and influence timing of radiographic assessments.

K. Venetis¹, A. Marra², D. Presti³, D. Trapani², V. Peruzzo¹, R. Adorisio¹, G. Castellano², A. Ranghiero¹, A. Borghi¹, D. Vacirca¹, E. Giordano², S. Perazzo², K. Qeraj², P. Zagami², M. Lombardi¹, E. Munzone³, G. Curigliano², E. Guerini Rocco¹, N. Fusco¹; ¹Division of Pathology, European Institute of Oncology IRCCS, Milan, ITALY, ²Early Drug Development for Innovative Therapies, European Institute of Oncology IRCCS, Milan, ITALY, ³Division of Medical Senology, European Institute of Oncology IRCCS, Milan, ITALY.

Introduction: Clinically actionable mutations in ESR1 and genes within the PI3K/AKT/PTEN pathway inform the use of selective estrogen receptor degraders (SERDs), PI3K inhibitors, and AKT inhibitors in patients (pts) with hormone receptor-positive, HER2-negative (HR+/HER2−) metastatic breast cancer (mBC). While the prevalence of such mutations detected through circulating tumor DNA (ctDNA) varies by treatment type and line, the broader landscape of co-occurring actionable mutations remains poorly characterized. Methods: Clinical and genomic data were retrospectively collected from patients with HR+/HER2− mBC who underwent ctDNA testing at the European Institute of Oncology (IEO) between November 2023 and June 2025, upon progression to endocrine therapy plus CDK4/6 inhibitors. Genomic profiling was performed using a 77-gene liquid biopsy (LB)-based next-generation sequencing (NGS) assay. Progression-free survival (PFS) and overall survival (OS) were compared between groups using Cox proportional hazards models. Results: Among the 256 patients who underwent ctDNA testing, 53% were tested after first-line (1L) therapy, 24% after second-line (2L), and 23% after third-line or beyond (≥3L). PIK3CA mutations were detected in 33% of pts (n=84), ESR1 in 32% of pts (n=82), PTEN in 5.9% of pts (n=15), and AKT1 in 3.9% of pts (n=10). PIK3CA mutations were more frequently identified in patients tested at ≥3L, while ESR1, PTEN, and AKT1 alterations were more frequently observed in patients tested at 1L. However, these distributional differences across treatment lines did not reach statistical significance. Other clinically relevant genes mutations were also observed in TP53 (n=47, 18.4%), ERBB2 (n=16, 6.3%), and RB1 (n=10, 3.9%). Co-mutations in both ESR1 and PIK3CA were identified in 30.4% (n=25) of ESR1-mutant patients, more frequently when tested after the 1L (31.2%) and ≥3L (42.1%). ESR1 co-mutation with AKT1 and PTEN was less frequent, occurring in 6.1% and 3.6% of ESR1-mutant patients, respectively. To explore the clinical relevance of these molecular profiles, PFS and OS were compared across three molecular subgroups: (i) ESR1-mutant/PI3K-pathway-wild-type, (ii) ESR1-wild-type/PI3K-pathway-mutant (PIK3CA, AKT1, or PTEN), and (iii) double-mutant (ESR1-mutant/PI3K-pathway-mutant). No associations between co-mutations and clinical outcomes were found. Conclusions: ctDNA-guided molecular profiling is a fundamental tool for guiding clinical decisions among patients with HR+/HER2− who experience progression on ET. In our cohort, co-mutations occurred in almost 1/3 of patients and seems not to influence prognosis. More data are needed to clarify the prognostic role of co-mutations and to determine the most appropriate molecular driver for therapeutic choices.

E. P. Taranto¹, W. Fawn¹, C. Desirae¹, E. Walinsky¹, S. Deluca¹, L. Bayne¹, E. P. Wileyto², A. Shirali³, C. J. Robbins³, Y. Abdou⁴, J. Anampa⁵, M. Balic⁶, J. V. Canzoniero⁷, N. C. Chen⁸, F. Howard⁹, L. Huppert¹⁰, A. Kahn¹¹, C. Mainor¹², J. Mouabbi¹³, V. Nair¹⁴, T. Degeer¹⁵, J. Weidler¹⁶, T. McNicholas¹⁵, C. Serway¹⁷, J. Reischl¹⁸, M. Bates¹⁶, R. Monroe¹⁷, D. L. Rimm³, A. DeMichele¹; ¹Division of Hematology/Oncology, University of Pennsylvania, Philadelphia, PA, ²Department of Biostatistics and Epidemiology, University of Pennsylvania, Philadelphia, PA, ³Department of Pathology, Yale School of Medicine, New Haven, CT, ⁴Division of Oncology, University of North Carolina School of Medicine, Chapel Hill, NC, ⁵Department of Oncology, Montefiore Einstein Comprehensive Cancer Center, New York, NY, ⁶Division of Oncology, Hillmans Cancer Center, University of Pittsburgh, Pittsburgh, PA, ⁷Division of Oncology, Johns Hopkins Sidney Kimmel Cancer Center, Baltimore, MD, ⁸Department of Medicine, Section of Hematology and Oncology, Baylor College of Medicine, Houston, TX, ⁹Biological Sciences Division, Section of Hematology/Oncology, University of Chicago, Chicago, IL, ¹⁰Department of Medicine, University of California San Francisco, San Francisco, CA, ¹¹Department of Medicine, Division of Medical Oncology and Hematology, Yale School of Medicine, New Haven, CT, ¹²Department of Internal Medicine, Division of Hematology/Oncology, Georgetown University School of Medicine, Washington, DC, ¹³Department of Breast Medical Oncology, UT MD Anderson Cancer Center, Houston, TX, ¹⁴Division of Hematology/Oncology, Fred Hutch/University of Washington School of Medicine, Seattle, WA, ¹⁵Open Innovation, Leica Biosystems, Richmond, IL, ¹⁶Medical/Scientific Affairs and Strategy, Oncology, Cepheid, Sunnyvale, CA, ¹⁷Oncology, Danaher Diagnostics, Washington, DC, ¹⁸External Innovation Diagnostics, Danaher Diagnostics, Washington, DC.

Background: Antibody drug conjugates (ADCs) can be effective therapeutic options for some patients (pts) with metastatic breast cancer (MBC). However, given serious potential toxicities, identifying pts most likely to benefit from ADCs is critical to optimizing quantity and quality of life. Trastuzumab-deruxtecan (T-DXd), a HER2-targeted ADC, is approved for pts with MBC expressing low/ultra-low levels of HER2 by immunohistochemistry (IHC 1+/2+, FISH-neg, or IHC 0 with faint membrane staining in 400 retrospective samples, ~50% of HER2 IHC 0 cases had HER2 QIF levels above the limit of quantification, with a significant relationship demonstrated between HER2 QIF and time-to-next-treatment in a retrospective cohort of pts treated with T-DXd. Retrospective data using Cepheid’s Xpert Breast Cancer STRAT4 mRNA assay have also demonstrated high ERBB2 concordance with HER2 IHC, with potentially greater range. Both assays hold promise to improve measurement of surface protein expression at lower limits of detection than traditional IHC and may better select pts likely to respond to T-DXd. The primary objective of QuantifyHER is to determine the relationship between HER2 QIF and/or ERBB2 mRNA levels and real-world objective response (rwOR) to T-DXd. Secondary objectives include evaluating the relationships between TROP2 and rwOR, HER2 QIF/mRNA and real-world progression-free survival, determining response thresholds and the relationship with ER status, and evaluating a joint QIF/mRNA approach. Methods: QuantifyHER (TBCRC066/NCT06551116) is a multi-site prospective cohort study. Key eligibility criteria include biopsy-proven HER2 IHC<2+ MBC, available tissue, measurable disease, and initiation of T-DXd within 30 days of registration. Pts receive T-DXd per standard of care and are followed for real-world endpoints. Up to 10 unstained FFPE/1 H&E section from an IHC <2+ archival specimen (most recent metastatic biopsy preferred) are sent to the Quantitative Diagnostics in Anatomic Pathology (QDAP) CLIA lab at Yale for analysis with both TROPLEX (returning HER2 and TROP2 QIF levels in amol/mm²) and the Cepheid Xpert STRAT4 Research Use Only (RUO) assay (returning ERBB2 mRNA delta Cycle threshold (dCt) levels). Patient data (demographics, clinical features, prior treatment, follow-up) are abstracted from the medical record, with clinical endpoints assessed by local investigators using real-world response categories and toxicities collected only as related to T-DXd discontinuation or modification. The planned sample size is 200 evaluable pts, providing 80% power (with a 2.5% 1-sided alpha), to detect a difference in proportions of 9-10 % per standard deviation QIF/mRNA increase, corresponding to an odds ratio of 1.6. A non-binding interim review at n=70 will assess preliminary biomarker distribution for any necessary sample size adjustment. QuantifyHER is open within the Translational Breast Cancer Research Consortium (TBCRC). As of 7/1/25, 10/12 planned sites are open and 23 pts have enrolled.

J. Bush¹, J. Essif¹, K. Lathrop², R. Kannaiyan², A. L. Risinger¹; ¹Cancer Pharmacology, UT Health Science Center San Antonio, San Antonio, TX, ²Oncology, UT Health Science Center San Antonio, San Antonio, TX.

Microtubule stabilizing taxanes are a mainstay of the treatment of breast cancer in both the adjuvant and neoadjuvant settings, including as backbone chemotherapy in combination with targeted agents. Despite their ubiquitous use and well-understood mechanism of action, there are currently no biomarkers to inform which tumors are most likely to respond to taxane treatment. A class of cytoskeletal associated small GTPases critical for multiple oncogenic processes, the septins, can directly interact with microtubules in a taxane-dependent manner with distinct septin9 isoforms having opposing effects. The longest isoform, sept9_i1, is associated with tumor progression, promoting oncogenic activity and taxane resistance through direct binding to microtubules. In contrast, the sept9_i2 isoform lacking this microtubule binding domain reduces migration with lower levels in tumors as compared to normal tissue. We hypothesized that the relative expression of these mechanistically distinct septin9 isoforms that differentially interact with microtubules could serve as a clinical biomarker of taxane response. We conducted a retrospective analysis of the relationship between septin9 isoform expression in tumor biopsies and the overall survival of taxane-treated breast cancer patients post-metastatic diagnosis. Septin9 isoform expression did not differ between primary and metastatic biopsies or between tumors of ductal versus lobular origin. However, tumors from triple-negative breast cancer (TNBC) patients expressed significantly lower levels of septin9_i1 and higher levels of septin9_i2 than ER and HER2 positive breast tumors. Elevated expression of septin9_i2 was correlated with a significant increase in overall survival specifically in TNBC patients with no correlation observed for septin9_i1. To further interrogate the role of septin9 isoform expression in the response of breast tumors to taxane-containing neoadjuvant chemotherapy we conducted a prospective clinical study in previously untreated patients where isoform expression was evaluated in pre and post-treatment biopsies and correlated with pathological response. Consistent with the retrospective study, we found a trend toward elevated levels of septin9_i2 expression in patients who had a pathologic complete response (PCR) to taxane-containing chemotherapy and a reduction in expression post-treatment in patients without a PCR. Comparative studies found a high correlation between septin9 isoform expression at the mRNA level using qRT-PCR and the protein level by immunohistochemistry using isoform-specific antibodies.To further interrogate the potential underlying mechanism of septin9 isoforms in the response of breast cancer cells to taxane treatment we generated isogenic breast cancer cell lines where specific isoforms were stably knocked down or overexpressed. Consistent with published results, we found that reduced septin9_i1 slightly improves the efficacy of taxane treatment in vitro however the beneficial role of septin9_i2 observed in the clinical studies was not evident in vitro. We further interrogated how alterations in septin9 isoform expression regulate the ability of taxanes to stabilize cellular microtubules and demonstrate that drugs that differentially alter microtubule structure and dynamics effectively shift the association of oncogenic septins between the microtubule and actin cytoskeleton. Together, these data support further interrogation of the role of septin9 isoforms in tumor models and clinical samples to determine whether septin9_i2 has a causal role in the response of tumors to taxane-containing chemotherapy and, if so, identify whether clinically approved chemotherapeutics that work through distinct mechanisms of action would have benefit in tumors with reduced expression of this isoform.

J. Vykoukal¹, Y. Chen¹, M. Zuo¹, R. Ballaro¹, M. Hong¹, H. Krishna¹, D. Rodriquez-Perera¹, H. Katayama¹, E. Irajizad², R. Wu¹, R. León-Letelier¹, J. Dennison¹, A. Gutierrez³, A. Paulucci-Holthauzen⁴, Y. Cai¹, F. Hsiao¹, S. Park¹, B. Arun³, S. Hanash¹, J. F. Fahrmann¹; ¹Clinical Cancer Prevention, UTMDACC, Houston, TX, ²Biostatistics, UTMDACC, Houston, TX, ³Breast Medical Oncology, UTMDACC, Houston, TX, ⁴Genetics, UTMDACC, Houston, TX.

Triple-negative breast cancer (TNBC) represents a heterogeneous subtype of breast cancer characterized by high pathological grade, strong invasiveness, local recurrence, high metastasis rate, and poor prognosis. Chemotherapy and in some cases, immunotherapy, remain the main choice of systemic treatment. However, response is variable, lacks durability, and acquired resistance to chemotherapy is a frequent manifestation. There remains an unmet need to establish novel alternative therapies that can both exert targeted effects on cancer cells as well as stimulate an anti-cancer immune response. We have uncovered an onco-metabolic biochemical feature in TNBC wherein cancer cells increase chemical conversion of sphingomyelin to ceramide with subsequent glycation to glycosphingolipid derivatives. Our mechanistic studies have established that this biochemical program is essential for survival mitophagy. The activation state of this program in cancer cells opens a metabolic vulnerability that is targetable through biochemical inhibition of glucosylceramide synthase (UGCG), a rate-initiating enzyme that mediates glycation of ceramides to glycosphingolipid derivatives. Our preclinical studies using orthotopic syngeneic mouse models of Brca-deficient TNBC as well as human cell line and patient-derived xenograft models of TNBC have demonstrated that targeting of cancer cell UGCG via repurposing eliglustat, an FDA-approved selective UGCC inhibitor for long-term use in patients with Gaucher’s disease, at clinically achievable doses elicits potent anti-cancer effects by targeting cancer cell mitochondria, resulting in accumulation of mitotoxic ceramides, induction of compensatory mitophagy, and induction of cyclic GMP-AMP Synthase (cGAS)-Stimulator of Interferon Genes (STING) pathway activities to promote an anti-cancer immune response. Importantly, assessment of eliglustat-associated toxicity revealed that extended-use of eliglustat was safe and tolerable in treated mice with histopathological examination of kidney and liver from treated mice revealing no remarkable differences compared to saline control. Clinical chemistry tests following eliglustat treatment indicated dose-dependent elevations in blood ALT and AST levels, consistent with clinical studies. Our study highlights utility of eliglustat as a potential anti-cancer agent for the treatment of TNBC.

M. M. Caffarel, U. Alvarez, P. Azcoaga, A. Romo, S. Manzano, P. Petit, I. Garmendia, M. Garzón, Z. Telletxea, M. Otaño, A. Urruticoechea, I. Alvarez-Lopez; Oncology, Biogipuzkoa Health Research Institute, Donostia / San Sebastian, SPAIN.

Inflammation is a natural immune response that can protect against cancer in early stages but, when becomes chronic, creates an immunosuppressive tumour microenvironment (TME) that promotes tumour progression. Recognized as a hallmark of cancer, chronic inflammation offers potential for new therapeutic strategies, particularly immunotherapy. However, some solid tumours, such as breast cancer, show partial resistance to these treatments. Cytokines are key regulators of inflammation, influencing cell communication and immune responses within the TME (1,2). Importantly, current clinical classification of BC often neglects the immune landscape, limiting the ability to predict responses to immunotherapy.Our hypothesis is that the cytokine profile of metastatic breast cancer possesses prognostic and predictive value and that cytokines can be potential biomarkers in breast cancer. Understanding how cytokine loops are coordinated within the breast TME could also lead to novel cytokine-based therapeutic strategies. Cytokine levels in serum and fresh tissue samples from metastatic BC patients (n=150) and in serum samples from metastatic cancer patients from other solid tumour types (n=186) were measured by Olink cytokine panel, which quantitatively assess 40 cytokines. Cytokine levels were correlated with clinical data, including overall and relapse free survival. The statistical analysis was done using different mathematical modelling methods. Our unpublished results show that the serum and tumour cytokine profiles do not correlate significantly, supporting the idea that some cytokines act locally in the TME. In addition, comparison with other metastatic tumour samples, led to the identification of some cytokines that are differentially expressed in BC. Integration of these data will shed light on the characterization of cytokine interactions in metastatic BC, helping to define its immune landscape. 1.Soler MF, Abaurrea A, Azcoaga P, Araujo AM, Caffarel MM. New perspectives in cancer immunotherapy: targeting IL-6 cytokine family. J Immunother Cancer. 2023;11(11):e007530; 2.Manzano S, Caffarel MM. Cytokine-centered strategies to boost cancer immunotherapy. Mol Oncol. 2025;19(3):579-583

C. J. Schwartz¹, D. Muldoon¹, H. Dopeso¹, A. Gazzo¹, P. Terraf¹, M. Repetto¹, B. Weigelt¹, D. Mandelker¹, L. Norton², H. Y. Wen¹, H. Zhang¹, D. Ross¹, B. Edi¹, A. Mamtani³, S. Shen², C. Bandlamudi¹, F. Pareja¹; ¹Pathology, Memorial Sloan Kettering Cancer Center, New York City, NY, ²Medicine, Memorial Sloan Kettering Cancer Center, New York City, NY, ³Surgery, Memorial Sloan Kettering Cancer Center, New York City, NY.

Background: Identification of genetic alterations in homologous recombination deficiency (HRD) and DNA damage response (DDR) genes in breast cancer (BC) is key for prognostication and potential selection for targeted therapies, such as PARPi. While HRD/DDR somatic oncogenic alterations are generally associated with ductal BCs, their roles in invasive lobular carcinoma (ILC) remain unclear. Here, we sought to investigate the spectrum of HRD/DDR somatic genetic alterations in ILC and their histologic features of these BCs. Design: We interrogated 1,076 ILC previously subjected to paired tumor-normal targeted sequencing (MSK-IMPACT) for oncogenic somatic genetic alterations in 22 HRD/DDR genes. Loss of heterozygosity (LOH) was assessed using FACETS, and mutational signatures using SigMA. HRD status was evaluated using an algorithm optimized for MSK-IMPACT. Histologic assessment was performed as per the WHO. ER/HER2 status was determined per ASCO/CAP guidelines, and follow-up data retrieved from medical records. Results: We identified 53 ILC, including 14 primary ILCs and 39 recurrent/metastatic ILCs harboring somatic oncogenic alterations in HRD/DDR genes, with BRCA2 (38%) and ATM (23%) being the most frequently affected. Other recurrently altered (≥2 cases) HRD/DDR genes were CHEK2, PALB2, BRCA1, BARD1, RAD50, RAD51C, NBN, RECQL and ABRAXAS1. LOH analysis revealed that 53% of HRD/DDR genetic alterations in ILCs were bi-allelic, including 75% of BRCA2 and 70% of ATM cases.Most (71%) primary ILCs were of pleomorphic type or exhibited pleomorphic features (Pleo-ILC), histologic grade 2/3 (93%) and ER+/HER2- (93%). Of the metastatic ILCs, 59% were pleomorphic, and most were ER+/HER2- (74%), followed by HER2+ (18%). Median follow-up for the HRD/DDR-mutant primary ILC cohort was 72 months (range 21-150); 6 patients (43%) had no evidence of disease, 5 were deceased (36%), and 3 were alive with disease (21%). Median follow-up time for the HRD/DDR-mutant metastatic ILC cohort was 84 months (range 19-288); 62% of patients were deceased. ILCs with biallelic inactivation of BRCA2 were classified as HRD by our algorithm in 60% of cases, and half of them showed a dominant HRD mutational signature, whereas the majority of those with ATM biallelic alterations (83%) were classified as non-HRD and exhibited either dominant APOBEC or clock signature (60%). The most frequent non-HRD/DDR affected genes in the cohort were CDH1 (76%), PIK3CA (48%), and FOXA1 (21%).Conclusion: Our findings show that, although at a lower frequency than that reported for ductal BC, DNA repair deficiency including HRD/DDR defects can occur in a subset of ILC. Somatic alterations affecting HRD/DDR genes in ILC are often biallelic, with BRCA2 and ATM being the most altered genes. While these tumors are phenotypically diverse, they often display pleomorphic histology and appear to be associated with more aggressive clinical behavior. These findings highlight the potential clinical relevance of HRD/DDR profiling in ILC, which may help inform treatment strategies.

S. D. Sisoudiya¹, J. S. Ross², R. B. Keller-Evans³, R. S. Huang³, A. B. Schrock³, G. Frampton¹, E. M. Ebot¹, E. S. Sokol¹, S. Sivakumar¹, A. M. Kahn⁴, M. Lustberg⁵; ¹Computational Discovery, Foundation Medicine Inc., Cambridge, MA, ²Medical Affairs, Foundation Medicine Inc., Cambridge, MA, ³Clinical Development, Foundation Medicine Inc., Cambridge, MA, ⁴Center for Breast Cancer, Yale School of Medicine, New Haven, CT, ⁵Yale Cancer Center, Yale School of Medicine, New Haven, CT.

Background: Recent advances in antibody-drug conjugates have demonstrated clinical benefit in patients with HER2-positive breast cancer as well as HER2-low and ultra-low expression tumors. However, prior studies have shown that only a fraction of patients positive by IHC will respond to HER2-targeted therapy. Orthogonal methods to determine HER2 activity such as gene expression have the potential to improve patient selection. We utilized gene expression profiling to build a multi-gene machine learning model to predict HER2 activity, which may aid the identification of patients most likely to benefit from HER2 targeted therapies, particularly for those with HER2-low tumors. Methods: Our cohort comprised 477 patients with breast cancer who underwent RNA-based gene expression profiling of 1,517 genes using FoundationOne®RNA (a laboratory developed test by Foundation Medicine, Inc.). HER2 IHC status was abstracted from pathology reports (IHC 0: 182; 1+: 137; 2+: 128; 3+: 30). Differentially expressed genes (DEGs) between HER2 IHC 0 (label: no activity) and 2+/3+ (label: high activity) groups were identified based on an FDR p-value <0.01 and log₂(fold change)>1. A random set of 25% of 0 and 2+ samples and 50% of 3+ samples were held out, prior to modeling. Using a 70:30 (training:test) split of the remaining samples, lasso regression was applied on the expression of DEGs (quantified as transcripts per million, TPM) to train the model. Held-out samples were used to assess the proportion of samples predicted to have high HER2 activity across all IHC categories. This was repeated for 50 iterations and summary metrics were calculated across all iterations. Results: A total of 44 DEGs were used to train the model; ERBB2, TTYH1 and GRB7 were some of the top predictors. Running the model on held-out samples across 50 iterations, IHC 3+ had the highest proportion of samples predicted as “high activity” (median proportion 87%, IQR [87%-93%]), with decreasing proportions in 2+ (58% [47%-63%]), 1+ (41% [35%-44%]) and 0 (22% [17%-28%]). Notably, HER2 3+ samples with no predicted activity had a significantly lower median ERBB2 expression (4,825 TPM [IQR: 4,502-5,167]) than high activity samples (34,146 [IQR: 21,568-44,604]) (P<0.0001) and were most frequently ERBB2 amplification negative (OR = 14.2, P < 0.0001). Additional performance assessment in an independent validation cohort will be presented. Conclusion: A multi-gene expression signature could identify signaling pathways that may better inform HER2 dependency. The machine learning-based model developed here showed a higher proportion of HER2 activity cases with increasing HER2 IHC staining, with potential utility to identify patients who may benefit from HER2 targeted therapies. Exploration in a clinical cohort is warranted.

M. Homsi¹, M. DaVila², R. Brentjens³, B. Betts⁴, S. Kabraji¹; ¹Department of Breast Medicine, Roswell Park Comprehensive Cancer Center, Buffalo, NY, ²Department of Medicine – Translational Research, Roswell Park Comprehensive Cancer Center, Buffalo, NY, ³Department of Medicine, Roswell Park Comprehensive Cancer Center, Buffalo, NY, ⁴Department of Medicine – Transplant & Cellular Therapy Center, Roswell Park Comprehensive Cancer Center, Buffalo, NY.

Background: We previously reported that CD83, a member of the immunoglobulin super family, is a putative target for CAR-T therapy in breast cancer based on preclinical activity against breast cancer cell lines and expression by IHC in a breast cancer tissue microarray. Here we use high-plex tissue imaging (cyclic immunofluorescence) to describe the single-cell and sample based expression of CD83 in tissue compartments and measure its association with canonical breast cancer biomarkers, estrogen receptor, HER2 receptor. Methods: Tissue microarrays underwent dewaxing and antigen retrieval using a BOND auto stainer in the Advanced Tissue Imaging Shared Resource (ATISR). Tissues then underwent bleaching followed by iterative staining of unconjugated or conjugated antibodies. We completed six cycles of staining for all tissues, comprising 18 markers per tissue. Images were acquired at 20x on Leica Thunder Tissue Imager (ATISR). Images were processed using the MCMICRO pipeline and downstream analysis performed in R and Scimap. Results: After filtering for cellularity and technical artifacts, we analyzed 81 TMA cores (41 patients) of which 48 cores were HER2+ (IHC 3+) and 33 were HER2-low (30 HER2 2+ FISH negative, 3 HER2 IHC 1+). 42 cores were ER+ while 28 were ER neg by IHC with 11 not classified. We analyzed a total of 5.0 x10⁵ cells of which 34% were tumor cells (pan-cytokeratin+), 37% were immune cells (CD45+), with the remainder being stromal cells(alpha-smooth muscle actin+). We found that single cell CD83 expression was positively correlated with pan-CK and CD45, consistent with CD83 expression (fluorescence) in both tumor and immune cells. Median CD83 single cell expression was negatively correlated with tumor ER expression (r -0.34, p= 0.0002) but not correlated with median tumor HER2 expression (r=0.03, p=0.7) by CyCIF or IHC (1+, 2+, 3+). Mean CD83 tumor cell expression was higher in ER neg (<1% by IHC) vs ER pos tumors (0.18 +/- 0.1 vs 0.13 +/- 0.7, p <0.001) while CD83 immune cell expression was higher in ER+ tumors (0.24 vs 0.21, p<0.001). Conclusions: Together, these data suggest that tumors that are ER negative may express highest CD83 in the tumor compartment. Given our previous work showing activity of CD83 CAR-T against LM2 TNBC murine tumor model, these data suggest that ER negative tumors may be more likely to respond to CD83-targeting CAR-T cells. Further spatial analysis and correlation of CD83 with outcome will be presented at the conference.

R. BOUCHELAGHEM, I. Boulouh, W. Bechairia, M. Ghassoul, S. Bouacha, S. Bouharoud, S. Mohammed Benali, R. Hachichi, A. Kitouni, N. Néche, H. Djedi; Medical Oncology, Anticancer Center Annaba, Annaba, ALGERIA.

Neoadjuvant chemotherapy impacts HER2 status as defined by Sataloff classification and consequent follow up in a lower middle-income country Rim Bouchelaghem ¹, Ibtissem Boulouh^1,2, Wafa Bechairia ¹, Marwa Ghassoul ¹, Salsabil Bouacha ¹, Safia Bouharoud ¹, Soumaya Benali ¹, Rayene Hachichi ¹, Ahmed Kitouni ¹, Nadia Néche ^1,2, Hanene Djedi ^1,2 1.Medical oncology Departement, Anticancer Center, Annaba, Algeria2.Faculty of medicine, Badji Mokhtar University, Annaba, Algeria Keywords: Breast cancer, HER2, Sataloff classification, Neadjuvant chemotherapy. Introduction: Neoadjuvant chemotherapy (NACT) followed by surgery remains a common therapeutic strategy for the management of breast cancer (BC) in Algeria. NACT can induce molecular heterogeneity of BC, which presents a challenge in achieving a complete response and impacts patient follow-up, particularly when it affects tumor HER2 expression and modifies patient eligibility for targeted therapies. This evidence can be notably demanding in Algerian public hospitals, where it presents a consequential issue in the face of a shortage of health products, especially for targeted therapies such as HER2 therapy. We therefore assessed post-NCAT HER2 status change in response to different protocols proposed to patients with BC.Methods: This study aimed to assess the change in HER2 status before and after NCAT, as determined by immunohistochemistry, for 61 patients managed at our single institution. According to guidelines, all patients received NACT, with or without anti-HER2 (Trastuzumab with or without Pertuzumab). Pathological response (pR) was evaluated according to Sataloff classification. All patients underwent post-NCAT surgery. Descriptive statistics and the Chi-square test were used to determine statistical differences of pR. Results: The mean age of patients was 50,01 ± 13.39 years (range 24-80). Among the 61 BC, 41 tumors were invasive carcinoma of no specific type (67%), 20 carcinomas (33%) were 10 ductal (16%), five lobular (8%), two mixed (3%), and three mucinous (5%). Hormone receptors (HR) were positive in 43 patients (70%) and negative in 12 patients (20%). Six patients were HER2-enriched. Of all evaluable tumors, the HER2 score was 0 in 18 patients (30%), low in 33 patients (54%), and 3+ in 10 patients (16%). There were 4 complete pR (TA) out of 61 (6,55%). Indeed, there were 7 grade TA (11%), 27 (44%) TB, 17 (28%) TC, and 10 (16%) TD. Among post-NACT-evolved HER2 tumors, 33 (54%) maintained their HER2 level, whereas 28 (46%) exhibited a modified HER2 level. Among 18 HER2-0 tumors (30%), 10 (16%) remained HER2-0, 7 (11%) transitioned to HER2-Low expression, and one to HER2-3+ expression. Among 33 HER2-low tumors (54%), 14 tumors (22%) remained HER2-low, 15 tumors (24%) transitioned to HER2-0, and 3 (9%) transitioned to HER2-3+. Among the 10 HER2-3+ tumors, 8 (80%) remained HER2-3+, while 2 (20%) transitioned to HER2-0 and Low status. Interestingly, there was a statistically significant difference in the percentage of tumors with better pR that maintained HER2 status grade TA (10%), TB (30%), TC (11%), TD (3%) in comparison to tumors that exhibited a changed HER2 status with no grade TA, TB (16%), TC (16%) and TD (13%) (p < 0,0001). Conclusion: Our results support the well-established post-NACT heterogeneity. This reflects a real issue for achieving a pR for patients and managing BC in our country. Indeed, we found for the first time that concomitant analysis of Sataloff response and HER2 changes can represent an endpoint surrogate for treatment achievement after NACT in BC, and a favorable prognosis is associated with HER2 upregulation. The assessment of HER2 after NCAT in combination with Sataloff response is a prerequisite for managing the follow-up of patients. Further studies are needed to characterize this finding.

I. Paris¹, G. Mantini², T. D’Angelo¹, A. Minucci³, G. Luciano², C. Parrillo², M. De Bonis⁴, D. Generali⁵, G. Raspaglio⁴, M. De Donato⁴, F. Pavese¹, C. Nero¹, A. Fabi⁶, A. Piermattei⁷, P. Fuso¹, V. Ricozzi¹, L. Carbognin⁶, S. Rotondaro¹, T. Pasciuto⁸, M. Marrazzo⁹, A. Mulè⁷, G. Garganese¹, A. Di Leone¹⁰, M. Sanchez¹⁰, A. Migliore¹¹, D. Giannarelli¹², G. Franceschini¹⁰, A. Fagotti¹³; ¹Gynecologic Oncology, Department of Woman and Child Health and Public Health, Fondazione Policlinico Agostino Gemelli IRCCS, Rome, ITALY, ²Bioinformatics Core Facility, Gemelli Science and Technology Park (G-STeP), Fondazione Policlinico Agostino Gemelli IRCCS, Rome, ITALY, ³Genomics Core Facility, Gemelli Science and Technology Park (G-Step), Fondazione Policlinico Agostino Gemelli IRCCS, Rome, ITALY, ⁴Genomics Core Facility, Gemelli Science and Technology Park (G-STeP), Fondazione Policlinico Agostino Gemelli IRCCS, Rome, ITALY, ⁵Department of Medical, Surgery and Health Sciences, Cremona Hospital, University of Trieste, Cremona, ITALY, ⁶Department of Woman and Child Health and Public Health, Fondazione Policlinico Agostino Gemelli IRCCS, Rome, ITALY, ⁷Gynecopathology and Breast Pathology Unit, Department of Woman and Child Health and Public Health, Fondazione Policlinico Agostino Gemelli IRCCS, Rome, ITALY, ⁸Data Collection Core Facility, Gemelli Science and Technology Park (G-STeP), Fondazione Policlinico Agostino Gemelli IRCCS, Rome, ITALY, ⁹Data Collection Core Facility, Gemelli Science and Technology Park (G-STeP),, Fondazione Policlinico Agostino Gemelli IRCCS, Rome, ITALY, ¹⁰Breast Unit, Department of Woman and Child Health and Public Health, Fondazione Policlinico Agostino Gemelli IRCCS, Rome, ITALY, ¹¹Department of Woman and Child Health and Public Health, Marche Polytechnic University and Fondazione Policlinico Universitario Agostino Gemelli IRCCS, Roma, ITALY, ¹²Facility of Epidemiology and Biostatistics, Fondazione Policlinico Agostino Gemelli IRCCS, Rome, ITALY, ¹³Head of Gynecologic Oncology Unit, Department of Woman and Child Health and Public Health, Fondazione Policlinico Agostino Gemelli IRCCS, Rome, ITALY.

BackgroundPathologic complete response after neoadjuvant treatment is a strong prognostic biomarker associated with improved survival in breast cancer patients, especially in triple negative breast cancer and HER2 positive subtypes. Despite high response rates to neoadjuvant chemotherapy, triple negative breast cancer has been linked to higher rates of recurrence and poorer outcome. The aim of this study is to evaluate the genomic landscape of tissue samples from locally advanced breast cancer patients treated in a neoadjuvant setting and to correlate the genomic analysis to response and clinical outcome. Methods We retrospectively evaluated 48 core needle biopsies before chemotherapy from a cohort of 136 patients with locally advanced triple negative breast cancer. Pathological chemotherapy response and not responders was assessed by surgery. DNA extracted from archival formalin-fixed paraffin-embedded samples was subjected to sequencing using the TruSight Oncology 500 panel (Illumina, San Diego, USA; 523 genes, size: 1.94 Mb), following the manufacturer’s protocol, which assesses microsatellite instability status, tumor mutation burden, recurrent somatic copy number variations, somatic sequence variants as single nucleotide variants and Insertion/Deletion. Libraries were sequenced on NovaSeq 6000 Illumina platform to reach a minimum of 500× read depth. 67 (51.3%) out of 136 patients received a standard anthracycline and taxane regimen and 69 (50.7%) patients received a platinum-containing chemotherapy regimen. Somatic mutations classified as Oncogenic or Likely Oncogenic according to OncoKB with a variant allele frequency ≥0.03 were included. Differences of mutational profile were assessed using Chi-squared test. Statistical analysis was performed with R Studio (version 4.2.2). ResultsIn total we analyzed samples from 48 patients. The median age was 46 years. 22 of 48 (46%) were HER2 low, 42 (87%) with G3, 27 (56%) with stage II and 16 (33%) with stage III. 28 out of 48 patients (58.3%) not obtained pCR.We showed statistically significant correlations between PIK3CA mutations and PI3K pathway mutations with the absence of pathological complete response with p-values of 0.013 and 0.024 respectively. Genomic analysis of the entire cohort revealed the most frequently altered genes to be TP53 (94%), PIK3CA (21%), BRCA1 (15%), along with amplifications of MYC (23%) and GATA3 (17%) and MCL1 (15%). The most aberrant pathways in not pCR population are PI3K and RTK-RAS signaling.Alterations in PIK3CA was observed in 39% of not responders and absent in responder patients (p-value 0.05, reflecting its key role in PI3K-AKT-mTOR signaling pathway. The lollipop plots show how missense (H1047R, E542K, R88Q, V344G) and inframe deletion (G106_E109del, E110del) mutations are presents in not responder patients after neoadjuvant chemotherapy.Interestingly, median TMB in not responder’s cohort is 5.5 while in chemo sensitive patients is 4.8 (p-value 0.93).ConclusionsIn this preliminary report, PIK3CA mutations and related pathways may impact the effectiveness of neoadjuvant chemotherapy in achieving pCR. A larger sample size is needed to further evaluate the significance of mutations and co-alterations in different subgroups. We would like to acknowledge the contribution of the Multispecialistic Biobank Research Core Facility of Fondazione Policlinico Universitario A. Gemelli Biobank, for providing the bioresources.

D. Nguyen¹, L. Chen², V. Radic³, C. Ma¹, F. Racke³; ¹Molecular Oncology, Quest Diagnostics, San Juan Capistrano, CA, ²R&D, Rarity Biosciences, Uppsala, SWEDEN, ³Hematology and Oncology, Quest Diagnostics, San Juan Capistrano, CA.

Introduction The ESR1 gene encodes the estrogen receptor alpha (ERα), a transcription factor that regulates gene expression in response to estrogen. Mutations in ESR1 can lead to constitutive activation of ERα, allowing estrogen receptor-positive (ER+) breast cancer cells to proliferate even in the absence of estrogen; this proliferation contributes to resistance against standard hormone therapies such as treatment with aromatase inhibitors (AI), which function by reducing estrogen levels. ESR1 mutations are typically acquired during treatment and are most frequently observed in metastatic breast cancer. Detecting ESR1 mutations, often through circulating tumor DNA (ctDNA) in liquid biopsies, is critical for tracking disease progression and optimizing therapeutic strategies. As reported at the recent ASCO 2025 meeting, treatment switch before clinical progression based on the detection of ESR1 mutations in ctDNA can potentially improve outcomes in HR+/HER2- advanced breast cancer. Thus, identifying ESR1 mutations enables clinicians to make informed treatment decisions and supports the development of targeted approaches to overcome endocrine resistance. Molecular assays such as digital PCR (dPCR), real-time PCR (RT-qPCR), and next-generation sequencing (NGS) are commonly employed for detection of ESR1 mutations, with each platform offering distinct advantages in sensitivity and mutation coverage. However, a common limitation of these technologies is their sensitivity, which is compounded by the limited quantity of ctDNA that can be isolated from human specimens. To address these limitations, we developed a novel assay specifically designed for the detection of ESR1 mutations; the assay leverages the high specificity and sensitivity of the SuperRCA® technology developed by Rarity to enable accurate identification of low-frequency ESR1 variants. Methods The SuperRCA® uses 2 consecutive rolling circle amplification (RCA) reactions, which enable highly specific genotyping and accurate quantification of low-frequency variants. The sensitivity of the assay for ESR1 was assessed by testing a quantitative, serial dilution of synthetic controls of commonly found resistant variants and numerical positive signals detected using flow cytometry. Assay specificity was assessed by the number of positive signals from cfDNA extracted from normal individuals. Results The assay was capable of detecting rare ESR1 mutations at a sensitivity of 0.01% VAF down to 0.001% VAF. The number of positive signals was essentially zero for normal individuals , indicating high specificity. Conclusions These results demonstrate the SuperRCA® can be used to detect clinically relevant ESR1 variants, which may offer a powerful tool for improving the management of endocrine-resistant breast cancer. References: [1] ESR1 mutations in HR+/HER2-metastatic breast cancer: Enhancing the accuracy of ctDNA testing. Konstantinos Venetis et al. Cancer Treat Rev. 2023. PMID: 37864956[2] ESR1 mutations and therapeutic resistance in metastatic breast cancer: progress and remaining challenges. Sarah K Herzog et al. Br J Cancer. 2022. PMID: 34621045 [3] ESR1 mutations-a mechanism for acquired endocrine resistance in breast cancer. Rinath Jeselsohn et al. Nat Rev Clin Oncol. 2015. PMID: 26122181 [4] Camizestrant + CDK4/6 inhibitor (CDK4/6i) for the treatment of emergent ESR1 mutations during first-line (1L) endocrine-based therapy (ET) and ahead of disease progression in patients (pts) with HR+/HER2 – advanced breast cancer (ABC): Phase 3, double-blind ctDNA-guided SERENA-6 trial. Nicholas C. Turner et al. ASCO 2025. Abstract LBA4

D. Kizub¹, A. Gutierrez², R. K. Murthy¹, C. Albarracin³, B. K. Arun¹; ¹Breast Medical Oncology, MD Anderson Cancer Center, Houston, TX, ²Breast Medical Oncology-Research, MD Anderson Cancer Center, Houston, TX, ³Pathology, Anatomical, MD Anderson Cancer Center, Houston, TX.

Background and objective: HER2 0 compared to HER2 low (1+ or 2+) is associated with worse survival in HER2-negative metastatic breast cancer (MBC). We explored the impact of HER2 status on survival in patients withe HER2-negative breast cancer (HER2-) and germline BRCA1, BRCA2, and PALB2 mutations. Methods: We evaluated outcomes among women with HER2- patients with breast cancer treated at MD Anderson Cancer Center from August 1996 to May 2024. Patient characteristics were analyzed using descriptive statistics and standard tests of association. Cox proportional hazard models evaluated factors associated with overall survival (OS). Results: PALB2. Among 93 patients with PALB2, median age at diagnosis was 49.5 years old (st.dev 10.5). 20 (21.5%) had Stage I, 21 (22.6% Stage II), 18 (19.4%) Stage III, 4 (4.3%) de novo Stage IV disease. ER was weakly positive in 4 (4.3%), negative in 28 (30.1%), positive in 61 (65.6%). PR was negative in 41 (45.2%), positive in 51 (54.8%). 48 (51.6%) were HER2 0, while 45 (48.3) were HER2 low (1+ or 2+). Of the 48 patients who were HER2 0, 5 died at last follow-up, with OS 89.6%. Of the 45 patients with HER2 low, 4 died with OS 91.1%. Median OS for HER2 0 patients was 130.06 months (95% CI 114.65, 145.465 months) compared to 261.96 months for HER2 low (95% CI 219.21, 297.89) (p < 0.05). BRCA1 and BRCA2. For 752 patients with BRCA1 and 2, median age was 44.5 years (st.dev 11.20). 224 (29.8%) had Stage I disease, 332 (44.1%) Stage II, 144 (19.1%) Stage III, and 42 (5.6%) de novo Stage IV. ER was negative in 374 (49.7%), positive in 377 (50.1%). PR was negative in 460 (61.2%) and positive in 289 (38.4%). HER2 was 0 in 443 (58.9%), 1+ in 208, 2+ in 101 (13.4%). BRCA1. Among 424 patients with BRCA1, 266 (62.7%) were HER2 0 and 158 HER2 low. Among 266 who were HER2 0, 49 died, with OS 81.6%. Among those who were HER2 low, 26 died, with OS 83.5%. Median OS was 289.34 (95% CI 229.66, 548.80) for patients with HER2 0 compared to 175.54 (95% CI 158.05,193.03) for HER2 low (p<0.05). BRCA2. Among 327 patients with BRCA2, 177 (54.1%) were HER2 0 and 150 HER2 low. Among 177 who were HER2 0, 26 died with OS 85.3%. Among those with HER2 low, 25 have died with OS 83.3%. For Her2 0, median OS was 384.46 months (95% CI 199.54, 441.99) compared to 172.927 for HER2 low (95% CI 107.98, 291.4) (p > 0.05). Conclusions: HER2 0 was associated with worse OS compared to HER2 low in patients with PALB2 mutations and with better OS compared to HER2 low in patients with BRCA1 and BRCA2, though the association for BRCA2 was not statistically significant. Future studies should address the impact of targeted therapy and tumor biology in these populations.

M. Matossian¹, M. Bungert¹, C. Nguyen¹, G. Yerradoddi², M. East³, D. Okumu³, R. Nanda¹, G. Johnson⁴, M. Rosner⁵; ¹Medicine, University of Chicago, Chicago, IL, ²Bioinformatics, University of Chicago, Chicago, IL, ³Pharmacology, University of North Carolina, Chapel Hill, NC, ⁴Pha, University of North Carolina, Chapel Hill, NC, ⁵Pathology, University of Chicago, Chicago, IL.

Background: Metastatic triple-negative breast cancer (mTNBC) is one of the most challenging forms of breast cancer to treat due to profound heterogeneity and drug resistance. There have been very limited advancements in approval of targeted treatment regimens that are not chemotherapy-based for mTNBC in the past decade. Kinases are tractable therapeutic targets, but while kinase inhibitors have improved outcomes in many metastatic solid tumors including other breast cancer subtypes, they have not improved patient outcomes in TNBC clinical trials. We posit this is because we do not leverage understanding of complex kinase signaling networks and how we utilize kinase inhibitors in TNBC is not optimized. Further, high doses of kinase inhibitors to achieve cytotoxicity have unacceptable side effects for patients limiting real-world use. Chronic stress states such as those initiated by hypoxia, poor nutrient delivery and higher metabolic demands promote pro-metastatic transcriptional programs in TNBC. Our preclinical research and secondary analyses from clinical trial data reveal stress kinase signaling networks, and not individual pathways, drives TNBC metastatic transformation. Our central hypothesis is the RKIP/BACH1 signaling axis is a regulator of the chronic stress response to suppress TNBC metastasis and can be leveraged to produce biomarkers for risk stratification and a novel combination regimens. Methods: Prior work showed activation of the physiologic metastasis suppressor RKIP effectively reduced TNBC metastasis in cell line models through inhibition of stress MAPK networks (p38, JNK, ERK1/2, MLK) resulting in decreased transcription of the pro-metastatic transcription factor BACH1. We utilized mass spectrometry-based techniques with a kinase inhibitor system (MIB-MS) to characterize the activity of kinase signaling networks in heavily treated mTNBC organoids and examined kinase networks prominent in BACH1-high TNBC from publicly available phosphoproteomics datasets. We utilize live/dead, proliferation assays and transwell invasion assays to test efficacy of our stress kinase inhibitor regimen. Results: We show that combinations of 3 kinase inhibitors used in early phase clinical trials (Ralimetinib (p38), JNK-in-8 (JNK) or FDA approved (Trametinib (MEK & ERK1/2)) inhibit stress kinase networks, pro-motility gene expression and TNBC cell invasion. Low dose inhibition of multiple kinase nodes of networks, rather than single kinase nodes, more effectively reduces oncogenic signaling output and prevents compensatory signaling activation and potentiates suppression of organoid tumor initiating capacity when used in combination with chemotherapy. Using patient-derived organoids that recapitulate heterogeneous mTNBC molecular characteristics we successfully generated stable BACH1-knockout organoids which will be used to further identify BACH1-dependent critical kinase networks. Conclusions: Inhibition of these stress kinase networks should enable effective reprogramming of metastatic or high-risk cancers to more benign states and make tumors more susceptible to approved anti-proliferative therapies such as chemotherapy or immunotherapy while limiting toxicities from kinase inhibitors using a low-dose, multi-drug combination approach.

N. Vidula¹, D. Giannarelli², A. Putur¹, L. Pontollilo³, E. Nicolo³, C. Reduzzi³, E. Podany⁴, G. Cunningham⁴, A. Davis⁴, E. Blouch¹, M. Cristofanilli³; ¹Hematology and Medical Oncology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, ²Oncology, Instituto Nazionale Tumori Regina Elena, Rome, ITALY, ³Hematology and Medical Oncology, Cornell University, New York, NY, ⁴Hematology and Medical Oncology, Washington University School of Medicine, St. Louis, MO.

Background: BRCA mutations may impact outcomes in metastatic breast cancer (MBC). While germline and somatic BRCA mutations have been studied, the incidence and impact of cell-free DNA (cfDNA) BRCA copy number loss and/or deletions (CNL/D) are not known. We evaluated cfDNA BRCA CNL/D, and the association with clinical/genomic features in MBC. Methods: Patients with MBC with cfDNA (Guardant360 next-generation sequencing) results at 3 academic centers in the Precision Medicine Action for Cancer consortium from 12/2021 (when Guardant360 began reporting BRCA CNL/D) to 1/2024 were identified. A retrospective analysis was conducted to identify clinical/genomic features associated with cfDNA BRCA1 and/or 2 CNL/D. Median therapy durations and overall survival (OS) after detection of cfDNA BRCA CNL/D were estimated with the Kaplan-Meier method. Results: Among 1,198 patients with MBC who had cfDNA testing, 65 (5.4%) had cfDNA BRCA CNL/D. In the cohort with cfDNA BRCA CNL/D, median age at MBC diagnosis was 60 years. Subtypes included 47/65 (72%) hormone receptor (HR)+/HER2-, 15/65 (23%) triple negative, and 3/65 (4.6%) HER2+. Prior to cfDNA testing, patients received a median of 1 hormone therapy (HR+/HER2-) and 0 chemotherapy agents (all subtypes) for MBC. Table 1 shows the types of cfDNA BRCA CNL/D that were detected. Of patients with cfDNA BRCA CNL/D, 16/65 (25%) were BRCA1, 37/65 (57%) BRCA2, and 12/65 (18%) BRCA1&2. Thirteen of 65 (20%) patients had a co-existing pathogenic cfDNA BRCA mutation, 40/65 (62%) had a cfDNA TP53 mutation, and 43/65 (66%) had cfDNA high tumor mutation burden (TMB) >/= 10 mut/Mb. Ten of 32 (31%) patients with tumor tissue genotyping results had a concomitant tissue BRCA1/2 mutation. Five of 45 (11%) patients with germline genetic testing results had a concomitant germline BRCA1/2 mutation. Median duration on first therapy post detection of cfDNA BRCA CNL/D was 94 days (95% confidence interval (CI): 54-134 days). Post detection of cfDNA BRCA CNL/D, 5 patients received platinum chemotherapy with a median response duration of 168 days (range 32-582 days), and 4 patients received a PARP inhibitor with a median response duration of 105 days (range 26-231). Median OS after detection of BRCA CNL/D was 490 days (95% CI: 232-748 days). Conclusions: CfDNA BRCA CNL/D were detectable in a subset of patients with MBC, with most present in non-germline BRCA carriers. CfDNA BRCA CNL/D co-occurred frequently with TP53 mutations and elevated TMB, suggesting the possibility that cfDNA BRCA CNL/D could arise from increased mutagenesis in MBC. Further research is needed to define the impact of platinum and PARP inhibitor therapy in this cohort.

W. Chen, Q. Shao, Z. Wang, B. Zhu; Department of Breast Surgery, Changzhou No.2 People’s Hospital, the Third Affiliated Hospital of Nanjing Medical University, Changzhou, CHINA.

Background: Macrophages, the predominant immune-related stromal cells within tumor microenvironment, could be polarized into M1 and alternative M2 macrophages. M1 macrophages are pro-inflammatory and eliminate cancer cells, whereas M2 macrophages are immunosuppressive and facilitate tumor growth. Mounting evidence has shown that M2 macrophages are responsible for a poor prognosis in triple negative breast cancer (TNBC) patients. Since neoadjuvant chemotherapy is becoming the foundation treatment for TNBC, the relationship between M2 macrophages and response to neoadjuvant chemotherapy remains to be elucidated. Methods: Infiltration of M2 macrophages were checked by analyzing the CD163 levels on surgical resection specimens from 42 Chinese TNBC patients treated with neoadjuvant chemotherapy. Evaluation of M2 macrophages together with other immune markers including CD8-positive cytotoxic T cells and PD-L1 expression in TNBC and its association with pCR after neoadjuvant chemotherapy were performed. Results: Pathological complete response (pCR) was significantly associated with younger age, higher tumor or stromal PD-L1 expression, higher levels of tumor or stromal CD8-positive cytotoxic T cells, but lower infiltration of CD163-positive M2 macrophages, with the level of CD163-positive M2 macrophages in stromal area as the strongest factor. Conclusions: Our study showed that high infiltration of CD163-positive M2 macrophages was negatively associated with the chemotherapeutic response to neoadjuvant chemotherapy in Chinese TNBC patients and could be used as a promising prognostic candidate for Chinese TNBC patients treated with neoadjuvant chemotherapy.

W. Chen, Q. Shao, Z. Wang, B. Zhu; Department of Breast Surgery, Changzhou No.2 People’s Hospital, the Third Affiliated Hospital of Nanjing Medical University, Changzhou, CHINA.

Background: Triple negative breast cancer (TNBC) is a highly aggressive subtype distinguished by the lack of hormone receptors and specific molecular targets, leading to limited treatment options, high recurrence rates, and poor clinical prognosis. The Warburg effect in cancer cells generates a tumor microenvironment characterized by hypoxia, low glucose, and high lactate, which severely promote chemo-resistance, immune escape, and distant metastasis. Nevertheless, prognostic significance of hypoxia and lactate metabolism-related genes in TNBC remains largely unclear, and more accurate prediction models are urgently needed. Methods: Clinical information and gene expression profiles of TNBC patients were obtained from the online database. LASSO and multivariate Cox regression analysis were used to establish the hypoxia and lactate metabolism-related prediction model, and prognostic value was further validated by Kaplan-Meier plotter, receiver operating characteristic curves, and nomogram. Gene set enrichment analysis was performed to evaluate the pathways and molecular functions of hypoxia and lactate metabolism-related gene signature. Tumor mutation, microsatellite instability, RNA expression-based stemness scores, immune infiltration, and drug susceptibility analysis were applied to identify potential therapeutic targets. A series of in vitro experiments were performed to evaluate the biological function of the selected gene. Results: Five genes including COL5A3, LRRC8D, IGFL1, SEPTIN3, and PEG10 were identified to establish the hypoxia and lactate metabolism-related risk score. TNBC patients divided into high-risk or low-risk group by risk score exhibited substantial differences in survival outcome, tumor mutation, immune infiltration, and drug sensitivity. Gene set enrichment analysis revealed that difference between these two groups may be related to extracellular matrix, intercellular links, intercellular information transfer, immune-related pathways, angiogenesis and metabolism. The constructed model showed strong predictive performance, with receiver operating characteristic curves showing area under the curve values of 0.82, 0.86, and 0.90 at 1, 3 and 5 years, respectively. In vitro studies indicated that hub gene COL5A3 could promote tumor proliferation, migration, and drug resistance via PI3K/AKT pathway in TNBC cells. Conclusion: A novel five hypoxia and lactate metabolism-related gene signature could be used for prognosis prediction in TNBC patients, and hub gene COL5A3 promotes TNBC progression via PI3K/AKT pathway. Further research is needed to explore the biological roles of these genes in TNBC aiming to refine therapeutic approaches.

A. Nakakami¹, Y. Tokumaru¹, Y. Niwa¹, R. Mori¹, M. Okawa¹, Y. Sato², H. Hatakeyama³, T. Hirotsu⁴, E. di Luccio³, N. Matsuhashi², M. Futamura¹; ¹Breast surgery, Gifu university Hospital, Gifu, JAPAN, ²Gastroenterological Surgery and Pediatric Surgery, Graduate School of Medicine, Gifu University, Gifu, JAPAN, ³R&D department, Hirotsu Bio Science Inc., Tokyo, JAPAN, ⁴Head office, Hirotsu Bio Science Inc., Tokyo, JAPAN.

(Background)Breast cancer (BC) remains a leading cause of cancer-related deaths despite advances in diagnosis and treatment. Accurate evaluation of response to neoadjuvant chemotherapy (NAC), especially in HER2-positive and triple-negative subtypes, is critical. In recent years, liquid biopsy techniques, which analyze tumor-derived materials in body fluids, have emerged as promising non-invasive tools for cancer detection and monitoring. Nematodes-NOSE (N-NOSE) is a novel cancer screening test that utilizes the chemotaxis index of the nematode Caenorhabditis elegans (C. elegans). The N-NOSE method is a non-invasive diagnostic approach that falls within the broader category of liquid biopsy, which analyzes tumor-derived materials in body fluids. C. elegans possesses a highly sophisticated olfactory system and has been reported to exhibit attraction toward the urine of cancer patients while avoiding that of healthy individuals. In the current study, we aimed to clarify the ability of the N-NOSE method to predict the response to NAC in BC patients. (Materials and Methods)A total of 36 BC patients scheduled to undergo NAC followed by surgery at our institution between August 2020 and May 2023 were prospectively enrolled in this study. All patients were histologically confirmed to have BC and were deemed eligible for NAC based on clinical staging and risk stratification. Urine samples were collected from each patient at three distinct time points: prior to the initiation of NAC, immediately before surgery, and four weeks after surgery. For chemotactic assays, 0.5 μL of 1 M sodium azide, an anesthetic used to minimize the effects of adaptation, was spotted onto four points on the agar plate (urine side and non-urine side), and 1 μL of urine sample diluted 100-fold with ultra-pure water was added to two points (urine side). Approximately 100 adult nematodes were placed in the center of the plate. After roaming for 30 minutes, the number of nematodes present in area A (urine side) and the number of nematodes present in area B (non-urine side) were counted. The chemotaxis index (CI) was calculated using the following equation: Index = (Number of nematodes in area A – Number of nematodes in area B)/Total number of nematodes. The change in CI (Index Reduction Score, IRS) was calculated for three intervals: before and after NAC (IRS1), before and after surgery (IRS2), and before NAC and after surgery (IRS3). The association between IRS and treatment efficacy was statistically evaluated using receiver operating characteristic (ROC) curve analysis. (Results)Among the 36 patients enrolled, the median age was 51 years (range: 35-77). Thirteen patients (36.1%) achieved pathological complete response (pCR), while sixteen patients (44.4%) achieved pathological partial response (pPR). When the treatment response included CR or PR, the area under the ROC curve (AUC) for IRS1, IRS2, and IRS3 were 0.53 (95% CI: 0.19-0.78), 0.76 (95% CI: 0.56-0.96), and 0.66 (95% CI: 0.37-0.96), respectively. Notably, when only pCR cases were analyzed, the AUCs were 0.58 (95% CI: 0.34-0.82) for IRS1, 0.64 (95% CI: 0.40-0.88) for IRS2, and 0.75 (95% CI: 0.54-0.95) for IRS3, indicating that IRS3 had the highest predictive value for pCR. In this study, some pCR cases exhibited minimal residual intraductal disease before surgery, which was subsequently removed during surgery, potentially resulting in a higher IRS3 compared to IRS1. (Conclusion)The IRS measured by the N-NOSE method may accurately reflect the efficacy of NAC in BC patients. Furthermore, N-NOSE may be able to detect minimal or subclinical residual tumors after NAC, which are usually grouped clinically but may differ in their biological characteristics. Further large-scale, multicenter prospective studies are warranted.

H. Y. Xiang¹, H. Y. Xiang², X. Li¹, C. Wang¹, T. A. Martin¹, E. Davies³, K. Mokbel¹, Y. H. Liu², W. G. Jiang¹; ¹School of Medicine, Cardiff University, CARDIFF, UNITED KINGDOM, ²Department of Thyroid and Breast Surgery, Peking University First Hospital, Beijing, CHINA, ³Wales Breast Centre, University Hospital Llandough, Cardiff, UNITED KINGDOM.

Introduction. Desmoplakin (DSP) is a membrane subcoat plaque protein, primary co-located within the desmosomes, which are membrane structures found in tissues such as skin and muscles. Known DSP interactive proteins include desomosome cadherin desmogleins (DSG)and desmocollins (DSC). Although an early indication of a possible loss of the DSP protein, particularly in high grade tumours in human breast cancer was reported (Davies et al 1999), the impact of DSP on the clinical course and prognosis in breast cancer has not been fully explored. Here, we investigated the expression of DSP together with the membrane integral demososome protein DSG1 (desmoglein-1), aiming to establish a possible prognostic significance in clinical breast cancer. Methods and Results. The expression of DSP gene transcript was quantitatively evaluated in a cohort of clinical breast cancer with a ten-year followup. The expression levels were fully analysed against the clinical and pathological factors and the long term clinical outcome of the patients. Where appropriate, the prognostic significance was conducted using Kaplan-Meier model and Bayes analyses. The most significant finding was that the expression levels of DSP is linked to the clinical outcome of the patients. Breast tumours which developed distant metastasis, local recurrence and in patients who died from breast cancer exhibited a significant reduction of DSP expression when compared with tumours from those who remained disease free (p=0.036, p=0.020 and p=0.022, respectively, versus disease free). The reduction of DSP was seen profoundly in ER positive tumours and HER2 positive tumours. Survival analysis indicated that DSP expression is a favourable factor for overall survival (>OS) with an AUC at 0.241 (p<0.0001) and hazard ratio (HR) at 0.063 (95% confidence internal 0.008-0.477, p=0.007 by Cox regression), a connection supported by an impressive Bayes factor (192.1). The connection with OS was observed across the molecular subtypes and when the molecular subtypes and hormone receptor status were considered, DSP remained a significant OS indicator in multivariant analysis. Very similarly, DSP expression is also a favourable indicator for disease free survival (DSF) (p=0.006), again seen in most subtypes. Significant correlation between expression pattern of DSP and DSG1 was not observed. Finally, evidence suggests that DSP expression level is a good indicator for patient response to both neoadjuvant and adjuvant chemotherapies in patients with breast cancer.Conclusion. Desmoplakin exhibits characteristics as a favourable prognostic indicator for patient overall and disease free survival in breast cancer patients. The connection, which is less influenced by the molecular subtypes, also has value in predicting the chemotherapeutic response of the patients. Reference:Davies EL et al. 1999. The immunohistochemical expression of desmoplakin and its role in vivo in the progression and metastasis of breast cancer. Eur J Cancer, 35, 902

B. Walbaum, A. Morell, J. Muñoz-Carrillo, L. Arenas, A. Del Val, M. Rey, P. Gimenez-Xavier, N. Chic, F. Brasó-Maristany, E. Segui, R. Gómez-Bravo, I. García-Fructuoso, T. Pascual, F. Schettini, M. Bergamino, M. González-Rodríguez, B. Adamo, M. Muñoz, E. Sanfeliu, A. Prat, O. Martínez-Sáez, M. Vidal; Oncology, Hospital Clinic Barcelona, Barcelona, SPAIN.

Background: Over half of women diagnosed with breast cancer (BC) before the age of 45 fall within the postpartum window, generally defined as a diagnosis made up to 10 years after last delivery. Postpartum has been associated with poorer outcomes and may be a key determinant of the worst prognosis described among young BC patients. Diagnostic delays and biological changes in tumor microenvironment could explain this increased risk. Moreover, delayed childbearing can intensify the postpartum risk and forestall the long-term protective effects of pregnancy. We herein present a retrospective analysis of postpartum-associated BC (PPBC) from a single institution.Methods: Retrospective analysis including a cohort of 223 women under 45 years diagnosed at the Hospital Clinic of Barcelona (2010-2023) with complete pregnancy history. Patients were stratified into PPBC (n=113) defined as BC diagnosed postpartum and up to 10 years after last delivery, and non-PPBC (n=110), either nulliparous or diagnosed after 10 years of last delivery. Analyses evaluated disease-free survival (DFS), distant disease-free survival (DDFS), and overall survival (OS) in early-stage patients. Subgroup analyses in PPBC explored the effect of time since delivery (30 years). PAM50/PROSIGNA was used to determine intrinsic subtypes and risk of recurrence (ROR). Clinicopathological features and survival outcomes were analysed using Chi-square tests, Kaplan-Meier, and Cox regression models.Results: PPBC cases were slightly older at diagnosis than non-PPBC (median 38.0 vs. 36.8 years, p=0.017). Stage distribution was similar, though stage IV disease at diagnosis was numerically higher in PPBC (9.7% [n=11] vs. 5.5% [n=6], p=0.262). Immunohistochemistry (IHC)-based subtypes were comparable (p=0.857), with 59.3% of PPBC and 57.3% of non-PPBC patients having estrogen receptor-positive /HER2-negative (ER+/HER2−) subtype. No other significant differences were observed regarding tumor size, nodal status, histologic type or grade, or Ki67 index. Among patients with available PROSIGNA (n=65; 35 PPBC and 30 non-PPBC), no significant differences for intrinsic subtypes (p=0.892) or ROR categories (p=0.344) were found.At a median follow-up of 85 months, 7-year DFS and DDFS were numerically lower in PPBC compared to non-PPBC (DFS: 75.1% vs. 83.3%; DDFS: 76.2% vs. 89.5%), though only DDFS approached statistical significance (hazard ratio [HR] 1.90, 95% CI 0.976-3.702; p=0.054). Noteworthy, OS was significantly reduced in PPBC patients (7-year OS: 89.6% vs. 95.6%; HR 3.01, 95% CI 1.00-9.79; p=0.042), with persisting differences in the full cohort including de novo stage IV cases (HR 3.4, 95% CI 1.23-9.38; p=0.01). Within PPBC, those diagnosed <5 years postpartum had numerically worse DFS (HR 1.72, 95% CI 0.84-3.53; p=0.136), while no survival differences were observed by age at first pregnancy (p=0.610).In multivariate Cox regression models excluding metastatic cases and adjusting for age at diagnosis, tumor size (cT 2-4 vs. 1), nodal status (cN 1-2 vs. 0), and IHC subtype, PPBC was independently associated with worse DFS (HR 2.63, 95% CI 1.18-5.89; p=0.019) and DDFS (HR 2.82, 95% CI 1.24-6.45; p=0.014). Tumor size was also an independent predictor of recurrence across all models. Regarding OS, PPBC remained an independent risk factor (HR: 6.96 95% CI 1.58-30.71; p=0.010), while increasing age was associated with better survival (HR: 0.81 95% CI 0.68-0.96; p=0.018).Conclusions: PPBC is an independent predictor of poor survival among young women. In the absence of differences in clinical stage or tumor subtype, these findings suggest that postpartum-specific biological mechanisms may underlie the worse outcomes observed. Further investigation is needed to refine risk stratification and improve care in this distinct population.

M. Pannone, A. Hettich, R. Vestermark, M. Karsdal, N. Willumsen; Oncology, Nordic Bioscience, Herlev, DENMARK.

Background: Extracellular matrix (ECM) remodeling is a hallmark of tumor progression, driven in part by cancer-associated fibroblasts (CAFs). The N-terminal pro-peptide of type III collagen (PRO-C3), released during collagen formation, is a circulating biomarker reflecting active tumor fibrosis (Nissen et al., 2022). Elevated pre-treatment PRO-C3 levels have been associated with poor outcomes in metastatic HER2+ breast cancer treated with Trastuzumab (Lipton et al., 2018). Additionally, in a Phase II trial of tetrathiomolybdate in high-risk triple-negative breast cancer, patients who remained disease-free exhibited lower PRO-C3 levels (Liu et al., 2021), suggesting a link between reduced collagen formation and improved outcomes.Methods: To investigate the cellular origin and regulation of type III collagen production in breast cancer, we quantified PRO-C3 in conditioned medium (by competitive ELISA) using primary breast CAFs and primary benign breast fibroblast cultured in Ficoll-containing media with ascorbic acid, according to the in vitro Scar-in-a-jar fibroblast model system (Chen et al., 2019). Fibroblasts were stimulated with profibrotic (TGF-β1, PDGF-AB) or inflammatory (IL-1α, IL-6) cytokines, and treated with antifibrotic agents (ALK5i, Fresolimumab) to assess potential changes in PRO-C3.Results: PRO-C3 was elevated intrinsically in CAFs compared to benign fibroblast. Breast CAFs exhibited high basal PRO-C3 levels and with minimal response to TGF-β1, while benign fibroblasts showed low basal levels with marked induction upon stimulation with TGF-β1. TGF-β1 was the dominant driver of PRO-C3 compared to the other cytokines. Both antifibrotic treatments effectively suppressed PRO-C3 to baseline in both CAFs and TGF-β1-induced benign fibroblast.Conclusions: PRO-C3 reflects both intrinsic and inducible fibrotic activity in breast CAFs and may serve as a surrogate marker of fibroblast activation in breast tumors. PRO-C3 is a clinically relevant biomarker of tumor fibrosis in breast cancer, with elevated circulating levels associated with poor prognosis. Our data support a fibroblast-derived origin of PRO-C3 and highlight TGF-β1 as a key upstream regulator. The observed findings underscore the need for individualized anti-fibrotic strategies in breast cancer management. Integration of PRO-C3 in clinical workflows could enable non-invasive monitoring of fibroblast activity and fibrotic burden, informing prognosis and therapeutic response.References: Nissen et al., 2022, PMID: 35159087Lipton et al., 2018, PMID: 29923614Liu et al., 2021, PMID: 34426581Chen et al., 2019, PMID: 19785660

D. J. Wiisniewski¹, D. J. Voeller¹, Y. A. Addissie¹, S. K. Deshmukh², S. K. Wu³, M. B. Lustberg⁴, D. Wangsa⁵, D. Wangsa⁵, K. Heselmeyer-Haddad⁵, G. Sledge⁶, S. Lipkowitz¹; ¹Women’s Malignancies Branch, Center for Cancer Research, NCI, NIH, Bethesda, MD, ²Clinical and Translational Research, Caris Life Sciences, Phoenix, AZ, ³Clinical and Translational Research, Caris Life Sciences, Irving, TX, ⁴Medical Oncology, Yale Univerrsity School of Medicine, New Haven, CT, ⁵Genetics Branch, Center for Cancer Research, NCI, NIH, Bethesda, MD, ⁶Chief Medical Officer, Caris Life Sciences, Irving, TX.

Background: Epidermal Growth Factor Receptor (EGFR) family signaling is commonly dysregulated in cancer by amplification or activating mutations. Although studies have investigated dual EGFR/PI3K inhibition in breast cancer, they have not determined biomarkers that predict success. Here, we present a subset of patients with EGFR amplification and PI3Kinase pathway mutations in breast cancer that can be synergistically targeted by dual EGFR/PI3K inhibition. Methods: Frequency of EGFR amplification and PI3K pathway mutations (PIK3CA, PIK3R1/2, PTEN, AKT1/2/3, TSC1/2/3) were studied through datasets from cBioPortal and Caris Life Sciences. Triple negative breast cancer (TNBC) cell lines used include BT20 (amplified EGFR, PIK3CA activating mutation), MDA-MB-468 (amplified EGFR, PTEN deletion), and MDA-MB-231 (no EGFR or PI3K alterations). EGFR amplification was determined by immunoblot and fluorescent in situ hybridization, and PI3K mutations by sequencing. Signaling was determined by immunoblot, drug synergy by cell viability, cell death by propidium iodide staining, cell cycle analysis by flow cytometry, and xenograft animal study. Caris Life Science CODEai was used to evaluate real-world overall survival (OS) obtained from insurance claims and calculated from date of tumor biopsy to last contact using Kaplan-Meier estimates. Statistical significances were determined by chi-square and Mann-Whitney U test, student’s t-test with 2 tailed comparisons assuming equal variance, one-way or two-way ANOVA were performed as indicated. P-values of ≤ 0.05 were considered significant. Results: EGFR amplification occurs in approximately 1-2.5% of breast cancer patients, more frequently in TNBC (2.45-6.7%) and ER-/HER2+ (1.3-6.5%) breast cancers, and in the molecular basal (2.33-8.1%) and HER2 enriched (1.87-5.4%) subtypes. EGFR amplified patients in cBioPortal and in the Caris datasets had worse median OS compared to those with non-amplified EGFR (mOS: 111.1 m vs 164.6 m, p=0.0269, and 21.6 m vs 32.9 m, HR 0.69, 95% CI 0.58 – 0.82, p<0.0001 for the cBioPortal and Caris data, respectively). cBioPortal and Caris databases showed that 54-71% of EGFR amplified tumors have activating mutations in the PI3K pathway. Our in vitro studies showed combination EGFR and PI3K inhibitors more dramatically reduced MAPK, mTOR and AKT signaling in the BT20 and MDA-MB-468 cells, whereas the inhibition of downstream signaling was less significant in MDA-MB-231 cells. Combination of EGFR and PI3K inhibitors reduced cell viability in these three cell lines, but inhibition was greater, statistically significant, and synergistic in MDA-MB-468 and BT20 compared to MDA-MB-231. Only MDA-MB-468 and BT20 cells had an increased fraction of apoptotic cells with combined pathway inhibition. EGFR or PI3K inhibition alone in a BT20 xenograft model reduced tumor volume, however the combination induced the only statistically significant reduction in tumor volume when compared to vehicle control. Conclusions: EGFR/PI3K inhibitor combination causes apoptosis and reduction in tumor growth in cells with EGFR amplification and PI3K alteration. EGFR amplification with coincident PI3K pathway mutations are drivers in a subset of breast cancers and identify a subgroup of breast cancers that are more likely to respond to dual targeted therapy.

H. Rugo¹, A. Schrock², J. Lee², R. Graf², M. Gearing³, A. Heilmann⁴, A. Gasco², N. Vasan⁵, J. Quintanilha²; ¹Department of Medical Oncology & Therapeutics Research, City of Hope, Duarte, CA, ²Clinical Development, Foundation Medicine, Inc, Boston, MA, ³Medical Affairs, Foundation Medicine, Inc, Boston, MA, ⁴Translational Oncology & Clinical Reporting, Foundation Medicine, Inc, Boston, MA, ⁵Department of Medicine, New York University, New York, NY.

Introduction: Genomic alterations in PIK3CA, AKT1 and PTEN are biomarkers for the AKT inhibitor Capivasertib (Capi), an approved treatment for metastatic BC (mBC) patients who progressed on ≥1 endocrine therapy or recurred during/within 12 months of adjuvant therapy. Several NGS assays include these genes in their panels; however, it is often unclear which types of alterations are evaluated, particularly PTENloss, and how the results compare across platforms. This study aimed to retrospectively compare two commercially available tissue NGS tests for the detection of short variant mutations (mut) in PIK3CA, AKT1, and PTEN, as well as homozygous PTEN copy loss (PTENloss) and PTEN rearrangements (re) in patients with BC: FoundationOne/FoundationOne®CDx (F1CDx®) and Caris NGS tests. Methods: This study included patients with BC who underwent F1CDx tissue comprehensive genomic profiling and had Caris tissue NGS testing. Data was obtained from the U.S.-wide de-identified Flatiron Health and Foundation Medicine real-world clinicogenomic breast database (CGDB), from ~280 U.S. cancer clinics (~800 sites of care) between 01/2011 and 12/2024. A comparison between F1CDx on-label Capi alterations and Caris NGS, abstracted from electronic health records (EHR)¹, was conducted for patients with both tests performed on tumor tissue specimens collected on the same day. Caris NGS specific mutations are not available, so we could not assess if they were on-label for Capi. Positive percentage agreement (PPA) was calculated using F1CDx as a reference. Caris IHC PTEN protein loss results abstracted from EHR were also assessed when available. Results: A total of 74, 85 and 80 patients had PIK3CA, AKT1, and PTEN NGS results available from both assays, respectively, and 68 patients also had a Caris IHC PTEN result available. For PIK3CAmut, we observed an agreement of 98.6 % (73/74), with 53 (71.6%) negative and 20 (27%) positive for both tests, and 1 patient F1CDx-/Caris+ (100% PPA). For AKT1mut, the agreement was 100% with all cases negative for both tests. For PTEN, a 95% (76/80) agreement was observed, with 75 (93.7%) being negative for both tests, 1 (1.25%) with both positive, and 4 (5%) F1CDx+/Caris- (20% PPA). No PTENloss or PTENre was reported by Caris NGS, while 3 patients (3.7%) were positive for PTENloss and 1 patient (1.25%) for PTENre by FMI (0% PPA). Of 8 patients with a PTEN alteration not reported by Caris NGS test and detected on F1CDx, 6 had an IHC PTEN protein loss reported by Caris. Additionally, 4 patients were identified to have PTEN protein loss by Caris’s IHC PTEN results, but no PIK3CA/AKT1/PTEN alterations were identified by either NGS test and two functional copies of PTEN were detected on F1CDx. Conclusions: 29% (8/28) BC patients with on-label Capi alterations detected with F1CDx were not reported by Caris NGS tissue testing. 89% (8/9) of PTEN alterations were not reported by Caris NGS tissue testing, including all PTENloss and PTENre. 75% (6/8) Caris IHC PTEN results reported PTEN protein loss (not a current Capi indication). Further investigation on variations in reporting of different NGS assays in both tissue and plasma and their ability to identify on-label alternations are needed to understand true differences between testing results. Footnotes¹ All data analyzed in this study related to the Caris tissue test is based on reported data only.

X. Ying, K. Zhang, X. Zhu, S. Jiang, X. Hu, C. Chen, Z. Wang; Department of Breast Surgery, Fudan University Shanghai Cancer Center, Shanghai, CHINA.

BACKGROUNDNeoadjuvant therapy (NAT) has emerged as a standard treatment strategy for locally advanced breast cancer (BC), yet effective predictive markers are still lacking. This study aimed to investigate genomic biomarkers of NAT efficacy in a Chinese BC cohort and characterize subtype-specific molecular signatures associated with differential treatment responses.METHODS A large-scale NAT cohort comprising 1,161 patients with primary BC was analyzed, including 1,145 patients who underwent comprehensive genomic profiling using the FUSCC-BC targeted panel. Mutational signatures predictive of pathological complete response (pCR) and metastatic recurrence were identified. Machine learning models were developed for NAT response prediction. RESULTSThe FUSCC-BC neoadjuvant cohort study integrated clinicopathological analysis and targeted sequencing of breast cancer subtypes (36.9% HER2+, 39.8% HR+/HER2-, 23.3% triple-negative) to identify genomic biomarkers associated with differential pCR rates. We revealed shared resistance mechanisms across subtypes, notably PIK3CA mutations and PI3K pathway activation predicting non-pCR in both HR+/HER2− and triple-negative breast cancer (TNBC), while ERBB2 and GRINA alterations emerged as a HER2-enriched subtype-specific predictor of therapeutic resistance. In patients who fail to achieve pCR, multivariate analysis revealed that multiple genomic alterations (e.g., TP53 and TOP2A) were independently associated with elevated risk of metastatic recurrence. To optimize clinical applicability, a machine learning framework integrating somatic mutation profiles and clinicopathological variables demonstrated robust predictive performance, achieving area under the curve (AUC) values of 0.82 in the training set and 0.81 in the test set for neoadjuvant therapy response prediction.CONCLUSIONOur study identified subtype-specific biomarkers predictive of the response of BC patients to neoadjuvant therapy and established a predictive framework to optimize NAT strategies, providing molecular guidance for personalized treatment regimens that may enhance clinical outcomes. 1

W. Cai, Y. Wang, Y. Qiu, S. Liu, Q. Cai, P. He, Y. Lin, M. Chen, L. Chen, F. Fu, C. Wang; Department of Breast Surgery, Fujian Medical University Union Hospital, Fuzhou, CHINA.

Background: Recently, perineural invasion (PNI) has emerged as a promising biomarker and a potential therapeutic target in cancer. PNI has been shown to provide prognostic value on locoregional recurrence and distant metastasis in the breast cancer. However, its role in the the progression of lymph node metastasis remains controversial. This study aims to explore the relationship between PNI and the extent of axillary lymph node metastasis (ALNM) in breast cancer. Methods: Patients with early-stage, operable invasive breast cancer who underwent surgery between June 2011 and June 2023 were enrolled. The patients were stratified into PNI-positive and PNI-negative groups based on postoperative pathology. To control for confounding factors and simulate a randomized controlled trial, we used propensity score matching (PSM), inverse probability of treatment weighting (IPTW), and overlap weighting (OW) methods between two groups. Multivariate logistic regression analysis was performed to evaluate the association between PNI and both ALNM and lymph node ratio (LNR). A meta-analysis was conducted to further validate the effect. Results: Among 4,156 early-stage operable breast cancer patients, 1,223 (29.4%) were PNI-positive. Multivariate logistic regression analysis revealed that PNI is independently associated with ALNM (P = 0.002) and a high-risk lymph node ratio (LNR > 0.2) (P = 0.012). These associations remained consistent after sensitivity analyses using PSM, IPTW, and OW methods (all P < 0.05). The meta-analysis included 16 eligible studies with 17,451 participants, further confirmed the significant association between PNI and ALNM (P < 0.001). Additionally, the meta-analysis demonstrated that PNI is linked to worse disease-free survival (DFS) (P < 0.001) and overall survival (OS) (P = 0.007) in patients with breast cancer. Conclusion: PNI is associated with a greater burden of ALNM in patients with early-stage operable invasive breast cancer, underscoring its potential value for risk stratification in the era of axillary surgery de-escalation.

M. Mego¹, J. Hanes², V. Parrak², G. Minarik³, M. Karaba⁴, B. Mravec⁵, P. Turcani⁶, K. Kalavska¹, L. Kucerova¹, N. Zilka⁷; ¹2nd Department of Oncology, Faculty of Medicine, Comenius University, National Cancer Institute, Bratislava, SLOVAKIA, ²Institute of Neuroimmunology, Institute of Neuroimmunology, Slovak Academy of Sciences, Bratislava, SLOVAKIA, ³Medirex Group Academy, Medirex Group Academy, Bratislava, SLOVAKIA, ⁴Department of Oncosurgery, National Cancer Institute Slovakia, Bratislava, SLOVAKIA, ⁵Department of Physiology, Faculty of Medicine, Comenius University, Bratislava, SLOVAKIA, ⁶1st Department of Neurology, Faculty of Medicine, Comenius University, Bratislava, SLOVAKIA, ⁷Institute of Neuroimunology, Institute of Neuroimunology, Slovak Academy Sciences, Bratislava, SLOVAKIA.

Background: Increased tumor innervation has been associated with more aggressive tumor behaviour and worse clinical outcomes across multiple cancer types. Emerging evidence suggests that heightened neural infiltration may lead to axonal damage. Neurofilaments (NfL), key structural components of neurons, have been proposed as biomarkers of neuronal damage, but their prognostic role in solid tumors remains unclear. This study aimed to assess the clinical significance of plasma neurofilament light chain levels in early-stage breast cancer.Methods: We prospectively analyzed plasma neurofilament light chain concentrations in 80 patients with non-metastatic breast cancer prior to systemic treatment that underwent surgery at National Cancer Institute, Slovakia from 2013 to 2015. NfL level was determined by ELISA. Associations with clinicopathological features and circulating tumor cells (CTCs) were evaluated using descriptive and inferential statistics. Survival outcomes were analyzed using Kaplan-Meier curves and log-rank tests, with a cut-off set higher than 95^th percentile of the normative distribution.Results: NfL light chain levels were not significantly associated with tumor size, nodal status, histological grade, hormone receptor, HER2, or Ki-67 status. However, higher NfL levels were significantly associated with negative Bcl-2 expression (mean 13.1 vs. 9.2 ng/mL, p = 0.012). Moreover, patients with high NfL plasma level had detectable CTCs with epithelial-mesenchymal transition phenotype compared to patients with low NfL level (66.7% vs. 25.4%, p = 0.01). Patients with low plasma neurofilament levels had significantly improved disease-free survival (HR = 0.23, 95% CI 0.05–1.08, p = 0.0006) and overall survival (HR = 0.19, 95% CI 0.03–1.16, p = 0.0005) compared to those with high levels. Multivariate analysis confirmed prognostic value of NfL light chain levels independently of tumor stage, nodal status, oestrogen receptor and/or HER2 status.Conclusions: Plasma NfL light chain level concentration is a promising prognostic biomarker in non-metastatic breast cancer, independently of conventional pathological features. Its association with poor survival and CTC EMT phenotype suggests a potential link to tumor aggressiveness and dissemination potential

O. A. KHARENKO¹, K. Norek¹, J. Kennard¹, K. Fuh¹, R. D. Shepherd¹, K. D. Rinker²; ¹Syantra Inc, Syantra Inc, Calgary, AB, CANADA, ²Syantra Inc, Syantra Inc, University of Calgary, Calgary, AB, CANADA.

Background: Tumour-induced immune reprogramming is a key mechanism by which cancers can evade immune surveillance and promote disease progression. In recent in vitro studies, we demonstrated that triple-negative breast cancer (TNBC) cells can modify gene expression in THP1 monocytes, through phenotypic and transcriptomic changes consistent with immune dysfunction. These effects included activation of oncogenic signalling pathways, upregulation of IL6, and enhancement of proliferative signals. Here we report on molecular changes in white blood cells and platelets in women with breast cancer. Methods: Blood samples were collected from participants in the IDBC clinical study (clinicaltrials.gov/study/NCT04495244) prior to biopsy. Samples were fractionated into white blood cells (WBCs) and platelets. RNA was extracted, transcriptomic analysis with RNA sequencing performed, and expression profiles compared between cancer and non-cancer samples. These samples were also compared to those from a model system of THP1 monocytes co-cultured with MDA-MB-231 TNBC cells that was used to identify key regulatory genes. Differential gene expression and pathway analyses were conducted using Ingenuity Pathway Analysis (IPA). Results: For clinical samples, 11,355 differentially expressed genes (DEGs) were identified between cancer and non-cancer: 1,741 were unique to the cancer WBC fraction and 2,070 were unique to the cancer platelet fraction. Key shared upstream regulators between clinical WBCs and platelets and in vitro model THP1 included TRAF2, ZNF281, BMP4, SMAD3, and ITGA11 (WBCs), as well as IFNG, STING1, CDK6, RIPK2, CD44, TGM2, and STAT1 (platelets). In WBCs, the upstream regulators point to a transcriptional landscape shaped by TGF-β signaling, cellular stress response, and integrin-mediated remodeling, consistent with immune tolerance and altered cell trafficking. In contrast, platelet-derived signatures were enriched for IFNG, STING1 and related factors, suggesting a shift toward innate immune sensing, inflammatory signaling, cell cycle activity and adhesion dynamics. Conclusions: WBCs and platelets exhibited differential expression of key immune-related transcripts in cancer versus non-cancer clinical samples, and were in alignment with tumor-induced changes in a in vitro co-culture model. These results demonstrate the presence of tumour-induced blood reprogramming that is detectable in blood, with WBCs reflecting chronic signaling modulation and platelets demonstrating acute immune activation and vascular interaction. This supports the biological basis for using tumor-educated immune changes as a foundation for enabling early detection and therapeutic strategies.

K. Akbary¹, L. Jiaojiao¹, R. Yayun¹, D. Tai¹, C. Hsiao², K. Huang²; ¹Research and Development, HistoIndex, Singapore, SINGAPORE, ²Dept of Surgery, National Taiwan University Hospital and National Taiwan University College of Medicine, Taipei, TAIWAN.

BACKGROUND: Triple-Negative Breast Cancer (TNBC) and HER2-positive (HER2+) breast cancer represent biologically distinct subtypes with differing molecular drivers, clinical behaviours, and responses to therapy. Despite these differences, both subtypes may be associated with significant tumor-stroma interactions and fibrotic remodelling, which may contribute to disease progression and therapeutic response differences. However, comparative spatial characterization of collagen architecture across different tissue compartments in these subtypes remains limited. Second harmonic generation/two-photon excitation (SHG/TPE) microscopy, coupled with AI-based image analysis, enables quantitative assessment of collagen morphometry beyond conventional histopathology. This study aims to compare region-specific collagen features between TNBC and HER2+ breast cancer biopsies and identify distinguishing fibrosis patterns across tumor and non-tumor regions. METHODS: Unstained breast cancer biopsies from TNBC (n=21) and HER2+ (n=158) patients from Dept of Surgery, National Taiwan University Hospital were imaged using SHG/TPE microscopy. Collagen morphometric features were identified using a proprietary AI-based analysis platform (Genesis®200, HistoIndex Pte Ltd, Singapore). Four anatomically distinct tissue regions were assessed: tumor, lobule and duct, stroma-fat, and stroma-fibrosis. Each region comprised of 65 collagen morphometric parameters, and was compared between TNBC and HER2+ groups within each region using normalized Median Relative Differences. Statistical comparisons were performed using the Wilcoxon rank-sum test to evaluate feature-level differences between the two subtypes. RESULTS: Region-specific collagen morphometric profiles demonstrated distinct differences between TNBC and HER2+ breast cancer biopsies. Within the tumor region, collagen parameters were largely comparable between the two groups, with no major differences observed. In the stroma-fibrosis compartment, TNBC samples showed elevations in parameters related to collagen fibres but these trends were not statistically significant. Statistically significant reductions (p < 0.05) of specific collagen parameters were observed in lobule and duct regions of HER2+ biopsies versus TNBC biopsies as follows; 25% decrease in number of thick strings per unit tissue area, a 17% reduction in string perimeter, and decrease of 20% and 18% in width of all strings and width of aggregated strings per unit tissue area, respectively. CONCLUSIONS: Quantitative SHG/TPE imaging combined with AI-based collagen morphometric profiling reveals distinct fibrosis patterns between TNBC and HER2+ breast cancer subtypes. The lobule and duct region in HER2+ biopsies demonstrates a unique collagen architecture when compared with rest of TNBC biopsies. These spatially localized differences in extracellular matrix structure may reflect subtype-specific stromal remodelling and could offer utility in subtype classification, prognosis, or therapeutic response prediction. Further validation of these imaging-derived biomarkers in larger cohorts is warranted.

J. Weidhaas¹, K. McGreevy², J. Le³, M. Alcaraz¹, E. Rietdorf¹, M. Ensenyat-Mendez⁴, J. Baker⁵, M. Dinome⁴, D. Marzese⁴, D. Telesca⁶; ¹Radiation Oncology, UCLA, Los Angeles, CA, ²Statistics, MiraDx, Los Angeles, CA, ³Surgery, UCSD, San Diego, CA, ⁴Surgery, Duke, Raleigh, NC, ⁵Surgical Oncology, UCLA, Los Angeles, CA, ⁶Statistics, UCLA, Los Angeles, CA.

Background: Germline variants in microRNAs (mirSNPs) have been found to shape tumor biology and to predict both outcome and response to therapy. Here, we tested the hypothesis that mirSNPs predict high versus low lymph node burden in breast cancer patients. Methods: DNA was isolated from microdissected tumor specimens from 73 breast cancer patients with invasive ductal carcinoma (IDC), who underwent axillary lymph node dissection. The mean age at diagnosis was 59.1 years (SD 12.9), and the cohort was racially diverse: 43.8% White, 23.3% Asian, 12.3% Black, 12.3% Hispanic, and 8.2% Middle Eastern. The average BMI was 27.3 (SD 7.3). The majority of patients had HR+/HER2- breast cancer (72.6%), with fewer HER2+ (15.1%) or triple-negative (12.3%) disease. Tumors were primarily grade 2 (41.1%) or grade 3 (46.6%), with a mean Ki67 proliferation index of 29.1% (SD 20.2). T2 was the most common clinical T stage (57.5%), and the mean tumor size at diagnosis was 28.3 mm (SD 19.7). Most patients were clinically node (cN) positive (91.8%), and 57.5% had four or more positive lymph nodes. Low nodal burden was defined as N1 (<3 axillary lymph nodes) and high nodal burden N2/N3 (>4 axillary lymph nodes).The germline mirSNP dataset was filtered prior to modeling, with 22 SNPs demonstrating an association with nodal burden at a Fisher exact p-value cutoff of 0.2. mirSNPs were further reduced by iteratively removing the bottom five predictors by AUC impact and re-fitting models, resulting in a final set of 17 mirSNPs. A set of clinical variables was selected by iterative boosted tree modeling, with features retained if they had non-zero median relative importance across five leave-one-out cross-validation (LOOCV) rounds. This resulted in retention of age at diagnosis, tumor size, BMI, race, Ki67, clinical T stage, cancer subtype (TNBC, HER2+, HR+/HER2-), tumor grade, IDC with lobular features, and use of adjuvant therapies (endocrine, chemotherapy, radiotherapy). These selected clinical features were then used alone and in conjunction with mirSNP genetic data in subsequent predictive modeling. Results: Elastic net regression using the 17 mirSNPs yielded strong predictive accuracy for high nodal burden, with AUCs ranging from 0.76 to 0.79 and specificity values between 0.81 and 0.87. Importantly, the predictive accuracy did not correlate with the number of nodes sampled, and misclassifications by the model were predominantly concentrated around the clinical threshold of four positive nodes- rather than being biased by the depth of nodal sampling. When mirSNPs were combined with clinical variables in the predictive models, the resulting performance was lower than for genetics alone, with AUCs ranging from 0.67 to 0.71. Models using only clinical variables performed the worst, achieving an AUC of 0.55. The most significant mirSNP in the model was in P2RX7, a gene whose expression has been previously associated with the risk of bone metastases in breast cancer, and the second most significant mirSNP in the model was in CSMD1, a gene that is linked to the spread of breast cancer to lymph nodes. Conclusions: Our findings support the potential of mirSNPs as an additional tool to help identify patients at increased risk of high nodal burden in breast cancer. Our mirSNP model outperformed clinical models in this data set. Studies investigating the benefit of incorporating mirSNPs with other biomarkers of lymph node burden, the association of this signature with outcome, as well as further validation of these findings are ongoing.

D. Mavroudis¹, D. Jinga², E. Razis³, S. Giassas⁴, N. Touroutoglou⁵, C. Bilir⁶, M. Ozdogan⁷, I. Ökten⁸, S. Karageorgopoulou⁹, E. Galani¹⁰, G. El Hachem¹¹, D. Tzanninis¹², T. Floros¹³, G. Lazaridis¹⁴, P. Vlachostergios¹⁵, I. Hacibekiroglu¹⁶, E. Maragkouli¹⁷, I. Gioulbasanis¹⁸, A. Tsantikidi¹⁹, S. Maxouri¹⁹, C. Florou-Chatzigiannidou¹⁹, V. Metaxa-Mariatou¹⁹, E. Papadopoulou¹⁹, D. Grigoriadis²⁰, A. Papathanasiou²⁰, G. Nasioulas¹⁹; ¹Department of Medical Oncology, University General Hospital of Heraklion, Heraklion, GREECE, ²Oncology Department,, Neolife Medical Center, Bucharest, ROMANIA, ³3rd Oncology Department, Hygeia Hospital, Marousi, Athens, GREECE, ⁴General Maternity and Gynecology Clinic, Iaso General Clinic, Marousi, Athens, GREECE, ⁵Oncology Department, European Interbalkan Medical Center, Thessaloniki, GREECE, ⁶Clinic of Medical Oncology, VM Medical Park Pendik Hospital, Istanbul, TURKEY, ⁷Clinic of Medical Oncology, Memorial Antalya Hospital, Antalya, TURKEY, ⁸Department of Medical Oncology, Göztepe Prof. Dr. Süleyman Yalçın City Hospital, Istanbul, TURKEY, ⁹3rd Oncology Clinic, Iaso General Clinic, Marousi, Athens, GREECE, ¹⁰2nd Oncology Clinic, Perseus Healthcare Group SA, Metropolitan Hospital, Neo Faliro, Athens, GREECE, ¹¹Department of Hematology and Medical Oncology, Saint George Hospital University Medical Center, Beirut, LEBANON, ¹²Oncology Department, Athens Medical Center, Marousi, Athens, GREECE, ¹³Department of Oncology, Athens Naval and Veterans Hospital, Athens, GREECE, ¹⁴Faculty of Medicine, Aristotle University of Thessaloniki, Thessaloniki, GREECE, ¹⁵Oncology Clinic, IASO Thessalias S.A. – General Hospital, Larissa, GREECE, ¹⁶Faculty of Medicine, Department of Medical Oncology, Sakarya University, Sakarya, TURKEY, ¹⁷Medical Oncology Dpt and Clinic, General Hospital of Trikala, Trikala, GREECE, ¹⁸Department of Oncology, Thessalia General Clinic, Larissa, GREECE, ¹⁹Molecular Oncology, Genekor Medical SA, Gerakas, Athens, GREECE, ²⁰Department of Bioinformatics, Genekor Medical SA, Gerakas, Athens, GREECE.

Introduction: The integration of broad genomic testing has become increasingly vital for personalized treatment planning in breast cancer patients. In this study, we evaluated the analytical performance and clinical relevance of a certified for in vitro diagnostics (CE-IVD), 1021-gene comprehensive genomic profiling panel in detecting actionable tumor biomarkers across a diverse cohort of breast cancer cases. The study aimed to validate the robustness of the assay in both tissue and liquid biopsy formats, and to explore its clinical utility in real-world patients with advanced breast cancer. Methods: Tumor and blood samples from 353 patients were analyzed, including 197 FFPE tumor and 156 plasma cfDNA. Concurrent molecular profiling of tissue and plasma was conducted in 23 cases to evaluate the efficacy of liquid biopsy although tissue samples may have originated from prior procedures. DNA extracted from leukocytes was used as a control to prevent the detection of false-positive results due to clonal hematopoiesis mutations. Targeted-capture sequencing was performed using the Oncology Multi-Gene Variant Assay (GenePlus) which is a qualitative in vitro diagnostic test for detection of variants in 1021 tumor-related genes, and gene fusions in 38 genes. Sequencing was carried out on the MGI DNBSEQ-G400 platform. NGS data were analyzed via a dedicated bioinformatics pipeline on the Gene Box platform (GenePlus), enabling detection of all major genomic alterations, including gene fusions, as well as assessment of tumor mutational burden (TMB) and microsatellite instability (MSI). Results: Regarding the analysis of 197 FFPE samples, at least one on-label actionable biomarker was identified in 54.8% of cases. Among these samples, PIK3CA mutations were the most frequently detected in 27.4% of biomarker-positivethe cases. ESR1 mutations were present in 16.2% Mutations in PTEN and AKT1 genes were also observed in 7.6% and 3% of cases, respectively. Additionally, 20% of patients harbored a suspected germline variant in BRCA1/2, PTEN and CDH1 genes. In the analysis of 156 plasma samples, 46% were positive for an on label-biomarker. Among these, PIK3CA alterations were identified in 17.9% and ESR1 mutations were present in 17.3% of the cases. At lower frequency, variants were also observed in BRCA1/2 (3.8%), PTEN (3.2%), AKT1(1.3%) genes. ERBB2 (1.3%) amplification and an NTRK fusion (0.6%) were also detected. Approximately 20% of plasma samples harbored verified germline variants, most frequently in BRCA1/2 and CHEK2 genes. In a subset of 23 mBC patients, concurrent analysis of tissue and liquid biopsy demonstrated a 70% overall concordance rate on on-label biomarker analysis. In two cases, FFPE analysis identified additional on-label mutations not detected in plasma, whereas in one case plasma analysis revealed a second ESR1 mutation not detected in the FFPE sample. In two patients, plasma analysis was negative, while tissue profiling identified alterations in ESR1 and PIK3CA genes. Importantly, in two cases, liquid biopsy surpassed tissue analysis by detecting an ESR1 variant and an NTRK fusion. Conclusions: This study demonstrates the clinical utility of comprehensive multigene NGS profiling in both tissue and plasma samples from patients with breast cancer by revealing a high rate of actionable variants. Parallel analysis of matched tissue and plasma samples revealed a high concordance rate, indicating complementary role of both approaches in genomic profiling. In conclusion, integrated NGS analysis of tissue and liquid biopsies provides a powerful platform for precision oncology in breast cancer, supporting therapeutic decisions, resistance monitoring, and hereditary cancer risk assessment. The combined use of both sample types maximizes diagnostic yield and clinical impact, especially in the context of advanced disease.

C. Carette¹, G. Zels¹, A. Pabba¹, M. Maetens¹, K. Borremans¹, K. Van Baelen¹, J. Van Cauwenberge¹, T. Geukens², M. De Schepper³, H. Nguyen¹, M. Nysen⁴, H. Hoogsteyn¹, T. Van Assche¹, A. Mahdami¹, S. Leduc¹, P. Vermeulen⁵, P. Neven⁶, H. Wildiers², W. Van Den Bogaert⁷, G. Floris³, F. Richard¹, S. Hatse⁴, C. Desmedt¹; ¹Laboratory for Translational Breast Cancer Research, Department of Oncology, KU Leuven, Leuven, BELGIUM, ²Department of General Medical Oncology, University Hospitals Leuven, UZ Leuven, Leuven, BELGIUM, ³Department of Pathology, University Hospitals Leuven, UZ Leuven, Leuven, BELGIUM, ⁴Laboratory of Experimental Oncology (LEO), Department of Oncology, KU Leuven, Leuven, BELGIUM, ⁵Translational Cancer Research Unit, ZAS Augustinus, Antwerp, BELGIUM, ⁶Department of Gynaecology and Obstetrics, University Hospitals Leuven, UZ Leuven, Leuven, BELGIUM, ⁷Department of Forensic Medicine, University Hospitals Leuven, UZ Leuven, Leuven, BELGIUM.

Background. In patients with metastatic breast cancer (BC), tumor cells can be observed in blood as well as in other fluids such as pleural fluid (PF), ascites (ASC), pericardial fluid (PERI) and cerebrospinal fluid (CSF). Here, we aimed at evaluating the association between presence of tumor cells in non-blood liquid biopsies (LB) and presence of metastases in surrounding tissues (aim 1), and assessing the heterogeneity of BC biomarkers (estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (HER2) and KI67) in these non-blood LB (aim 2) from patients with metastatic BC included in our post-mortem tissue donation program, UPTIDER (NCT04531696, PMID: 38658604). Methods. 38 patients were included in aim 1. When possible, non-blood LB samples were collected during autopsy and instantly centrifuged. Cell pellets were formalin-fixed, paraffin-embedded, and evaluated histologically for tumor cellularity. Associations between the presence of tumor cells in LB and solid metastases in surrounding tissues were explored with a Fisher’s exact test. All LB with ≥5% tumor cellularity and primary tumor samples were considered for aim 2 and immunohistochemically stained for ER, PR, HER2 and KI67. Results. Aim 1. 155 samples (52 PF, 30 ASC, 34 PERI and 39 CSF) were analyzed with a median of 4 LB per patient (range 2-7). Malignant PF was observed in 29 out of 31 patients with collected PF. All except one patient (28/29) had pleural, lung and/or mediastinal metastases. Malignant ASC was observed in 18 from 27 patients with collected ASC. 17/18 had peritoneal, liver and/or gastro-intestinal metastases. Malignant CSF was detected in 6 out of 36 patients with collected CSF. Four (4/6) had brain and/or meningeal metastases. Malignant PERI was observed in 19 out of 34 patients with collected PERI, of whom 5/19 had pericardial or heart metastases. No statistical evidence was found for association of presence of tumor cells in PF, ASC or CSF and solid metastases in surrounding tissues (Fisher’s exact p-value (p) = 0.13, 1.00 and 0.37, respectively). However, a trend was seen for PERI and pericardial/heart involvement (p = 0.053). Aim 2 was evaluated on 69 samples from 27 patients. Among them, 20 presented with primary hormone receptor (HR)+/HER2- BC (18 ER+/PR+, 1 ER+/PR- and 1 ER-/PR+), and 7 patients had primary HR-/HER2- BC. Regarding intra-patient heterogeneity across different LB, the median variation in expression was 5% for ER, 0.5% for PR, 0% for HER2, and 13% for KI67 with maximum differences of 90%, 50%, 60% and 70% respectively. Of note, 10/19 (53%) patients diagnosed with primary ER+ BC, had at least 1 LB that lost ER-expression. Similarly, 14/19 (74%) patients with primary PR+ BC, lost PR-expression in at least 1 LB. All patients had HER2 non amplified primaries, but 11/27 (41%) had at least 1 LB with HER2-low and/or HER2-ultralow expression, while the other 16 patients (59%) showed no detectable HER2 expression in any of their LB. KI67 value from the primary was available for 17 patients. Twelve out of 17 patients presented with low KI67 (≤15%) at primary diagnosis of whom 5 had at least one LB with high KI67 (>15%). The remaining 5 patients presenting with high KI67 at primary diagnosis all had at least one LB with low KI67. Conclusions. This study reports on the detection of tumor cells and expression of clinically relevant biomarkers in LB from patients with metastatic BC. Our findings suggest that a multi-fluid biopsy approach, when possible, may improve the characterization of the metastatic disease heterogeneity and aid in selecting the best next-line treatment options.

Y. Lin¹, S. Lin², Y. Zhang³, J. She¹, R. Zhao¹, A. Qiu¹, L. Zhang¹, Q. Yang¹; ¹Department of Breast Surgery, the First Affiliated Hospital of Xiamen University, School of Medicine, Xiamen University, Xiamen, CHINA, ²School of Medicine, Xiamen University, Xiamen, CHINA, ³Medical college, Guangxi University, Nanning, CHINA.

Background: Breast cancer is one of the most common malignancies in women. Although previous genome-wide association studies (GWAS) have identified multiple susceptibility loci, they account for only a small portion of the heritable risk. Transcriptome-wide association studies (TWAS), which integrate GWAS and expression quantitative trait loci (eQTL) data, offer a more effective strategy for uncovering functional genes involved in complex traits. This study aims to systematically identify breast cancer susceptibility genes through multiple TWAS frameworks and to elucidate their biological roles using multi-omics integration. Methods: Firstly, we performed cross-tissue TWAS for breast cancer using both UTMOST and JTI frameworks. Next, single-tissue associations were evaluated via FUSION and validated using MAGMA. We further prioritized key genes through Mendelian Randomization (MR) and colocalization analyses, followed by experimental validation. Single-cell and spatial transcriptomic data were employed to delineate gene expression patterns, intercellular communication, and spatial heterogeneity. Additionally, drug resistance profiles were constructed using TCGA transcriptomic data and verified across multiple pharmacogenomics databases. Results: Cross-tissue TWAS analysis identified 29 candidate susceptibility genes for breast cancer. The FUSION method detected 1,768 genes with FDR < 0.05 in at least one tissue, while MAGMA analysis revealed 354 breast cancer-associated genes. By integrating results from four analytical approaches, we prioritized 13 high-confidence susceptibility genes, including EFR3B, CASP8, and XBP1. MR and colocalization analyses further confirmed EFR3B and CASP8 as causal genes with shared genetic architecture in breast cancer. EFR3B, a susceptibility gene for both ER-positive and ER-negative breast cancer, was experimentally validated to be overexpressed in breast cancer cell lines and tumor tissues using Western blot, real-time PCR, and immunohistochemistry. Single-cell transcriptomic analysis revealed that EFR3B was predominantly expressed in specific epithelial and endothelial subpopulations within breast tumors, exhibiting notable expression heterogeneity. Analysis of VEGFA ligand-receptor signaling indicated enhanced intercellular communication between EFR3B-positive epithelial and endothelial cells. Spatial transcriptomics demonstrated heterogeneous expression of EFR3B in tumor tissues and showed that EFR3B-positive epithelial cells are spatially co-localized with vascular endothelial cells, suggesting a potential role for EFR3B in modulating the tumor microenvironment. Finally, drug sensitivity analysis revealed that high EFR3B expression is significantly associated with resistance to paclitaxel and anthracycline-based chemotherapies. Conclusions: This study identified 13 breast cancer susceptibility genes through integrative TWAS approaches, providing new insights into the genetic architecture of the disease. EFR3B emerged as a key risk gene involved in tumor microenvironment modulation and chemotherapy resistance, representing a potential therapeutic target for personalized treatment strategies in breast cancer.

W. Shijing, W. Hongrui, X. Mao; Breast Surgery, The First Affiliated Hospital of China Medical University, Shenyang, CHINA.

Objective: Identifying effective predictors of response to neoadjuvant chemotherapy (NAC) is essential for personalized breast cancer treatment. Our previous observations suggested that a higher tumor-to-gland density ratio (TGDR) on mammography may predict poorer therapeutic response. This study evaluated the predictive value of TGDR in NAC and developed models based on various pathological complete response (pCR) criteria. Methods: This retrospective study included 303 breast cancer patients who received NAC at the First Affiliated Hospital of China Medical University from January 2020 to July 2024. pCR was defined by four criteria: pCR1 (ypT0/is), pCR2 (ypT0/is ypN0), pCR3 (ypT0), and pCR4 (ypT0 ypN0). Patients were randomly divided into training and test sets (7:3). Logistic regression analyses identified independent predictors for each pCR criterion, and multiple logistic regression (LR) models were constructed. Model performance was assessed using AUC, F1 score, DeLong test, DCA, NRI, and IDI. Four machine learning algorithms (SVM, K-nearest neighbor, XGBoost, and LightGBM) were applied to compare predictive performance. Results: For pCR1, 140 patients (46.2%) achieved response. TGDR, ADC, ER, HER2, histological grade, and molecular subtype were independent predictors. Combined-variable models (AUCs: 0.840-0.852) outperformed single-variable models and demonstrated good clinical utility. Model 3 was selected for its simplicity and MRI independence; the LR algorithm performed best. For pCR2 (n=127), similar predictors were identified. Combined models (AUCs: 0.847-0.859) showed strong performance; SVM performed best. For pCR3 (n=118), model 2 (TGDR + pathological features) outperformed TGDR alone (AUC: 0.848 vs. 0.705); XGBoost performed best. For pCR4 (n=109), model 2 (TGDR, ER, HER2) outperformed model 1 (TGDR alone) in both training and test sets; the LR model performed best. Conclusion: TGDR is a novel imaging biomarker with predictive performance comparable to ADC and is an independent predictor of NAC response. Across all pCR criteria, TGDR-based models combined with clinicopathological features consistently outperformed TGDR alone. TGDR offers a practical alternative for NAC response prediction, particularly in settings with limited MRI access.

M. Vetter¹, T. Kanthaiah¹, E. Chiru¹, C. Francesc², I. Krol², A. Wicki³, A. Ring⁴, B. Nguyen-Straeuli⁵, V. Heinzelmann-Schwarz⁵, W. Weber⁶, C. Kurzeder⁶, N. Aceto², F. Schwab⁵; ¹Cancer Center and Center for Oncology and Hematology, Cantonal Hospital Baselland, Liestal, SWITZERLAND, ²Institute of Molecular Health Sciences, ETH Zürich, Zürich, SWITZERLAND, ³Comprehensive Cancer Center, University Hospital Zürich, Zürich, SWITZERLAND, ⁴Medical Oncology, University Hospital Zürich, Zürich, SWITZERLAND, ⁵Gynecology and Obstetrics, University Hospital Basel, Basel, SWITZERLAND, ⁶Breast Clinic, University Hospital Basel, Basel, SWITZERLAND.

Background: CTC clusters are established mediators of metastasis in breast cancer and are associated with unfavorable clinical outcomes. However, their prevalence and phenotypic characteristics in older patient populations remain inadequately studied. Systemic inflammation, as measured by the NLR, may influence CTC dynamics. Methods: A prospective translational analysis was conducted in women with metastatic breast cancer (mBC), stratified into two cohorts: Group A (age ≥70 years) and group B (age <70 years). Peripheral blood samples were processed using the FDA-cleared Parsortix® system for CTC enrichment, followed by immunofluorescence staining (EpCAM/HER2/EGFR/CD45) to identify single CTCs and CTC clusters (homotypic and heterotypic). Baseline clinical characteristics, including histological subtype, stage, tumor grade, Ki-67 index, and sites of metastasis, were documented. NLR ≥4 was defined as elevated. Primary endpoints included CTC cluster frequency and size; secondary endpoints assessed correlations with age group and NLR status. Results: A total of 155 patients were evaluable, including 58 patients in group A (mean age: 78 years). A trend for elevated NLR being more frequent in Group A compared to Group B was visible (53.2% vs 36.8%, P = 0.075). No statistically significant differences were observed between groups regarding histological subtype, stage at diagnosis, or tumor grade. The Ki-67 proliferation index was significantly elevated in group B (mean 37%) compared to Group A (mean 21%, P < 0.001). The pattern of metastatic spread was comparable, although a trend toward increased brain metastases was observed in Group B. A trend showing a higher proportion of patients exhibiting ≥1 CTC in Group B was detected (43.5% vs. 29.3%, P = 0.082). CTC clusters were identified in 6.9% of Group A and 12.5% of Group B (P = 0.275). Elevated NLR (≥4) was not significantly associated with CTC cluster presence (11.1% vs. 8.2%, P = 0.594). Conclusions: While older patients with mBC (Group A) exhibited a trend for higher systemic inflammation as indicated by NLR, no statistically significant differences in CTC cluster frequency or size were observed compared to younger patients (Group B). Moreover, elevated NLR was not independently predictive of increased cluster prevalence. These findings underscore the need for further investigation into age-associated CTC phenotypes and their potential clinical implications in larger patient populations.

M. W. Daniels¹, J. L. Wittliff²; ¹Bioinformatics and Biostatistics, University of Louisville, Louisville, KY, ²Biochemistry & Molecular Genetics and Institute for Molecular Diversity & Drug Design, University of Louisville, Louisville, KY.

Background: Clinical assessment of breast cancer progression and therapeutic response typically relies on qualitative immunohistochemistry results for ER, PR, and HER2 proteins, which may be improved by adherence to ASCO/CAP Guidelines. To identify more robust, clinically relevant molecular markers, we conducted in silico analyses of a comprehensive, de-identified database that integrates quantified protein biomarker levels, qPCR, and microarray gene expression data, as well as long-term clinical outcomes. Because TNBC represents an aggressive disease that is difficult to treat, our multifarious investigation focuses on differentiating biomarkers for this subtype that improve clinical outcomes. Methods: Biomarker Triads were created from ER and PR proteins quantified by radio-ligand binding (n = 13,966) and enzyme immunoassays (n = 4,408), expressed as fmol/mg cytosol protein. HER2 protein was quantified in 504 and 901 specimens using either ELISA or EIA, respectively, and expressed as HNU/mg membrane protein. Clinical data were available for 1,194 ER/PR cases, and complete ER/PR/HER2 protein Triads were available for 189 of these. However, survival outcomes (PFS and OS) were only available for a subset of these patients. Expression levels of these genes and 82 other cancer-associated genes were validated by RTqPCR of 580 specimens after RNA extraction and purification. Microarray analysis (~22,000 genes) was conducted on RNA from each of 247 LCM-isolated carcinoma samples. Biomarker Triads were defined as ER+/PR+/HER2+ (TPBC) or ER-/PR-/HER2- (TNBC), using either protein levels or cognate gene expression (ESR1, PGR, ERBB2). Differential gene expression was assessed using nonparametric tests and Spearman correlations in R (version 4.4.2). For the subset of patients with survival data, exploratory Kaplan-Meier and Cox proportional hazards models were used to assess associations between gene expression and progression-free survival (PFS) or overall survival (OS). Results: Neither ER/PR assay platform influenced ER+/PR+ or ER-/PR- status, and neither ER nor PR levels were associated with HER2 expression. TPBC biopsies showed elevated expression of BL2 (CADM1), CAXII, ERBB2, and ERBB4 compared to TNBC, supporting their use in subtype definition. TPBC also exhibited increased IL10 and VEGFATV2, while TGFB1 was downregulated in both TPBC and ER+ cases. BRCA1/2 expression was reduced in ER-/PR- tumors, while NAT1/2 and COMT were elevated in ER+, ER+/PR+, and TPBC samples. FOXA1 and ART3 emerged as potential targets from gene-based Triad definitions using microarray data. In a subset of cases with available clinical outcomes, exploratory survival analysis suggested that TPBC triads were associated with more favorable PFS and OS compared to TNBC, and elevated expression of certain genes (e.g., ESR1, FOXA1) trended with improved survival outcomes. Conclusions: This integrative dataset, which links rigorously collected clinical biopsies, precise biomarker protein quantification, and gene expression profiles from LCM-procured carcinoma cells, enabled the discovery of genes differentially expressed across biomarker-defined subtypes. Biomarker Triads based on protein and gene expression levels in primary breast tissue biopsies revealed a landscape of genomic diversity in carcinomas, which improved molecular subtype classification. Select genes, including the transcription factor FOXA1 and ART3, reported to be involved in cell migration, represent probable targets for drug development and/or companion diagnostics in ER+ and triple-negative breast carcinomas.

Y. Gao¹, M. Weitz², S. Fragkogianni², K. Layng³, H. Mai¹, J. Chang¹; ¹Neal Cancer Center, Houston Methodist Hospital, Houston, TX, ²Data Science, Tempus AI, Chicago, IL, ³Medical Affairs, Tempus AI, Chicago, IL.

Background Recent molecular profiling advances have transformed breast cancer treatment. These breakthroughs have largely benefited patients with hormone receptor-positive and triple-negative breast cancers, while HER2-positive (HER2+) breast cancer remains less understood molecularly. We evaluated the tumor immune microenvironment of HER2+ breast cancer, assessing immune cell infiltration, TMB, PD-L1 status, and HLA allele prevalence, to uncover biomarkers for treatment guidance. MethodsWe used Tempus Lens (Tempus AI, Inc., Chicago, IL) to identify 455 female patients with HER2+ breast cancer and xT and xR testing. Patients were classified as having localized (stage 1-3, N = 124) or de novo metastatic disease (stage 4, N = 331). Wilcoxon rank sum, Fisher’s exact, and Pearson’s Chi-squared tests were used to compare groups. Immune cell proportions were estimated using quanTIseq. TMB, PD-L1, HLA, MSI, and MMR were also analyzed. Real-world progression-free survival (rwPFS) was time from therapy (neoadjuvant or first-line) start to progression. Real-world overall survival (rwOS) was time from first-line therapy start to death or loss to follow up. Median rwOS and rwPFS were estimated with Kaplan-Meier (KM) and hazard ratios (HR) using Cox proportional hazard models (CoxPh). Results In patients with localized disease, the median proportion of tumor-infiltrating B cells was significantly higher than in those with metastatic disease (p < 0.001). A similar pattern was observed for natural killer (NK) cells, CD8⁺ T cells, and regulatory T cells (Tregs), with enrichment in localized stage patients (p<0.001, p<0.001, and p 10 mutations/Mb) was observed in 6.7% of patients with localized disease and 10% of patients with de novo metastatic disease. 29% of patients with localized disease were PD-L1 IHC positive, as were 11% of patients with de novo metastatic disease. Interestingly, in metastatic patients who received standard chemotherapy and/or anti-HER2 therapy as first line treatment, positive PD-L1 status and high TMB were significantly associated with worse median rwOS (p<0.021 and p<0.001, respectively). Across all HER2+ breast cancer patients with patient-reported race information available, the HLA-A *02:01 allele exhibited the highest overall frequency (43%). Stratified by race, this allele was most prevalent in White patients (48%), with significantly lower frequencies observed in Asian (25%) and Black (24%) patients (p=0.018), consistent with previous studies. In contrast, the HLA-A *33:03 allele demonstrated a markedly higher frequency in Asian patients (33%), while its prevalence was substantially lower in Black (17%) and White (1.4%) patients (p<0.001). Similarly, the HLA-A *23:01 allele was more frequent in Black patients (36%) compared to White (5.2%) and Asian (0%) patients (p<0.001). Conclusions Our findings suggest a potential therapeutic benefit of incorporating immunotherapy into the treatment paradigm for HER2+ breast cancer, in both localized and metastatic disease settings. Moreover, the ethnically enriched HLA polymorphisms within our patient cohort may help elucidate observed disparities in clinical outcomes and offer valuable insights for the development of population-tailored immunotherapeutic strategies.

J. K. Nag, G. Subedi, I. Zin, P. L. Fox; Cardiovascular and Metabolic Sciences, Cleveland Clinic Research, Cleveland Clinic, Cleveland, OH.

Breast cancer remains a formidable clinical challenge, underscored by the persistence of cancer stem cells (CSCs) that drive tumorigenesis and therapeutic resistance. CSCs are characterized by elevated expression of pluripotency-associated transcription factors, including SOX2, OCT4, and NANOG (SON factors), which orchestrate self-renewal and tumor maintenance programs. The downstream transcriptional targets of the SON factors have been investigated in detail; however, the upstream regulation of the factors is less well understood. We previously reported that activation of AKT, or chemical or genetic inhibition of PTEN, induces nuclear translocation of EPRS1 (glutamyl-prolyl tRNA synthetase) a cytoplasmic protein essential for accurate interpretation of the genetic code during translation (1). The contribution of EPRS1 to breast cancer progression is sup­port­ed by reports that EPRS1 mRNA and protein are elevated in human breast tumors, and associ­ated with reduced overall survival of breast cancer patients (2, 3). In view of reports that AKT activation is a principal determinant of expression of SOX2 (4), the master switch of pluripotency factors, we considered the potential role of AKT-induced nuclear localization of EPRS1 in SON factor expression. siRNA-mediated knockdown of EPRS1 in PTEN^– MDA-MB-468 and HCC38 cells markedly inhibited expression of the mRNAs encoding the three SON factors, as shown by RT-qPCR; protein expression was similarly diminished. Knockdown of other aminoacyl-tRNA synthetases, e.g., LARS1 (leucyl-tRNA synthetase), was ineffective, thereby showing specificity. Ablation of EPRS1 in breast cancer cells reduced stemness, migration, invasion, and clonogenic potential. Moreover, limited dilution assays in mice showed EPRS1 knockdown significantly reduced tumor mass and tumor-initiating cell number. A proteomic analysis of EPRS1-interacting proteins in nuclear lysates of MDA-MB-468 cells gave an insight into the mechanism underlying regulation of SON expression by EPRS1 (1). Mass spectrometric analysis revealed interaction of nuclear EPRS1 with RNA cytidine acetyltransferase (NAT10), the only enzyme known to acetylate RNAs, including mRNAs, at the 4-cytidine position, i.e., ac⁴C modification (5). The interaction was validated by co-immunoprecipitation. Knockdown of NAT10 reduced ac⁴C modification of SON mRNAs as determined by ac⁴C-RIP (RNA immunoprecipitation) as well as SON mRNA expression, indicating an important role for ac⁴C modification in SON expression. Importantly, EPRS1 knockdown likewise inhibited ac⁴C modification of SON mRNAs, revealing for the first time a role for EPRS1 in mRNA modification. These results implicate EPRS1 and ac⁴C modification of SON factors as key drivers of cancer stem cell activity and suggest new targets for breast cancer treatment. References: 1. I. Zin et al., AKT-dependent nuclear localization of EPRS1 activates PARP1 in breast cancer cells. Proc. Natl. Acad. Sci. U. S. A. 121, e2303642121 (2024). 2. I. Katsyv et al., EPRS is a critical regulator of cell proliferation and estrogen signaling in ER+ breast cancer. Oncotarget 7, 69592-69605 (2016). 3. L. Qi et al., Significant prognostic values of differentially expressed-aberrantly methylated hub genes in breast cancer. J. Cancer 10, 6618-6634 (2019). 4. Z. Wang et al., AKT drives SOX2 overexpression and cancer cell stemness in esophageal cancer by protecting SOX2 from UBR5-mediated degradation. Oncogene 38, 5250-5264 (2019). 5. D. Arango et al., Acetylation of cytidine in mRNA promotes translation efficiency. Cell 175, 1872-1886 e1824 (2018).

K. Yuan¹, S. Humble², A. Mabry¹, R. Kladney¹, L. Maggi¹, C. X. Ma¹, J. D. Weber¹, G. A. Colditz³; ¹Division of Oncology, Washington University in St. Louis, Saint Louis, MO, ²Division of Public Health Sciences Department of Surgery, Washington University in St. Louis, Saint Louis, MO, ³Division of Public Health Sciences in the Department of Surgery, Washington University in St. Louis, Saint Louis, MO.

Background: Triple-negative breast cancer (TNBC) accounts for approximately 15-20% of all breast cancers, characterized by aggressive clinicopathology, limited targeted therapies, and high rates of early recurrence. Immune activation, particularly type I interferon (IFN) signaling, has emerged as a key modulator of tumor behavior, response to therapy, and patient outcomes in TNBC. However, the impact of IFN pathway regulators on long-term prognosis remains unclear.As modulators of immune activation, ADAR1 and ISG15 are IFN pathway genes that may influence the tumor-immune microenvironment and clinical outcomes in TNBC. This study investigates whether ADAR1 and ISG15 expression are associated with disease-free survival in TNBC, aiming to clarify their prognostic significance and potential interplay as immune-modulatory biomarkers in this high-risk breast cancer subtype. We hypothesized that the expression of ADAR1 and ISG15 would be linked to DFS events.Methods: This cohort was prepared by the St. Louis Breast Tissue Registry, derived from tissue microarrays (TMAs) constructed from surgical tumor blocks of 326 patients with TNBC diagnosed between 1973 and 2022 who underwent curative surgery as the immediate therapy at the Barnes-Jewish Hospital and Siteman Cancer Center. All clinical data and tissue specimens were collected by the Washington University Institutional Review Board (IRB #201102394). Information was deidentified prior to distribution of TMA to investigators. For this study, disease-free survival (DFS) is defined as the time from diagnosis to recurrence, second primary malignancy, death or last disease-free follow-up. ISG15 and ADAR1 expression were examined by immunohistochemistry and scored by Allred and HALO. Clinical and pathological data were systematically annotated and analyzed using custom R scripts. Descriptive statistics were employed to summarize clinical and molecular variables. Cox proportional hazards models were generated to assess the association of biomarker expression with DFS. Kaplan-Meier survival analysis was used to evaluate clinical outcome and additional models were constructed to explore the impact of clinical stage and race on DFS.Results: The median age was 56.5 years, with 225 White and 101 Black patients. 221 patients were disease-free at follow-up. Median follow-up time was 10 years. In this cohort, stage was prognostic for DFS in multivariate analysis. Cytoplasmic ADAR1 and ISG15 expression were significantly higher in Black patients compared to White patients (6.49 vs. 5.95, 0.92 vs. 0.84, p = 0.044, 0.041). However, race was not independently predictive of disease-free survival in this model (p = 0.80).Higher ISG15 expression was associated with improved DFS through univariate analysis (p = 0.03) but not in multivariate analysis that included stage. Interestingly, patients who were disease-free at follow-up had significantly higher ADAR1 expression through univariate analysis (mean difference = 0.571, 95% CI: [-1.09, -0.05], p = 0.030). Kaplan-Meier analysis confirmed that high expression was correlated with DFS (p = 0.012). However, in this cohort, patients with higher ADAR1 expression also have a higher rate of adjuvant chemotherapy (p=0.01) and those treated with chemotherapy are more likely to be disease-free than patients who did not undergo chemotherapy (p=0.012).Conclusion: To our knowledge, this study is the first to examine the association between IFN pathway regulators, ISG15 and ADAR1, and TNBC prognosis. The finding of increased ISG15 in association with improved DFS is consistent with the likely activated IFN pathway state for these tumors. Additional analysis is ongoing to examine the association of these markers with immune cell infiltration and the expression of immune checkpoints.

J. Cheun¹, C. Lee², D. Shin², S. Park³, T. Ahn⁴, W. Han², H. Kim²; ¹Surgery, SMG-SNU Boramae Medical Center, Seoul, KOREA, REPUBLIC OF, ²Surgery, Seoul National University Hospital, Seoul, KOREA, REPUBLIC OF, ³Integrated science, Handong Global University, Pohang, KOREA, REPUBLIC OF, ⁴Life Science, Handong Global University, Pohang, KOREA, REPUBLIC OF.

Emerging evidence suggests that tumor-educated platelets (TEPs) undergo dynamic RNA splicing changes in response to cancer-derived signals, offering a minimally invasive avenue for cancer detection. In this study, we leveraged platelet RNA sequencing (RNA-seq) data to identify breast cancer-associated exon-exon splice junctions whose expression is significantly altered in cancer patients compared to healthy controls. Using a stepwise feature selection method guided by the Akaike Information Criterion (AIC), we identified 11 discriminatory splice junctions from platelet RNA profiles. These junctions were used to train a logistic regression classifier, which achieved strong predictive performance with an AUC of 0.873 on public test data and 0.949 on newly generated clinical samples, highlighting the model’s robustness and clinical applicability. Dimensionality reduction (t-SNE) and heatmap visualizations confirmed distinct separation between breast cancer and control samples based on these junction features. Gene Ontology enrichment analysis of the corresponding genes revealed significant enrichment in pathways such as endothelial cell development, establishment of endothelial barrier, and regulation of protein localization to the cell surface (FDR < 0.05), suggesting that platelet splicing alterations may reflect tumor-induced vascular and immune interactions. Our findings demonstrate that platelet-derived exon-exon junction expression profiles serve as promising biomarkers for breast cancer detection.

B. E. Nascimento¹, L. P. Lacerda², P. F. Silva¹, R. M. Goveia¹, I. M. Calassa¹, K. M. Veiga¹, R. M. Rahal³, W. D. Oliveira¹, D. C. Maia⁴, E. P. Silveira-Lacerda⁵; ¹Genetics, Federal University of Goiás, Goiânia, BRAZIL, ²Medicine Faculty, UNICEPLAC – Universidade do Planalto Central, Gama, BRAZIL, ³CORA, Clinical Hospital, Federal University of Goiás, Goiânia, BRAZIL, ⁴oncology, Federal University of Ceará, Fortaleza, BRAZIL, ⁵Center for Human Genetics, Institute of Biological Sciences, Federal University of Goiás, Goiânia, BRAZIL.

Variants of Uncertain Significance (VUS) represent a major challenge in clinical genetics, particularly in the context of hereditary breast and ovarian cancer (HBOC). The underrepresentation of Latin American populations in global genomic databases contributes to a higher prevalence of inconclusive results, hindering genetic counseling and risk-reduction interventions. Efforts to improve variant interpretation through population-specific evidence and clinical integration are urgently needed. To reclassify VUS identified in patients suspected of HBOC by applying the structured ACMG/AMP guidelines, integrating computational prediction, population data, and clinical context; and to develop and implement a dedicated institutional platform (CLASSIVUS) to record, standardize, and track the classification of genetic variants in a Brazilian clinical genetics center. We analyzed 105 patients with a confirmed diagnosis of breast cancer who met the NCCN criteria for HBOC. Genetic testing was performed using next-generation multigene panel sequencing covering key susceptibility genes. VUS were selected based on ACMG criteria and evaluated using in silico predictors (PolyPhen-2, SIFT, MutationTaster), population frequency databases (gnomAD, ABraOM), and clinical annotation platforms (ClinVar, Franklin by Genoox, Varsome). Each variant was scored according to the ACMG criteria, and a final classification was determined. A web-based institutional database, CLASSIVUS, was developed to register variants, document the applied criteria, associate the responsible professionals, and enable filtered queries and reanalyses. Thirteen VUS were identified in 10 patients (9.5%) across seven genes (ATM, BRCA1, BRCA2, BRIP1, BARD1, CHEK2, TP53). Five variants (38.5%) were reclassified as likely benign based on ACMG scoring, while eight (61.5%) remained classified as VUS. The CLASSIVUS platform enabled centralized storage of variant records, real-time retrieval of previous classifications, and standardized documentation of scoring justification. This system reduced analytical redundancy and ensured traceability and reproducibility among professionals, offering a scalable solution for routine clinical workflows. The integration of ACMG guidelines with computational tools, clinical annotations, and local population frequency data enabled reliable reclassification of VUS in a Brazilian HBOC cohort. The CLASSIVUS platform proved to be a valuable tool for standardizing variant interpretation and supporting safe, reproducible, and transparent clinical decisions. This approach strengthens precision medicine practices in underrepresented populations and may serve as a model for other institutions seeking to enhance their genetic curation capabilities.

H. Kaplan¹, A. Dowdell², J. Wagner³, J. Welle³, S. Reynolds³, C. Carney¹, D. Page⁴, A. Conlin⁴, K. McCormick⁵, F. Yan⁵, E. Johnston⁶, D. File⁵, T. Wahl⁷, J. Lopez⁶, Z. Topp⁴, K. Paulson¹, B. Xie⁷, S. Montgomery⁶, A. Bartlett⁸, C. Bifulco⁹, B. Piening⁹; ¹Medical Oncology, Swedish Cancer Institute and Paul G Allen Research Center, Seattle, WA, ²Molecular Genomics Lab, Earle A Chiles Research Institute and Providence Genomics, Portland, OR, ³Molecular Genomics Lab, Providence Genomics, Portland, OR, ⁴Medical Oncology, Earle A Chiles Research Institute and Providence Cancer Center, Portland, OR, ⁵Medical Oncology, Swedish Cancer Institute, Seattle, WA, ⁶Medical Oncology, Swedish Cancer Institute, Edmonds, WA, ⁷Medical Oncology, Swedish Cancer Institute, Issaquah, WA, ⁸Medical Oncology, Earle A Chiles Research Institute and Providence Genomics, Portland, OR, ⁹Medical Oncology, Earle A Chiles Research Institute and Providence Genomics, Portland, OR.

Background Despite treatment advances, long-term survival for patients with metastatic breast cancer (mBC) remains low. While many tumors exhibit an initial response to conventional or targeted therapies, this is typically followed by the acquisition of resistance. While a subset of these acquired resistance events are driven by known mechanisms (e.g. ESR1 mutation on hormone therapy), others are driven by mechanisms that remain unsolved. Moreover, the trajectory and timeline by which tumors acquire resistance remains challenging to predict. To address this, we developed a study following the molecular evolution of mBC in tissues and circulating tumor DNA (ctDNA) combining both DNA sequencing and DNA methylation analysis. Methods We developed a protocol enrolling patients with a new diagnosis of de novo or first recurrence HER2-negative mBC. For enrolled patients, primary tumor and one or more metastatic sites are profiled using comprehensive genomic profiling (CGP) (ProvSeq Solid Tumor) and DNA methylation (DNAm) (Twist Human Methylome Panel). Patient blood samples are collected using Streck tubes at enrollment and at 12-week intervals for 3 years. Blood ctDNA is extracted and profiled using CGP (ProvSeq Liquid) and DNA methylation (Twist Human Methylome Panel). Mutations and methylation signatures are assessed in primary and metastatic tissue, and over time in ctDNA. Results While recruitment and longitudinal sample collection is ongoing, we report preliminary analyses from the first 40 participants enrolled in the study. Using CGP testing across solid tumor and liquid biopsy samples, we found 735 unique reportable variants, with 165 variants annotated as clinically-significant (OncoKB database), with 16 patients harboring a pathogenic alteration in PIK3CA. Solid tumor variant detection via CGP was highly accurate down to 5% variant allele fraction and down to 0.5% variant allele fraction in liquid CGP. For patients with both primary and metastatic tissue samples (n=20) mean concordance of variants was 86%. For variants deemed clinically significant in the tissue, 36% were also found in the closest ctDNA and 40% were found in ≥ 1 ctDNA sample over the monitoring period. Thus far, two patients (both receiving hormone therapy) have demonstrated ESR1 mutations. One was detected in the first ctDNA and progressed within 3 months. The other was detected 3 months after the diagnosis of metastatic disease, but did not progress clinically until 5 months later when the VAF had increased from 3% to 13%. Both solid tissue and liquid methylation analysis yielded high coverage across genome-wide CpG sites assessed in the panel, and current efforts are focused on investigating methylation changes in non-mutational acquired therapeutic resistance. Conclusions The combination of comprehensive genomic profiling and genome-wide methylation profiling is highly feasible in both tissue and liquid biopsies and may represent a synergistic strategy for maximizing diagnostic yield. Given the incomplete portrait of genomic evolution contributing to therapeutic response in mBC, our ongoing efforts seek to provide more comprehensive maps of mutational and epigenomic changes that may yield better treatment strategies in this population.

S. Jacob¹, C. Wu², A. Glas³, C. Yau⁴, L. Brown-Swigart⁵, G. Hirst⁶, T. Haddad⁷, K. Giridhar⁷, A. Elias⁸, R. Mukhtar⁹, L. Huppert¹⁰, K. Albain¹¹, D. Yee¹², L. van’t Veer⁵, L. Esserman⁹, J. Chien¹³; ¹Medicine, University of Pennsylvania, Philadelphia, PA, ²Radiology and Biomedical Imaging,  University of California San Francisco, San Francisco, CA, ³Biomarkers and Innovation, Quantum Leap Healthcare Collaborative, San Francisco, CA, ⁴Surgery,  University of California San Francisco, San Francisco, CA, ⁵Laboratory Medicine, University of California San Francisco, San Francisco, CA, ⁶Surgery, Div of Surgical Oncology, University of California San Francisco, San Francisco, CA, ⁷Medical Oncology, Mayo Clinic Comprehensive Cancer Center, Rochester, MN, ⁸Department of Medicine, Medical Oncology, University of Colorado, Aurora, CO, ⁹Surgery, University of California San Francisco, San Francisco, CA, ¹⁰Medicine,  University of California San Francisco, San Francisco, CA, ¹¹Medicine, Loyola University Chicago Stritch School of Medicine, Maywood, IL, ¹²Medicine and Pharmacology, University of Minnesota, Minneapolis, MN, ¹³Medicine, University of California San Francisco, San Francisco, CA.

Intro: The 70-gene MammaPrint (MP) assay is prognostic for distant recurrence and predictive of benefit from adjuvant chemotherapy (CT) benefit established on surgical tissue of systemically untreated patients (pts). The MP index range is from -1 to +1 with higher index indicating lower risk of early recurrence. Risk is categorized as MP High Risk, MP Low Risk, and MP UltraLow Risk, and results can guide treatment decisions. Pts with MP index >0 are classified as Low Risk and derive little benefit from CT. There are clinical scenarios in which the only available tissue for MP or other genomic testing has been exposed to neoadjuvant endocrine therapy (NET). Few studies have investigated the impact of ET exposure on MP index. Methods: Pts were identified through the I-SPY Low Risk Registry (LRR), an observational sub-study which enrolled pts with hormone receptor positive (HR+)/HER2-negative early breast cancer ineligible for the I-SPY2 trial due to MP Low Risk or UltraLow Risk status, between 2010-2020. Pts were treated at the discretion of their provider. We conducted a retrospective analysis of MP index from paired tumor tissue obtained from systemically untreated diagnostic core biopsies and surgical specimens after NET. MP categories and index before and after NET were compared. Change in MP index was assessed for an association with changes in clinicopathologic factors obtained through the LRR using the paired-t test. Results: 28 pts with paired tumor samples were included in the analysis. Median age was 48 (range 31-75) and 18 (64%) were premenopausal. At baseline, 13 (46%) were clinically node-positive and 22 (76%) had ductal histology. Tumor grade was 1 in 6 pts (21%), 2 in 20 pts (71%), and 3 in 2 pts (7%). ER expression was >95% in 25 pts (range 60-100%) and PR expression was >90% in 14 pts (range 0-100%). NET regimens included tamoxifen +/- ovarian function suppression (OFS) in 3 pts (10.7%), aromatase inhibitor (AI) +/- OFS in 15 pts (53.6%), tamoxifen followed by or in combination with AI +/- OFS in 8 pts (28.6%), fulvestrant in 1 pt (3.6%) and OFS alone in 1 pt (3.6%). 11 (61%) of premenopausal pts received OFS. 96% (27/28) of pts were classified as Low Risk both pre- and post-NET. Only one pt had MP change from Low Risk to High Risk. Three pts had change from Low Risk to UltraLow Risk or vice versa. As a continuous variable, a small average increase in MP index was observed after NET; but, there was a range of responses. 19 pts (68%) had an increase in MP index after NET, while 9 pts (32%) had a decrease. In those with a decrease in MP index, the average ER expression decreased by -13.6%, whereas in those with an increase in MP index, ER expression was unchanged (-0.18%) (p=0.049). There was no significant difference in changes in PR expression between tumors with increased vs decreased MP indices. At a median follow-up of 35 months, 4 distant recurrences were observed. Of the 9 pts with a decrease in MP index, 3 pts (33%), including the one who changed from Low Risk to High Risk, experienced a distant recurrence. In contrast, in those with an increase in MP index, 1 of 19 (5%) experienced a distant recurrence. Conclusion: In pts with early-stage HR+/HER2- MP Low Risk or UltraLow Risk breast cancer, NET had minimal impact on the tumor MP classification. These data suggest that the MP assay could be used for tissue exposed to NET if untreated tissue is not available. Further investigation in larger cohorts is needed to validate these findings. The observation that there were more distant recurrences in pts with tumors that had an increase in MP after NET is hypothesis generating, and potentially clinically relevant if validated in a larger study. The potential of MP as a dynamic risk marker to stratify recurrence risk following NET is being evaluated in the ongoing I-SPY2 Endocrine Optimization Pilot.

A. K. Souto¹, F. C. Koyama², S. S. Koide², B. B. Souza¹, J. D. Massaro², L. J. Oliveira², L. Yamamoto², T. A. Santana³, M. L. Bulcão⁴, C. B. Fernandes², C. A. Resende¹, M. S. Mano⁵, L. C. Landeiro⁶, G. O. Bretas¹, C. B. Miranda⁷, R. Dienstmann², R. L. Sandoval⁸; ¹Oncology, Oncoclínicas&Co, Brasília, BRAZIL, ²Oncology, Oncoclínicas&Co, São Paulo, BRAZIL, ³Oncology, Oncoclínicas&Co, Aracaju, BRAZIL, ⁴Oncology, Oncoclínicas&Co, Rio de Janeiro, BRAZIL, ⁵Oncology, Hospital Albert Einstein, São Paulo, BRAZIL, ⁶Oncology, Oncoclínicas&Co, Salvador, BRAZIL, ⁷Oncology, Oncoclínicas&Co, Vitória, BRAZIL, ⁸Oncology, Hospital Sírio Libanês, Brasília, BRAZIL.

Introduction: Breast cancer (BC) is the most prevalent malignancy among women, and 5-10% are associated with germline pathogenic/likely pathogenic variants (PVs) in high-penetrance BC genes, mostly BRCA1/2. These genes encode proteins involved in the homologous recombination repair (HRR) pathway, whose dysfunction leads to homologous recombination deficiency (HRD). Germline PVs, somatic alterations, and/or epigenetic modifications in HRR pathway genes can induce the HRD tumor phenotype. While HRD is well-characterized in ovarian cancer and is a predictive biomarker for Parp inhibitors in this setting, data in BC remains scarce, especially in the Brazilian population. Objective: To evaluate HRD status in breast cancer samples from a Brazilian cohort. Methods: This retrospective single-center study included women with BC treated between 2020-2023 whose formalin-fixed paraffin-embedded tumor samples were available for analysis at Oncoclínicas Precision Medicine laboratory (OCPM). The laboratory database was used to identify eligible samples. All tumor samples were submitted to pathology review. HRD analysis assessed genomic scars (loss of heterozygosity, telomeric allelic imbalance, and large-scale state transitions) by sequencing 13,809 single nucleotide polymorphisms, plus mutational profiling of 15 genes involved in the HRR pathway. HRD status was defined using a validated assay by OCPM. Genomic instability score cutoff of 65 or higher. HRD profiles were described in relation to clinicopathological and molecular features. Results: Among 134 identified samples, 55 patients were consecutively selected for tumor analysis according to the study budget. Eight samples were excluded due to inadequate material for molecular analysis. Among 47 cases eligible for HRD analysis, 9 samples were excluded due to post-analytical failures. HRD profile was performed in 38 samples. The mean age of BC diagnosis was 51.4 years (range 49,8-52,2 years); 52.6% were premenopausal. Regarding molecular subtype distribution: 47.4% were Luminal B-like, 31.6% triple-negative, 15.8% Luminal A-like, and 2.6% HER2-positive. Half of the samples were from primary tumor sites and the remaining from metastatic sites. Most tumors were stage III (31.6%) at diagnosis followed by stage II (26.3%), IV (18.4%), and I (15.8%). Regarding germline status, 26.3% of patients harbored germline PVs, among these, 40% in BRCA1/2, 40% in other HRR genes, and 20% in non-HRR genes. The overall HRD prevalence was 31.6% (12/38). TNBC had a higher rate of HRD (66.7%; 8/12), followed by Luminal B-like (25%) and Luminal A-like (8.3%). None of the HER2-positive tumors had HRD phenotype. Patients with HRD tumors (n=12) were predominantly premenopausal (75%), had a younger age at BC diagnosis (mean age 49.8 years), presented more locally advanced disease (41.7% stage III), and in 75% of these cases, tumor samples were from the primary site. Thirty-three percent of patients with HRD tumors harbored germline PVs in HRR genes (2BRCA1,1BLM;1RAD51C); 60% of those were truncating variants (nonsense and frameshifts). Conclusions: This is the first Brazilian study to evaluate HRD in breast cancer, revealing a 31.6% prevalence consistent with international data. HRD was more common in aggressive tumors subtypes, younger patients, and those with germline PVs in HRR genes. The identification of alterations beyond classical BRCA1/2 mutations expands the molecular spectrum of HRD.

L. In Hee¹, L. Soo Jung¹, K. Byeongju², L. Jeeyeon², P. Ho Yong², M. Joon Suk², L. Yangsoo³, Y. Jung Dug⁴, K. Eun Ae⁵, J. Seolhwa⁵, K. Jieun⁵, C. Yee Soo¹; ¹Oncology/Hematology, Kyungpook National Chilgok University Hospital, Daegu, KOREA, REPUBLIC OF, ²Breast Surgery, Kyungpook National Chilgok University Hospital, Daegu, KOREA, REPUBLIC OF, ³Surgery, Kyungpook National Chilgok University Hospital, Daegu, KOREA, REPUBLIC OF, ⁴Plastic and Reconstructive Surgery, Kyungpook National Chilgok University Hospital, Daegu, KOREA, REPUBLIC OF, ⁵Breast Cancer Precision Medicine Institute, Kyungpook National Chilgok University Hospital, Daegu, KOREA, REPUBLIC OF.

Background: MAGED1 (Melanoma Antigen Gene Family, Member D1) is a member of the MAGE gene family, which is notable for its involvement in various cellular processes, including apoptosis, cell cycle regulation, and differentiation. Furthermore, previous studies showed that MAGED1 is required for double-strand breaks (DSB) repair via homologous recombination repair (HRR) and downregulation of MAGED1 sensitizes cancer cells to DNA-damaging agents. In this study, we aimed to investigate the role of MAGED1 in HRR in TNBC and BRCA1/2 mutations breast cancer. Methods: In this study, we validated MAGED1 function in HRD using siRNA in TNBC and BRCA mutant breast cancer cell lines. Cell viability was assessed by CytoFLEX after treatment with drugs (cisplatin, olaparib and niraparib) at several concentrations (0–300 µM) in Hs578T, HCC-1937 (BRCA1 5382insC), and BT-474 (BRCA2 c.0391C>A) breast cancer cell lines. Cells were transfected with control siRNA (si-Cont) or two independent siRNAs (si- MAGED1 #1, si-MAGED1 #2) targeting MAGED1. Result: Silencing of MAGED1 by siRNA in Hs578T (TNBC), HCC-1937 (BRCA1 5382insC), and BT-474 (BRCA2 c.0391C>A) cells led to enhanced sensitivity to the PARP inhibitors olaparib and niraparib, as assessed by flow cytometry using CytoFLEX. The synergistic effect between MAGED1 suppression and PARP inhibitors was more pronounced in BRCA-mutant cell lines (HCC-1937 and BT-474) compared to the TNBC Hs578T cells. In contrast, MAGED1 knockdown did not enhance sensitivity to cisplatin. Combined treatment with cisplatin and MAGED1 knockdown resulted in an unexpected increase in cell viability, warranting further investigation. To further elucidate the role of MAGED1 in HRD, additional functional studies are being conducted involving RNF8, BARD1, RAD51, and ATM. Conclusions: Our study demonstrates that silencing of MAGED1 enhances sensitivity to PARP inhibitors in BRCA1/2-mutant breast cancer cell lines. These results suggest that MAGED1 is involved in HRD and could be a therapeutic target to enhance PARP inhibitor response in breast cancer.

M. Hou¹, M. Pan², S. MOI², F. Chen¹, C. Luo¹; ¹Division of Breast Oncology and Surgery, Department of Surgery, Kaohsiung Medical University Hospital, Kaohsiung city, TAIWAN, ²Graduate Institute of Clinical Medicine, Kaohsiung Medical University, Kaohsiung city, TAIWAN.

Background: Pathologic complete response (pCR) to neoadjuvant chemotherapy (NAC) is a strong prognostic marker in triple-negative breast cancer (TNBC). However, predictive biomarkers for pCR remain limited. This study aimed to identify gene expression profiles associated with pCR in TNBC to aid treatment stratification. Methods: Tumor samples were obtained from 16 TNBC patients treated with NAC at Kaohsiung Medical University Hospital (KMUH), including 5 pCR and 11 non-pCR cases. RNA sequencing (RNA-seq) was performed, and differentially expressed genes (DEGs) were identified using DESeq2 (|log₂FC| ≥ 2, adjusted p < 0.05). Expression profiles were compared with a validation cohort of 27 chemotherapy-responsive TNBC cases from The Cancer Genome Atlas (TCGA). Overlapping DEGs were identified via Venn diagram analysis. Drug-gene interaction databases were queried to assess therapeutic relevance. Results: A total of 175 DEGs were found in the KMUH cohort, with 146 upregulated and 29 downregulated in non-pCR tumors. Fifteen DEGs showed consistent expression patterns between KMUH non-pCR tumors and both KMUH pCR and TCGA responder cohorts. These genes were enriched in pCR samples and may serve as predictive biomarkers. Several candidates were identified as potentially druggable, suggesting relevance for therapeutic targeting. Conclusions: We identified a 15-gene signature associated with pCR in TNBC, validated across independent cohorts. These biomarkers may inform treatment decisions, improve patient stratification, and offer new directions for targeted therapy development in chemoresistant TNBC.

A. VELASCO¹, I. URDANIBIA¹, G. RODRIGUEZ¹, N. TUSET², G. ARIADNA³, S. MORALES³; ¹Medical Oncology, IRB LLEIDA, LLEIDA, SPAIN, ²Medical Oncology, HOSPITAL ARNAU DE VILANOVA DE LLEIDA, LLEIDA, SPAIN, ³Medical Oncology, HospItal arnau de vilanova de lleida, LLEIDA, SPAIN.

Background: Accurate HER2 status assessment is crucial for guiding treatment decisions in breast cancer. MammaTyper® is a CE-marked in-vitro diagnostic device utilizing real-time quantitative PCR (qRT-PCR) to quantify mRNA transcripts of ERBB2, ESR1, PGR, and MKI67. It provides an objective and standardized classification of breast cancer into intrinsic subtypes (Luminal A, Luminal B, HER2-enriched, and Triple-Negative), offering prognostic and predictive information that complements traditional immunohistochemistry (IHC). MammaTyper®’s high reproducibility and sensitivity are particularly valuable when subjective interpretation of IHC presents challenges.Objective: This study aimed to analyses whether the MammaTyper® platform improves the classification of HER2-equivocal breast cancer cases and to assess its concordance with FISH-based methods.Methods: We analysed 93 breast cancer patients. Estrogen receptor (ER) and progesterone receptor (PR) status, along with Ki67 index, were determined by IHC. HER2 status was initially assessed by IHC and FISH. MammaTyper® was then performed to classify intrinsic subtypes and determine HER2 status based on ERBB2 mRNA levels. Concordance between MammaTyper®, IHC, and FISH was evaluated.Results: Of the 93 patients, 83 (89%) were ER-positive, 42 (45%) PR-positive, and 51 (55%) had a high Ki67 index by IHC. Forty-two cases (45%) were classified as HER2-positive by IHC and FISH. MammaTyper® confirmed HER2-positivity in 25 (60%) of these cases. Among the 51 HER2-negative cases by IHC and FISH, MammaTyper® was positive in only 8 (15%). The overall concordance between MammaTyper® and the IHC/FISH combined assessment was 73%. Specifically, concordance was 60% for IHC/FISH-positive cases and 84% for IHC/FISH-negative cases. Of the 32 cases classified as HER2-positive by MammaTyper®, 23 (71%) had a HER2 FISH ratio of ≥2. Conversely, among the 55 cases classified as HER2-negative by MammaTyper®, 17 (31%) had a HER2 FISH ratio of ≥2. MammaTyper® showed a 72% concordance with IHC for HER2 phenotype determination, identifying a lower proportion of HER2-positive tumors (34%) compared to IHC (45%). This concordance was notably higher (84%) in HER2-negative cases.Conclusion: MammaTyper® demonstrates substantial concordance with IHC and FISH for HER2 status determination, particularly in HER2-negative cases. It identifies a smaller subset of HER2-positive tumors compared to IHC, suggesting its potential to refine the selection of patients for HER2-targeted therapies. Further clinical utility should be assessed by correlating MammaTyper® results with treatment response and survival outcomes.

A. Yarunin¹, A. Martin², M. Gomez-Rey², A. Baizan³, M. Sesé⁴, S. Clavé⁵, L. Camacho⁵, C. Saura⁶, J. Longshore¹, J. Hernández-Losa⁷, B. Bellosillo⁵, A. Vivancos²; ¹Global Oncology Diagnostics, AstraZeneca UK Limited, Cambridge, UNITED KINGDOM, ²Cancer Genomics, Vall d’Hebron Institute of Oncology, Barcelona, SPAIN, ³Breast Cancer Unit, Vall d’Hebron Institute of Oncology, Barcelona, SPAIN, ⁴Pathology, Hospital Universitari Vall d’Hebron, Barcelona, SPAIN, ⁵Pathology, Hospital del Mar, Barcelona, SPAIN, ⁶Oncology, Hospital Universitari Vall d’Hebron (HUVH), Barcelona, SPAIN, ⁷Pathology, Hospital Universitari Vall d’Hebron (HUVH), Barcelona, SPAIN.

Background: Activating mutations in the ESR1 gene represent a well-known mechanism of acquired resistance to aromatase inhibitors (AIs) in patients with hormone receptor-positive (HR+) metastatic breast cancer. Liquid biopsy, through the analysis of circulating free DNA (cfDNA), provides a non-invasive and dynamic approach to detect ESR1 mutations, enabling real-time monitoring of tumor evolution. This is especially relevant in the context of novel oral selective estrogen receptor degraders (SERDs), where ESR1 mutation profiling will inform treatment selection and act as a biomarker of resistance and response. As these agents become part of routine clinical care, the ability to accurately detect and monitor ESR1 mutations in plasma is critical. As several next-generation sequencing (NGS)-based assays are commercially available, studies comparing their analytical concordance in detecting ESR1 mutations in real-life plasma samples are needed. Methods: This study aims to assess the analytical concordance between two amplicon-based NGS assays (Oncomine™ Precision Assay and Pillar Biosciences™ OncoReveal Essential LBx) and a hybrid-capture based NGS test, VHIO360 (tech transfer to VHIO of the Guardant360® assay) for the detection of ESR1 mutations in cfDNA. VHIO360, with established clinical validation is considered as the reference gold standard. One hundred plasma samples from patients with known ESR1 mutation genotypes as per VHIO360 testing (from the VHIO Molecular Prescreening Program), containing 64 pathogenic/ likely pathogenic variants in ESR1 (59 corresponding to hotspot Y537S/N/C, D538G, and E380Q/K) will be additionally tested with the aforementioned amplicon-seq based approaches. Mutant variant frequencies in the sample set are present at a wide range (0.02-69%). Concordance is measured using metrics such as overall agreement, positive and negative percent agreement, and analytical sensitivity across variant allele frequencies. Discordant results will be further investigated to identify assay-specific limitations or systematic biases, including variant dropout or differential coverage. Results: Results provide a comparative assessment of analytical performance across platforms, focusing on concordance in detection of clinically relevant ESR1 mutations and determination of assay-specific limits of detection. Particular attention will be paid to samples with low variant allele frequency (VAF) mutations, where differences in platform sensitivity may impact clinical interpretation. Conclusions: This study will provide critical evidence for the analytical performance and clinical reliability of commonly used NGS-based assays for detecting ESR1 mutations in plasma. As SERDs and other ESR1-targeted therapies become integrated into standard-of-care for HR+ metastatic breast cancer, accurate and validated mutation testing is essential in order to guide clinicians in selecting appropriate local testing approaches.

C. Lee¹, T. Hoque², S. Qazi², S. saund³, C. Hsiao², T. Joh²; ¹R&D, Sapu Bioscience, LLC, San Diego, CA, ²Clinical, Sapu Bioscience, LLC, San Diego, CA, ³Manufacturing, Sapu Bioscience, LLC, San Diego, CA.

Background Everolimus (an mTORC1 inhibitor) packaged as Sapu003, an injectable deciparticle formulation, has shown superior cytotoxic activity compared with paclitaxel in glycolysis-addicted tumors. Leveraging retrospective pan-cancer transcriptomic datasets, we explored whether expression of core mTOR-complex scaffolds and metabolic modulators defines a subgroup most likely to benefit from Sapu003 and complementary metabolic therapy. Methods Publicly available survival cohorts (METABRIC, KM-Plotter, TCGA) were mined for overall survival (OS) and relapse-free survival (RFS) according to mRNA quartiles of mTORC1 (RPTOR, AKT1S1) and mTORC2 (RICTOR, MAPKAP1) components plus downstream metabolic genes (SKP2, LDHA, MK, STK11). Differential survival was validated across >9,000 samples spanning breast, lung, gastric, ovarian, RCC and AML. A composite biomarker was derived by recursive partitioning. Results High RICTOR expression—reflecting low baseline mTORC1 activity—was independently associated with prolonged OS (110 months vs 66 months) and RFS (48 vs 31 months) in breast cancer (P = 0.0029 and 4.7 × 10⁻⁷, respectively). Low RPTOR amplified OS gains conferred by low-AKT3, low-TSC1 or low-RHEB status in lung adenocarcinoma (median 118-139 vs 57-69 months; P < 10⁻⁵). Across seven tumour types, simultaneous low expression of ≥3 scaffold genes (RPTOR, MLST8, AKT1S1, MAPKAP1, PRR5L, RICTOR, PRR5, DEPTOR) predicted a ≥1.8-fold improvement in overall survival (OS). In glycolysis-addicted U-87 MG xenografts, Sapu003 achieved durable tumour regression with no weight loss. Conclusions A High RICTOR / Low RPTOR ± metabolic-gene-low signature enriches for tumors with intrinsic dependence on adaptive metabolism and may predict enhanced responsiveness to Sapu003. A Phase I/Ib study will prospectively enroll biomarker-positive patients (high-RICTOR / low-RPTOR breast and non-breast solid tumors) and incorporate pharmacodynamic confirmation of mTORC1 blockade and metabolic rewiring.

M. Lin¹, T. Du², J. Tang²; ¹Department of ultrasound, Sun Yat-sen University Cancer Center, Guangzhou, CHINA, ²Department of breast oncology, Sun Yat-sen University Cancer Center, Guangzhou, CHINA.

Background: Neoadjuvant immunotherapy combined with chemotherapy has shown promising potential in breast cancer management, but predicting pathological complete response (pCR) remains challenging. This study aimed to develop and validate a predictive model for pCR in breast cancer patients receiving neoadjuvant immunotherapy. Methods: We retrospectively analyzed breast cancer patients treated with neoadjuvant immunotherapy at Sun Yat-sen University Cancer Center (SYSUCC) between November 2019 and April 2025. pCR was defined as Miller-Payne grade 5, while grades 1-4 were classified as non-pCR. Clinical, pathological, and laboratory variables—including age, body mass index (BMI), menopausal status, TNM stage, histological type, Ki67 expression, chemotherapy regimen, immune checkpoint inhibitor type, immunotherapy cycles, treatment pattern, neutrophil-to-lymphocyte ratio (NLR), albumin, C-reactive protein, and serum amyloid A—were analyzed. Univariate logistic regression identified variables significantly associated with pCR (p < 0.05), which were further analyzed by multivariate logistic regression. Variables with p < 0.1 in the multivariate model were included in the final predictive model. The cohort was randomly divided (2:1) into a training and validation set. Model performance was assessed using receiver operating characteristic (ROC) curves, calibration plots, and decision curve analysis (DCA). Results: A total of 233 patients were included, and 52.8% achieved pCR. The final model incorporated pretreatment clinical T stage, pretreatment Ki67 expression, pretreatment NLR, chemotherapy regimens, and number of immunotherapy cycles. The model demonstrated good discrimination, with an AUC of 0.782 in the training set and 0.818 in the validation set. Calibration curves indicated good agreement between predicted and observed pCR probabilities. DCA showed the model provided a net clinical benefit over “treat all” and “treat none” strategies across a range of threshold probabilities. Conclusions: We developed and validated a practical predictive model for pCR in breast cancer patients receiving neoadjuvant immunotherapy. The model may assist clinicians in individualizing treatment strategies and optimizing patient outcomes. Prospective validation is warranted.

C. Helal¹, N. Schintu², J. Jin², L. Pinon², E. Montaudon³, L. Sourd³, H. Derrien³, M. Nurmik⁴, M. Simon², M. Parrini⁴, L. Cabel¹, S. Descroix²; ¹Paris, Department of Medical Oncology, Institut Curie-Institut of Women’Cancers, Paris, FRANCE, ²Paris, Institut Curie, CNRS UMR168, Laboratoire Physico Chimie Curie, Institut Pierre-Gilles de Gennes, PSL Research University, Paris, FRANCE, ³Paris, Laboratory of Preclinical Investigation, Translational Research Department, Institut Curie, Paris, FRANCE, ⁴Paris, Institut Curie, Inserm U1339, CNRS UMR3666 CNRS, PSL Research University, Paris, FRANCE.

Background: Metastatic breast cancer (mBC) is the leading cause of cancer mortality in women, yet treatment selection in later lines remains empirical. Patient-derived xenografts (PDX) and organoids (PDO) have been shown to predict patient’s drug response, however their clinical utility for real-time clinical decision-making is limited by lengthy turnaround times and a low tumor take for luminal breast cancers, and rapid functional assays are urgently needed. We developed a microfluidic tumor‑on‑chip (ToC) for rapid exvivo chemosensitivity profiling to demonstrate the potential of ToC using cancer cells from BC PDX and a fresh patient’s tumor sample. Methods: After tumor dissociation, tumor cells were isolated and embedded in 3D in collagen type I inside microfluidic device chips (AIMBiotech). Viability of tumor cells was assessed using a live/dead assay after several days of drug exposure (carboplatin, paclitaxel, 5‑fluorouracil, and trastuzumab‑deruxtecan if applicable) and compared to tumor growth measured invivo in the corresponding PDX model using a caliper. A panel of 6 PDX models was selected to reflect distinct BC subtypes and chemosensitivity profiles: 4 TNBC, 1 HER2 3+ and 1 ER+. To mimic clinical biopsies, we engineered a miniaturized chip by casting PDMS in 3D-printed silanized molds optimized for low cell input, and evaluated its performance using simulated core needle biopsies (18G) on PDX tumors and a patient sample. Results: We first investigated the response of ToC derived from fresh PDX samples. Differential drug responses were observed after 4 days of on-chip culture, suggesting that this duration is required to discriminate between sensitive and resistant models. The #152 PDX model (TNBC) showed a strong invivo response to carboplatin and paclitaxel, mirrored by a significant drop in cell viability on the ToC at D4. In contrast, neither mouse nor ToC responded to 5-FU. We performed live imaging to validate our endpoint measurement method, and comparable patterns of drug sensitivity were obtained. Overall, when a drug was identified as effective in the PDX model, we observed ToC sensitivity at D4 in 78 % of cases (in 7 cases out of 9, sensitivity was correctly detected on ToC in 6 PDX models using 4 drugs). Conversely, when the drug was resistant in PDX, we confirmed resistance in 100% of cases (11/11 using 3 drugs). Altogether, these findings highlight the correlation between the two models and demonstrate the ability of ToC platform to deliver rapid functional readouts within a clinically actionable timeframe of 4 days. Next, we successfully generated a functional ToC model using a fresh human primary tumor of TNBC; exposure to paclitaxel suggested tumor sensitivity. We finally developed an innovative microfluidic device compatible with patient biopsy samples. From three 18G core needle biopsies per tumor, we consistently isolated 100,000 to 200,000 viable tumor cells, allowing the generation of up to 12 individualized ToC units and demonstrating the feasibility of establishing ToC from limited biopsy material. Conclusions: Microfluidic ToC replicates PDX drug responses within 4 days using minimal tissue, a timeframe compatible with clinical decision-making. Optimization for cell use and ease of automation supports future clinical-scale implementation and high-throughput workflows, addressing a key gap in functional precision oncology. Prospective studies are warranted to validate ToC-guided therapy in BC.

H. Fritsch¹, L. Borchert¹, L. Paul², K. Hide-Moser², C. Vogel-Minea³, C. Eichler¹; ¹Breast Cancer Centre, St. Franziskus Hospital Münster, Münster, GERMANY, ²Breast Cancer Centre, University Hospital Cologne, Cologne, GERMANY, ³Breast Cancer Centre, Rottal Inn Kliniken, Eggenfelden, GERMANY.

Improved Diagnostic Accuracy for Breast Implant Rupture Through Strain Elastography – Results of a User-Based Evaluation Using Standardized Image Sets Background: Ultrasound is a widely used tool for evaluating breast implant integrity, but its diagnostic accuracy remains limited by operator dependency. This study investigated whether the addition of strain elastography can improve diagnostic sensitivity and specificity in detecting implant rupture. Methods: Based on a standardized animal model, 15 silicone breast implants (intact and ruptured) were scanned using both B-mode ultrasound and strain elastography. From these scans, 30 paired images were selected for an online evaluation. In total, 99 participants reviewed the image sets and were asked to determine the presence or absence of implant rupture — first using B-mode images alone, and then with the corresponding elastography images. Results: With B-mode alone, participants achieved a diagnostic sensitivity of 81.4% and specificity of 71.0%. When strain elastography was added, sensitivity rose to 90.9%, and specificity increased to 87.6%.Conclusion: The integration of strain elastography significantly enhances diagnostic performance in assessing breast implant rupture. In particular, it reduces false-positive rates while increasing detection of true ruptures. Using standardized image data from an animal model enables reproducible evaluation and may serve as a basis for future training and quality assurance in breast imaging.

S. E. Rudd¹, J. Van Zuylekom², B. Blyth², M. Harris³, E. Sztangret⁴, P. S. Donnelly¹; ¹School of Chemistry and Bio21 Molecular Science and Biotechnology Institute, University of Melbourne, Melbourne, AUSTRALIA, ²Models of Cancer Translational Research Centre, Peter MacCallum Cancer Centre, Melbourne, AUSTRALIA, ³R&D, Clarity Pharmaceuticals, Eveleigh, AUSTRALIA, ⁴Medical Affairs, Clarity Pharmaceuticals, Eveleigh, AUSTRALIA.

Background and Purpose: Human epidermal growth factor receptor 2 (HER2) overexpression occurs in several cancers including breast cancer. HER2 targeted therapies such as trastuzumab have significantly improved outcomes for HER2-positive breast cancer patients, however resistance and recurrence remain challenging. Radioimmunotherapy (RIT) utilizing trastuzumab could provide an additional treatment option for HER2+ breast cancer patients. We produced the theranostic pair ^64/67Cu-SAR-trastuzumab and assessed the PET imaging and therapeutic potential of these agents in a HER2-posive mouse model. Methods: A bifunctional variant of the macrobicyclic cage amine sarcophagine (SAR) was conjugated to the monoclonal antibody trastuzumab. Proof of concept studies were performed using SKOV-3 tumour bearing BALB/c nu/nu mice, intravenously injected with a single dose of ⁶⁴Cu-SAR-trastuzumab or ⁶⁷Cu-SAR-trastuzumab to compare biodistribution of the theranostic pair. Tumor uptake was assessed using PET/CT imaging and/or ex vivo biodistribution studies. In therapy studies, a single dose of saline, cold SAR-trastuzumab, or ⁶⁷Cu-SAR-trastuzumab was intravenously administered and endpoints assessed. Results: PET/CT imaging 48 hours after administration with ⁶⁴Cu-SAR-trastuzumab demonstrated excellent tumor uptake, with mean SUV_max = 21.0 ± 2.5. ⁶⁴Cu-SAR-trastuzumab and ⁶⁷Cu-SAR-trastuzumab showed comparable high tumour-specific uptake in biodistribution studies, with tumour uptake at 48 hours of 61 ± 6 % IA/g for ⁶⁴Cu-SAR-trastuzumab and 59 ± 7 % IA/g for ⁶⁷Cu-SAR-trastuzumab. Uptake of ≤10 % IA/g was observed for both agents in all other organs measured. Mice treated with ⁶⁷Cu-SAR-trastuzumab tolerated the treatment well with limited side effects or AEs. A transient reduction in animal body weight (≤7 % of baseline) was observed in all cohorts but this did not differ significantly from the vehicle control. Treated mice also showed significant tumor growth inhibition and extended survival in a dose dependent manner compared with the control groups. The median survival was 32 days in the control groups and 102 days in the 9 MBq group. Conclusion: These results demonstrate the potential of ^64/67Cu-SAR-trastuzumab as a new theranostic approach for the treatment of HER2-positive cancer. This theranostic pair exhibit similar biodistribution in the same model, with the high tumor uptake of ⁶⁴Cu-SAR-trastuzumab resulting in high quality PET images. ⁶⁷Cu-SAR-trastuzumab effectively increased the survival of all treated groups. The outstanding therapeutic outcomes demonstrate the potential of ⁶⁷Cu-SAR-trastuzumab as an RIT candidate for translation to clinical trials.

C. Nelson, J. Flores, M. Lopez, A. Simonson, A. Nunez, S. Johnson, A. Rojas, A. Stackpole, J. Lund, P. Forofontov, L. Wilson, J. Garcia, I. Sturgill, A. Cunningham, M. Wick; XenoSTART, The START Center for Cancer Research, San Antonio, TX.

Background: Therapies targeting the PI3-K/Akt pathway are an active field of research. Capivasertib in combination with fulvestrant was recently approved in endocrine therapy-resistant, advanced or metastatic ER+/HER2- breast cancer patients harboring hotspot mutations in AKT, PIK3CA or PTEN genes. However, sensitivity to capivasertib therapy in ER+ breast cancers without these specific mutations is unclear. To this end, we established a panel of ER+ breast XPDX models with or without PI3-K/Akt pathway mutations and evaluated each in vivo with capivasertib alone or in combination with fulvestrant. Methods: Fifty-six previously developed XPDX models representing ER+ breast cancer were evaluated in this study. Models were grown subcutaneously in female athymic nude mice supplemented with estradiol in drinking water when necessary and ER expression confirmed at multiple passes; mutations in and out of the PI3-K/Akt pathway were determined by WES and RNA_seq. For in vivo studies, capivasertib was administered twice daily by oral gavage at 100 mg/kg/dose on a four-days on/ three-days off cycle through study completion. Endpoints included tumor volume and time from treatment initiation with %T/C values and tumor regression reported at study completion; a T/C of ≤ 20% versus control was considered sensitive. Tumor regression (%T/C<0%) versus Day 0 tumor volume was also reported. Results: Identified PI3-K/Akt pathway mutations included AKT1^E17K (N=3), PIK3CA^{E542X/E545X/H1047X} (N=17) and other PIK3CA variants (N=8); alterations in PTEN were also reported in some models. In this study, capivasertib was found active in 12/56 models, including partial or complete tumor regressions in 3/12 and tumor stasis in 2/12 models. 1/12 models with AKT1^E17K, 5/17 models with PIK3CA^{E542X/E545X/H1047X} and 3/8 models with other PIK3CA variants were found sensitive to single agent capivasertib. Interestingly, the most sensitive model tested harbored only a PTEN deletion in the PI3-K/Akt pathway and 2/3 of the most sensitive models were established from HER2+ patients. In addition, co-treatment with fulvestrant antagonized efficacy in several of the sensitive models. Conclusion: We have benchmarked a panel of ER+ breast XPDX models with capivasertib alone and in combination with fulvestrant and compared activity of models based on mutations in and out of the PI3-K/Akt pathway. This data is a valuable tool in developing PI3-K/Akt pathway therapies in ER+ breast cancer.

T. Usui¹, T. Miyake¹, T. Watabe², K. Hiroki³, K. Abe¹, T. Ryu¹, S. Yasufumi¹, N. Masunaga¹, C. Mishima¹, M. Tsukabe¹, Y. Sota¹, T. Yoshinami¹, T. Tanei¹, K. Shimazu¹; ¹Breast and endocrine surgery, The University of Osaka, Suita, JAPAN, ²Nuclear Medicine and Tracer Kinetics, The University of Osaka, Suita, JAPAN, ³Advanced Radioisotope Medicine, Institute for Radiation Sciences, The University of Osaka, Suita, JAPAN.

Background: Accurate detection of sentinel lymph node (SLN) metastasis remains a diagnostic challenge in breast cancer. This study aimed to evaluate the feasibility of immuno-positron emission tomography (PET) using ⁶⁴Cu-labeled cetuximab for detecting SLN metastasis and to assess its preliminary therapeutic effect in a preclinical model of epidermal growth factor receptor (EGFR)-positive breast cancer. Methods: The SLN metastasis model was established using EGFR-overexpressing MDA-MB-468 breast cancer cell line. [⁶⁴Cu]Cu-PCTA-cetuximab was administered either intravenously (IV; 5.8 ± 0.9 MBq; n=12) or intradermally / subdermally (ID/SD; 4.3 ± 0.4 MBq; n=11) into the parapapillary region of the tumor-containing mammary gland. PET/computed tomography scans were performed 24 h after tracer injection, with delayed imaging at 48 h to evaluate retention and contrast. For comparison, ¹⁸F-FDG PET was also performed via IV (9.1 ± 1.4 MBq, n = 4) or ID/SD (5.4 ± 2.2 MBq, n = 3) routes in the same cohort. SLNs were identified using blue dye, and were pathologicaly evaluated for metastases. To evaluate therapeutic efficacy, additional mice received IV [⁶⁴Cu]Cu-PCTA-cetuximab at 9, 18, or 37 MBq (n = 3 per group), and tumor diameters and body weight were monitored over time in comparison to untreated controls. Results: After intravenous administration of [⁶⁴Cu]Cu-PCTA-cetuximab (n=12), radiotracer accumulation was detected in the primary tumor in all mice and in the axilla of eight mice (67%, SUV_max 1.24 ± 0.51), all of which were found to have SLNs with histologically confirmed metastasis. The sensitivity, specificity, accuracy, and negative and positive predictive values for PET imaging in this group were 89%, 100%, 92%, 75%, and 100%, respectively. In the ID/SD group , all 11 mice showed high tracer accumulation in both the primary tumor and axillary nodes (SUV_max 4.28 ± 1.19); however, only six mice (55%, SUV_max 5.01 ± 1.12) had histologically confirmed SLN metastasis. The sensitivity, specificity, accuracy, and positive predictive values for this route were 100%, 0%, 55% and 55%, respectively. SLN metastasis was not detectable by ¹⁸F-FDG PET regardless of the administration route. Regarding therapeutic evaluation, all IV-treated groups showed tumor growth suppression compared to controls. The 37 MBq group exhibited the most pronounced tumor reduction but also showed a marked decrease in body weight, indicating potential systemic toxicity at higher doses. Conclusions: Immuno-PET with IV-administered [⁶⁴Cu]Cu-PCTA-cetuximab demonstrated high precision for diagnosis of SLN metastasis and showed promising therapeutic efficacy in an EGFR-positive breast cancer model. While ID/SD administration showed lower diagnostic specificity, its strong accumulation in SLNs indicates potential utility for localized therapy. These findings support further exploration of ⁶⁴Cu-labeled cetuximab for theranostic applications in breast cancer.

T. Sahin¹, C. Orhan², I. Ozercan³, H. Muglu⁴, B. Er⁵, A. Akyildiz¹, S. Tunbekici⁶, A. Oruc⁷, M. Aykan⁸, B. Koylu⁹, O. Akdogan¹⁰, I. Deliktas Onur¹¹, O. Ates¹¹, O. Yazici¹⁰, F. Selcukbiricik⁹, N. Karadurmus⁸, M. Araz⁷, D. Erdem¹², E. Goker⁶, A. Bilici⁴, S. Aksoy¹, D. Guven¹; ¹Department of Medical Oncology, Hacettepe University, Ankara, TURKEY, ²Department of Animal Nutrition, Fırat University Faculty of Veterinary Medicine, Elazig, TURKEY, ³Department of Medical Pathology, Fırat University Faculty of Medicine, Elazig, TURKEY, ⁴Department of Medical Oncology, Istanbul Medipol University, Istanbul, TURKEY, ⁵Department of Biology, Faculty of Science, Firat University, Elazig, TURKEY, ⁶Department of Medical Oncology, Ege University, Izmir, TURKEY, ⁷Department of Medical Oncology, Necmettin Erbakan University, Konya, TURKEY, ⁸Department of Medical Oncology, Health Science University, Gulhane School of Medicine, Ankara, TURKEY, ⁹Department of Medical Oncology, Koç University School of Medicine, Istanbul, TURKEY, ¹⁰Department of Medical Oncology, Gazi University, Ankara, TURKEY, ¹¹Department of Medical Oncology, Dr. Abdurrahman Yurtaslan Ankara Oncology Training and Research, Ankara, TURKEY, ¹²Department of Medical Oncology, Samsun Medical Park Hospital, Samsun, TURKEY.

Background: Trastuzumab deruxtecan (T-DXd) has demonstrated significant clinical efficacy in metastatic breast cancer (mBC) with varying HER2 expression, including HER2-low and HER2-ultralow. Statins have been shown to increase HER2 membrane localization and stability, potentially enhancing responsiveness to HER2-targeted therapies. Given this mechanistic rationale, we investigated the potential of statins to enhance T-DXd activity through a translational approach combining a preclinical HER2-negative breast cancer model and a real-world clinical cohort of patients with HER2-positive mBC treated with T-DXd. Methods: In the preclinical arm, 60 female Wistar albino rats were induced with hormone receptor-positive, HER2-negative mammary tumors using N-methyl-N-nitrosourea (MNU). Animals were randomized into five groups: Control, MNU-only, MNU+T-DXd, MNU+statin, and MNU+T-DXd+statin. Tumor volume, number, and survival were monitored. HER2 expression, apoptotic markers (Bax, Bcl-2, Caspase-3/9), and proliferative signaling pathways (p-AKT, p-ERK) were assessed via immunohistochemistry (IHC) and Western blotting. In the clinical arm, 109 patients with HER2-positive mBC who received T-DXd were retrospectively analyzed. Patients were stratified based on concomitant statin use. Progression-free survival (PFS) and overall survival (OS) were evaluated using Kaplan-Meier analysis and log-rank testing. Results: The combination of T-DXd and statin significantly reduced tumor volume compared to either monotherapy (p < 0.0001), with a strong trend toward reduced tumor count (p = 0.051). HER2 protein levels were elevated in the statin and combination groups, as demonstrated by both IHC and Western blot analyses. The combination group exhibited significantly higher levels of pro-apoptotic markers (↑Bax, ↑Caspase-3), decreased anti-apoptotic Bcl-2, and stronger suppression of p-AKT and p-ERK (p < 0.0001 for all vs. T-DXd). In a real-world cohort of 109 heavily pretreated patients with HER2-positive mBC receiving T-DXd, concomitant statin use was associated with significantly improved mPFS (21.83 months vs. 14.02 months; p = 0.049) and median OS (not reached vs. 19.68 months; p = 0.023) compared to non-users. Conclusions: Our findings revealed that statin use significantly enhanced the efficacy of T-DXd in both preclinical and clinical settings. These results support the potential role of statins as cost-effective and widely available agents to potentiate HER2-targeted therapies. Prospective clinical trials are warranted to validate these observations.

H. Mori¹, H. Moon², Q. Chen¹, A. Patino³, K. Ramamurthy³, J. Choi⁴, J. D. McPherson⁵, R. Cardiff¹, J. Snyder³, A. D. Borowsky¹; ¹PATHOLOGY, UC DAVIS, SACRAMENTO, CA, ²Research, Eli Lilly, South San Francisco, CA, ³Cell and Molecular Biology, Duke University, Durham, NC, ⁴Department of Biomedical Sciences, Seoul National University College of Medicine, Seoul, KOREA, REPUBLIC OF, ⁵Biochemistyr, UC DAVIS, SACRAMENTO, CA.

Introduction: Estrogen receptor positive (ER+) subtypes of mammary adenocarcinoma comprise 79% of all breast cancer diagnosis and 67% of all breast cancer mortality. The paucity of models of ER+ mammary cancer that mimic human disease and response to treatment has limited critical preclinical study of mechanisms and new therapies for ER+ breast cancer. The Stat1 knockout, 129S6/SvEvTac-Stat1tm1Rds (Stat1-/-), females develop luminal type FoxA1+, ER+, and PR+ mammary carcinomas after prolonged latencies. Initial studies showed that a cell line derived from a Stat1-/- mammary carcinoma was tumorigenic in syngeneic mice, but non-tumorigenic in ovariectomized (Ovx) mice. Methods: Syngeneic tumor transplants, either from the SSM2 cell line established as a primary culture from a single mammary tumor in a Stat1-/- female, or from direct biopsy transplants of two different mammary tumors in Stat1-/- female mice, were allowed to establish growth under intact ovary syngeneic (129SvEv) female mice. After 3 weeks, these tumor bearing mice were Ovx’ed and growth was monitored, and tumor tissues harvested for analysis.Results: Ovx performed after SSM2 tumors establish growth results in ovarian hormone independent growth. The viable post-Ovx tumors were primarily composed of metaplastic CK14+ basal type cells with a high percentage p53 immunohistochemistry (IHC) positive “mutation pattern”, rather than the original luminal type tumors with low percent “wild type” pattern p53. Comparing whole exome sequences of ER+ Stat1-/- mammary tumors before and after Ovx, revealed basal keratins, mesenchymal (EMT) phenotypes, and unique mutation profiles in genes, including Trp53 and Prlr, in the estrogen-independent tumors (Prlr mutations were seen in both estrogen dependant and independant tumors). Our experimental findings are consistent with the clinical evidence of tumor heterogeneity of ER+ breast cancers in patients in recent whole genome sequencing studies. Similarly, spontaneous Stat1-/- tumors with high percentage p53 “mutation pattern” were more basaloid and grew rapidly after Ovx, while retaining high expression of ER and FoxA1. Conclusions: This study demonstrates that the STAT1-/-, ER+ estrogen dependent breast cancers can become resistant to through clonal selection of mammary cells comprised of metaplastic p53+/CK14+ basaloid cells. This implies that p53 mutation in human breast cancers may be a gain-of-function with respect to driving a more basal and/or metaplastic phenotype contributing to this highest mortality subtype of breast cancers (high grade/ Luminal B-type ER+ carcinomas).

T. J. Kalantzakos¹, Y. Zhou¹, X. Liu¹, J. Proehl¹, C. Durfee¹, I. Tamayo¹, B. Troness¹, A. Soni¹, H. B. Gupta¹, R. S. Harris²; ¹Biochemistry & Structural Biology, UT Health San Antonio, San Antonio, TX, ²Biochemistry & Structural Biology, Howard Hughes Medical Institute, UT Health San Antonio, San Antonio, TX.

Breast cancer is the most common cancer diagnosis in women. Clinical studies with triple-negative breast cancer (TNBC) are encouraging for immunotherapy (αPD-1) combined with chemotherapy (paclitaxel and/or carboplatin). However, additional clinical advances could be pursued more rapidly with preclinical TNBC models including syngeneic mammary tumor cell lines. Here, we generate two mammary tumor cell lines and show differential responsiveness to immunotherapy in vivo. Spontaneous mammary tumors from C57BL/6J MMTV-Cre Tp53^fl/+ animals were passaged serially in cell culture and in vivo in the mammary fat pad of fully wildtype animals. The resulting lines, MM001i and MM008i, lost Trp53 and formed 1000 mm³ tumors in the mammary fat pad within 21-28 days. Despite originating from the same genetic background, MM001i and MM008i exhibit differential responses to immunotherapy. For αPD-1 immunotherapy, MM001i is poorly responsive and MM008i is strongly responsive with near-complete tumor regression. In comparison, both MM001i and MM008i respond rapidly to αCTLA-4 therapy. Both models express unique tumor antigens as evidenced by immunity to subsequent engraftment. Primary MM008i tumors exhibit greater T cell infiltration, and CD8-positive T lymphocytes are required for αPD-1 responses. These TNBC models are promising for further mechanistic studies and testing future single and combinatorial therapies.

S. Inoue¹, M. Miki¹, M. Yamamoto², H. Tanino³, T. Kunihisa¹, S. Inubushi¹; ¹Division of Breast Surgery, Kobe University Graduate School of Medicine, kobe, JAPAN, ²Department of Breast and Endocrine Surgery, Kobe University Hospital, kobe, JAPAN, ³Division of Breast Surgery, Naga Municipal Hospital, Wakayama, Japan, Kinokawa, JAPAN.

Background: Triple-negative breast cancer (TNBC) is known to be highly heterogeneous compared to other subtypes. Establishing reliable preclinical models is essential for personalized treatment strategies, such as selecting effective drugs for each patient. Patient-derived xenograft (PDX) models, created by transplanting patient tumors into immunodeficient nude mice, are considered to more accurately replicate tumor diversity and drug responses. This study aimed to evaluate the retention of clinically relevant biomarkers in TNBC-PDX models.Methods: With approval from the Kobe University Institutional Review Board (Approval number: B200006), tumor tissues were collected from TNBC patients who met the inclusion criteria and provided informed consent. The tissues were implanted into nude mice to establish TNBC-PDX models. Expression of key biomarkers—EGFR, AR, CK5/6, HER2, and Trop-2—was assessed in both the original patient tumors and corresponding PDX tumors using immunohistochemistry.Results: EGFR and Trop-2 expression was retained in all models. However, AR expression was not preserved. CK5/6 and HER2 showed variable expression across cases.Conclusion: TNBC-PDX models are promising tools for drug evaluation in personalized medicine. However, since some biomarkers, such as AR and CK5/6, may not be consistently retained, prior validation of biomarker expression is essential for accurate therapeutic evaluation.

J. Yang, M. Ku; Department of Radiology, Yonsei University, Seoul, KOREA, REPUBLIC OF.

With breast cancer incidence rising worldwide—including in younger Korean women in their 40s—there is an urgent need for advanced disease models that capture mechanobiological features unique to this population. Post-surgical hormonal alterations and the high prevalence of bone metastases further complicate outcomes, underscoring the need for strategies addressing both primary tumor biomechanics and skeletal microenvironments. To meet this need, we developed a three-dimensional (3D) breast cancer spheroid model using biomimetic extracellular matrix materials to replicate native tissue mechanics and architecture, supported by collagen-producing stem cells. High-resolution confocal microscopy, immunofluorescent labeling of F-actin, β-Tubulin, and Vimentin, and computational analysis enabled detailed visualization of cytoskeletal organization and intracellular changes. Mechanobiology-driven evaluation identified membrane-type 1 matrix metalloproteinase (MT1-MMP) as a key regulator of invasive behavior in TNBC spheroids, mediating actin-based protrusions such as filopodia and lamellipodia. Functional assays showed that targeted MT1-MMP inhibition altered cellular mechanics and significantly enhanced chemotherapy efficacy in the 3D model.To validate this platform, dose-dependent accumulation and spatial distribution of doxorubicin were analyzed using confocal Z-stack imaging. Quantitative radial intensity profiling showed approximately 20% decrease in fluorescence signal from the spheroid edge to core at higher concentrations, indicating measurable penetration barriers. Channel-to-channel distance analyses confirmed heterogeneous but quantifiable drug distribution patterns. We also assessed drug distribution and viability using Calcein AM and fluorescence imaging. MT1-MMP inhibitor alone maintained high viability (~80% live-cell area) with spheroid diameters of ~180-200 µm and minimal drug uptake. Chemotherapy monotherapy reduced viability to ~45% with high drug uptake and decreased spheroid integrity (~160-180 µm). Simultaneous combination improved uptake but reduced viability to ~40%, with similar spheroid size. Notably, sequential MT1-MMP inhibitor pre-treatment preserved higher viability (~15%) while maintaining comparable drug uptake and spheroid size (~180 µm), indicating improved penetration with moderated cytotoxicity.To further characterize the model, XF Seahorse extracellular flux analysis was used to evaluate metabolic responses under different treatment conditions. By quantifying glycolytic activity and oxidative phosphorylation (OXPHOS), we identified shifts in ATP production pathways across monotherapy, combination, and sequential regimens. Sequential MT1-MMP inhibitor pre-treatment induced a more balanced metabolic profile, confirming that MT1-MMP inhibition can modulate and enhance control over OXPHOS activity. Protein marker analysis further supported the mechanism, showing that MT1-MMP-centered mechano-regulation can modulate chemosensitivity in this 3D spheroid system.Therefore, this advanced 3D spheroid system represents an essential platform for Mechanomedicine research and evaluation, closely replicating clinical tumor architecture and mechanical microenvironments. It enables precise study of mechanobiological drivers of metastasis and drug response, providing a foundation for identifying and validating Mechanomedicine targets like MT1-MMP and developing Mechanomedicine candidate therapeutics to improve precision oncology outcomes in breast cancer.

C. Wei, L. Mengbo, Z. Hui; Department of Breast Surgery, Shengli Clinical Medical College of Fujian Medical University, Fuzhou, CHINA.

Background: Estrogen receptor-positive (ER+) breast cancer (BC) is the most common BC subtype, which is prevalent in women and seriously threatens their health. However, the present chemotherapy for ER+ BC exists many challenges, suggesting that there is an urgent need for novel drug targets to ER+ BC.Methods: The protein quantitative trait loci (pQTLs) data of plasma proteins and genome-wide association study (GWAS) summary data of ER+ BC were searched from the Integrative Epidemiology Unit (IEU) OpenGWAS database, which were included in Mendelian randomization (MR) analysis. Five methods were employed to perform the causal inference between plasma proteins and ER+ BC, including MR-Egger, Weighted median, Simple mode, Weighted mode, and Inverse variance weighted (IVW). Moreover, the sensitivity analysis, steiger test and co-localization were carried out to further reinforce the causality between plasma proteins and ER+ BC. Finally, the phenotype scanning was implemented to search for potential diseases associated with druggable genes, and the molecular docking was also proceeded to ascertain the potential drugs for druggable genes.Results: Based on IVW, a total of 31 plasma proteins were significantly causally effected on ER+ BC, such as VEGFR-3, IL3RA, ISLR2, VEGFR-2 and MICB. The reliability of their causal relationships was validated using sensitivity analysis and steiger test. Specifically, VEGFR-3 (rs635634), IL3RA (rs115478735), ISLR2 (rs115478735), VEGFR-2 (rs635634) and MICB (rs34158769) shared the same variant with ER+ BC (rs532436) with the help of co-localization analysis (coloc.abf-PPH4>95%). In addition, only three symbol genes (FLT4, KDR and IL3RA) corresponding to above exposure factors existed potential drugs, such as both of FLT4 and KDR binding to Axitinib. With respect to phenotype scanning, we discovered that VEGFR-3, IL3RA and VEGFR-2 as protective factors were causally related to nausea and vomiting, BC and diabetes mellitus, respectively. Whhereas IL3RA was a risk factor for lung cancer, as well as VEGFR-2 for colorectal cancer.Conclusion: Our study supported that plasma proteins were causally connected with ER+ BC risk, especially VEGFR-3, IL3RA, ISLR2, VEGFR-2 and MICB, which might be promising drug targets for ER+ BC.

A. Krysztofiak, A. Brown, A. Minnah, F. Rogers; Therapeutic Radiology, Yale School of Medicine, New Haven, CT.

Human epidermal growth factor receptor 2 (HER2), which gene is often amplified in breast cancer (BC), promotes metastatic phenotype and confers aggressive tumor behavior. Progress made in recent years by therapeutic strategies using monoclonal antibodies has increased survival rates of HER2-positive BC patients. However, current clinical trials in BC combining immunotherapy (inhibition of PD-1/PD-L1) with anti-HER2 antibodies, resulted only in a relatively modest overall response rate. This underlines that rise of resistance, recurrence, and metastatic processes remain the main hindrance for successful therapy and complete cancer remissions. Thus, there is an urgent need to bridge a therapeutic gap for the development of targeted therapies, especially for HER2-positive metastatic BC. Direct targeting of amplified cancer genes to invoke apoptosis through the activation of DNA damage response offers an alternative mechanism to induce tumor-specific cell death. The technology recently developed in our laboratory utilizes triplex-forming oligonucleotides (TFOs) to form triplex structures at amplified oncogenic loci. Excessive distortion to the helical DNA structure caused by the TFOs leads to immoderate DNA damage and apoptosis, but only in response to the formation of multiple triplex structures. As a result, normal cells having only two copies of the gene can efficiently handle a low level of triplex-induced DNA damage by employing nucleotide excision repair (NER). This phenomenon provides an opportunity to specifically target and induce apoptosis in BC cells with amplifications of the HER2 gene. Intriguingly, recent studies provide strong evidence that genomic DNA damage leads to activation of the cyclic GMP-AMP synthase (cGAS)-stimulator of interferon genes (STING) pathway and facilitates antitumor immune responses. HER2 also directly participates in a disruption of STING signaling, suppressing antitumor immunity and provides a survival advantage, conferring resistance to trastuzumab. Therefore, the crosstalk between stimulation of innate immune signaling and DNA damage response opens an interesting perspective of combination therapy strategies focusing on both DNA-damaging and immune system-activating therapies in BC. In this work, we discovered that TFO-induced DNA damage leads to activation of innate immune signaling in BC cell line models. First, we characterized HER2 amplification and expression of cGAS-STING pathway in numerous BC cell lines and assessed levels of triplex-induced DNA damage and apoptosis. Next, we evaluated activation of the cGAS-STING pathway in BC cell lines by analyzing STING downstream targets following TFO treatment. We discovered that TFO-induced DNA damage response in BC cells had limited effect on canonical STING signaling as measured by stimulation of TBK1-IRF3 downstream targets. However, TFO treatment induced phosphorylation of NF-κB p65 transcription factor. We validated STING-dependency in the stimulation of innate immune system following TFO treatment and confirmed that this activation can also be achieved with TFOs targeting various genomic regions including exons and introns within the HER2 gene. To reveal mechanistic landscape of the TFO-induced activation of innate immune signaling we performed analysis of gene expression using bulk RNA sequencing. In summary, we have validated that HER2-targeted TFOs can induce activation of the innate immune system signaling in BC cell lines. Our work contributes to understanding of the crosstalk between DNA repair, apoptosis, and innate immune signaling. This will aid in the design of novel combinations of therapeutic approaches to increase responsiveness of BCs to immune-targeted treatments through activation of DNA damage response, ultimately overcoming therapy resistance mechanisms.

N. Priedigkeit¹, J. Koob², A. Lebrón-Torres³, J. Liao³, Y. Wang³, M. E. Hughes⁴, D. L. Abravanel⁵, S. M. Tolaney⁵, N. U. Lin⁵, S. Oesterreich⁶, A. V. Lee⁷, F. Chen⁸, T. R. Golub³; ¹Medical Oncology, Dana-Farber Cancer Institute / Broad Institute, Boston, MA, ²Molecular and Cellular Biology, Broad Institute of MIT & Harvard, Cambridge, MA, ³Cancer Program, Broad Institute of MIT & Harvard, Cambridge, MA, ⁴Yale Cancer Center, Yale, New Haven, CT, ⁵Medical Oncology, Dana-Farber Cancer Institute, Boston, MA, ⁶Pharmacology and Chemical Biology, University of Pittsburgh School of Medicine, Pittsburgh, PA, ⁷Department of Pharmacology and Chemical Biology, University of Pittsburgh School of Medicine, Pittsburgh, PA, ⁸Department of Stem Cell and Regenerative Biology, Broad Institute of MIT & Harvard, Cambridge, MA.

Background: The challenge of cancer heterogeneity and the limitations of protein-targeting therapeutics underscore an urgent need for innovative, patient-scalable strategies that can precisely eliminate tumor cells while sparing healthy tissues. Towards this aim, breast cancer is dominated by genomic structural variation-which generates hundreds of tumor-specific nucleic acid sequences per cancer. We have previously shown that up to one-third of metastatic breast cancers harbor highly expressed fusion RNAs as a result of these genomic aberrations (Priedigkeit, SABCS 2023). Although fusions have been successfully targeted at the protein-level such as NTRK and FGFR-family fusions, many remain undruggable with conventional therapies. We hypothesized that advances in nucleic acid technology could be applied to therapeutically exploit these cancer-specific biomolecules. Methods: To test this hypothesis, we applied a modular nucleic acid targeting approach (RADARS, Jiang, Koob et al 2023) that utilizes engineered nucleic acid “sensors” designed to selectively hybridize to target RNA sequences. Upon hybridization, the sensors create a double stranded RNA duplex and recruit an endogenous protein called ADAR, which deaminates an intentional A:C mismatch in the sensor-triggering translation of a linked protein cargo only in the presence of a target RNA. We designed and tested 91 sensors against cancer fusion RNA targets with two broad designs—linear sensors that hybridize directly across the fusion RNA breakpoint sequence and gap sensors that target distinct regions of the 5′ and 3′ fusion partners on the same guide. We tested the selective expression of protein payloads using both designs against two clinically relevant fusion RNAs: (1) an ESR1 fusion and (2) an endogenous fusion RNA in MCF7 cells. Sensor activity was assessed with microscopy for fluorescent protein payloads, and cell viability readouts including CellTiter-Glo, protein quantification of apoptosis markers, and live-cell imaging for cell death payloads. Results: We first confirmed that outlier-expressed fusion RNAs are prevalent across luminal, HER2+, and basal-like cell models—mirroring our prior data on human tumors. We then tested sensors targeting an ESR1::SOX9 fusion RNA. In cells co-transfected with the sensors and fusion RNA, we observed robust mNeon expression, whereas cells without the fusion showed only minimal background fluorescence (mean 25-fold increase vs. no target controls, p < 0.0001), validating the sensors’ specificity. Next, we targeted an endogenous ARFGEF2::SULF2 fusion RNA in the MCF7 breast cancer cell line, deploying sensors with iCaspase9; a pro-apoptotic cargo. Transfection of both linear and gap fusion-targeting sensors into MCF7 cells and subsequent treatment with iCaspase9 inducer drug AP1903 resulted in a significant reduction in cell viability (up to 84% 36-hours post sensor transfection vs. non-targeting controls, p < 0.0001). Live-cell imaging and probing for markers of apoptosis confirmed robust sensor-induced cell death in MCF7 cells, while fusion-negative cells exhibited no toxicity when treated with the same fusion-targeting sensors, demonstrating high selectivity. Conclusions: This study demonstrates the successful design and in vitro proof-of-principle of a biomolecule sensing system that selectively triggers a therapeutic protein effector in the presence of a cancer-specific nucleic acid. These findings have the potential to establish a novel therapeutic paradigm-in which an RNA medicine can repurpose cancer-defining nucleic acid sequences into a targeted, modular, and precise cell-killing mechanism. Our ongoing efforts focus on optimizing sensor design, interrogating mechanisms of sensor-mediated cell killing with diverse protein payloads, and addressing challenges of in vivo delivery.

C. Bose¹, S. Awasthi², T. Hutson², F. Sardela de Miranda³, M. F. Mahecha³, M. W. Melkus³, R. Layeequr Rahman³, S. P. Singh²; ¹Neuropharmacology, TTUHSC, Lubbock, TX, ²Internal Medicine, TTUHSC, Lubbock, TX, ³Surgery, TTUHSC, Lubbock, TX.

Background: Rlip (aka RLIP76 protein encoded by RALBP1) is a Ral-regulated, membrane-associated ATPase enzyme that facilitates Phase 2 detoxification, plays a role in clathrin-dependent endocytosis (CDE), and is vital for cancer cell formation and survival. DNA repair is essential for both normal and cancer cell survival, as it protects cancer cells from DNA damage caused by chemotherapy or radiation. Poly (ADP-ribose) polymerases (PARPs) play a crucial role in the repair of DNA damage. By targeting these repair mechanisms in cancer cells, it is possible to improve the effectiveness of cancer treatments. A notable example is dysfunctional homology-directed repair (HDR), which results from inactivating mutations in the BRCA1 and BRCA2 genes, contributing to radiation resistance in human cancer cells. Therefore, identifying new targets within double-strand break (DSB) repair pathways is crucial. A core concept in using PARP inhibitors (PARPi) for treating patients with BRCA1/2 mutations is the concept of synthetic lethality. Our research found that Rlip depletion or inhibition, combined with PARPi, works synergistically in both BRCA1/2-mutated and BRCA1/2-wild-type breast cancer cells. Methods: Two PARP inhibitors, AZD2281 (Olaparib) and AZD2461, were tested on the proliferation and survival of different cancer cell lines, including MDA-MB-231 (triple negative), MCF-7 (ER+), SKBR-3, AU565 (HER-positive), and MDA-MB-436 (BRCA1 mutated triple negative), with and without Rlip depletion using an antisense locked nucleic acid (LNA). For all experiments, 2.5 µM AZD2461 and 10 µM AZD2281 were used for 24 hours in complete growth medium. Cell cycle analysis was performed by flow cytometry. Expression levels of γH2AX and Rad51 proteins were assessed via Western blotting and immunofluorescence staining. The comet assay was used to measure DNA damage and fragmentation. Results: Rlip depletion significantly increased the reduction in cell survival caused by PARPi treatments across all tested cell lines. Combination index (CI) values were calculated using the Hill equation and Chou-Talalay’s method with CompuSyn software. The effects were highly synergistic (CI 1), but the doses of both inhibitors were significantly reduced (p < 0.001) after Rlip depletion. Cell cycle analysis with Rlip depletion showed either an increase in the sub-G1 population (cell death) or G2/M phase arrest, suggesting cells attempt to undergo mitosis with unrepaired DNA damage, leading to elevated DNA double-strand break signaling, cell cycle arrest, and apoptosis. Alkaline comet assay confirmed more DNA damage with Rlip depletion compared to treatments with the inhibitor alone. Western blot analysis for Rad51 and γ-H2AX showed defects in protein levels following Rlip depletion. The number of γ-H2AX foci, characteristic of DNA damage in nuclei, increased significantly (p < 0.01) after Rlip depletion compared to inhibitor treatment alone. Conclusion: Our findings suggest that drugs that inhibit or deplete Rlip could help sensitize breast cancer cells, regardless of their BRCA status or whether they are wild-type ER+ or triple-negative breast cancers, to PARP inhibitors. The immunogenicity of HRD-associated malignancies and the immunomodulatory effects of PARP inhibition could be combined with Rlip inhibitors to treat HRD cancers. These findings clarify the underlying mechanisms and are expected to lead to the development of first-in-class therapies.

S. Min¹, K. Forero¹, J. Anderson², W. Putnam², C. Evizi², C. McCallum², R. Hoff¹, C. Hsieh², K. Ho²; ¹Formulation, Sapu Bioscience, LLC, San Diego, CA, ²Manufacturing, Sapu Bioscience, LLC, San Diego, CA.

Background: Everolimus is approved for advanced renal-cell carcinoma following VEGF-targeted therapy, for metastatic HR-positive/HER2-negative breast cancer in combination with exemestane, and for other oncology indications. However, everolimus lacks an intravenous option due to its extreme hydrophobicity, despite its low oral bioavailability of only 15-20%. Here, we describe the successful development of an intravenous formulation of Everolimus using our Deciparticle™ platform. Formulation: A 38-member mPEG-block polymer library was screened as potential candidates for encapsulating Everolimus. Nine polymers demonstrated the ability to form stable Deciparticles™. Among these candidates, VT111 was selected based on its favorable biocompatibility. VT111 and its variants were able to formulate both peptides and polyketides, including Everolimus, Sirolimus, Tacrolimus, and Cyclosporine. Manufacturing: Clinical lots were produced using a 7-day process at our cGMP facility. The manufacturing process began with the dissolution of VT111 and Everolimus in ethanol, which was subsequently solubilized into lactose and Water for Injection to form Deciparticles™. The process was conducted under yellow light and utilized amber vials to minimize Everolimus degradation. The bulk drug was filter sterilized. Fill/finish operations were in a temperature-controlled environment with ≥98% filling weight accuracy. The drug product was reconstituted to a final concentration of 4 mg/mL for intravenous administration. Upon reconstitution, the product yielded monodisperse Deciparticles™ that met sterility and particle-size specifications and remained within specifications for at least one month at 5 °C storage and for at least 24 hours of in-use stability at 25 °C. The same process has been successfully scaled to over 10 g, sufficient for Phase 1 clinical supply. In Vivo Activity: Sapu003 exhibited potent antitumor activity in vivo (IV, BIW × 4) and demonstrated tumor growth inhibition (TGI) percentages of: U-87 MG glioblastoma (TGI 97-98%), HT-29 colorectal cancer (71-73%), OVCAR-3 ovarian cancer (66%), and MDA-MB-231 breast cancer (48%). No phlebitis was observed during repeated tail-vein administrations. Sapu003 sensitivity correlated with glycolytic dependence (U-87 ≫ HT-29 > OVCAR-3 ≈ MDA-MB-231). In the glycolysis-dependent U-87 model, Sapu003 was cytotoxic and outperformed paclitaxel, without inducing body weight loss. Conclusions: We screened a 38-member mPEG-block polymer library for self-assembling capacity and identified a leading candidate, VT111, which reproducibly formed stable Deciparticles™ under 20 nm in mean particle diameter and demonstrated potent antitumor activity across multiple xenograft models. VT111 uniquely defines the chemical space for Deciparticle™ technology and enables a reproducible cGMP process yielding a long-term stable drug product. These data could support the ongoing first-in-human study of weekly Sapu003 and open the Deciparticle™ platform to other water-insoluble oncology agents.

D. Cecil, J. Childs, M. Disis; Cancer Vaccine Insitute, University of Washington Medicine, Seattle, WA.

Background: The breast tumor immune microenvironment is characterized by a Type II signature which is an early event in breast cancer and portends more aggressive disease. We have previously demonstrated that bacteria-specific T-cells, derived from the gut, also recognize tumor antigens with which they share a high degree of homology at the level of Class II epitopes presented in MHC molecules. These peripheralized bacterial-tumor antigen cross reactive T-cells are tolerogenic, secrete high levels of IL-10 and, in mouse models, when adoptively transferred, traffic specifically to breast tumors and accelerate tumor growth. We questioned if the Type II signature present in human breast cancer is due to these bacterial-tumor antigen (BAC-TA) cross reactive T-cells. Methods: PBMC and stool were collected from newly diagnosed, treatment naive breast cancer patients (n= 56) or volunteer aged matched donors (n=14). PBMC were screened with a panel of nine previously identified BAC-TA epitopes by IL-10 ELISPOT. Tumor biopsies were collected from three patients at the time of surgery. BAC-TA-specific T-cell lines were generated from PBMC from these three patients that responded with high magnitude IL-10 in the initial screen to BIRC5-p13-30 for Donor 1 (mean, 178±59 SPW / 10⁶; p=0.0002) and Donor 2 (mean, 1612±234 SPW / 10⁶; p=0.0011) and PRL3-p104-118 for Donor 3 (mean, 365±100 SPW / 10⁶ T-cells; p=0.03). Extracted DNA from the PBMC, T-cell lines and tumor biopsy was subjected to TCRb sequencing. Metagenomic sequencing was performed on stool samples. Results: A higher precursor frequency of BAC-TA cross reactive IL-10-secreting T-cells were observed in 60% of breast cancer patients (median, 1:160,000 PBMC) as compared to 28% of the volunteer donors (median 50% homology to BIRC5-p13-30 were enriched in the stool of Donor 1 as compared to the levels in the stool of the 56 breast cancer patients evaluated in this study (L. lactis; enriched by 186-fold and S. oralis and P. fluorescens; both enriched by 3-fold). Similar trends were observed in the other two donors. Eleven BIRC5-p13-specific T-cell clones from Donor 2 were found in the matching tumor biopsy, with enrichment of two bacterial species with high BIRC5-p13-30 sequence homology (K. pneumoniae; 256-fold increase and P. distasonis; 4-fold increase). 25 PRL3-p104-specific T-cell clones in Donor 3 were found to be identical between the tumor biopsy and T-cell line. The stool of Donor 3 demonstrated a 15-fold and 3.5-fold increase in P. copri and P. fluorescens, respectively, two bacteria with at least 50% homology to PRL3-p104-118. Conclusion: IL-10-secreting BAC-TA T-cells can be found in high numbers in the peripheral blood of of newly diagnosed breast cancer patients and can be identified infiltrating the tumor. IL-10 is a key immunosuppressive cytokine that plays a central role in dampening immune responses and maintaining immune homeostasis. IL-10 inhibits the production of pro-inflammatory cytokines such as TNF-α and IFN-γ and can limit antigen-presenting cell function, leading to the reduction of Th1, Th17 and cytotoxic CD8+ T-cell activation. These data lay the foundation for a study that will follow newly diagnosed breast cancer patients correlating levels of BAC-TA cross reactive T-cells in blood with prognosis or response to therapy.

K. Nassar¹, J. Sanchez¹, D. Panneerdoss¹, E. Behnam¹, X. Yang¹, U. Pratap¹, M. Mahajan¹, S. Alejo¹, P. Subbarayalu², D. Zhou³, G. Sareddy¹, M. Rao², S. Viswanadhapalli¹, R. K. Vadlamudi¹; ¹ObGyn, UT Health SanAntonio, SAN ANTONIO, TX, ²GCCRI, UT Health SanAntonio, SAN ANTONIO, TX, ³Department of Biochemistry & Structural Biology, UT Health SanAntonio, SAN ANTONIO, TX.

Background: Triple-negative breast cancer (TNBC) is an aggressive breast cancer subtype associated with poor prognosis and high rates of metastasis, contributing to 15-24% of breast cancer-related deaths. The oncogene proline-, glutamic acid-, and leucine-rich protein 1 (PELP1) is frequently overexpressed in TNBC and is associated with poor survival. This study aimed to elucidate the critical role of PELP1 in DNA repair mechanisms of TNBC and to evaluate a novel therapeutic strategy using a PELP1 small-molecule inhibitor, SMIP34, in combination with topoisomerase inhibitors (TIs). Methods: PELP1-inducible knockdown (iKD) TNBC models (MDA-MB-231, BT-549, HCC-1806, SUM-149) were generated via lentiviral transduction using IPTG-inducible shRNA constructs. PELP1 signaling was inhibited pharmacologically using SMIP34. A drug screen of 140 FDA-approved compounds was performed using MTT assays to identify synergistic agents with SMIP34. Functional assays, including colony formation and Annexin V/PI apoptosis assays, were used to assess therapeutic efficacy. Mechanistic studies utilized RT-qPCR, Western blotting, comet assays, and DNA Damage Response Phosphorylation Arrays. Combination treatments were validated in cell line-derived xenografts , patient-derived organoids , and patient-derived xenografts . Results: Western blot analyses confirmed that IPTG induction resulted in a dose-dependent suppression of PELP1 expression in all four PELP1-iKD TNBC cell lines. PELP1 knockdown significantly impaired cell proliferation and colony formation while inducing apoptosis. In parallel, pharmacological inhibition using SMIP34 produced similar phenotypic outcomes, validating the functional relevance of PELP1 in TNBC cell survival. Importantly, PELP1 suppression led to a delayed DNA damage response (DDR), marked by elevated γ-H2AX and activation of key DDR kinases including p-ATM, p-ATR, and DNA-PKcs. SMIP34 treatment replicated these effects, confirming that both genetic and pharmacologic inhibition of PELP1 compromises DNA repair efficiency. A high-throughput MTT-based screen of 140 FDA-approved agents identified topoisomerase inhibitors (TIs) as potent synergistic agent with SMIP34. Co-treatment with SMIP34 and TIs resulted in significantly enhanced growth inhibition and apoptosis compared to monotherapy. This synergy was confirmed in PELP1-iKD cells, where knockdown further sensitized cells to TIs, providing genetic validation of PELP1’s role in TI resistance. Mechanistic studies showed that combination treatments (PELP1-iKD+TIs or SMIP34+TIs) markedly increased DNA damage and apoptotic signaling. Western blotting and phospho-antibody arrays revealed upregulation of DDR and checkpoint proteins, including γ-H2AX, p-CHK2, p-ATR, p-CHK1, p-BRCA1, and p21, indicating enhanced replication stress and cell cycle arrest. Comet assays demonstrated significantly greater DNA fragmentation with combination therapy compared to single agents. In preclinical models, SMIP34+TI treatment significantly suppressed organoid growth in patient-derived organoid cultures. In vivo, this combination led to marked tumor growth inhibition in both cell line-derived xenografts and patient-derived xenografts, outperforming either agent alone. No significant toxicity was observed, supporting the therapeutic potential of this approach. Conclusion: These findings highlight PELP1 as a critical modulator of DNA damage response in TNBC and support the development of combination strategies using SMIP34 and topoisomerase inhibitors. This approach offers a promising therapeutic avenue for improving outcomes in patients with TNBC.

M. A. Perez¹, C. S. Fermaintt¹, A. Risinger²; ¹Department of Chemistry and Biochemistry, University of the Incarnate Word, San Antonio, TX, ²Department of Pharmacology, University of Texas Health San Antonio, San Antonio, TX.

Breast cancer affects approximately 1 in 8 women in the United States, with triple-negativebreast cancer (TNBC), representing 10–15% of all cases. TNBC is characterized by lowexpression of estrogen, progesterone, and HER2 receptors, which limits the efficacy ofapproved therapeutics targeted against these receptors. The molecular subtyping of TNBCshas been valuable in the search for new genetic and pharmacological vulnerabilities against thisheterogeneous disease and includes basal-like 1 (BL1), basal-like 2 (BL2), mesenchymal (M),and luminal androgen receptor (LAR) subtypes. We previously screened natural productextracts for selective cytotoxic efficacy among TNBC cell lines representing each of thesesubtypes, which led to the identification of the diterpenoid yuanhuacine as highly selectiveagainst the BL2 subtype, both in vitro and in vivo antitumor studies. This is notable as BL2tumors represent 13% of TNBCs and have the lowest pathological complete response rate toneoadjuvant chemotherapy among these molecular subtypes at 0%. Further studies havedemonstrated that structurally and functionally related ingenane diterpenoids in clinicaldevelopment for the topical treatment of skin malignancies, including tigilanol tiglate andingenol-3-angelate, retain BL2 subtype-selective activity with over one-thousand-fold selectivity,providing a unique opportunity for drug repurposing.To support these efforts, we have taken a multipronged approach to identify the mechanism ofselectivity of these compounds against BL2 TNBC cells, perform structure-activity relationship(SAR) and pharmacokinetic studies to inform on repurposing this drug class for systemicadministration, and determine their efficacy in combination with standard of care chemotherapy.We have demonstrated that the selective activity of yuanhuacine and ingenane diterpenoids isdue to activation of PKCβ, which is elevated in cell lines representing the BL2 TNBC subtype.The evaluation of sixteen structurally related compounds with potency in BL2 cell lines rangingfrom 0.07 – 50,000 nM with 3 – 300,00-fold selectivity for this subtype has informed on the SARfor BL2-specific cytotoxicity. In vitro pharmacokinetic studies have further demonstratedpotential liabilities of this drug class that need to be taken into account with systemicadministration that will be used in combination with SAR studies to inform on medicinalchemistry efforts to prioritize compounds for antitumor efficacy studies in BL2 tumor models anddevelop compounds with a greater therapeutic index than yuanhuacine. We have also identifiedoptimal combinations of these highly BL2-selective compounds with standard of carechemotherapeutics to identify drug pairings that would have optimal efficacy in heterogenousBL2-enriched TNBC tumors. Finally, we have identified additional tumor types that expressPKCβ, including pediatric cancers, and determined the potential for ingenane diterpenoids forrepurposing in a broader group of cancers through molecular profiling. Together, these studiesinform on the potential to repurpose a class of natural products developed for skin cancer for thetreatment of a molecularly defined subtype of TNBC to improve targeted treatment options forthese patients.

S. C. Stanfield¹, C. S. Fermaintt², A. Risinger³; ¹Biology, University of the Incarnate Word, San Antonio, TX, ²Chemistry and Biochemistry, University of the Incarnate Word, San Antonio, TX, ³Pharmacology, University of Texas Health San Antonio, San Antonio, TX.

Triple-negative breast cancer (TNBC) accounts for approximately 15–20% of all breast cancercases and disproportionately affects younger women. Characterized by the absence of estrogen(ER), progesterone (PR), and HER2 receptors, TNBC lacks the molecular targets utilized in mostcurrent therapies, resulting in limited treatment options and a heavy reliance on cytotoxicchemotherapy. TNBC is associated with a more aggressive clinical course, early metastasis, andpoor overall survival as compared to other types of breast cancer, highlighting the urgent needfor novel and effective therapeutic strategies. To identify new therapeutic targets, TNBC hasbeen classified into four molecular subtypes: basal-like 1 (BL1), basal-like 2 (BL2),mesenchymal (M), and luminal androgen receptor (LAR). Among these, the BL2 subtypeexhibits particularly poor prognosis and resistance to standard of care treatments. In our study,we identified yuanhuacine and other related diterpenoid natural products as compounds withpotent (0.07 – 7.7 nM) cytotoxicity against BL2 TNBC cells. These compounds are over one-thousand-fold more selective for the BL2 subtype of TNBC as compared to other TNBCsubtypes or immortalized normal breast lines supporting the identification of a novel targetablevulnerability in this breast cancer subtype. Most notably, we demonstrated this exquisite potencyand selectivity for BL2 TNBC is also a feature of compounds of this class that are undergoingclinical development for the topical treatment of skin cancers demonstrating an opportunity fordrug repurposing.Mechanistic studies revealed that the efficacy of these compounds is mediated by activation ofProtein Kinase C beta (PKCβ), which is elevated in BL2 TNBC cells. Pharmacologic inhibitionof PKCβ significantly reduced the potency of yuanhuacine and ingenol-3-angelate in BL2 TNBCcells, indicating that PKCβ activation is necessary for their selective activity. Geneticexperiments are being conducted to interrogate whether PKCβ expression is necessary andsufficient for sensitivity to yuanhuacine with reduced potency expected in in BL2 lines uponPKCβ knockdown and sensitization of non-BL2 cell lines when ectopically expressed. Cellbiological experiments demonstrated treatment of BL2 cells with yuanhuacine promotesphosphorylation of PKCβ at Ser660 and activation of PKC substrates within one hour oftreatment, suggesting rapid activation and engagement of downstream signaling. Additionalimmunofluorescence studies demonstrate the effect of this activation on the intracellularlocalization and degradation of PKCβ as well as induction of apoptosis.Our findings demonstrate that yuanhuacine and other ingenane diterpenoids, includingcompounds in clinical development for other indications, have the potential for the targetedtreatment of the BL2 subtype of TNBC. Additionally, the identification of PKCβ expression as abiomarker of sensitivity to this class of drugs opens the potential for efficacy in other cancerswhere PKCβ is expressed. Given the lack of targeted therapies for BL2 TNBC, the developmentof PKCβ-activating compounds like yuanhuacine represents a promising and urgently neededadvancement in personalized oncology.

A. Kamal¹, A. S. Chin¹, S. Mukherjee¹, J. Li¹, K. Peregrina¹, D. Lewis¹, H. Sethi¹, L. Xu¹, D. Patel², M. Roy¹, A. Bardia³; ¹R&D, Actinium Pharmaceuticals, New York, NY, ²Clinical, Actinium Pharmaceuticals, New York, NY, ³Geffen School of Medicine, University of California Los Angeles, Los Angeles, CA.

Background: Hormone receptor positive (HR+) is the most common subtype of breast cancer (BC) and comprises >70% of BC diagnosis. Tamoxifen is the most widely prescribed endocrine therapy (ET) for modulating estrogen receptors, however, ~20-30% of BC patients have disease recurrence despite ET. Similarly, resistance to HER2-targeted therapy trastuzumab (Herceptin) can develop in HER2+ BC. While antibody drug conjugates (ADCs) like trastuzumab deruxtecan (Enhertu) show strong efficacy in HER2+ and HER2-low BC, their use is limited by interstitial lung disease toxicities. In contrast, Actinium-225-based radioconjugates may provide potent tumor killing with less off-target lung toxicity. Here, we report the development of ATNM-400, a novel antibody radioconjugate using Actinium-225 to target a protein that is overexpressed in BC. The target is linked to disease progression and metastasis, with overexpression correlating with poor clinical outcomes in BC patients. Furthermore, target expression is elevated in patients who develop ET and HER2 therapy resistance. Here, we evaluated the anti-tumor efficacy of ATNM-400 in preclinical BC models and evaluated ATNM-400’s potential to overcome current therapeutic limitations and improve outcomes beyond what is achievable with tamoxifen or trastuzumab. Methods: ATNM-400 was synthesized by conjugating ATNM400 antibody with p-SCN-Bn-DOTA, followed by radiolabeling with Ac-225 and characterized by radio-iTLC, HPLC and mass spectrometry. Target expression, binding affinity, and internalization were assessed using flow cytometry and radioactivity quantification in parental and drug-resistant BC cell lines, including MCF-7 (HR+), MCF-7 Tam1 (tamoxifen-resistant), BT-474 (HER2+), BT-474 Clone 5 (trastuzumab-resistant), and MDA-MB-468 (triple-negative). In vivo biodistribution and efficacy were evaluated in murine xenograft models. Efficacy of ATNM-400 alone and in combination with standard-of-care (SOC) therapies was assessed in both in vitro and in vivo resistance models. Results: ATNM-400 demonstrated over 98% radiochemical purity. ATNM-400 exhibited strong and specific binding and internalized in target-positive BC cell lines. Target expression was significantly elevated in tamoxifen- and trastuzumab-resistant lines, supporting its mechanistic link to resistance. In vitro, ATNM-400 induced potent, dose-dependent cytotoxicity as monotherapy, which was further enhanced when combined with SOC agents. In vivo, ATNM-400 achieved tumor growth inhibition (TGI) of 76.4% and 111.5% in MCF-7 xenografts at 20 and 40 μCi/kg, respectively, with the lower dose being well tolerated. In MDA-MB-468 xenografts, TGI reached 66.3% and 103.4% at corresponding doses, both of which were well tolerated. Notably, ATNM-400 retained efficacy in both tamoxifen- and trastuzumab-resistant models, demonstrating its potential to overcome resistance mechanisms. Conclusions: ATNM-400 shows strong anti-tumor efficacy and favorable tolerability across multiple BC subtypes, including models resistant to tamoxifen and trastuzumab. These findings support further development of ATNM-400 as a novel therapeutic approach for patients with limited options following endocrine or HER2-targeted therapy failure-both as monotherapy and rational combination regimens.

J. S. Manavalan¹, A. Ahmad², D. Mor¹, J. Davis¹, R. Chakrabarti³, A. Pollack², E. Davis¹; ¹Oncotope Platform, Cancer Antibodies Inc, Longwood, FL, ²Radiation Oncology, University of Miami, Miller School of Medicine, Miami, FL, ³Surgery, University of Miami, Miller School of Medicine, Miami, FL.

Background: Antibody-drug conjugates (ADCs), bispecific antibodies, and CAR-T cells offer tumor-selective cytotoxicity with reduced off-target toxicity. Their success depends on identifying cancer-specific surface antigens absent in normal tissues. The Oncotope Platform, developed by Cancer Antibodies Inc., enables high-throughput antigen discovery through rabbit immunization with cancer cell lines, filtration to remove cross-reactive antibodies, and protein microarray screening of >20,000 human proteins. This platform shortened discovery timelines and demonstrated selectivity by generating antibodies that killed breast cancer cells (Hs578T) while sparing matched normal cells (Hs578Bst). One lead antigen was SLC39A7 (ZIP7), a zinc transporter normally confined to the ER/Golgi but aberrantly expressed on the surface of malignant cells. In contrast to PSMA, which is expressed in salivary glands and kidneys and associated with significant toxicities in radioligand trials (e.g., 81% salivary, 25% renal), ZIP7 showed restricted expression in normal tissues, making it a promising target for safer therapeutics in breast and prostate cancer, including castration-resistant disease. Methods: ZIP7 was validated using in vitro and in situ approaches. Breast cancer cell lines were evaluated by flow cytometry for ZIP7 surface expression and by cytotoxicity assays for selective tumor killing. ZIP7 mRNA expression was assessed in 10 breast cancer cell lines representing Luminal A, HER2+, and TNBC subtypes. Immunofluorescence (IF) was performed on TNBC samples (n=5; grades 2-3) and prostate cancer samples (n=6; Gleason 3+4, 4+3, including CRPC). Each case included adjacent benign tissue as control. CyTOF imaging analyzed co-localization of ZIP7 with CD4+ and CD8+ T cells. IHC-DAB was used to evaluate ZIP7 expression in PSMA-negative (PC3) and PSMA-positive (LNCaP) xenografts and in normal kidney, spleen, and brain tissues. Results: In vitro, filtered antibodies killed ZIP7-expressing Hs578T cells while sparing Hs578Bst cells. ZIP7 was detected in 8/10 breast cancer cell lines, and ZIP7-specific antibodies bound selectively and induced cytotoxicity. Flow cytometry confirmed surface expression. In FFPE samples, IF showed ZIP7 staining in all TNBC tumors (5/5), absent in adjacent normal epithelium. In prostate cancer samples, ZIP7 expression was observed across all grades and CRPC, with enrichment at invasive margins and immune-infiltrated areas. CyTOF showed ZIP7+ cells in proximity to CD4+/CD8+ T cells. IHC-DAB confirmed ZIP7 in both PC3 and LNCaP xenografts. No ZIP7 expression was detected in normal kidney, spleen, or brain tissues. Conclusions: ZIP7 is a tumor-selective surface antigen with aberrant expression in breast and prostate cancers, including resistant subtypes. Its absence from essential normal tissues supports its safety as a therapeutic target. The Oncotope Platform enabled the rapid discovery and validation of ZIP7, demonstrating its power to uncover high-value, tumor-specific targets for antibody-based cancer therapies.

H. Fang¹, W. Jiang²; ¹Department of Breast Cancer Center, Chongqing University Cancer Hospital, Chongqing, CHINA, ²Department of Radiology, The Second Affiliated Hospital of Chongqing Medical University, Chongqing, CHINA.

AbstractBackground: Breast cancer is one of the most common malignant tumors worldwide, and its treatment faces challenges such as the complexity of the immune microenvironment and the side effects associated with conventional therapies. Doxorubicin (DOX) is often confronted with issues of resistance and significant side effects, while Resiquimod (R848) has been shown to enhance anti-tumor immune responses. This study presents a liposome-based nanodrug delivery platform that combines the advantages of liposomes with the protective properties of chitosan, aiming to co-deliver DOX and R848 to target tumor cells and tumor-associated macrophages (TAMs), thereby improving the therapeutic efficacy in breast cancer treatment.Methods: In this study, folic acid (FA) and chitosan (CS) were conjugated to form an FA-CS complex for incorporation into liposomes, enhancing their stability and targeting ability. FA-CS-R848/DOX@Lip liposomes were prepared via the rotary evaporation method, and their morphology, particle size, and zeta potential were characterized using transmission electron microscopy (TEM) and dynamic light scattering (DLS). The drug encapsulation efficiency, loading capacity, and in vitro release were assessed. In vivo, a mouse breast cancer xenograft model was established to evaluate the anti-tumor efficacy and biocompatibility of the drug delivery system. Flow cytometry was used to quantify immune cell populations in the tumor immune microenvironment, including CD8+ T cells, TAMs, regulatory T cells (Tregs), and myeloid-derived suppressor cells (MDSCs), to analyze changes in the immune response. Results: FA-CS-R848/DOX@Lip liposomes were successfully prepared with a particle size of approximately 211 nm and a zeta potential of 5.3 mV. The encapsulation efficiencies of R848 and DOX were found to exceed 80% and 70%, respectively, with drug loading capacities of 6.8% and 4.6%. Drug release studies indicated that the release profile of FA-CS-R848/DOX@Lip liposomes was relatively slow. Cellular uptake assays demonstrated significantly higher uptake efficiency in RAW 264.7 macrophages and EO771 tumor cells compared to non-targeted liposomes. Cytotoxicity evaluations showed that FA-CS-R848/DOX@Lip exhibited superior anti-tumor activity, effectively inducing tumor cell death while simultaneously stimulating immune responses. In vivo experiments revealed that FA-CS-R848/DOX@Lip significantly inhibited tumor growth, while maintaining favorable biocompatibility and safety profiles. Flow cytometry analysis indicated a notable increase in the proportion of CD8+ T cells with enhanced functionality, as well as a significant shift in the M1/M2 ratio of TAMs and a reduction in the numbers of Tregs and MDSCs. Collectively, these results suggest that this nanodelivery platform effectively activates immune responses and remodels the tumor immune microenvironment.Conclusion: The FA-CS-R848/DOX@Lip liposomes developed in this study synergistically combine the cytotoxic effects of DOX with the immune activation properties of R848, achieving a cooperative interaction between chemotherapy and immunotherapy. This delivery platform effectively targets tumor cells and TAMs, significantly enhancing anti-tumor immune responses while overcoming the limitations of conventional therapies. By optimizing the drug delivery system and enabling precise immune modulation, FA-CS-R848/DOX@Lip represents a novel and effective strategy for breast cancer treatment. This study provides a theoretical foundation for the combined application of chemotherapy and immunotherapy, offering promising prospects for advancing therapeutic approaches in other cancer types.

D. Singh¹, B. Onyeagucha², D. Medina¹, P. Subbarayalu¹, R. Mojidra³, P. Venkata¹, J. Huang¹, J. Prochnau¹, P. Do¹, M. Rao¹; ¹Cell System & Anatomy, UTHSA, Greehey Children’s Cancer Research Institute, San Antonio, TX, ²Department of Science & Mathematics, Mississippi University for Women, Columbus, MS, ³Department of Biochemistry and Structural Biology, University of Texas Health San Antonio, San Antonio, TX.

Introduction: Antibody-based therapy is one of the most successful strategies for treating patients with cancers. This is also true for breast cancers as trastuzumab (a monoclonal antibody against Her2) is used as the first line therapy for Her2+ breast cancer. Despite substantial progress, there remains a notable gap in effective targeted treatments for TNBC. In the landscape of cancer treatment, chemotherapy is the common treatment option for TNBC. However, resistance and tumor relapse remain the major obstacles hindering the effectiveness of chemotherapeutic agents in cancer patients. Therefore, the development of new therapeutics, especially targeted approaches like antibody-based therapies, holds immense significance. This study aims to identify and test the role of secretory protein/s that may support the growth and progression of TNBC cells. Additionally, it seeks to determine if targeting these oncogenic secretory proteins with antibodies can provide a novel therapeutic approach for TNBC patients. Our study identified secretory protein “Signal peptide CUB domain EGF-like 3 (SCUBE3)” as a critical factor that promotes survival and therapy resistance of TNBC cells. Importantly, using antibody-development platform, our study presents a novel therapeutic approach targeting SCUBE3 using anti-SCUBE3 monoclonal antibody, offering new hope for TNBC patients with limited treatment options. Methods: Unbiased high-throughput loss of function genomic screen was performed on TNBC cells in the presence or absence of doxorubicin. TNBC cells were transfected either with SCUBE3 overexpression plasmid or shRNA specific to SCUBE3 to perform functional assay, in vivo tumor xenograft assays, RNA sequencing and Mass spectrometry (MS). Syngeneic mouse model was used to study the effect of SCUBE3 in shaping tumor immune microenvironment of TNBCs. Hybridoma culture technique was used to generate anti-SCUBE3 monoclonal antibodies. Screening of top anti-SCUBE3 antibody producing clone was done based on ELISA, long-term and short-term colony formation assay. Therapeutic efficacy of antibodies was evaluated in tumor xenograft models. Result: Through a loss-of-function genomic screening approach, we identified secretory SCUBE3 as a critical player that promotes the survival, therapy resistance and metastasis of TNBC cells. Our MS data shows that SCUBE3 interacts with several oncogenic proteins including EFGR, TGFβR and mutant CALR. Transcriptomic analysis demonstrates that SCUBE3 depletion leads to inhibition of pro-tumorigenic signaling as well as activation of anti-tumor immunity pathways. Our results reveal that SCUBE3 regulates the expression of FOXR2 which in turn stabilizes Myc and recruits DNMT to form a repressor complex on the promoter of IRF1. Inactivation of IRF1 gene suppresses the expression of MHC class I and class II genes which leads to immune suppression. Further, our syngeneic mouse model study shows that suppression of SCUBE3 leads to increased infiltration of CD8+ T cells and NK cells into the tumor. Importantly, using multiple pre-clinical models including PDX models, we show that antibody targeting SCUBE3 blocks tumor growth. Conclusion: This study will set the stage for a new paradigm of treating TNBCs by inhibiting SCUBE3 activity. Our study is first to identify SCUBE3 as a novel ligand of EGFR1 and CALR^mut. Notably, our study shows that SCUBE3 promotes pro-tumorigenic signaling cascade while inhibiting anti-tumor immunity to promote TNBC growth and render them resistant to therapy. Importantly, we provide compelling evidence that antibody targeting SCUBE3 may serve as a potent therapeutic for treating TNBCs.

K. Liu, J. Geng, Z. Yu, Y. Yeh, H. Wei, W. Sun, W. Li, J. Lu, J. Deng, C. Yang, L. Geng, X. Luo, Y. Liang, Z. Liu, Z. Gao, Y. Lou; R&D Department, NiKang Therapeutics Inc, Wilmington, DE.

CDK4 is the pivotal driver of HR+HER2- breast cancer and CDK4/6 inhibitors in combination with endocrine therapy have revolutionized the treatment landscape. However, several challenges remain for the approved CDK4/6 inhibitors, including severe hematotoxicity associated with CDK6 inhibition and eventual resistance, often mediated by CDK2/cyclin E pathway activation. Therefore, simultaneously and selectively targeting CDK4 and CDK2 represents a promising therapeutic strategy. The high homology between CDK1 and CDK2, as well as between CDK4 and CDK6, poses a significant challenge for the discovery of selective CDK2/4 dual inhibitors. To overcome these limitations, we employed the targeted protein degradation approach and have discovered NKT5097, a first-in-class, selective, and orally bioavailable CDK2/4 dual degrader. Herein, we disclose the evaluation of NKT5097 activity across a panel of human cancer cell lines to assess its ability to induce CDK2 and CDK4 protein degradation. Its selectivity over other CDKs, especially CDK1 andCDK6, was confirmed through a series of cellular and biochemical assays. Additionally, anti-tumor efficacy and potential off-target effects of NKT5097 were assessed and benchmarked against CDK2 or CDK4 specific inhibitors currently in clinical development. NKT5097 degrades CDK2 and CDK4 rapidly and effectively (DC50: 0.6-15 nM), leading to potent inhibition of Rb phosphorylation and proliferation in CDK2- or CDK4-dependent cancer cells and has up to 50-fold greater potency than PF-07104091 (tagtociclib, a CDK2i) or PF-07220060 (atirmociclib, a CDK4i). In addition, NKT5097 demonstrates strong selectivity for CDK2/4 degradation over CDK1/6/7/9 in various cellular assays. Indeed, NKT5097 treatment predominantly arrests CDK2- and/or CDK4-dependent cells at G1 phase of the cell cycle and has minimal impact on G2 distribution, suggesting a lack of CDK1 inhibition. NKT5097 also demonstrates a 68-fold increase in selectivity for CDK2 over CDK1 in comparison to PF-07104091, and a 33-fold increase in selectivity for CDK4 over CDK6 compared to PF-07220060 in functional cellular assays. The selectivity of NKT5097 for CDK2/4 is further confirmed in a Kinome screen as well as by global proteomic profiling, in which CDK2 and CDK4 are the top hits among the >8,000 proteins quantified. Oral administration of NKT5097 dose-dependently degrades CDK2 and CDK4 and shows robust anti-tumor activities across multiple tumor xenograft models including HR+HER2- breast cancer and cyclin E1-high cancer models. Additionally, combination of NKT5097 with fulvestrant leads to deeper degradation of CDK2 and CDK4 in vitro, resulting in better tumor growth inhibition in HR+ breast cancer models. These findings establish NKT5097 as a potent, selective and orally bioavailable dual degrader of CDK2 and CDK4. NKT5097 offers distinct advantages over the combined use of CDK2 and CDK4 inhibitors, including more comprehensive pathway inhibition and a lower risk of overlapping toxicity, owing to its high potency against CDK2/4 and strong selectivity over CDK1/6. Therefore, as the first-in-class CDK2/4 dual degrader, NKT5097 presents a unique opportunity to serve as a backbone therapy for HR+HER2- breast cancer and other cancers with CDK2 and/or CDK4 dysregulation. NKT5097 is currently being investigated in a phase I clinical study.

H. Jung, S. Kim, S. Kim, Y. Lee, J. Yoo, J. Kwon, S. Han, D. Park, D. Shin, H. Kim, J. Kim, Y. Ko, S. Kim; R&D, Voronoi Inc, Incheon, KOREA, REPUBLIC OF.

Many systemic therapeutic options for advanced metastatic HER2-positive breast cancer have been approved, including antibody-based therapies, chemotherapies, and tyrosine kinase inhibitors (TKIs). However, brain metastasis and acquired resistance are still major unmet medical needs for HER2-driven metastatic breast cancer and advanced solid cancers. VRN101099 is an orally bioavailable, highly selective HER2 kinase inhibitor, demonstrating excellent specificity over other kinases, including EGFR. This compound potently inhibits HER2 by covalently binding to Cys805 within the ATP-binding pocket, leading to robust catalytic inhibition of HER2 kinase. In vitro studies show that VRN101099 exhibits potent anti-proliferative effects against HER2-driven breast cancer cell lines, such as BT-474 and SK-BR-3. Furthermore, VRN101099’s irreversible binding properties enable effective inhibition of various HER2 activating mutants, including L755S, V777L, D769H/Y and L869R. Notably, VRN101099 demonstrates brain penetrability. Given its distinct mechanism of action—kinase catalytic inhibition—which differs from that of antibody-based HER2 therapies (e.g., Trastuzumab, Pertuzumab, T-DM1, T-DXd, Zanidatamab), VRN101099 holds promise for synergistic therapeutic effects when combined with these agents, particularly in the treatment of metastatic HER2-positive breast cancer. Moreover, VRN101099 induces internalization of the HER2 receptor, promoting internalization of the antibody-drug conjugate. In BT474 or SKBR3, HER2 amplified breast cancer cells, xenograft models with subcutaneous or intracranial implantation, VRN101099 showed combinatory and synergistic anti-tumor effects with Trastuzumab, Pertuzumab, T-DM1, T-DXd, and Zanidatamab. The first-in-human dose escalation study of VRN101099 monotherapy will determine the maximal tolerable dose, and an optimal dose finding study for the combination will be followed.

A. Patnaik¹, M. Block², M. Cristofanilli³, Q. Liu⁴, H. Majeski⁴, A. Shakya⁴, P. Rajagopalan⁴, C. Deshpande⁴, T. Marmon⁴, S. Padam⁴, D. Powell⁴, A. H. Chourasia⁴, R. Urbanski⁴, M. Goetz⁵, R. M. Layman⁶; ¹The START Center for Cancer Research, The START Center for Cancer Research, San Antonio, TX, ²Nebraska Cancer Specialists, Nebraska Cancer Specialists, Omaha, NE, ³Weill Cornell Medicine/New York-Presbyterian Hospital, Weill Cornell Medicine/New York-Presbyterian Hospital, New York, NY, ⁴Biotheryx, Biotheryx, San Diego, CA, ⁵Department of Oncology, Mayo Clinic, Rochester, MN, ⁶Department of Breast Medical Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX.

Background: BTX-9341 is a first-in-class bifunctional degrader of cyclin dependent kinase (CDK)4 and CDK6 that also inhibits the transcription of CDK2 and Cyclin E. Preclinical data highlight its superiority compared with approved CDK4/6 inhibitors (CDK4/6i) in inhibition of phosphorylation of retinoblastoma (p-RB), cell cycle arrest, and in vivo efficacy in breast cancer (BC) xenografts. BTX-9341 is active in CDK4/6i‑resistant models with p-RB half-maximal inhibitory concentrations (IC₅₀s) of 1‑15 nM and can overcome key mechanisms that drive CDK4/6i resistance. Here we report early monotherapy data from the first-in-human study of BTX-9341 in patients (pts) with advanced and/or metastatic HR+/HER2- BC. Methods: BTX-9341-101 is a multicenter, nonrandomized, open-label trial to assess the safety, tolerability, pharmacokinetics (PK), pharmacodynamics (PD), and preliminary efficacy of orally administered BTX-9341 in pts with RB1 wild type advanced and/or metastatic HR+/HER2- BC who received prior CDK4/6i therapy. An accelerated titration and Bayesian Optimal Interval design (BOIN) were used. The primary objective of dose escalation is to determine the maximum tolerated dose (MTD) or maximum evaluable dose (MED) of BTX-9341 monotherapy and in combination with fulvestrant. Secondary objectives include the characterization of BTX-9341 on PK and efficacy; exploratory objectives include the assessment of BTX-9341 on PD and PK/PD relationships. Results: As of June 9, 2025, 12 pts were evaluated at the following BTX-9341 once daily (QD) dose levels: 50 mg (dose level [DL]1), 100 mg (DL2), and 75 mg (DL3) monotherapy, and 50 mg in combination with fulvestrant (DLA). Pts received up to 4 lines of therapy in the metastatic setting prior to enrollment. No dose‑limiting toxicities (DLTs) were observed at any dose levels during the 28-day DLT period. The most common adverse events (AEs) during the DLT period were decreased neutrophil and white blood cell counts (Grades 1-4). There were no dose reductions in DL1. The starting dose was reduced in all pts in DL2; DL3 is currently enrolling. At least 2-fold accumulation was observed after QD dosing, and steady-state trough concentrations were in the range of in vitro IC₅₀ values at DL1 and higher dose levels. Serum thymidine kinase activity (sTKa), a clinical biomarker for tumor response to CDK4/6 inhibition, showed >50% reduction at Cycle 2 Day 1 (C2D1) compared with baseline for 7 of 8 evaluable monotherapy pts, with 6 pts below the sTKa detectable level. “On target” PD reductions in CDK4, CDK6, CDK2, and Cyclin E levels were observed in peripheral blood mononuclear cells. Of the 8 evaluable monotherapy pts, the best overall responses in pts with measurable disease included 1 pt with a confirmed partial response (PR) and 3 pts with stable disease ([SD] first assessed at 8 weeks; i.e., the first time point for assessment in the protocol), and in pts with non-measurable disease, 2 pts had non-complete response (CR)/non‑progressive disease. For all pts, the longest duration on treatment was 10 cycles. A clinical benefit rate (defined as a best overall response of CR, PR, or SD for pts with measurable disease or non-CR/non-progressive disease for pts with non‑measurable disease, per RECIST v1.1 at Week 24) of 66.7% was observed in the completed cohorts (DL1 and DL2). Conclusion: BTX-9341 monotherapy shows a favorable safety profile, with the most common AEs being decreased neutrophil and white blood cell counts, at doses that show encouraging preliminary PK/PD and efficacy, enabling further evaluation of monotherapy and in combination with fulvestrant. The trial is actively enrolling with anticipated completion of dose escalation enrollment by the end of 2025 (Clinical trial: NCT06515470).

W. Chang¹, E. Olivar¹, G. Pai¹, J. Low¹, S. Min², R. Hoff², N. Chang³, T. Hoque³, A. Park³, C. Lee⁴; ¹QC, Sapu Bioscience, LLC, San Diego, CA, ²Formulation, Sapu Bioscience, LLC, San Diego, CA, ³QA, Sapu Bioscience, LLC, San Diego, CA, ⁴R&D, Sapu Bioscience, LLC, San Diego, CA.

Background Oral everolimus (Afinitor) is an established therapy for hormone receptor-positive, HER2-negative breast cancer. However, its clinical utility is limited by low and variable oral bioavailability (~15%), gastrointestinal toxicity, and a narrow therapeutic window driven by first-pass gut extraction and CYP3A/P-gp interactions. Sapu003 is an intravenous Deciparticles™ formulation of everolimus designed to circumvent intestinal loss and toxicity while delivering consistent systemic exposure. Methods Pharmacokinetics (PK), tissue distribution, metabolism, and excretion of Sapu003 were characterized after a single 3 mg/kg IV bolus in Sprague-Dawley rats and a 0.6 mg/kg 30-minute IV infusion in beagle dogs, with oral everolimus (1-3 mg/kg) used as a comparator. Whole blood, organs, feces, and urine were quantified for everolimus by a validated LC-MS/MS method. In vitro plasma, microsome, and hepatocyte stability were assessed across multiple species, including mouse, rat, dog, monkey, and human. Results In rats, IV Sapu003 achieved a mean Cmax of 133 ng/mL and an AUC₀-∞ of 339 h·ng/mL, compared with 2.4 ng/mL and 52 h·ng/mL after a 3 mg/kg oral PEG-400 everolimus suspension (representing 55- and 6.5-fold increases, respectively). In dogs, the IV infusion yielded a Cmax of 51-76 ng/mL and an AUC₀-∞ of 203-395 h·ng/mL, compared with 8.4 ng/mL and 94-123 h·ng/mL after 1 mg/kg oral Afinitor (representing 6-9- and 2-4-fold increases, respectively), while maintaining comparable half-lives. Plasma, microsome, and hepatocyte stabilities were unchanged relative to everolimus. Sapu003 rapidly distributed to high-flow organs but, unlike oral everolimus, showed low deposition in the gastrointestinal tract. Elimination remained primarily biliary/fecal (>94% of the dose, completed within ≤48 h), with renal clearance accounting for <2%. No intestinal toxicity was observed in dogs administered Sapu003. Conclusions IV administration of Sapu003 overcomes the absorption bottleneck of oral everolimus, providing predictable exposure with a controllable Cmax while preserving the native metabolic fate. By bypassing the GI tract, Sapu003 exhibits consistent blood PK with no intestinal toxicity. These data support the ongoing Phase I evaluation of weekly 30-minute infusions of Sapu003 in metastatic breast cancer, aiming to enhance mTOR pathway suppression, including effects on mTORC2 targets.

Y. Chen¹, J. Zhang¹, H. Guo², Y. Sun³, Y. Wang⁴, Z. Zhang⁵, T. Yao⁶, Y. Shi⁷, H. Zhang⁸, Y. Chen⁹, Y. Lu¹⁰, J. Song¹⁰, P. Li¹⁰, D. Fang¹⁰, W. Sha¹⁰, C. Z. Ding¹⁰; ¹Breast Cancer, Fudan University Shanghai Cancer Center, Shanghai, CHINA, ²Breast Cancer, Henan Cancer Hospital, Zhengzhou, CHINA, ³Breast Cancer, Shandong Cancer Hospital and Institute, Jinan, CHINA, ⁴Medical, Shandong Cancer Hospital and Institute, Jinan, CHINA, ⁵Breast Cancer, College of Clinical Medicine of Henan University of Science and Technology, Luoyang, CHINA, ⁶Breast Cancer, Hospital of Bengbu Medical College, Bengbu Anhui, CHINA, ⁷Breast Cancer, Tianjin Clinical Research Center for Cancer, Tianjin, CHINA, ⁸Breast Cancer, Nanyang Central Hospital, Nanyang Henan, CHINA, ⁹Breast Cancer, Hospital of Zhejiang University School of Medicine, Hangzhou, CHINA, ¹⁰Medical, Yangli Pharmaceutical Technology, Hangzhou, CHINA.

Background: TFX06, a third-generation oral CERAN/SERD, was evaluated in a multicenter Phase 1a/b study (NCT05927779) in patients with ER+/HER2- locally advanced and/or metastatic breast cancer (Zhang J, et al., 2024 SABCS annual meeting). The recommended dose was established at 150mg QD. Dose expansion study at 150mg had been initiated, and this abstract summarizes our key findings. Patients and Methods: Eligible patients regardless of their menopausal status, who had received ≥1 lines of endocrine therapy (ET), prior treatment with CDK4/6i and fulvestrant (fulv); and ≤2 line of chemotherapy for metastatic disease were allowed to enroll. TFX06 tablets were orally administered 150mg once daily (QD) every 28 days of treatment cycle until disease progression or intolerable toxicity. As of cut-off date June 30, 2025, 57 patients received TFX06 monotherapy at 150mg dose, median age of 60 years (range 34-74) with ECOG score of 0-1. 78.9% of patients enrolled had visceral metastasis in liver, lung or both, and 14% of patients had brain metastasis. Median number of prior endocrine therapies were 2 (1-4) including fulvestrant (56.1%), CDK4/6i (61.4%) and combination of both (38.6%) and a median number of chemotherapies was 1 (0-3). Tumor assessments by investigators according to RECIST v1.1 were performed every eight weeks. Results: TFX06 was well tolerated. The most common (≥20% of patients) treatment-emergent adverse events (TEAEs) were hypertriglyceridemia, AST increased, anemia, lactate dehydrogenase increased and hypoalbuminaemia. The incidence of nausea, diarrhea and vomiting was low (7.0%, 5.3% and 1.8%, respectively). In addition, no ocular toxicity reported. Among 55 evaluable patients, there were 6 partial responses (ORR was 10.9%); clinical benefit rate (CBR) was 45.5% and mPFS for the overall population was 5.5 months. mPFS for ESR1m detectable and ESR1m-undetectable patients were 5.7 months and 4.6 months respectively. In addition, among the 13 patients who had prior CDK4/6i treatment and with detectable ESR1 mutation, the CBR was 46.2% and mPFS was 5.6 months. Among 21 patients who had prior fulvestrant+CDK4/6i combo treatment, CBR was 42.9% and mPFS was 3.7 months, while ESR1 detectable mutation patients, CBR was 55.6% and mPFS was 5.6 months. Notably, of four patients with measurable brain lesions, three were evaluable: two showed >70% brain lesion reduction (PR), one had stable disease, suggesting TFX06 had intracranial activity. Conclusions: TFX06 at 150 mg QD dose expansion trial enrolled endocrine therapy heavily treated patients, including prior treatment with combination of fulvestrant and CDK4/6 inhibitor progressed patients. TFX06 demonstrated favorable safety and clinical efficacy even in this hard-to-treat patient population.

I. Lanzagorta-Calvillo¹, B. Kaghazchi¹, E. Poradosu², N. O. Carragher¹, A. Unciti-Broceta¹; ¹Institute of Genetics and Cancer, University of Edinburgh, Edinburgh, UNITED KINGDOM, ²R&D, Nuvectis Pharma Inc., Fort Lee, NY.

Background. Most breast cancer patients present tumors with high estrogen receptor levels (ER+), which make endocrine therapies (ET) the standard of care for this breast cancer subtype. However, resistance to treatment inevitably develops over time, which demands optimal combination strategies. The combination of CDK4/6 inhibitors with ET has emerged as the foremost therapeutic modality for hormone receptor-positive/HER2-negative advanced breast cancer. However, resistance to these therapies eventually develops with the clinical benefit of subsequent lines of therapy being limited and short lasting. Deregulation of the Hippo pathway leading to YAP1 nuclear translocation has been shown to promote resistance to CDK4/6 inhibitors. NXP900 (eCF506), is a novel, highly potent, YES/SRC kinase inhibitor with high selectivity that uniquely binds the kinase in the autoinhibited closed conformation, thereby blocking its catalytic and scaffolding properties [1]. Preclinical studies have demonstrated that NXP900 effectively inhibits YAP1 nuclear translocation and proliferation of FAT1 mutated cells [2]. In addition, NXP900 demonstrated potent inhibition of cell proliferation of induced endocrine resistance in vitro [3], thus, establishing the hypothesis that NXP900 may represent a therapeutic option for the treatment of endocrine and/or CDK4/6 resistance breast cancer. The aims of this project were to (i) investigate the sensitivity of luminal A cell lines to NXP900 as a single agent, and (ii) assessing NXP900’s efficacy in combination with endocrine therapy in vitro and in vivo in comparison to current approved combinations of endocrine therapy and CDK4/6 inhibitors. Methods. Cell proliferation and viability were determined by automated image analysis of confluence and Presto Blue assay, respectively. Dose response studies were performed for NXP900, dasatinib and fulvestrant in a library of ER+ luminal A cell lines. Western blot analysis was performed to study target inhibition. Dose-ratio combinations of; NXP900 and fulvestrant; dasatinib and fulvestrant, and CDK4/6 inhibitors and fulvestrant, were used to compare the effect of each combination. In vivo xenograft studies were performed for the combination of fulvestrant and NXP900. Results. NXP900 showed higher antiproliferative activity than dasatinib against the luminal A breast cancer cell lines tested. Notably, in these cell lines NXP900 demonstrated far superior SRC/YES inhibition than dasatinib, correlating with the cancer cell viability data. Dose ratio matrix drug combination analysis using SynergyFinder+ demonstrated significantly stronger synergistic interactions in ER+ cell lines treated with endocrine therapy in combination with NXP900, rather than in combination with dasatinib or CDK4/6 inhibitors. Conclusions. Our results demonstrate that luminal A cancer cell lines are generally sensitive to NXP900 treatment and that in combination with endocrine therapy it is more synergistic than the standard-of-care combination of fulvestrant and CDK4/6 inhibitors. This study supports the potential translation of NXP900 into breast cancer patients alongside endocrine therapies to improve the treatment of ER+ tumors. References [1] Temps et al. A Conformation Selective Mode of Inhibiting SRC Improves Drug Efficacy and Tolerability. Cancer Res. 2021, 81 (21), 5438-5450. [2] Kaghazchi et al. NXP900, a novel YES1/SRC kinase inhibitor currently in clinical development, blocks YAP1 signaling in NSCLC cell lines. Cancer Res. 2025, 85 (8, Suppl_1), 5599. [3] Lanzagorta-Calvillo et al. Endocrine therapy-resistant luminal A breast cancer cell lines are sensitive to the novel YES1/SRC tyrosine kinase inhibitor, NXP900. Cancer Res. 2025, 85 (8, Suppl_2), LB292.

A. J. Montero¹, M. R. Patel², A. I. Spira³, B. Bashir⁴, D. L. Richardson⁵, S. Thamake⁶, D. Trone⁶, B. Veresh⁶, P. LoRusso⁷; ¹Department of Hematology and Oncology, University Hospitals/Seidman Cancer Center, Case Western Reserve University, Cleveland, OH, ²Drug Development, Florida Cancer Specialists/Sarah Cannon Research Institute, Sarasota, FL, ³Medical Oncology, Clinical Research, NEXT Oncology-Virginia, Fairfax, VA, ⁴Department of Medical Oncology, Division of Solid Tumors, Sidney Kimmel Comprehensive Cancer Center at Thomas Jefferson University, Philadelphia, PA, ⁵Section of Gynecologic Oncology, Stephenson Cancer Center/Sarah Cannon Research Institute, University of Oklahoma Health Sciences Center, Oklahoma City, OK, ⁶Avenzo Therapeutics, Avenzo Therapeutics, San Diego, CA, ⁷Medical Oncology and Hematology, Yale Medicine, New Haven, CT.

Background: Resistance to CDK4/6 inhibitors (CDK4/6i) in HR+/ HER2− metastatic breast cancer (mBC) is frequently associated with CCNE1 overexpression and CDK2 activation. Targeting CDK2 may overcome this resistance and enhance the efficacy of CDK4/6i-based regimens. AVZO-021 is a potent CDK2i with high selectivity for CDK2 over CDK1, thus minimizing off-target toxicity. We present preliminary results from the Phase 1 segment of the ongoing first-in-human Phase 1/2 study of AVZO-021 (NCT05867251). Methods: The ongoing Phase 1 open-label, dose-escalation segment of this study is evaluating the safety, pharmacokinetics (PK), and preliminary efficacy of AVZO-021 alone and in combination with endocrine therapy [(ET) fulvestrant], CDK4/6i (ribociclib/abemaciclib) + ET (fulvestrant/letrozole), an antibody-drug conjugate (sacituzumab govitecan-hziy), or chemotherapy (carboplatin) in patients with advanced solid tumors, including HR+/HER2− mBC and CCNE1-amplified solid tumors. AVZO-021 is administered orally once daily (QD) across escalating dose levels (20 to 220 mg in monotherapy dose escalation). Results: As of May 14, 2025, 47 patients have received at least 1 dose of AVZO-021 (27 in monotherapy; 20 in combination). The median age was 63 years for patients who received AVZO‑021 in monotherapy and 59 years for patients who received AVZO‑021 in combination with ET ± CDK4/6i, sacituzumab govitecan-hziy, or chemotherapy. Overall, 96% of patients were female, 68% had breast cancer, 15% had ovarian/fallopian tube cancer, and 6% had endometrial/uterine cancer. Patients who received AVZO-021 in monotherapy had a median of 4 prior lines of systemic therapy, and those treated with combination therapy had a median of 6 prior lines. All patients with HR+/HER2− mBC (n=30) had received prior CDK4/6i and hormonal therapy. No dose-limiting toxicities (DLTs) or treatment discontinuations due to adverse events (AEs) were observed with AVZO-021 monotherapy up to a dose of 220 mg. The most common (>25%) all-grade treatment-emergent AEs (TEAEs) with monotherapy were fatigue (41%), nausea (37%), anemia (26%), and vomiting (26%); 2 patients required dose reductions due to AEs (Grade 2 thrombocytopenia [n=1, 180 mg dose level]; Grade 2 worsening fatigue [n=1, 120 mg dose level]). No new safety signals or DLTs were reported in the combination cohorts. AVZO-021 monotherapy demonstrated a favorable PK profile with a low peak-to-trough ratio, robust target coverage, and mean half-life of 14 hours. Updated clinical data will be presented at the conference, including ctDNA analysis and preliminary efficacy of monotherapy and combination therapy in HR+/HER2− mBC and CCNE1-amplified solid tumors. Conclusions: AVZO-021 is a novel CDK2i with high selectivity for CDK2 over CDK1 and a favorable preliminary clinical profile. Together with the mechanistic rationale, the emerging clinical data support the continued development of AVZO-021 as a monotherapy and in combinations, including with ET ± CDK4/6i and selective CDK4i.

Y. Lin¹, H. Sang¹, A. Ku¹, D. Thomas¹, L. Jia¹, A. Talab¹, D. Zhou², B. Lim³, Y. Li¹; ¹Lester & Sue Smith Breast Center, Baylor College of Medicine, Houston, TX, ²Department of Biochemistry & Structural Biology, University of Texas Health Science Center at San Antonio, San Antonio, TX, ³Department of Breast Medical Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX.

Triple-negative breast cancer (TNBC) accounts for approximately 10-15% of all breast cancer cases and has a poorer prognosis compared to other breast cancer subtypes. However, there are only limited therapy options for TNBC. The anti-apoptotic protein BCL-xL, a member of the BCL-2 family, has been implicated in promoting tumor cell survival and resistance to chemotherapeutic agents. However, clinical application of current small-molecule BCL-xL inhibitors has been hindered by dose-limiting toxicities, particularly thrombocytopenia resulting from on-target effects in platelets. DT2216 is a novel proteolysis-targeting chimera (PROTAC) designed to degrade BCL-xL in tumor cells while sparing platelets. Here, we aim to evaluate the potential use of DT2216 in the treatment of TNBC patients. Bioinformatics analysis revealed that the BCL-xL gene is highly expressed in TNBC and associated with poor clinical outcomes in TNBC patients. Consistently, BCL-xL expression was also elevated in TNBC cell lines relative to non-TNBC subtypes. BCL-xL knockdown significantly reduced viability of TNBC cell lines (MDA-MB-231 and HS578T) while showing little impact in estrogen receptor (ER)-positive breast cancer cell lines (MCF-7 and T47D), indicating a subtype-specific dependency on BCL-xL. In accordance with these genetic findings, DT2216 markedly inhibited growth of TNBC cell lines while exhibiting minimal effects on ER+ breast cancer cell lines. We further conducted a comprehensive evaluation of the combined effects of DT2216 and standard chemotherapy in TNBC models. DT2216 significantly potentiated the cytotoxic effects of two commonly used chemotherapeutic agents (paclitaxel and carboplatin) across a range of concentrations, producing a robust synergistic inhibition of TNBC cell proliferation. In vivo experiments validating these in vitro data are ongoing. Together, these data indicate that the selective degradation of BCL-xL by DT2216, either as monotherapy or in combination with standard-of-care chemotherapy, could offer a novel therapeutic strategy to improve outcomes for patients with TNBC.

H. R. Luz¹, L. P. Franchi², A. C. Tedesco³, E. d. Lacerda⁴; ¹Genetics, Federal University of Goiás, Goiânia, BRAZIL, ²Institute of Biological Sciences, Federal University of Goiás, Goiânia, BRAZIL, ³Department of Chemistry, University of São Paulo, Ribeirão Preto, BRAZIL, ⁴Center for Human Genetics, Institute of Biological Sciences, Federal University of Goiás, Goiânia, BRAZIL.

Triple-negative breast cancer is a highly aggressive subtype with the poorest prognosis among breast cancers. It is defined by the absence of estrogen (ER), progesterone (PR), and human epidermal growth factor receptor 2 (HER2), limiting the effectiveness of conventional hormone-based therapies. Around 20% of the 2.3 million annual breast cancer cases are triple-negative, and this subtype is linked to high metastasis and mortality rates. Its molecular heterogeneity and therapeutic resistance highlight the urgent need for alternative treatment strategies. Photodynamic therapy is a promising, minimally invasive technique already employed in Brazil’s public health system (SUS) for the treatment of basal cell carcinoma. It relies on the activation of a photosensitizer (PS) by a specific light source, leading to the generation of reactive oxygen species (ROS) that trigger oxidative stress, cell death, and immune response activation. This in vitro study assessed the cytotoxicity of photodynamic therapy using a nanoemulsion containing chloroaluminum phthalocyanine (ClAlPc/Ne) in breast cancer cell lines exposed to varying PS concentrations and light doses. The MCF-7 and MDA-MB-231 cell lines, representing luminal A and triple-negative breast cancer respectively, were incubated with ClAlPc/Ne for 24 hours before photoactivation, followed by a further 24-hour incubation before metabolic viability analysis using the MTT assay. Three experimental conditions were tested: (I) nanoemulsion without PS, (II) ClAlPc/Ne without irradiation, and (III) photoactivated ClAlPc/Ne. PS concentrations ranged from 0.01 µM to 2.0 µM, with light doses from 1 J/cm² to 16 J/cm². Absorbance was measured at 545 nm. Results showed selective phototoxicity of ClAlPc/Ne against the triple-negative breast cancer cell line MDA-MB-231, with IC50 (50% inhibitory concentration) values of 1.96 µM at 1 J/cm², 0.19 µM at 4 J/cm², and 2.0 µM at 1 J/cm², 0.14 µM at 4 J/cm², and <0.06 µM at 16 J/cm². These findings indicate that phototoxicity is dose-dependent and more pronounced in triple-negative breast cancer cells. The differential response between cell lines suggests that triple-negative breast cancer may be more sensitive to oxidative stress induced by photodynamic therapy. These findings reinforce the importance of molecular profiling in guiding treatment decisions and support the potential of photodynamic therapy as a targeted, less invasive approach for hard-to-treat breast cancer subtypes.

S. Bhandari, A. Alhalwani, I. Terrazas, R. Sridharan, S. Rice, H. L. McArthur, I. S. Chan; Internal Medicine, UT Southwestern Medical Center, Dallas, TX.

Background: Triple negative breast cancer (TNBC) is a biologically distinct subtype with a high risk of early relapse, particularly in patients with residual disease following neoadjuvant chemotherapy (NAC). Failure to achieve a complete response may reflect underlying immune evasion mechanisms, including upregulation of immune checkpoints such as programmed cell death protein 1 (PD-1), which impair effective T cell-mediated antitumor responses. Cryoablation (cryo) can stimulate T-cell priming and reduce MDSCs to overcome resistance to immune checkpoint inhibition (ICI). Preclinical and early clinical data suggest that the combination of cryo with ICI may synergistically improve immune activation. We conducted a multi-center, single arm, phase 2 study to evaluate the impact of cryo with standard-of-care pembrolizumab (pembro) in women with residual TNBC after taxane-based neoadjuvant chemotherapy (NAC), a group at high risk of early relapse (NCT03546686). Methods: Eligible women are ≥18y, with ER < 10%, PR < 10%, HER2 negative (per ASCO/CAP definition), ≥ 1.0 cm, residual operable disease after taxane-based NAC. Patients underwent percutaneous, image-guided cryo with concurrent research core biopsy 7-10 days prior to surgery and receive peri-operative pembro per standard-of-care. Peripheral blood was collected pre- and post-cryo, and flow cytometry was performed on paired peripheral blood mononuclear cells (PBMCs) from an initial cohort (n=5) to identify changes in immune response. Results: Post-cryo, we observed an overall trend in the decrease of PD1-positive T cells and myeloid cells. PD1+CD8+ T cells decreased by 2.9-fold, PD1+CD4+ T cells decreased by 4.6-fold and PD1+CD3- decreased by 6.3-fold. Notably, we observed a decrease in CD11b+CD3- MDSCs by 14.92-fold. Additional exploratory correlative studies on tumor and blood, including tumor infiltrating lymphocytes and tumor PD-1 IHC, are ongoing to fully characterize the immunologic impact of the intervention and to explore predictors of efficacy. Conclusion: These initial results suggest that the addition of cryo to pembro may act synergistically to reduce the overall immunosuppressive immune environment, potentially resulting in improved immunity in patients with high-risk residual TNBC. Funding sources: Susan G. Komen, B-SEARCH, Conquer Cancer Foundation, Breast Cancer Research Foundation, Boston Scientific.

M. Lobba¹, D. Trinter¹, M. Nguyen¹, S. Brady¹, C. Symister¹, A. Lau¹, D. Garcia-Almedina¹, S. Johri¹, F. Matthew², R. Kendall¹; ¹Research and Development, CatenaBio, Berkeley, CA, ²Dept. of Chemistry, University of California at Berkeley, Berkeley, CA.

Introduction: Antibody-Drug Conjugates (ADCs) have a tremendous impact on patient outcomes in breast and other cancers. Dato-DXd, for example, is second-line therapy for stage IV, HR positive/HER2 negative metastatic breast cancer. Sacituzumab govitecan (SG), a TROP2 targeted ADC, is approved as a 3^rd line therapy in locally advanced or metastatic HER2 negative breast cancer, and in combination with pembrolizumab, is likely to see approval in 1^st line therapy for PD-L1 positive tumors. However, many patients fail to respond or relapse after treatment with ADC therapies due to tumor heterogeneity and eventual resistance to the ADC payload. Combination therapies have historically outperformed mono chemo approaches across most solid tumors to date, pointing to a potential avenue for improvement of ADC efficacy. We are developing next generation Multi-Payload Conjugates™ (MPCs™) that deliver targeted combination chemotherapies within a single molecule to address shortcomings in current ADCs. Method: CatenaBio has developed highly stable, dual-payload ADC combination therapies, with tunable payload ratios. Our selective conjugation platform allows the attachment of distinct payloads targeting different mechanisms of action at three unique sites on antibody scaffolds. Results: Catena’s lead TROP2 targeting MPC, CATB-101, features an optimized combination and ratio of tubulin and Topo1 inhibitors. CATB-101 demonstrates superior inhibition and excellent tolerability in mouse models of tumor growth in TROP2 expressing xenograft models including multiple CDX and PDX models of TNBC (triple negative breast cancer). In head-to-head comparisons with T-DXd, SG and Dato-DXd, MPCs out-perform the current approved ADCs in breast cancer and demonstrate full tumor elimination at low mg/kg doses. Furthermore, CATB-101 rescues CDX models after progression on SG treatment, demonstrating potential applications in late-line treatment. Non-GLP Non-Human Primate (NHP) toxicology trials with CATB101 demonstrate remarkable safety profile, surpassing benchmark mono-payload ADCs in tolerability. Conclusion: While advances have been made in the design of ADCs to expand to previously unaddressed populations, high patient relapse and the failure of recent mono-payload ADCs in late-stage trials indicate a need for novel multi-payload conjugates. Catena’s MPCs™ offer the next step in ADC design and allow for the targeted delivery of multiple mechanisms of action with a single MPC™ are highly efficacious at eliminating tumors across multiple CDX and PDX xenograft models with a range of target expression. Now validated in early NHP toxicology studies with a significantly enhanced therapeutic window, these molecules offer the potential to circumvent tumor resistance pathways to deliver deeper and more durable patient responses.

J. F. Rodriguez-Orengo¹, J. Duconge², M. Acosta³, J. Wang⁴, D. Yardley⁵, S. Dharmawardhane⁶, N. Sankar¹; ¹Cancer Drug Development, MBQ Pharma, Inc., San Juan, PR, ²Pharmaceutical Sciences, University of Puerto Rico Medical Sciences Campus, San Juan, PR, ³Oncology, FDI Clinical Research, San Juan, PR, ⁴Cancer Treatment, Florida cancer Specialists and Research Institute, Tallahassee, FL, ⁵Oncology Therapy, SCRI Oncology Partners, Nashville, TN, ⁶Biochemistry, University of Puerto Rico Medical Sciences Campus, San Juan, PR.

MBQ-167 is a first-in-class dual inhibitor of Rac and Cdc42 GTPases with potent preclinical anti-metastatic activity and is currently under evaluation in a Phase 1 clinical trial (NCT06075810) in patients with advanced breast cancer. This study aimed to characterize the pharmacokinetic (PK) profile of MBQ-167 and its active metabolite M6 in patients treated with 20 mg, 40 mg and 80 mg BID doses (cohorts 0, 1 and 2, respectively). Plasma concentration-time data were analyzed from eight patients using non-compartmental analysis. MBQ-167 was rapidly absorbed with a median T_max of 1–4 hours and a C_max ranging from 139 to 573 ng/mL (20 mg BID), 520 to 843 ng/mL (40 mg BID) and 578 to 688 ng/mL (80 mg BID). The elimination half-life (t½) of MBQ-167 was approximately 3–4.5 hours, and the mean residence time (MRT) was <7 hours. Dose-normalized systemic exposure (AUC/D) did not increase proportionally with dose, suggesting a possible reduction in bioavailability at higher doses, potentially due to enterohepatic recycling or saturation of absorption. Importantly, there was no evidence of systemic overexposure or meaningful accumulation (Rac < 2.0) of MBQ-167 or M6, indicating a favorable safety margin. Volume of distribution (Vdss) exceeded 30L for both compounds, reflecting extensive tissue distribution consistent with their lipophilicity. M6 demonstrated rapid formation and elimination with t½ values ranging from 3.1 to 4.4 hours, further supporting efficient metabolic conversion. These findings confirm that MBQ-167 exhibits linear PK at clinically relevant doses with manageable exposure and clearance. The observed pharmacokinetics support continued clinical development and provided essential dosing guidance for Phase 2 trials. Further analyses with additional cohorts will be presented up to 400 mg BID.

S. Piha-Paul¹, S. Lavasani², F. Dayyani³, S. Goel⁴, A. Tsimberidou¹, N. Ibrahim¹, C. Barcenas¹, A. Ilheme¹, J. Rodon¹, E. E. Dumbrava¹, K. Hennessy⁵, P. Peng⁶, C. Sun⁷, H. Wang⁷, R. Xu⁷, J. Fan⁵, F. Meric-Bernstam¹; ¹Department of Investigational Cancer Therapeutics, MD Anderson Cancer Center, Houston, TX, ²Medical Oncology, University of California, Irvine, Orange, CA, ³Medical Oncology, City of Hope National Medical Center, Duarte, CA, ⁴Oncology, Rutgers Cancer Center Institute of New Jersey, New Brunswick, NJ, ⁵Clinical Operations, TransThera Sciences (US), Inc, Gaithersburg, MD, ⁶Preclinical, TransThera Sciences (Nanjing), Inc., Nanjing, CHINA, ⁷Clinical Operations, TransThera Sciences (Nanjing), Inc., Nanjing, CHINA.

Background: Patients (pts) with advanced/metastatic hormone receptor positive /human epidermal growth factor receptor 2 negative (HR+/HER2-) breast cancer (BC) often develop resistance to endocrine therapy and CDK4/6 inhibitors, while those with triple-negative BC (TNBC) usually lack effective targeted therapies after progression on chemotherapy or antibody drug conjugates. Tinengotinib, a novel multi-kinase inhibitor targeting FGFRs, VEGFRs, Aurora A/B, JAK1/2, and CSF1R, demonstrated preclinical activity in BC models, including those resistant to standard therapies. Here we present the final efficacy, safety and correlative biomarker data of tinengotinib monotherapy in pts with HR+/HER2- and TNBC from the Phase Ib/II trial TT420X1103 (NCT04742959).Methods: Eligible pts with HR+/HER2- BC or TNBC with no effective standard therapeutic treatment options were enrolled. TNBC pts included acquired TNBC (aTNBC) disease, defined as pts initially diagnosed with HR+/HER2- BC but found to be transformed to TNBC at study entry. Tinengotinib was given as monotherapy at 10 mg or 12 mg daily dose (QD) dose levels, or 6 mg twice daily (BID) dose level . Response was assessed per RECIST v1.1, and safety was evaluated per CTCAE v5.0. Biomarker sampling for circulating tumor DNA (ctDNA) analysis was completed at baseline, cycle 3 day 1 (C3D1), the time of response, the time of progression and end of treatment (EOT). Results: As of 20Feb2024, 21 pts with advanced/metastatic BC were treated with tinengotinib monotherapy at 10 mg QD (n=3), 12 mg QD (n=16), and 6 mg BID (n=2) dose levels. All pts were female; median age 58 years (IQR 48-63), ECOG PS 1 in 100% of pts, 18 (85.7%) of pts had Stage IV disease, and 18 (85.7%) of pts had ≥3 lines of prior systemic therapy.Of 20 efficacy evaluable pts, objective response rate (ORR) was 30% (n = 6), duration of response was 4.81 months (mos) (95% CI 3.15, 6.01) and disease control rate (DCR) was 75% (n=16). Treatment related adverse events (TRAEs) were reported in 20 (95.2%) pts, 10 (47.6%) pts had G3/4 TRAEs, and no G5 TRAEs were reported. Eighteen pts were evaluable for ctDNA analysis. Among the most frequent alterations (TP53, DNMT3A, APC, PIK3CA, FGFR2 and PTEN), the incidence of ROS1 mutations was higher in pts with partial response (PR) (p=0.03), as well as in pts with PFS>3.5 mos (p=0.02). Of these 18 pts, 8 had TNBC and 6 had aTNBC. It was observed that pts with aTNBC had a longer PFS vs TNBC pts (4.3 vs 2.6 months, p=0.011); ORR was 67% vs 0% and DCR was 100% vs 50%. The incidence of ROS1 and DNMT3A mutations was higher in aTNBC vs TNBC (p=0.015 and p=0.026, respectively). Analysis showed a reduction in variant allele frequency (VAF) of 7 pts from baseline to C3D1 (p<0.0001), indicating a molecular response with treatment. One pt with progressive disease showed increased VAF in mutations of PIK3CA, PTEN, TP53, and RAD51D from baseline to EOT; 3 pts with acquired resistance to the therapy detected new gene mutations (DNMT3A, NF1, CTNNA1, ERBB3, MLL2, CDKN2A and PDGFRB) at EOT. Conclusions: Tinengotinib exhibited promising clinical benefit and manageable safety profile for the treatment of HR+/HER2- BC or TNBC. The biomarker correlative analysis revealed that ROS1 mutations could be a potential predictive biomarker. Early reduction in ctDNA VAF may serve as a pharmacodynamic biomarker. Baseline genomic alterations in PIK3CA, PTEN, TP53, and RAD51D may contribute to primary resistance, while acquired resistance may be associated with emergence of mutations in pathways of RAS/MAPK, PI3K/AKT, RTK, MET, and cell cycle regulation. The improved clinical outcomes and enrichment of ROS1/DNMT3A mutations in aTNBC pts define this transformed subtype as a subgroup warranting further investigation.

V. Guarneri¹, G. Griguolo², C. Vernieri³, M. Dieci¹, Y. Silvestri⁴, F. Giovanardi⁵, F. Ligorio³, G. Fotia⁶, S. Rapisarda⁴, A. Chirco⁷, A. Zambelli⁸, M. De Laurentiis⁹, A. Gennari¹⁰, S. Marsoni¹¹, S. Piccolo¹², M. Foiani¹³; ¹Department of Surgery, Oncology and Gastroenterology – University of Padova, Oncology 2, Istituto Oncologico Veneto IRCCS, Padova, ITALY, ²Department of Surgery, Oncology and Gastroenterology – University of Padova, Oncology 2, Istituto Oncologico Veneto IRCCS, Padova, ITALY, ³Medical Oncology Department – Fondazione IRCCS Istituto Nazionale dei Tumori, IFOM ETS, the AIRC Institute of Molecular Oncology, Milan, ITALY, ⁴Precision Oncology Unit, IFOM ETS, the AIRC Institute of Molecular Oncology, Milan, ITALY, ⁵Medical Oncology, Comprehensive Cancer Centre, AUSL-IRCS di Reggio Emilia, Reggio Emilia, ITALY, ⁶Medical Oncology Department – Fondazione IRCCS Istituto Nazionale dei Tumori, Oncology and Hematology-Oncology Department, University of Milan, Milan, ITALY, ⁷Medical Oncology Unit, ASST Papa Giovanni XXIII, Bergamo, ITALY, ⁸Department of Biomedical Sciences, Humanitas University, Pieve Emanuele, Milan, ITALY, ⁹Department of Breast and Thoracic Oncology, Division of Breast Medical Oncology, Istituto Nazionale Tumori IRCCS Fondazione Pascale, Naples, ITALY, ¹⁰Department of Translational Medicine – University of Piemonte Orientale, Novara, IT, Division of Medical Oncology, Maggiore University Hospital, Novara, ITALY, ¹¹Precision Oncology Unit, IFOM ETS, the AIRC Institute of Molecular Oncology, Padova, ITALY, ¹²Department of Molecular Medicine, University of Padova, IFOM ETS, the AIRC Institute of Molecular Oncology, Padova, ITALY, ¹³Istituto di Genetica Molecolare, CNR, IFOM ETS, the AIRC Institute of Molecular Oncology, Pavia, ITALY.

Background: Preclinical studies have reported that Ataxia telangiectasia and Rad3-related (ATR) inhibition results in increased nuclear envelope fragility and promotion of cGAS-STING activation, thus potentially resulting in an enhanced immune response. The phase II investigator-initiated ATRiBRAVE trial evaluates the efficacy of the ATR inhibitor ceralasertib priming followed by durvalumab-nab-paclitaxel in patients with TNBC relapsing after immune checkpoint inhibitors (ICIs) and chemotherapy (CT) treatment in the curative setting. This report presents results from the safety run-in phase (SRP) of the trial, representing the first assessment of the safety and feasibility of the triple combination of ceralasertib, durvalumab and nab-paclitaxel in human. Methods: This single-arm trial enrolls patients with unresectable/metastatic TNBC (ER<10% HER2-), previously treated with ICIs and CT in the early setting (adjuvant, neoadjuvant or both) and not previously treated in the metastatic setting. Treatment consists of: Ceralasertib 240mg orally BID (DL 0) for 7 days before cycle 1, then days 22-28 of each cycle (with planned dose de-escalation if needed), Durvalumab 1500 mg IV, day 1 (q28), Nab-paclitaxel 100mg/m² IV, days 1, 8, 15 (q28). Treatment continues until disease progression or unacceptable toxicity. Primary endpoint of the study: Progression-free survival (PFS). Study sample size: 37 patients (34 evaluable patients + 10% drop-out) SRP design: a SRP with a 3+3 de-escalation schema was preplanned in the study to assess safety. If fewer than 2 dose-limiting toxicities (DLTs) occur in the first 6 patients at DL 0, SRP is successfully completed. Otherwise, ceralasertib will be de-escalated to 160 mg BID (DL-1) and, if necessary, to 120 mg BID (DL-2). Results: Up to date, 18 patients have been enrolled in the study, 9 of which were treated in the SRP (3+6). Among the first 3 patients, 2 DLTs (prolonged neutropenia leading to significant dose omissions) were observed, prompting ceralasertib reduction to 160 mg BID (DL-1). No DLTs were observed in the first 6 (3+3) patients at DL-1 (ceralasertib 160 mg BID for 7 days), thus successfully completing SRP. Among patients treated in the SRP, no grade 5 AEs occurred. All grade 3-4 AEs observed were hematologic but did not qualify as DLTs. No patients discontinued treatment due to toxicity during SRP. Durvalumab safety profile was consistent with previous reports. Conclusions: Ceralasertib priming at a 160 mg BID for 7 days, followed by durvalumab and nab-paclitaxel, is feasible and well tolerated. Nine additional patients have already been enrolled in the phase II portion of the trial (total enrolment: 18 patients) and patient enrolment is currently ongoing. Updated safety data for patients who completed treatment in the study will be presented at the meeting. This study was funded by Fondazione AIRC under 5 per mille 2019 (ID. 22759 program—group leader VG and SM). This study was conducted with the unconditional support of AstraZeneca.

Y. Yin¹, A. Spira², H. Lan³, M. Zhao⁴, Y. Mao⁵, W. Yao⁶, Y. Sun⁷, J. Wu⁸, X. Sun⁹, S. Liu⁹, H. Mu⁹, K. Wang⁹, Y. Qiu¹⁰, E. Hamilton¹¹; ¹Department of Oncology, The First Affiliated Hospital of Nanjing Medical University, Nanjing, CHINA, ²Clinical Research, Next Oncology, Fairfax, VA, ³Department of Medical Oncology, Sichuan Provincial People’s Hospital, Sichuan Academy of Medical Sciences, School of Medicine UESTC, Chengdu, CHINA, ⁴Department of Medical Oncology, The First Affiliated Hospital of China Medical University, Shenyang, CHINA, ⁵Department of Oncology, The Affiliated Hospital of Jiangnan University, Jiangsu, CHINA, ⁶Department of Thoracic Medical Oncology, Sichuan Cancer Hospital, Chengdu, CHINA, ⁷Phase I Clinical Research Center, The Affiliated Cancer Hospital of Shandong First Medical University, Jinan, CHINA, ⁸Department of Radiotherapy, The First Affiliated Hospital of Hainan Medical University, Haikou, CHINA, ⁹Clinical Development, Dualitybiologics, Shanghai, CHINA, ¹⁰R&D, Dualitybiologics, Shanghai, CHINA, ¹¹Breast Cancer Research Program, Sarah Cannon Research Institute, Nashville, TN.

Background: HER3-targeted antibody-drug conjugates (ADCs) have shown benefits in breast cancer (BC). DB-1310 is a novel ADC comprised of a humanized anti-HER3 IgG1 monoclonal antibody with a highly internalizing unique Domain I epitope, a cleavable peptide linker, and a potent DNA topoisomerase I inhibitor P1021. Initial clinical data demonstrated a manageable safety profile and encouraging anti-tumor activity, particularly in patients with EGFR-mutant NSCLC (Lisberg ASCO 2025). Here, we present the efficacy and safety results of DB-1310 in pretreated hormone receptor-positive (HR+), HER2-negative (HER2-) BC patients. Methods: In the Phase 1 part of this trial (NCT05785741), patients with advanced solid tumors, including HR+/HER2- BC, received DB-1310 monotherapy intravenously every three weeks (Q3W). Safety was the primary endpoint, with efficacy endpoints including objective response rate (ORR), duration of response (DoR), disease control rate (DCR), progression-free survival (PFS), and overall survival (OS) as secondary endpoints. Results: As of July 04, 2025, a total of 207 patients with solid tumors were enrolled and treated with DB-1310 monotherapy (1.5-6.5 mg/kg, Q3W) in Phase 1 (ECOG PS 1, 84.1%; White, 26.1%; Asian, 69.1%). Among these, 16 HR+/HER2- BC patients were enrolled (3.0-5.5 mg/kg, Q3W), with 15 of them being efficacy-evaluable, and a median prior line of systemic therapy was 2 (range, 1-7). Efficacy Results (N=15 HR+/HER2- BC patients from Phase 1): nine achieved a partial response (one pending confirmation), and four achieved stable disease, resulting in a confirmed ORR of 53.3% (95%CI: 26.59, 78.73) and a DCR of 86.7% (95%CI: 59.54, 98.34). Notably, all partial responses occurred in patients treated with 5-5.5 mg/kg (N=13), yielding an unconfirmed ORR of 69.2%, a confirmed ORR of 61.5%, and a DCR of 92.3% for this dose range. The median DoR is not reached. Median PFS is 14.78 months (95% CI: 11.60, not evaluable) and median OS is 14.69 months (95% CI: 11.89, not evaluable). Safety Results (N=207, all enrolled patients from Phase 1): The safety profile was assessed in the entire Phase 1 part. Treatment-related adverse events (TRAEs) of Grade (G) ≥ 3 occurred in 89 (43.0%) patients, and 25 patients (12.1%) had treatment-related serious adverse events. TRAEs led to dose reduction in 38 (18.4%) and discontinuation in 9 (4.3%) patients. No TRAE-related deaths were reported. Most common TRAEs (any grade/≥G3; >20% overall) were neutrophile count decreased (49.3%/ 26.1%), anemia (47.8%/ 6.8%), platelet count decreased (45.4%/ 15.9%), white blood cell count decreased (43.5%/ 11.1%), nausea (40.6%/ 1.0%), decreased appetite (29.5%/ 0.5%), aspartate aminotransferase increased (26.6%/ 0), alopecia (26.1%/ 0), vomiting (24.6%/ 0), and hypoalbuminemia (21.7%/ 0). Interstitial lung disease/pneumonitis occurred in 10 pts (4.8%; one Grade 2, all others G1). The safety profile in the 16 HR+/HER- BC patients was consistent with the overall Phase 1 part. Conclusion: DB-1310 showed promising anti-tumor activity in heavily pretreated HR+/HER2- BC patients. The manageable safety profile observed in the larger Phase 1 part, involving 207 patients, is consistent with that of the BC subpopulation, supporting its further development.

E. J. Diego¹, J. A. Foldi², P. F. McAuliffe¹, Q. Sabih¹, L. A. Emens³, A. Salazar⁴, O. J. Finn⁵; ¹Department of Surgery, University of Pittsburgh, Pittsburgh, PA, ²Department of Internal Medicine, University of Pittsburgh, Pittsburgh, PA, ³Department of Internal Medicine, Kaiser Permanente South Sacramento, Sacramento, CA, ⁴Oncovir, Oncovir, Inc, Winchester, VA, ⁵Department of Immunology, University of Pittsburgh, Pittsburgh, PA.

Early Results from a Phase 1 Clinical Trial of a Cancer Vaccine Targeting Tumor Associated Antigen MUC1 for Immune Interception of Breast Cancer in Patients with Ductal Carcinoma in SituIntroductionHypoglycosylated tumor MUC1, a transmembrane glycoprotein, is recognized by human T-cells and antibodies as a tumor-associated antigen. It is overexpressed in premalignant precursor lesions, including ductal carcinoma in situ (DCIS), thereby serving as a potential potent DCIS rejection target. Vaccine-induced immune response to MUC1 may halt DCIS recurrence or progression to invasive disease and may offer a future strategy for primary disease prevention in high-risk individuals. We hypothesize that our MUC1 peptide vaccine will be both safe and immunogenic in patients with DCIS and that the elicited systemic immune response will affect changes in the microenvironment of the DCIS from pro- to anti-tumor. MethodsThis single institution, open label, randomized phase I clinical trial (NCT06218303) is designed for 50 post-menopausal women with previously untreated estrogen-receptor positive DCIS histologically confirmed on core needle biopsy (CNB). Patients with autoimmune diseases were excluded. Patients are randomized 2:1 to the vaccine group. The vaccine is composed of a 100aa long MUC1 peptide corresponding to 5 tandem repeats of 20 amino acids from the MUC1 variable number of tandem repeats region (VNTR), admixed with the poly-ICLC adjuvant Hiltonol. The vaccine is administered subcutaneously (SC) or intramuscularly (IM). The vaccine group receives the MUC1 peptide vaccine series pre-operatively (at 0, 2, and 10 weeks) with an optional aromatase inhibitor (AI). The control group receives only optional AI. All undergo surgery at 12 weeks. Research blood is drawn in the vaccine group at baseline and 2 weeks after each vaccine, and in the control group at baseline and at week 12. Tissue from the pre-treatment CNB and the post-treatment operation are also collected. An optional booster is available to vaccine responders 6 months post-surgery.The primary endpoint is vaccine immunogenicity, measured as anti-MUC1 IgG level. Vaccine responders are identified based on a > 2-fold increase in serum anti-MUC1 IgG from baseline, using enzyme-linked immunosorbent assay (ELISA). Adverse events are monitored per Common Terminology Criteria for Adverse Events (CTCAE) version 5.0.Fisher’s exact test will compare primary immunogenicity between the study arms with one-sided α=0.05. Correlative studies include extensive phenotyping of circulating as well as DCIS-infiltrating immune cells and increases in anti-MUC1 cellular immunity post vaccination. ResultsTo date,13 patients completed the study: 7 in the vaccine and 6 in the control group. Mean age is 66 years. There were no serious adverse events. The primary vaccine series elicited or boosted anti-MUC1 IgG in 6 of 7 vaccinated women, for a response rate of 86%. Only 1 patient to date was eligible to receive the booster vaccine and showed a strong immune memory response. There was no anti-MUC1 IgG present in the control group. Extensive phenotyping of peripheral blood mononuclear cell samples from both groups and evaluation of MUC-specific cellular responses are in progress, as are tissue analyses. ConclusionMUC1 peptide vaccine given in the premalignant DCIS setting is safe in the first 7 patients to receive the vaccine. The 86% immune response rate predicts a high rate of response, which will allow us to study immune mechanisms activated by the vaccine, including changes in the DCIS microenvironment.

J. M. Johnson¹, E. E. Rhoades², H. B. Levengood¹, A. Ali², H. Gilmore³, M. L. Kruse², E. E. Roesch², T. Onger², B. Elliott², E. Haury², C. Porvasnik², T. Young⁴, T. Coutee⁵, J. A. Fitzgerald⁶, T. S. Stappenbeck¹, G. T. Budd²; ¹Department of Inflammation & Immunity, Lerner Research Institute, Cleveland Clinic, Cleveland, OH, ²Taussig Cancer Institute, Cleveland Clinic, Cleveland, OH, ³Robert J. Tomsich Pathology and Laboratory Medicine Institute, Cleveland Clinic, Cleveland, OH, ⁴n/a, Previvorsandsurvivors.com, Inc., Delray Beach, FL, ⁵n/a, DiepCJourney Foundation, Duvall, WA, ⁶n/a, Sisters4Prevention, West Palm Beach, FL.

Background: α-Lactalbumin (aLA) is expressed in lactating breasts and 70% of triple-negative breast cancer (TNBC) but not at other times or in other tissues. Based on the “retired protein hypothesis” vaccination with aLA provided protection from development of autochthonous tumors in transgenic murine models of breast cancer and inhibited growth of established 4T1 transplantable breast tumors in BALB/c mice. Methods: We completed accrual to a Phase I trial of recombinant human aLA with GMP-grade zymosan adjuvant in Montanide ISA 51 VG vehicle in 3 cohorts of subjects: Ia) patients with high-risk TNBC who have completed all standard treatment; Ib) patients with BRCA1, BRCA2, or PALB2 mutations who are undergoing risk-reducing mastectomies; and Ic) patients with TNBC who have residual cancer after primary chemo-immunotherapy and are receiving post-operative treatment with pembrolizumab +/- capecitabine. Three vaccinations were given once every 2 weeks. Events of Common Terminology Criteria for Adverse Events (CTCAE) Grade ≥ 2 are considered dose-limiting toxicities (DLTs). Blood was drawn prior to therapy as well as 14, 28, and 56 days after the first vaccination to assess cellular response using enzyme-linked immunosorbent spot (ELISPOT) assays of interferon-gamma and interleukin-17 production in response to aLA and for humoral response by enzyme-linked immunosorbent assay (ELISA). The breast tissue of participants in the Phase Ib cohort were or will be examined for occult lactational foci and inflammatory changes. Results: We vaccinated 26 patients in Cohort Ia, 4 in Cohort Ib, and 5 in Cohort Ic. CTCAE toxicity by dose level (DL) is summarized in Table 1 by grade for each study cohort. All adverse events (AEs) were injection site reactions, with ulceration and need for incisional drainage representing the Grade 3 DLT events. 19 of 26 patients assayed to date across all cohorts met protocol specified definitions of an immune response based on ELISPOT assays that quantify frequencies of T cells producing IFNγ or IL-17 in response to recombinant aLA; these data included 4 of 6 subjects in Cohort Ia at DL1 and 6 of 11 subjects in all cohorts treated at DL1 and tested to date. ELISPOT and ELISA data for all subjects will be available in August 2025 and will be presented. Conclusions: DL1 is the maximum tolerated dose (MTD) by protocol definition, produced at worst Grade 1 toxicity in a total of 14 of 15 subjects, and produced an immune response in most patients, based on the criteria prospectively defined in the study protocol. DL1e was tolerable in 5 of 5 subjects but was not determined the MTD per our 3+3 trial design. Immune data for all subjects in the trial, including new data across all cohorts, will be reported and will inform the design of subsequent Phase II trials. Funding Source: Department of Defense (W81XWH-17-1-0592 and W81XWH-17-1-0593).

*Diagnosed while pregnant

K. Nakamura¹, K. Saito², R. Ito¹, M. Yoshida¹, E. Hatakawa¹, R. Yamakado¹, M. Yoshikawa¹, S. Watanabe¹, N. Imai¹, M. Shibusawa¹, A. Noro¹, Y. Kozuka³, K. Kawaguchi¹; ¹Breast Surgery, Mie University Hospital, Tsu, JAPAN, ²Medical Oncology, Mie University Hospital, Tsu, JAPAN, ³Pathology, Mie University Hospital, Tsu, JAPAN.

Background: Circadian rhythms regulate immune responses, but their clinical relevance in cancer immunotherapy remains underexplored. We investigated whether the timing of immune checkpoint inhibitor (ICI) administration, in conjunction with tumor-infiltrating lymphocyte (TIL) status, influences treatment efficacy in breast cancer. Methods: We conducted a retrospective analysis of two cohorts: a perioperative cohort receiving neoadjuvant ICI (n=22, early-stage) and a metastatic cohort treated with ICI in the advanced setting (n=18). For each patient, the median time of ICI administration was determined and dichotomized as morning (<12:00) or afternoon (≥12:00). In the perioperative cohort, pathological complete response (pCR) was assessed; in the metastatic cohort, progression-free survival (PFS) was analyzed. Pre-treatment stromal TIL levels were evaluated and stratified by median value (TIL-high vs. TIL-low). Associations between administration time, TIL status, and treatment outcomes were explored. Results: In the perioperative cohort, neither administration timing nor TIL level alone showed a significant association with pCR. However, when stratifying by both variables, an interaction emerged: among 9 patients with high TILs, those treated in the morning (n=4) achieved a 100% pCR rate, while those treated in the afternoon (n=5) had a 55.6% non-pCR rate. In contrast, patients with low TILs had high pCR rates regardless of administration time (morning: 83.3%, afternoon: 80.0%). In the metastatic cohort, PFS did not significantly differ between morning and afternoon administration, nor by PD-L1 or TIL status, although trends suggested numerically shorter PFS in afternoon-treated patients with high PD-L1 expression. Across both cohorts, no meaningful association was observed between time of administration and immune-related adverse events. Conclusion: While overall efficacy was not significantly influenced by ICI administration time, subgroup analysis revealed a paradoxical pattern among TIL-high patients in the perioperative setting: afternoon administration was associated with reduced pCR rates. This finding suggests that the functional phenotype of TILs—and possibly their interaction with circadian immune modulation—may shape response to immunotherapy. Further studies incorporating spatial and functional profiling of the tumor immune microenvironment are warranted to clarify the clinical implications of circadian timing in ICI treatment.

C. Benvenuti¹, L. Perotti², A. Martin³, C. Gaudin³, C. Jouannaud⁴, M. Fournier⁵, C. Kaderbhaï⁶, A. Kieffer⁷, F. Cherifi⁸, M. Campone⁹, F. Lerebours¹⁰, P. Cottu¹¹, F. André¹, A. Bertaut¹², D. Antonio¹, S. Michiels², B. Pistilli¹, T. Grinda¹; ¹Medical Oncology, Gustave Roussy, Villejuif, FRANCE, ²Biostatistics and Epidemiology, Gustave Roussy, Villejuif, FRANCE, ³Data and Partnerships, UNICANCER, Paris, FRANCE, ⁴Medical Oncology, Institut Godinot, Reims, FRANCE, ⁵Medical Oncology, Institut Bergonié, Bordeaux, FRANCE, ⁶Medical Oncology, Centre Georges-François Leclerc, Dijon, FRANCE, ⁷Medical Oncology, Institut de cancérologie de Lorraine, Vandœuvre-lès-Nancy, FRANCE, ⁸Medical Oncology, Centre François Baclesse, Caen, FRANCE, ⁹Medical Oncology, Institut de Cancérologie de l’Ouest, Saint-Herblain, FRANCE, ¹⁰Medical Oncology, Institut Curie, Saint-Cloud, FRANCE, ¹¹Medical Oncology, Institut Curie, Paris, FRANCE, ¹²Medical Oncology, Centre Georges François Leclerc, Dijon, FRANCE.

Background: TNBC exhibits a high risk of early relapse, yet predictors of time to relapse remain poorly defined. Early relapses (while on or within 12 months after ending curative therapies) are associated with chemotherapy-resistant disease and poor prognosis (median overall survival <12 months; Grinda et al., Eur J Cancer 2023). Identifying key factors associated with early and late relapse may help guide therapeutic interventions. This study explores associations between clinical factors and timing of relapse in patients (pts) with TNBC from the prospective CANTO cohort (NCT01993498). Methods: Early-stage TNBC pts enrolled in the CANTO cohort between 2012 and 2023 were included. Descriptive statistics were used to summarize clinical variables. Invasive disease-free survival (IDFS) was estimated with Kaplan-Meier curves from the time of study entry. The proportional hazards assumption was assessed with Schoenfeld residuals and cumulative hazard plots. A time-varying fixed-effect multivariable Cox model was built using stratification at 24 months for distinguishing early and late relapses. To identify the best-fitting multivariable model for IDFS, stepwise selection (stepAIC) was applied starting from a priori covariates, including age, grade, surgery type, and TNM stage , with additional candidates tested based on univariate results and clinical relevance. Results: Among the 999 pts included, median age was 52.9 years (IQR: 45.1-61.9). Most pts presented with stage II disease (56.3%), followed by stage I (34.4%) and stage III (9.3%). Chemotherapy was administered in 93.5% of pts: adjuvant in 51.3% and neoadjuvant in 42.2% of cases. Breast-conserving surgery was performed in 71.7% pts, mastectomy in 28.3%. Around half of the pts underwent axillary dissection (47.6%).Median follow-up was 74.3 months (IQR: 51.2-99.4). During this period, 239 iDFS events occurred, 49% within the first 24 months. The best final time-varying fixed effect multivariate model for predicting iDFS included age, BMI at diagnosis, smoking, type of surgery, grade, stage and chemotherapy exposure. TNM stage showed a strong prognostic role across the entire follow-up, with stage III disease consistently associated with higher relapse risk compared to stage I (HR: 2.44; 95% CI: 1.44-4.14; p < 0.001). For early relapse, underweight status (BMI < 18.5 Kg/m²) was the strongest predictor (HR: 3.35; 95% CI: 1.58-7.11; p = 0.002). Axillary dissection (HR: 1.60; 95% CI: 1.00-2.55; p = 0.05) and receipt of (neo)adjuvant chemotherapy (HR: 1.51; 95% CI: 0.95-2.40; p = 0.078) trended toward a signifcant increased risk of early relapse. Despite adjustment for TNM stage to account for disease severity, these treatment-related variables may act as proxies for disease burden, and some residual counfounding cannot be excluded. Late relapse was associated with older age (HR: 1.03; 95% CI: 1.01-1.04; p = 0.003), while smokers had a borderline increased risk versus non-smokers (HR: 1.61; 95% CI: 0.98-2.63; p = 0.06). Conclusions: This large cohort study shows that early and late relapses in early-stage TNBC are driven by distinct clinical factors. Our findings integrate clinicopathological variables to support the development of models reflecting individual relapse risk and timing. Early relapses, accounting for half of the events, were strongly associated with underweight status, possibly reflecting patient frailty and comorbidities, and high disease burden at diagnosis, as indicated by stage III disease and high treatment intensity, including axillary dissection and chemotherapy. Late relapses were predicted by older age and stage III disease. These data support time-aware prognostic models to guide individual-risk adapted strategies and underscore the need for biological investigations to better understand early relapse in TNBC.

M. Antonini¹, A. Mattar², F. P. Cavalcante³, F. P. Zerwes⁴, E. C. Millen⁵, F. P. Brenelli⁶, A. L. Frasson⁷, M. D. Teixeira², M. Madeira⁸, G. Facina⁹, H. L. Couto¹⁰, G. T. Tosello¹¹, R. Arakelian¹², A. D. Lima¹³, G. N. Garcia¹⁴, F. Bagnoli¹⁵, A. G. Amorim¹⁶, M. F. de Figueiredo², L. R. Soares¹⁷, R. Freitas Júnior¹⁸, L. H. Gebrim¹⁹; ¹Breast Surgery, Hospital do Servidor Publico Estadual de São Paulo, São Paulo, BRAZIL, ²Breast Surgery, Hospital da Mulher SP, São Paulo, BRAZIL, ³Breast Surgery, Hospital Geral de Fortaleza, Fortaleza, BRAZIL, ⁴Breast Surgery, Pontifícia Universidade Católica do Rio Grande do Sul, Porto Alegre, BRAZIL, ⁵Breast Surgery, Americas Oncologia, Rio de Janeiro, BRAZIL, ⁶Breast Surgery, Universidade Estadual de Campinas, Campinas, BRAZIL, ⁷Breast Surgery, Hospital Israelita Albert Einstein, São Paulo, BRAZIL, ⁸Breast Surgery, Faculdade Israelita de Ciências da Saúde Albert Einstein, São Paulo, BRAZIL, ⁹Breast Surgery, Universidade Federal de São Paulo, São Paulo, BRAZIL, ¹⁰Breast Surgery, Redimama – Redimasto, Belo Horizonte, BRAZIL, ¹¹Breast Surgery, Unoeste, Presidente Prudente, BRAZIL, ¹²Oncology, Hospital da Mulher SP, São Paulo, BRAZIL, ¹³Breast Surgery, Universidade Federal do Rio de Janeiro, Rio de Janeiro, BRAZIL, ¹⁴Breast Surgery, Oncoclínicas, São Paulo, BRAZIL, ¹⁵Breast Surgery, Santa casa de misericórdia de São Paulo, São Paulo, BRAZIL, ¹⁶Breast Surgery, Hospital do Servidor Público Estadual Francisco Morato de Oliveira, São Paulo, BRAZIL, ¹⁷Breast Surgery, Universidade Federal de Goiás, São Paulo, BRAZIL, ¹⁸Breast Surgery, Universidade Federal de Goiás, Goiania, BRAZIL, ¹⁹Breast Surgery, Hospital Beneficiência Portuguesa de São Paulo, São Paulo, BRAZIL.

Background: Triple-negative breast cancer (TNBC) accounts for 15-20% of breast cancers and lacks expression of estrogen receptor, progesterone receptor, and HER2, limiting targeted options. Recent neoadjuvant therapy (NAT) advances include immune checkpoint inhibitors (ICIs) and PARP inhibitors, improving pathologic complete response (pCR) and survival but raising concerns about severe toxicity. Despite broader adoption of these regimens, a systematic analysis of Grade ≥3 adverse events (AEs) from contemporary phase III trials is lacking. Methods: We conducted a PRISMA-guided systematic review and meta-analysis (PROSPERO: CRD42024562785). MEDLINE, Embase, and Cochrane were searched for phase III randomized trials (Jan 2010-Jun 2025) evaluating NAT in non-metastatic TNBC and reporting Grade ≥3 AEs (CTCAE). One phase II trial (WSG-ADAPT-TN; n=336) with robust comparative design and AE reporting was also included. Pooled prevalence and risk ratios (RR) with 95% CI were calculated using DerSimonian-Laird random-effects models. Subgroup analyses included: Group A (PARP inhibitors), B (immunotherapy), and C (platinum-based regimens). Heterogeneity was assessed via I². Results: Six trials with 3,316 TNBC patients were included (BrighTNess, IMpassion031, KEYNOTE-522, NeoTRIP, PARTNER, WSG-ADAPT-TN). Neutropenia was the most common Grade ≥3 AE, highest with immunotherapy (33.5%), followed by PARP inhibitors (23.7%) and platinum regimens (16.0%) (Q=71.8, p<0.0001). PARP inhibitors significantly increased neutropenia risk (RR=1.84), while immunotherapy showed no increase (RR=1.00, p=0.94). Grade ≥3 anemia was higher with PARP inhibitors (20.6%) vs. immunotherapy (9.9%), with doubled risk vs. placebo (RR=2.08, p=0.02). Thrombocytopenia was most prevalent with immunotherapy (7.9%), but PARP inhibitors showed the highest risk (RR=3.56, p=0.0001). Febrile neutropenia was more common with ICIs (14.9%) than PARP inhibitors (3.0%). ALT elevation was more frequent in platinum regimens (6.4%), without significant differences across groups. Fatal AEs were <1% in all trials. Conclusions: Modern NAT regimens improve outcomes in TNBC but confer significant Grade ≥3 toxicity. PARP inhibitors pose the highest hematologic risk, while ICIs increase febrile neutropenia. Platinum regimens showed the most favorable toxicity profile. Findings highlight the need for biomarker-driven, individualized treatment and vigilant AE monitoring in early TNBC.

R. Barroso-Sousa¹, L. Testa², P. Mandó³, M. C. Tavares⁴, N. C. Nunes⁵, G. Cordoba³, F. Waisberg³, F. C. Balint⁴, I. M. de Sousa⁶, M. O. Andrade¹, M. C. Gouveia⁷, F. Madasi⁸, J. Bines⁸, R. P. Ferreira⁹, D. D. Rosa⁹, C. L. Santos¹⁰, M. R. Monteiro¹¹, Z. S. de Souza¹², D. Assad-Suzuki¹², C. dos Anjos¹³, D. M. Gagliato¹⁴, A. U. Gomes¹⁴, B. M. Zucchetti⁷, A. Ferrari¹⁵, M. L. de Brito¹⁶, M. F. Monteiro¹⁷, P. A. Signorini¹⁸, N. J. Gomes¹⁹, C. Gallina³, S. Sanches⁴, P. M. Hoff²⁰, M. Estevez-Diz²¹, R. C. Bonadio²⁰; ¹Oncology, Brasilia Hospital, Oncologia Américas, Brasilia, BRAZIL, ²Oncology, Instituto D’Or de Pesquisa e Ensino, Brasilia, BRAZIL, ³Oncology, SUMA (Grupo Cooperativa Argentino para el estudio y la investigación del Cáncer de Mama), Buenos Aires, ARGENTINA, ⁴Oncology, A.C.Camargo Cancer Center, São Paulo, BRAZIL, ⁵Oncology, Grupo Américas, Rio de Janeiro, BRAZIL, ⁶Oncology, A.C.Camargo Cancer Center, Brasilia, BRAZIL, ⁷Oncology, Hospital 9 de Julho, Grupo Américas, São Paulo, BRAZIL, ⁸Oncology, Instituto D’Or de Pesquisa e Ensino, Rio de Janeiro, BRAZIL, ⁹Oncology, Hospital Moinhos de Vento, Porto Alegre, BRAZIL, ¹⁰Oncology, Instituto D’Or de Pesquisa e Ensino, Recife, BRAZIL, ¹¹Oncology, Grupo Américas, Hospital Samaritano, São Paulo, BRAZIL, ¹²Oncology, Hospital Sírio-Libanês, Brasilia, BRAZIL, ¹³Oncology, Hospital Sírio-Libanês, São Paulo, BRAZIL, ¹⁴Oncology, Hospital Beneficência Portuguesa, São Paulo, BRAZIL, ¹⁵Oncology, Hospital Santa Paula, Grupo Américas, São Paulo, BRAZIL, ¹⁶Oncology, Clínica AMO, Salvador, BRAZIL, ¹⁷Oncology, Instituto do Câncer do Ceará, Fortaleza, BRAZIL, ¹⁸Oncology, Centro Integrado de Pesquisa da Amazônia (CINPAM), Manaus, BRAZIL, ¹⁹Oncology, Hospital São Domingos, DASA Oncologia, São Luiz, BRAZIL, ²⁰Oncology, Instituto D’Or de Pesquisa e Ensino, São Paulo, BRAZIL, ²¹Oncology, Instituto do Câncer do Estado de São Paulo, Faculdade de Medicina da Universidade de São Paulo, São Paulo, BRAZIL.

Rare but Resistant: Neoadjuvant Chemoimmunotherapy in Special Histological Subtypes of Triple-Negative Breast Cancer — Insights from the Neo-Real/GBECAM-0123 Study Background Special histological types (SHT) of triple-negative breast cancer (TNBC) exhibit distinct biological behavior and treatment responses. However, due to their rarity, patients with SHT are underrepresented in clinical trials. We aimed to evaluate the effectiveness of pembrolizumab (P) combined with neoadjuvant chemotherapy (NCT) in patients with SHT of TNBC. MethodsWe analyzed data from patients with TNBC and SHT enrolled in the Neo-Real/GBECAM-0123 study—a multicenter, real-world study including patients treated with neoadjuvant P+NCT across ten cancer centers since July 2020. Effectiveness outcomes included pathologic complete response (pCR) and event-free survival (EFS). ResultsOf the 727 patients included to date in the Neo-REAL study, 646 (88.9%) had TNBC of no special type (NST), 51 (7%) SHT, 4 (0.6%) undifferentiated, and 26 (3.6%) unknown histology. Among the SHT group, 20 metaplastic, 16 lobular, and 15 other subtypes (9 apocrine, 3 micropapillary, 2 neuroendocrine, and 1 medullary). Patients with lobular, metaplastic, and other SHT tumors were older than those with NST (median ages: 58, 51, 49, and 44 years, respectively; p = 0.001). Grade 3 tumors were more frequent in NST (76%) and metaplastic (84%) subtypes than in lobular (50%) and other SHT tumors (50%) (p = 0.019). A high Ki-67 index (≥50%) was also more common in NST tumors (76%) compared to metaplastic (52%), lobular (53%), and other SHT tumors (36%) (p < 0.001). Tumor stage distribution was similar across groups. Disease progression during P+NCT was significantly higher in the metaplastic group (33%) compared to NST (3%) and lobular tumors (0%) (p = 0.001). However, all but one patient with progression in the metaplastic group underwent surgery. pCR and EFS rates are presented in the Table. pCR rates were markedly lower in metaplastic tumors compared to NST and lobular subtypes. With a median follow-up of 22 months, 83 events of disease recurrence or death were recorded. Lobular and metaplastic tumors were associated with significantly lower 2-year EFS compared to NST. These differences remained significant in a multivariable Cox regression including tumor grade and stage. ConclusionPatients with SHT of TNBC had worse outcomes with neoadjuvant pembrolizumab plus chemotherapy compared to those with NST tumors. Metaplastic carcinoma was associated with lower pCR and a trend tower lower EFS rates. Although lobular carcinoma had pCR rates similar to NST, an unfavorable EFS was also observed in this group. These results underscore the urgent need for developing new tailored therapeutic strategies for these rare and aggressive subtypes of TNBC.

B. Kuntal¹, X. Stern¹, T. Semba², K. Nakaoka¹, M. Fujimoto¹, M. William¹, K. Leonard¹, D. Rampa¹, J. Lee¹, N. Ueno³, X. Xie¹; ¹Cancer Biology Program, University of Hawaii Cancer Center, Honolulu, HI, ²Division of Carcinogenesis, The Cancer Institute, Japanese Foundation for Cancer Research, Tokyo, JAPAN, ³Cancer Biology Program and Translational and Clinical Research, University of Hawaii Cancer Center, Honolulu, HI.

Background: Triple-negative breast cancer (TNBC) is characterized by an immunologicallycold tumor microenvironment (TME), which facilitates immune evasion and tumorprogression. Our recent study showed that c-Jun N-terminal kinase (JNK) signaling promotesan immunosuppressive TME via macrophage-derived CCL2, thereby driving TNBC growthand metastasis. JNK has three isoforms, among which JNK1 and JNK2 have been implicatedin the development of various cancers. However, the distinct roles of JNK1 and JNK2 inregulating the TME immune landscape and aggressiveness in TNBC remain poorly understood.Building on our previous findings, in this study, we tested our hypothesis that both JNK1 andJNK2 promote TNBC aggressiveness by fostering an M2 macrophage–drivenimmunosuppressive TME.Methods: The effect of JNK singling inhibition by the pan-JNK inhibitor JNK-IN-8 and theeffect of JNK1 or JNK2 knockout (KO) on the growth of murine PyMT-M TNBC cells wereassessed in vitro using CellTiter-Blue cell viability and clonogenic assays. The roles ofintratumoral JNK1 and JNK2 in regulating the TME immune landscape and tumor growth wereexamined using an immunocompetent syngeneic C57BL/6 mouse model bearing PyMT-Mtumors derived from JNK1 or JNK2 KO cells. The contributions of JNK1 and JNK2 in theTME was examined using jnk1^-/- and jnk2^-/- immunocompetent syngeneic C57BL/6 mousemodels bearing PyMT-M tumors. The effects of the loss of JNK1 and JNK2 in tumors or in theTME on immune cell tumor infiltration were analyzed by flow cytometry. Key cytokinesinvolved in JNK-mediated modulation of the TME immune landscape were identified bycytokine array analysis.Results: Inhibition of JNK signaling by JNK-IN-8 and KO of either JNK1 or JNK2significantly suppressed the proliferation of murine PyMT-M TNBC cells in vitro, suggestingthat both JNK1 and JNK2 are essential for TNBC cell growth. However, compared to the Cas9control cells, intratumoral KO of either JNK1 or JNK2 in PyMT-M tumor cells did not affecttumor initiation and growth or immune cell tumor infiltration in the wild-typeimmunocompetent syngeneic C57BL/6 mice, demonstrating that intratumoral JNK1 and JNK2are not required for PyMT-M tumor initiation and progression. In contrast, regardless of theabsence or presence of JNK1 and JNK2 in tumor cells, tumor initiation was markedly delayedand tumor growth was significantly reduced in jnk1^-/- and jnk2^-/- immunocompetent syngeneicC57BL/6 mice, with a more pronounced effect observed in jnk1^-/- mice. This result highlightsthat both JNK1 and JNK2 in the TME play a critical role in PyMT-M tumor initiation andprogression. Interestingly, compared to tumors from wild-type mice, those from both jnk1^-/-and jnk2^-/- mice exhibited a marked increase in M1 macrophages and a reduction in M2macrophages, with a greater effect observed in jnk1^-/- mice. This result indicates that JNKdeficiency in the TME induces the reprogramming of M2 macrophages toward an M1phenotype. Indeed, cytokine array analysis revealed elevated expression of CCL12, CCL5, andCXCL9 in JNK1 KO tumors from jnk1^-/- mice, as well as increased CXCL9 expression in JNK2KO tumors from jnk2^-/- mice. These cytokines play critical roles in macrophage recruitmentand polarization. Taken together, these findings support our hypothesis that both JNK1 andJNK2 promote TNBC aggressiveness via fostering an M2 macrophage–drivenimmunosuppressive TME.Conclusion: Our findings demonstrate that both JNK1 and JNK2 in the TME drive TNBCinitiation and progression via promoting an M2 macrophage–driven immunosuppressive TMEin the macrophage-enriched TNBC mouse model. These findings establish JNK1 and JNK2 askey regulators of the immunosuppressive TME in TNBC, highlighting their therapeuticimportance in treating TNBC.

F. Braso-Maristany¹, E. Ciruelos², E. Seguí³, J. Montes², P. Tolosa², M. Rey¹, F. Pardo⁴, R. Sánchez-Bayona², L. Lema², M. Cobos², P. Blasco¹, P. Galván¹, L. Parilla², R. Moreno², G. Riu⁵, I. Garcia-Fructuoso³, R. Gómez-Bravo³, M. Bergamino³, F. Schettini³, A. Madariaga², B. Adamo³, M. Muñoz³, M. Marín-Aguilera⁴, L. Manso², L. Paré⁴, T. Pascual³, M. Vidal³, A. Prat³, M. Alva²; ¹Translational Genomics and targeted therapies in solid tumors lab, Fundació de Recerca Clínic Barcelona, Barcelona, SPAIN, ²Oncology, Hospital 12 de Octubre, Madrid, SPAIN, ³Oncology, Hospital Clínic Barcelona, Barcelona, SPAIN, ⁴Reveal Genomics, Reveal Genomics, Barcelona, SPAIN, ⁵Pharmacy, Hospital Clínic Barcelona, Barcelona, SPAIN.

Background: TNBCDX is a 15-gene composite biomarker (10-gene immune module, 4-gene proliferation signature, ERBB2 and clinical variables) for predicting pathological complete response (pCR) and recurrence risk in early-stage triple-negative breast cancer (TNBC). In Martin et al. (Ann Oncol 2024), TNBCDX was retrospectively validated in 527 patients (pts) from the WSG-ADAPT-TN, MMJ-CAR-2014-01 and NeoPACT cohorts—each receiving neoadjuvant taxane-based therapy without anthracycline/cyclophosphamide (AC)—where it independently stratified pCR rates, distant recurrence-free survival (DRFS) and overall survival (OS) after adjustment for pCR status and tumor-infiltrating lymphocytes (TILs). We now assess TNBCDX in 164 consecutive pts from two Spanish academic centers, 127 treated with sequential AC→taxane-carboplatin± pembrolizumab. Methods: Baseline pre-treatment formalin-fixed, paraffin-embedded tumor samples were collected retrospectively from pts with stage I-III TNBC treated at Clinic Barcelona Comprehensive Cancer Center (n=108) and Hospital 12 de Octubre, Madrid (n=56). Keynote-522 backbone regimen (AC-taxane-carboplatin) with and without pembrolizumab was used in 47 pts and 80 pts, respectively. Gene expression profiling was performed on the TNBCDX standardized platform. TNBCDX pCR and risk scores were calculated centrally, blinded to outcomes. Associations with DRFS, OS, and pCR were evaluated using Kaplan-Meier analyses and multivariable Cox and logistic regression models adjusted for age, stage, and TILs. Results: Median age was 56.2 years (range 27.3-84.9). Ductal morphology predominated (88.4%), followed by metaplastic (3.7%), apocrine (1.8%), and lobular (1.2%). Most tumors were high grade (68%), cT2 (59%), and cN0 (56%). Median TILs was 15% (range 0-90%). Nearly all pts (97.6%) received neoadjuvant taxane-carboplatin regimens—79.3% with AC and 36.6% with immunotherapy—achieving a 53% overall pCR rate. At a median follow-up of 44.5 months, there were 16 DRFS events and 15 deaths. Among the 127 pts treated with Keynote-522 backbone regimen, those receiving pembrolizumab (n=47) had a pCR rate of 53.7% versus 60.0% in those who did not (n=80) (p=0.562). The distribution of the TNBCDX pCR score was 32.9% low, 28.7% medium, and 38.4% high. High versus low TNBCDX pCR score was associated with higher pCR rates (65.5% vs 38.8%; odds ratio=6.51; 95% CI 1.45-3.41; p=0.019). The TNBCDX risk score classified 45.7% as low-risk and 54.3% as high-risk. High versus low TNBCDX risk score was independently associated with shorter DRFS (hazard ratio [HR]=5.12; 95% CI 1.58-16.50; p=0.006) and shorter OS (HR=5.56; 95% CI 1.51-20.5; p=0.010). Higher TILs predicted pCR (OR=1.03; 95% CI 1.01-1.05; p=0.005) but were not significantly associated with DRFS or OS (both p>0.05). Conclusions: In this independent cohort, TNBCDX pCR and risk scores were significantly associated with pCR and differential survival outcomes—independently of pCR status and systemic therapy. These data underscore the value of TNBCDX for tailoring systemic therapy intensity in early-stage TNBC.

R. Jaber Chehayeb¹, P. Coutifaris², F. White¹, J. Heintz³, S. L. Jacob², J. Q. Zhang⁴, O. M. Fayanju⁴, E. S. Lebrow², N. Taunk⁵, K. Lee², H. Knollman², I. Makhlin², A. Clark², R. Jankowitz², A. Nayak⁶, S. Domchek², A. DeMichele², P. D. Shah²; ¹Perelman School of Medicine at the University of Pennsylvania, Hospital of the University of Pennsylvania, Philadelphia, PA, ²Division of Hematology Oncology, Perelman Center for Advanced Medicine, University of Pennsylvania, Philadelphia, PA, ³Biostatistics Analysis Center, Biostatistics Analysis Center, Center for Clinical Epidemiology and Biostatistics, Perelman School of Medicine University of Pennsylvania, Philadelphia, PA, ⁴Department of Surgery, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, ⁵Department of Radiation Oncology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, ⁶Department of Pathology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA.

Background: The KEYNOTE522 (KN522) regimen of neoadjuvant chemotherapy with pembrolizumab yields statistically significant and clinically meaningful improvements in pathologic complete response (pCR) rate, event-free and overall survival compared to chemotherapy alone in patients with early-stage triple-negative breast cancer (eTNBC). However, one-third of patients (pts) have residual disease, and of these, many develop distant recurrence that contributes to mortality; understanding characteristics of these high-risk pts is a critical unmet need. This study aimed to understand factors associated with (1) achieving pCR with KN522 and (2) in the absence of pCR, developing distant recurrence. Methods: We conducted a retrospective analysis of pts with eTNBC treated at the University of Pennsylvania with the KN522 regimen between September 2021 and December 2023. Multivariate logistic regression analysis was performed to identify variables associated with pCR using the generalized linear model function in R4.5.0. Given sample size restrictions, Fisher’s exact test was used to evaluate variables associated with recurrence. For these analyses, exact odds ratios and 95% confidence intervals are reported. Results: 225 pts with eTNBC initiated KN522 during the study period. 222 had surgery; 120 (54%) had pCR (no invasive cancer) and 102 (46%) had residual disease. At a median follow-up of 28 months, 26 pts (12%) including 2 with pCR developed distant recurrence. Factors significantly associated with achieving pCR were tumor grade 3 (vs 2; OR 3.42, 95% CI 1.31-9.84, p=0.016) and development of an immune-related adverse event (IRAE vs no IRAE, OR 2.26, 95% CI 1.05-5.07, p=0.041). Factors negatively associated with pCR included clinical stage III (vs I/II, OR 0.50, 95% CI 0.26-0.92, p=0.028) and older age at diagnosis (OR 0.97, 95% CI 0.94-0.99, p=0.047). Among patients who did not achieve pCR, factors associated with distant recurrence were pathologic (p)N stage (N1+ vs N0, OR 19.83, 95% CI 5.58 – 91.85, p<0.0001), pT stage (T2+ vs T0/T1, OR 4.50, 95% CI 1.54-13.53, p=0.003), and clinical stage III (vs I/II, OR 4.25, 95% CI 1.42-14.58, p=0.005). Conclusions: This real-world analysis reinforces established factors associated with pCR including tumor grade and clinical stage. Moreover, these results add to growing evidence across solid tumors that the development of IRAEs correlates with improved response. Among patients with residual disease after KN522, those with node positivity, pT2 or larger tumors, and stage III disease may represent target populations for studies investigating enhanced surveillance and treatment strategies, such as with circulating tumor DNA and antibody-drug conjugates.

B. Flood¹, C. Nawrocki², L. Neiman³, K. Treuner⁴, Y. Zhang⁴, D. Sgroi⁵; ¹Pathology, Massachusetts General Brigham, Boston, MA, ²Krantz Family Cancer Center, Massachusetts General Hospital, Boston, MA, ³Krantz Family for Cancer Research, Massachusetts General Hospital, Boston, MA, ⁴Onocology, Biotheranostics Inc., a Hologic Company, San Diego, CA, ⁵Pathology, Massachusetts General Hospital, Boston, MA.

Background: HOXB13 gene expression has been extensively studies in hormone receptor positive (HR+) post-menopausal breast cancer patients. Recently, we demonstrated that HOXB13 is expressed in a subset of triple negative breast cancer (TNBC) patients and that ectopic expression of HOXB13 in a mouse mammary model of TNBC confers a tumoral growth advantage by impeding antitumor T-cell immunity. Here, we performed comparative spatial transcriptomic profiling of the immune microenvironment of HOXB13+ and HOXB13– human TNBCs. Methods: Using the Nanostring CosMx platform, we performed 1000-plex spatial transcriptomic profiling of 15 treatment-naïve human HOXB13+ TNBCs and 8 treatment-naïve HOXB13– TNBCs. Cells were segmented with Cellpose, then filtered based on RNA counts. Sparse areas of tissue were filtered entirely to retain structured tissue. Expression was normalized, log-transformed, and scaled for dimensionality reduction with PCA and UMAP. Cell-typing was performed using an immune-oncology reference profile. Cell types were validated via marker genes identified by differential expression. Differential expression analysis was conducted with generalized linear mixed modeling. Analytical neighborhood enrichment analysis for each cell type was conducted using a spatial neighborhood graph constructed by Delaunay triangulation.Results: Using spatial transcriptomics, we find two immune “neighborhoods” within TNBC patients denoted by an association of multiple cell types that preferentially are found adjacent to one another. The first “cytotoxic” neighborhood is associated with CD8+ T cells, antigen presenting cells, and CD4+ T cells all of which preferentially are found near one another but are not found in proximity to other immune cell types. The second “plasmablast-rich” neighborhood is comprised mostly of plasmablasts with B cells and plasma cells also present. HOXB13+ patients had significantly more plasmablasts (p= 0.017) than HOXB13– patients and these plasmablasts were typically found associated with B cells or plasma cells in a plasmablast-rich neighborhood and not associated with CD8+ T cells or any of the other cytotoxic neighborhood members.Conclusions: A plasmablast-rich immune neighborhood is more frequently seen in HOXB13+ rather than HOXB13– TNBCs, and such a neighborhood is not associated with CD8+ T cells. We have recently shown that HOXB13 expression in a preclinical model of TNBC impedes anti-tumor CD8 T-cell immunity. Our findings raise the possibility that plasmablasts may directly suppress CD8+ T cells or that the formation of these plasmablast-rich neighborhoods occupies a niche that comes at the cost of anti-tumor/cytotoxic neighborhoods. Understanding the molecular underpinnings of this process could provide for novel immunotherapeutic strategies.

D. Vazquez-Juarez¹, A. Aranda-Gutierrez², Y. Chavarri-Guerra², F. Petracci³, O. Peña-Curiel¹, F. Acevedo⁴, E. Zamudio Lozoya⁵, L. Gonzalez Gonzalez⁵, W. Mantilla⁶, S. Franco⁷, C. Arce Salinas⁸, K. Centelles López⁸, H. Gomez⁹, P. Carreon¹⁰, E. Aguirre Alvarez¹¹, A. Lopez-Galindo¹², B. Martinez-Cannon², M. Garcia Garces¹³, E. Korbenfeld¹⁴, A. Benitez-Cruz¹⁴, E. Willars¹⁵, M. Lema¹⁶, C. Lema¹⁷, C. Villarreal-Garza¹; ¹Breast Cancer Center, Hospital Zambrano Hellion TecSalud, Tecnologico de Monterrey, San Pedro Garza Garcia, MEXICO, ²Departamento de Hemato-Oncología, Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán, Mexico City, MEXICO, ³Departamento de Oncología Clínica, Instituto Alexander Fleming, Buenos Aires, ARGENTINA, ⁴Hematology and Oncology, Pontificia Universidad Católica de Chile, Santiago, CHILE, ⁵Unidad de Cancer de Mama, Centro Estatal de Cancerologia de Chihuahua, Chihuahua, MEXICO, ⁶Breast Cancer, Luis Carlos Sarmiento Angulo Center for Research and Cancer Treatment, Bogota, COLOMBIA, ⁷Breast Cancer, Hospital Zambrano Hellion TecSalud, Tecnologico de Monterrey, Bogota, COLOMBIA, ⁸Unidad funcional de Cancer de mama, Instituto Nacional de Cancerología, Tlalpan, MEXICO, ⁹Oncology, Oncosalud-AUNA, Lima, PERU, ¹⁰Oncology, Oncosalud-AUNA, Lima, MEXICO, ¹¹Breast Cancer Unit, SOHEC – Sociedad de Oncología y Hematología del Cesar, Valledupar, COLOMBIA, ¹²Oncology, Doctors Hospital-AUNA, Monterrey, MEXICO, ¹³Oncology, ISSEMYM Hospital Regional de Toluca, Toluca, MEXICO, ¹⁴Oncology, Hospital Británico Buenos Aires, Buenos Aires, ARGENTINA, ¹⁵Oncology, Centro Oncológico Lic. Salvador Chavarría Delgado, Saltillo, MEXICO, ¹⁶Oncology, Clínica de Oncología Astorga, Medellin, COLOMBIA, ¹⁷Oncology, Clínica de Oncología Astorga, San Pedro Garza Garcia, COLOMBIA.

Tumor-infiltrating lymphocytes (TILs) are an established predictive and prognostic biomarker in early triple-negative breast cancer (TNBC), associated with higher pathologic complete response (pCR) rates and favorable outcomes following chemoimmunotherapy. However, in real-world settings, access limitations, immune-related adverse events, and logistical challenges can result in early discontinuation of neoadjuvant immunotherapy. Previously, our collaborative group reported that an incomplete number of neoadjuvant pembrolizumab cycles was associated with a significantly lower pCR rate. Whether high baseline TIL levels can mitigate the impact of incomplete pembrolizumab exposure on pCR remains unknown. Patients with early TNBC who received neoadjuvant chemotherapy and pembrolizumab within the PETRHA retrospective cohort were included. Stromal TILs were assessed on pre-treatment core biopsies by local pathology teams and analyzed using two predefined cut-offs: >20% and >30%. Patients were stratified by pembrolizumab treatment completion: those who received all planned neoadjuvant cycles versus those who did not. The primary outcome was pCR rate, defined as ypT0/is ypN0. Descriptive statistics were used to summarize clinicopathological features. Associations between TILs levels, neoadjuvant pembrolizumab completion, and pCR rate were evaluated using chi-square or Fisher’s exact tests, with logistic regression used to explore associations and estimate odds ratios (OR) with 95% confidence intervals (CI). Of the 304 patients included in the PETRHA cohort, TILs data were available for 111 patients (36.5%). Among these, 59 patients (53.2%) had TILs <20% and 52 (46.8%) had TILs ≥20%, while 70 (63.1%) had TILs <30% and 41 (36.9%) had TILs >30%. Thirty-eight (34.2%) received an incomplete number of neoadjuvant pembrolizumab cycles, while 73 (65.8%) completed the regimen. Among patients who did not complete the neoadjuvant pembrolizumab cycles, those with higher TILs, whether defined as ≥20% or >30%, had significantly higher pCR rates (76.9%) compared to those with lower TILs (32.0%) (p=0.016 for both cutoffs). Similarly, among patients who completed all neoadjuvant cycles, higher TILs were again associated with increased pCR rates (>20% cut-off: 71.8% vs. 38.2%, p=0.004; and >30% cut-off: 85.7% vs. 37.8%, p<0.0001). In a multivariable logistic regression model including variables with a p-value <0.10 in univariate analysis, only TILs >30% (OR 12.74, 95% CI 2.68-60.49; p=0.001) and completion of neoadjuvant chemotherapy (OR 4.95, 95% CI 1.04-23.69; p=0.045) were independently associated with higher probability of achieving pCR. In contrast, TILs >20%, completion of neoadjuvant anthracyclines, and completion of neoadjuvant pembrolizumab were not significantly associated with pCR (p=0.633, p=0.318, and p=0.866, respectively). In our study, higher baseline TILs levels, particularly >30%, were associated with increased pCR rates, irrespective of neoadjuvant pembrolizumab completion. Notably, patients with TILs >30% who did not complete all pembrolizumab cycles had higher pCR rates than those with TILs <30% who completed the full neoadjuvant regimen (76.9% vs. 37.8%), suggesting that robust immune infiltration may partially mitigate reduced immunotherapy exposure. However, the highest pCR rate was observed in patients with both high TILs and complete neoadjuvant pembrolizumab treatment (85.7%), supporting a potential synergistic effect between host immune activation and full immunotherapy administration.

R. C. Bonadio¹, M. C. Tavares², F. C. Balint², G. Ferreira³, C. dos Anjos⁴, D. Gagliato⁵, M. L. de Brito⁶, D. Assad-Suzuki⁷, D. D. Rosa⁸, N. J. Gomes⁹, N. C. Nunes¹⁰, L. Testa¹, M. Rigesti³, V. Baro³, I. M. de Sousa², M. O. Andrade¹¹, M. Gouveia¹², F. Madasi¹³, J. Bines¹³, R. P. Ferreira⁸, C. L. Santos¹⁴, M. Tavares¹⁵, M. Monteiro¹⁶, Z. S. de Souza¹⁷, A. U. Gomes⁵, B. M. Zucchetti¹², A. Ferrari¹⁸, M. F. Monteiro¹⁹, P. A. Signorini²⁰, A. Aguilar³, S. Sanches², P. G. Hoff¹, M. Estevez-Diz²¹, R. Barroso-Sousa¹¹; ¹Oncology, Instituto D’Or de Pesquisa e Ensino (IDOR), São Paulo, BRAZIL, ²Oncology, A.C.Camargo Cancer Center, São Paulo, BRAZIL, ³Oncology, SUMA (Grupo Cooperativa Argentino para el estudio y la investigación del Cáncer de Mama), Buenos Aires, ARGENTINA, ⁴Oncology, Hospital Sírio-Libanês, São Paulo, BRAZIL, ⁵Oncology, Hospital Beneficência Portuguesa, São Paulo, BRAZIL, ⁶Oncology, Clínica AMO, Salvador, BRAZIL, ⁷Oncology, Hospital Sírio-Libanês, Brasilia, BRAZIL, ⁸Oncology, Hospital Moinhos de Vento, Porto Alegre, BRAZIL, ⁹Oncology, Hospital São Domingos, DASA Oncologia, São Luiz, BRAZIL, ¹⁰Oncology, Grupo Américas, Rio de Janeiro, BRAZIL, ¹¹Oncology, Brasilia Hospital, Oncologia Américas, Brasília, BRAZIL, ¹²Oncology, Hospital 9 de Julho, Grupo Américas, São Paulo, BRAZIL, ¹³Oncology, Instituto D’Or de Pesquisa e Ensino (IDOR), Rio de Janeiro, BRAZIL, ¹⁴Oncology, Instituto D’Or de Pesquisa e Ensino (IDOR), Recife, BRAZIL, ¹⁵Oncology, Instituto D’Or de Pesquisa e Ensino (IDOR), Salvador, BRAZIL, ¹⁶Oncology, Grupo Américas, Hospital Samaritano, São Paulo, BRAZIL, ¹⁷Oncology, Hospital Sírio-Libanês, Brasília, BRAZIL, ¹⁸Oncology, Hospital Santa Paula, Grupo Américas, São Paulo, BRAZIL, ¹⁹Oncology, Instituto do Câncer do Ceará, Fortaleza, BRAZIL, ²⁰Oncology, Centro Integrado de Pesquisa da Amazônia (CINPAM), Manaus, BRAZIL, ²¹Oncology, Instituto do Câncer do Estado de São Paulo, Faculdade de Medicina da Universidade de São Paulo, São Paulo, BRAZIL.

Background: Tumor-infiltrating lymphocytes (TILs) are established prognostic and predictive biomarkers in early-stage triple-negative breast cancer (eTNBC). The addition of pembrolizumab to neoadjuvant chemotherapy (NAC) has demonstrated improved outcomes in this setting (KEYNOTE-522 trial). The role of TIL levels in this context needs to be further explored. Methods: The Neo-Real / GBECAM-0123 study is a retrospective multicenter cohort including patients (pts) with stage II-III TNBC treated with pembrolizumab plus NAC across institutions in Brazil and Argentina since 2020. Baseline stromal TILs were assessed on pre-treatment biopsies according to international guidelines and categorized as <30%, 30-49%, or ≥50%. Associations with pCR and event-free survival (EFS) were evaluated using logistic and Cox regression models, respectively. Results: A total of 248 pts were included (median age: 44 years; 70.8% with stage II disease). TILs <30%, 30-50%, and ≥50% were observed in 72.6%, 12.9%, and 14.5% of pts, respectively. In multivariable logistic regression including TILs, Ki-67 index, clinical stage, tumor grade, and number of neoadjuvant pembrolizumab cycles, both TILs ≥50% (OR 6.96, 95% CI 1.88-25.65, P=0.004) and Ki-67 ≥50% (OR 4.88, 95% CI 2.31-10.32, P<0.001) were strongly associated with pCR. Nearly all pts with both high TILs and high Ki-67 achieved pCR (Table).With a median follow-up of 24 months, pts with pCR had significantly higher 2-year EFS compared to those with residual disease (95.9% vs. 76.5%, P<0.001). In the multivariable Cox model, pCR and clinical stage remained independent predictors of EFS, whereas TILs were not. Pts who achieved pCR had an 80% lower risk of recurrence or death compared to those with residual disease (HR 0.20, 95% CI 0.07-0.54, P=0.002). Conversely, pts with stage III disease had significantly worse EFS compared to those with stage II (HR 3.68, 95% CI 1.58-8.53, P=0.002).Among pts with pCR, 2-year EFS was 97.0% in TILs < 30%, 95.2% in TILs 30-50% (HR 1.26, 95% CI 0.12-12.42, P=0.839), and 93.2% in TILs ≥50% (HR 1.74, 95% CI 0.28-10.53, P=0.544). Among those with residual disease, 2-year EFS was 73.3% in TILs < 30%, 90% in TILs 30-50% (HR 0.33, 95% CI 0.04-2.51, P=0.288), and 100% in TILs ≥50% (HR 1.39^e-15, 95% CI NA, P=1.000; analysis limited by small number of pts in this group of TILs ≥50% with residual disease). Conclusion: In this real-world cohort of pts with eTNBC treated with the KEYNOTE-522 regimen, high baseline TILs and Ki-67 index were strongly associated with higher pCR rates. These findings support the role of TILs as a relevant biomarker for treatment response. Long-term outcomes were primarily driven by pathologic response and disease stage, underscoring the importance of achieving pCR.

M. C. Tavares¹, F. C. Balint¹, R. B. Barroso-Sousa², S. Sanches¹, L. Testa³, G. Colucci⁴, N. C. Nunes⁵, B. G. Giner⁴, M. Gill⁴, I. S. Martins¹, M. Andrade⁶, M. Gouveia⁷, F. Madasi⁸, J. Bines⁸, R. Ferreira⁹, D. Rosa⁹, C. Santos¹⁰, M. Monteiro¹¹, Z. S. Souza¹², D. Assad-Suzuki¹², C. Anjos¹³, D. M. Gagliato¹⁴, A. Gomes¹⁴, B. Zucchetti¹⁵, A. Ferrari¹⁰, M. Brito¹⁶, M. F. Monteiro¹⁷, P. A. Signorini¹⁸, N. J. Gomes¹⁹, S. Mazzotta⁴, P. M. Hoff²⁰, M. Estevez-Diz²¹, R. R. Bonadio²²; ¹Oncology, AC Camargo, SAO PAULO, BRAZIL, ²Oncology, Brasilia Hospital, Oncologia Américas, Brasília, Brazil., SAO PAULO, BRAZIL, ³Oncology, Instituto do Câncer do Estado de São Paulo, Faculdade de Medicina da Universidade de São Paulo, São Paulo, Brazil, SAO PAULO, BRAZIL, ⁴Oncology, SUMA (Grupo Cooperativa Argentino para el estudio y la investigación del Cáncer de Mama), Buenos Aires, ARGENTINA, ⁵Oncology, Brasilia Hospital, Oncologia Américas, Brasília, Brazil., BRASILIA, BRAZIL, ⁶Oncology, Brasilia Hospital, Oncologia Américas, Brasília, Brazil, BRASILIA, BRAZIL, ⁷Oncology, Grupo Américas, Hospital 9 de Julho, São Paulo, Brazil., SAO PAULO, BRAZIL, ⁸Oncology, Instituto D’Or de Pesquisa e Ensino (IDOR), Rio de Janeiro, Brazil, RIO DE JANEIRO, BRAZIL, ⁹Oncology, Serviço de Oncologia, Hospital Moinhos de Vento, Porto Alegre, Brazil, PORTO ALEGRE, BRAZIL, ¹⁰Oncology, Grupo Américas, Hospital Santa Paula, São Paulo, Brazil., SAO PAULO, BRAZIL, ¹¹Oncology, GRUPO AMERICAS, HOSPITAL SAMARITANO, SAO PAULO, BRAZIL, ¹²Oncology, Medical Oncology Department, Hospital Sírio-Libanês, Brasília, Brazil, BRASILIA, BRAZIL, ¹³Oncology, Medical Oncology Department, Hospital Sírio-Libanês, São Paulo, Brazil, SAO PAULO, BRAZIL, ¹⁴Oncology, Centro de Oncologia – Hospital Beneficência Portuguesa, São Paulo, Brazil, SAO PAULO, BRAZIL, ¹⁵Oncology, Grupo Américas, Hospital 9 de Julho, São Paulo, Brazil, SAO PAULO, BRAZIL, ¹⁶Oncology, DASA Oncology – Clínica AMO, Salvador, Brazil, SALVADOR, BRAZIL, ¹⁷Oncology, Instituto do Câncer do Ceará, Fortaleza, Brazil, SAO PAULO, BRAZIL, ¹⁸Oncology, Centro intergrado de pesquisa da Amazonia(CINPAM), MANAUS, BRAZIL, ¹⁹Oncology, Hospital São Domingos – DASA, SAO PAULO, BRAZIL, ²⁰Oncology, Instituto D’Or de Pesquisa e Ensino (IDOR), SAO PAULO, BRAZIL, ²¹Oncology, Instituto D’Or de Pesquisa e Ensino (IDOR),, SAO PAULO, BRAZIL, ²²Oncology, Instiituto Rede d’or pesquisa e ensino Oncologia, SAO PAULO, BRAZIL.

Background The KEYNOTE-522 study established a new standard of care for early-stage triple-negative breast cancer (TNBC), demonstrating that the addition of pembrolizumab to neoadjuvant chemotherapy, followed by adjuvant pembrolizumab, significantly improves pathological complete response (PCR) rates, event-free survival (EFS), and overall survival (OS). BRCA1/2 mutations define a distinct molecular subgroup of TNBC, characterized by impaired DNA repair and potential differential sensitivity to platinum agents and immunotherapy. Real-world data on this subgroup remain limited. Methods Neo-Real/GBECAM-0123 is a multicenter, real-world study that included patients with early-stage TNBC treated with the KEYNOTE-522 regimen in Brazil and Argentina since July 2020. Patients were categorized as BRCA mutation carriers (mBRCA) or wild-type/unknown (wt/unkBRCA). Clinical, pathological, and treatment-related outcomes were compared using descriptive and inferential statistics. Results Among 726 patients, 104 (14.3%) were mBRCA and 622 (85.7%) were wt/unkBRCA (495 [84.2%] wild-type and 93 [15.8%] unknown). Patients with mBRCA were significantly younger at diagnosis (mean age: 40.2 vs. 45.3 years; p < 0.001). In both groups, the majority of patients had stage II (73.0% and 72.3%) and stage 3 (80.2% and 74.2%) tumors. The pCR rate was significantly higher in the mBRCA group (74.0% vs. 61.4%; p = 0.018). Mastectomy was performed more frequently in mBRCA patients (88.3% vs. 34.7%; p < 0.001), and radiation therapy was also administered less frequently (64.4% vs. 87.1%; p < 0.001). With a median follow-up of 22 months, 82 patients experienced disease recurrence or death. Two-year EFS rates were 92.3% in the mBRCA group and 86.0% in the wt/unkBRCA group (HR 0.51, 95% CI: 0.23-1.11; p = 0.091). Discussion This real-world analysis provides strong evidence that patients with triple-negative breast cancer (TNBC) with mBRCA experience distinct clinical characteristics and superior outcomes when treated with the KEYNOTE-522 regimen. The 12.3% absolute increase in the pCR rate translated into favorable EFS trends. These findings are clinically relevant and may guide future strategies for risk-adapted treatment de-escalation.

A. Aranda-Gutierrez¹, D. Vazquez-Juarez², Y. Chavarri-Guerra¹, F. Petracci³, O. Peña-Curiel², F. Acevedo⁴, E. Zamudio Lozoya⁵, L. Gonzalez Gonzalez⁵, W. Mantilla⁶, S. Franco⁶, M. Bravo⁶, P. Herrera Ríos⁷, C. Arce Salinas⁸, K. Centelles López⁸, H. Gomez⁹, P. Carreon⁹, E. Aguirre Alvarez¹⁰, B. Martinez-Cannon¹, A. Lopez-Galindo¹¹, M. Garcia Garces¹², E. Willars¹³, E. Korbenfeld¹⁴, A. Benitez-Cruz¹⁴, M. Lema¹⁵, C. Lema¹⁵, C. Villarreal-Garza²; ¹Departamento de Hemato-Oncología, Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán, Mexico City, MEXICO, ²Breast Cancer Center, Hospital Zambrano Hellion TecSalud, Tecnologico de Monterrey, San Pedro Garza Garcia, MEXICO, ³Departamento de Oncología Clínica, Instituto Alexander Fleming, Buenos Aires, ARGENTINA, ⁴Hematology and Oncology, Pontificia Universidad Católica de Chile, Santiago, CHILE, ⁵Unidad de Cancer de Mama, Centro Estatal de Cancerologia de Chihuahua, Chihuahua, MEXICO, ⁶Breast Cancer, Luis Carlos Sarmiento Angulo Center for Research and Cancer Treatment, Bogota, COLOMBIA, ⁷-, UNEME-DEDICAM Servicios de Salud de Nuevo Leon, Guadalupe, Nuevo León, MEXICO, ⁸Unidad funcional de Cancer de mama, Instituto Nacional de Cancerología, Mexico City, MEXICO, ⁹Oncology, Oncosalud-AUNA, Lima, PERU, ¹⁰Breast Cancer Unit, SOHEC – Sociedad de Oncología y Hematología del Cesar, Valledupar, MEXICO, ¹¹Oncology, Doctors Hospital-AUNA, Monterrey, MEXICO, ¹²Oncology, ISSEMYM Hospital Regional de Toluca, Toluca, MEXICO, ¹³Oncology, Centro Oncológico Lic. Salvador Chavarría Delgado, Saltillo, MEXICO, ¹⁴Oncology, Hospital Británico Buenos Aires, Buenos Aires, ARGENTINA, ¹⁵Oncology, Clínica de Oncología Astorga, Medellin, COLOMBIA.

Pembrolizumab combined with neoadjuvant chemotherapy has become the standard of care for most patients with early-stage triple-negative breast cancer (TNBC), typically administered at a dose of 200 mg every three weeks (q3w), as studied in the KEYNOTE-522 trial. A 400 mg every six weeks (q6w) dosing schedule was subsequently approved by regulatory agencies based on pharmacokinetic modeling and exposure-response analyses demonstrating comparable drug exposure and similar expected efficacy and safety to the standard q3w regimen. Although the q6w schedule was not evaluated in prospective neoadjuvant trials, its adoption has increased in real-world practice, particularly in settings with logistical constraints. However, comparative clinical data on the safety and efficacy of q6w versus q3w dosing in early TNBC remain limited. Patients with early TNBC who received neoadjuvant chemotherapy and pembrolizumab within the PETRHA retrospective cohort were included. Pembrolizumab was administered either 200 mg every three weeks (q3w) or 400 mg every six weeks (q6w), based on physicians’ choice and local practice. Patients receiving q3w and q6w dosing were compared in terms of treatment completion (defined as receipt of all planned neoadjuvant pembrolizumab cycles), immune-related adverse events (irAEs), and pathologic complete response (pCR) rates, defined as ypT0/is ypN0. Descriptive statistics were used to summarize baseline characteristics. Categorical variables were compared using chi-square or Fisher’s exact tests. A two-sided p-value <0.05 was considered statistically significant. A total of 304 patients were included, of whom 232 (76.3%) received pembrolizumab 200 mg q3w, and 72 (23.7%) received 400 mg q6w. Most patients had stage II (53.9%) or stage III (42.1%) disease, with a similar clinical stage distribution across dosing groups. The use of the q6w schedule was significantly more common in public healthcare settings, with only 0.6% of patients in private institutions receiving pembrolizumab q6w compared to 51.1% in public centers (p<0.001). Treatment completion rates were comparable, with 75.7% of patients in the q3w group and 76.4% in the q6w group completing all planned neoadjuvant pembrolizumab cycles (p=0.60). The pCR rate was higher in the q3w group (62.5%) compared to the q6w group (51.4%), although this difference did not reach statistical significance (p=0.093). irAEs occurred in 37.6% of patients receiving q3w and 25.0% of those receiving q6w dosing (p=0.051), with no statistically significant differences in grade 3 or higher irAEs (12.8% vs. 0%, p=0.205), In this real-world cohort of Hispano-American women with early TNBC, the use of a q6w neoadjuvant pembrolizumab dosing schedule was associated with similar treatment completion rates and a non-significantly lower pCR rate compared to the standard q3w neoadjuvant regimen. The numerically lower incidence and severity of irAEs with q6w dosing suggest a potentially favorable toxicity profile, although the difference was not statistically significant. These findings support the feasibility of q6w neoadjuvant pembrolizumab dosing in clinical practice.

A. Longobardi¹, R. Buonaiuto², C. Calderaio², A. Caltavituro¹, V. Cantile¹, G. Crimaldi¹, F. Puglisi³, L. Del Mastro⁴, C. De Angelis⁵, M. De Laurentiis², M. Giuliano¹, G. Arpino¹; ¹Oncology Unit, Department of Clinical Medicine and Surgery,, University of Naples “Federico II”, Naples, ITALY, ²Department of Breast and Thoracic Oncology, Istituto Nazionale Tumori IRCCS “Fondazione G. Pascale”, Naples, ITALY, ³Department of Medical Oncology, CRO Aviano, National Cancer Institute, IRCCS, Aviano, Italy, Aviano, ITALY, ⁴Department of Medical Oncology, IRCCS Ospedale Policlinico San Martino, Genova, Italy., Genova, ITALY, ⁵Clinical and Translational Oncology, Scuola Superiore Meridionale, Naples, ITALY.

Background: Immune checkpoint inhibitors (ICIs) in combination with chemotherapy, have markedly transformed the therapeutic alogorithm of high risk early-stage triple-negative breast cancer (eTNBC). Although the safety and efficacy of ICIs have been extensively evaluated in randomized clinical trials (RCTs), their toxicity and effectiveness profiles may differ in real-world clinical practice. This systematic review and meta-analysis aimed to compare the incidence of all-grade adverse events (AEs), grade ≥3 AEs, treatment related discontinuation, and pathological complete response (pCR) rates across phase I-III RCTs and real-world evidence (RWE) in patients with eTNBC receiving neoadjuvant ICIs. Methods: Fourteen RCTs and ten RWE studies were included in the analysis. Incidence of all-grade AEs, grade ≥3 AEs, treatment related discontinuation and pCR rate were collected. Proportions were logit-transformed to stabilize variances and pooled using a random-effects model with Hartung-Knapp adjustment. Subgroup differences between RWE and RCTs were evaluated using the Q-test for subgroup analysis. Separate analyses were conducted for pCR, overall AEs, grade ≥3 AEs and treatment discontinuation to explore consistency of findings across different outcomes. Results: The pooled incidence of all-grade AEs was higher in RCTs compared to RWE, with estimates of 93% (95% CI: 78-98%) and 80% (95% CI: 27-98%), respectively. However, the difference between subgroups did not reach statistical significance in the random-effects model (Q = 1.18, p = 0.28). Similarly, the incidence of grade ≥3 AEs was 49% in RCTs (95% CI: 27-71%) versus 30% in RWE (95% CI: 14-52%), with no significant difference observed between groups (Q = 1.85, p = 0.17). Treatment discontinuation rates were consistent across study types, with both RCTs and RWE showing a pooled incidence of 19%, with no difference between subgroups (Q = 0.00, p = 0.97). Regarding pCR, the pooled rate was 58% (95% CI: 53-62%) in RCTs and 54% (95% CI: 43-65%) in RWD, with no statistically significant subgroup difference (Q = 0.46, p = 0.50). Conclusions: The incidence and severity of adverse events and pCR rates, were comparable between randomized clinical trials and real-world evidence, indicating that immunotherapy maintains consistent safety and effectiveness profiles across both contexts. These results reinforce the external validity of trial-derived data for application in routine clinical practice. Ongoing post-marketing surveillance and standardized reporting remain crucial to ensure patient safety, optimize outcomes, and inform evidence-based treatment decisions.

T. Al-Batsh, F. Tamimi, M. Horani, B. Sharaf, H. Bani Hani, L. El Saket, A. Issa, Z. Muhanna, O. Almuhaisen, O. Mahafdah, A. Shammout, M. El-Atrash, M. Abunasser, H. Abdel-Razeq; Internal Medicine, King Hussein Cancer Center, Amman, JORDAN.

Introduction: Triple-negative breast cancer (TNBC) represents an aggressive subtype of breast cancer, often diagnosed at a younger age and associated with limited therapeutic options. Despite advances in systemic therapy, long-term outcomes remain suboptimal, particularly in resource-limited settings. This study aims to characterize the clinical features, germline mutation patterns, and survival outcomes of TNBC patients treated in Jordan, with a focus on real-world response to neoadjuvant chemo-immunotherapy and the prognostic value of germline genetic findings. Methods: This retrospective cohort study included patients diagnosed with TNBC between 2003 and 2023 at King Hussein Cancer Center. Demographic, clinical, pathological, and treatment-related data were extracted from medical records. All patients underwent germline genetic testing as part of their workup. Pathological complete response (pCR; defined as ypT0/is, ypN0), event-free survival (EFS), and overall survival (OS) were estimated using the Kaplan-Meier method. Multivariable logistic regression was performed to identify independent predictors of pCR. Results: A total of 526 patients with TNBC were included, predominantly Jordanian (473, 89.9%), with a median age at diagnosis of 45 years (range: 36-54). Notably, 165 patients (31.4%) were diagnosed before the age of 40, indicating a high proportion of young patients. A personal history of non-breast malignancy was reported in 35 patients (6.6%), and 11 (2.1%) had a prior diagnosis of breast cancer with a different hormone receptor profile. A positive family history of cancer was documented in 407 patients (77.4%). At presentation, 494 patients (94.0%) had non-metastatic disease, including 283 (52.7%) with locally advanced tumors. Histopathological analysis revealed invasive breast carcinoma of no special type (NST) in 434 patients (82.5%), with grade 3 histology in 422 cases (80.2%). HER2-low status was identified in 95 patients (18.1%). Germline genetic testing revealed pathogenic or likely pathogenic (P/LP) variants in 117 patients (22.2%), most commonly BRCA1 (77 patients, 65.8%) and BRCA2 (31 patients, 26.5%). Among the 494 non-metastatic patients, 337 (68.2%) received neoadjuvant therapy, including 74 (21.9%) who received chemo-immunotherapy. This was associated with a significantly higher pCR rate compared to chemotherapy alone (59.5% vs. 39.9%; OR: 2.2, p = 0.003). In multivariable analysis, both neoadjuvant immunotherapy (OR: 2.18, p = 0.006) and the presence of P/LP mutations (OR: 1.95, p= 0.017) were independently associated with increased odds of achieving pCR, while age, HER2 status, tumor grade, and clinical stage were not significant predictors. Mastectomy was performed in 281 patients (56.9%), and among those with P/LP variants, 60.8% underwent risk-reducing contralateral mastectomy and 51.6% had bilateral salpingo-oophorectomy. After a median follow-up of 55.8 months, the 5-year event-free survival (EFS) among patients who presented with non-metastatic disease was 74.2% (95% CI: 70.0%-78.7%). In subgroup analysis, 5-year EFS was similar in patients with P/LP variants and those without; 77.5% (95% CI: 68.6%-87.6%) compared to 73.1% (95% CI: 68.3%-78.2%), p= 0.205. Conclusion: This study highlights the clinicopathologic diversity and outcomes of TNBC patients in a Middle Eastern population. A notable proportion carried BRCA1/2 mutations, supporting routine genetic testing. Adding immunotherapy to neoadjuvant chemotherapy improved pCR rates. Long-term outcomes were favorable, with no survival differences by mutation status.

J. Tchou¹, A. Nayak², H. N. Xu³, J. Kollmar¹, E. H. Smith¹, M. Z. Mazur¹, L. Keele¹, A. Clark⁴, N. Taunk⁵, J. Fraietta⁶, on behalf of the BreastVax Study Team; ¹Surgery, University of Pennsylvania, Philadelphia, PA, ²Pathology, University of Pennsylvania, Philadelphia, PA, ³Radiology, University of Pennsylvania, Philadelphia, PA, ⁴Medicine, University of Pennsylvania, Philadelphia, PA, ⁵Radiation Oncology, University of Pennsylvania, Philadelphia, PA, ⁶Microbiology, University of Pennsylvania, Philadelphia, PA.

Background Neoadjuvant chemotherapy combined with immune checkpoint blockade (ICB) improves outcomes in high-risk TNBC, but this regimen is not approved and may be overtreatment for cT1N0 TNBC. Preclinical studies show that low dose radiotherapy (RT) can enhance antitumor immunity. In this study, we evaluated whether a neoadjuvant chemotherapy-free regimen of a single pembrolizumab (pembro) dose + a single subablative RT fraction (7Gy) may elicit antitumor immune response and increase stromal tumor infiltrating lymphocytes (sTILs) in TNBC to high levels, defined here as > 50%, which is a biomarker known to correlate with excellent outcomes without chemotherapy. Methods In this open-label phase 1b/2 trial (NCT04454528), participants with histologically confirmed early stage invasive breast cancer planned for standard of care (SOC) upfront surgery were enrolled. Each participant received a single preop dose of pembro either alone or in combination with RT given before or after pembro. The co-primary endpoints were demonstration of 1) feasibility of completing RT + pembro regimen within 21 days without delaying planned surgery > 14 days and 2) immune activation, defined as a) increase in % sTILs, and, b) increase in % proliferating (Ki67⁺) circulating T cells. Secondary objectives included assessment of tumor response by change in size from baseline. Responders are defined as those with tumor size reduction > 30% (TΔ30). Study sample size was powered for TΔ30. Assuming 40% of treated subjects achieve TΔ30 versus 0% in historical controls, 15 subjects yield 80% power at α=0.05. Exploratory analyses included immune profiling of peripheral blood mononuclear cells (PBMC) by flow cytometry for markers of T-cell activation, proliferation, differentiation and exhaustion, T-cell receptor repertoire (TCR) dynamics, and digital spatial profiling of the tumor immune microenvironment pre and post treatment. All participants received SOC adjuvant therapy after surgery. Findings Between January 6, 2021, and February 18, 2025, 30 participants (12 with HR+HER2-, 17 TNBC and one HER2+) enrolled and completed the preop regimen. All participants proceeded to surgery without delay. Median follow-up was 29 months (IQR 14 – 45). There were no adverse events (AEs) ≥ grade 2 which included one grade 2 immune-related AE (hypothyroidism). TΔ30 rate was significantly higher in RT + pembro arms (14 of 25, 56%, including 3 with pCR) compared to pembro only arm (0 of 5, 0%), p = 0.024. Post-treatment, sTILs shifted rightward from 20% (IQR 5 – 50 ) to 50% (IQR 10 – 60) overall and from 40% (IQR 20 – 55) to 60% (IQR 50 – 70) in TNBC. The proportion of TNBC with high sTILs (>50%) increased from 3 of 11 (27%) to 7 of 13 (54%) post treatment. Immune profiling of PBMC from responders demonstrated increased proportions of early-memory CD4⁺ and CD8⁺ T cells at baseline, marked expansion of activated, Ki67⁺ CD4⁺ and CD8⁺ T cells post treatment, and a higher and more diverse TCR repertoire both pre- and post-treatment. Spatial transcriptomic analyses demonstrated that tumor regions directly interacting with T-cells in responders were enriched for gene signatures of cGAS-STING-pathway activation, antigen processing and presentation, and response to ICB. Interpretation In this small single institution study, we demonstrate that a single pembro dose + a single sub-ablative RT fraction can elicit anti-tumor immune response and increase sTILs from low to high levels in TNBC. As only ~20% TNBC have high sTILs, this well tolerated neoadjuvant chemotherapy-free regimen may be evaluated in future studies as an immune induction strategy to expand the number of patients eligible to enroll in adjuvant chemotherapy omission trials such as the OPTImaL study.

U. Pratap¹, M. Mahajan¹, K. Nassar¹, T. Adeniran¹, N. Karinel², G. Sareddy¹, N. Mukherjee¹, S. Viswanadhapalli¹, S. McHardy², A. Brenner¹, R. Vadlamudi¹; ¹OBGYN, UT Health San Antonio, San Antonio, TX, ²OBGYN, UTSA, San Antonio, TX.

Background: Breast cancer (BC) is the most common cancer in women, with ~40,000 annual deaths in the U.S. Triple-negative breast cancer (TNBC), comprising 15-24% of cases, is aggressive and linked to high mortality. TNBC lacks ER alpha (ERa), PR, and HER2 but uniquely expresses ER beta (ERb), a tumor suppressor. However, ERb-targeted therapies remain underdeveloped due to limited selective agonists and epigenetic silencing in advanced disease. We identified CIDD-0149897, a novel ERb agonist, and evaluated its activity alone and with HDAC inhibitors (HDACi) to enhance ERb expression and therapeutic efficacy in TNBC. Methods: CIDD-0149897 was evaluated in seven TNBC cell lines (SUM159, BT549, MDA-MB-468, HCC1806, HCC70, MDA-MB-231, E0771). ERb expression was measured by RT-qPCR and Western blot. Functional assays included cell proliferation, colony formation, invasion, and apoptosis following treatment with CIDD-0149897, HDACi, or combination therapy. Mechanistic studies involved RNA sequencing, reporter assays, immunohistochemistry, and pathway analysis. In vivo efficacy was assessed using TNBC xenograft and syngeneic mouse models. Results: Expression analysis of ERb across seven TNBC cell lines (SUM159, BT549, MDA-MB-468, HCC1806, HCC70, MDA-MB-231, and E0771) revealed detectable but variable levels of ERb mRNA and protein. Treatment with the selective ERb agonist CIDD-0149897 led to a dose-dependent decrease in cell viability across multiple TNBC models. Treatment with CIDD-0149897 significantly induced apoptosis. ERb reporter assays confirmed increased transcriptional activity following CIDD-0149897 exposure. Overexpression of ERb further sensitized cells to treatment, while ERb knockout attenuated response. HDAC inhibitors (HDACi) upregulated ERb mRNA and protein expression, with chromatin immunoprecipitation confirming increased acetylation at the ERb promoter. Combination treatment with CIDD-0149897 and HDACi synergistically inhibited proliferation, reduced clonogenic survival, and suppressed invasion more effectively than either agent alone. RNA-seq and RT-qPCR showed enhanced activation of ERβ target genes involved in apoptosis and cell cycle arrest. In vivo, monotherapy with CIDD-0149897 or HDACi modestly inhibited tumor growth in HCC-1806 xenograft and E0771 syngeneic models. However, combination therapy significantly suppressed tumor volume over time, prolonged survival, and reduced tumor cell proliferation (Ki-67 staining). Immunohistochemical analysis revealed upregulation of ERb and increased cleaved caspase-3 in tumors treated with the combination therapy. In the immunocompetent E0771 model, combination therapy not only inhibited tumor progression but also enhanced infiltration of immune cells, suggesting favorable modulation of the tumor immune microenvironment. Conclusion: These findings demonstrate that CIDD-0149897, in combination with HDAC inhibition, effectively reactivates ERb signaling and modulates the tumor microenvironment. This combinatorial strategy offers promising therapeutic potential for TNBC and provides critical insights that may guide the development of more effective, targeted treatments.

R. Costa, A. Soyano, A. Armaghani, N. Abdo, S. Wallace-Morrison, M. Al-Jumayli,, L. Loftus, E. Abraham, J. Whiting,, Q. Mo, Z. Jameel, T. O’Connor, K. Dvir, S. Hoover, J. Kiluk, M. Lee, C. Laronga, N. Khakpour, H. Han, H. Soliman, B. Czerniecki; Breast Oncology, H. Lee Moffitt Cancer Center, New Tampa, FL.

Title : Results of the Dose-Expansion Cohort of a Phase 1 Trial of Intratumoral HER2- and HER3-Primed Dendritic Cells Injections for the Treatment of Early-Stage TNBC and HR Low Positive Breast Cancer.DecipHER trial Authors : Ricardo L B Costa, Aixa E. Soyano, Avan Armaghani, Neveen Abdo, Shere Wallace-Morrison, Mohammed Al-Jumayli, Loretta Loftus, Edith Abraham, Junmin Whiting, Qianxing Mo, Zena Jameel, Tracey O’Connor, Kathrin Dvir, Susan Hoover, John Kiluk, Catherine Lee, Christine Laronga, Nazanin Khakpour, Hyo S. Han, Hatem H. Soliman, MD, Brian J. Czerniecki. Abstract Background : Patients (pts) with breast cancer (BC) harboring low expression of hormone receptors (HR) and human epidermal growth factor receptor-2 (HER2) have poorer outcomes compared to other subsets of BC. Results from the KEYNOTE-522 trial showed that activation of the immune system using a PD1/PD-L1-targeted approach leads to clinically meaningful improvement in the outcomes of patients with these high-risk tumors. Dendritic cells (DCs) are antigen presenting cells which are pivotal for robust cytotoxic responses via broader activation of the adaptive immune system against tumor-associated antigens. Methods : DecipHER is a dose-escalation, dose-expansion phase 1 trial designed to assess the safety and the preliminary efficacy of autologous, HER2- and HER3-primed DCs in combination with KEYNOTE 522 regimen. Pts with clinical stage cT1cN1/2 or cT2-4cN0-2, HR 20%, HER2-negative BCs are eligible. Patients with inflammatory BC and uncontrolled immune-mediated diseases are excluded. After collection through apheresis, autologous DCs are primed ex vivo against 6 HER2 and 8 HER3 immunogenic peptides. Pts receive alternating ultrasound-guided intratumoral HER2 and HER3 DCs injections administered twice a week for 8 doses starting 2 weeks prior to neoadjuvant therapy with KEYNOTE 522 regimen. The dose-escalation phase of the study had a classic 3+3 design (ie, DL1-3 [10-20, 30-50, 80-100 million], n=12). Additionally, 12 pts were treated at the maximum tolerated dose (MTD) of 80-100 million in the dose-expansion cohort. Endpoints included the absolute risk of adverse events, complete pathological response rate (PCR) and recurrence-free survival. Tumor tissue, blood and stool samples were collected for correlative analyses. Herein we report the safety results of the dose-expansion cohort ; 6 pts treated at DL3 were added for analysis of PCR. Results : A total 12 patients were enrolled into the dose-expansion cohort between 06/2024 and 09/2024. The median age in years was 53.5 (32-76) and 16.7% of pts were black; 33.3% had HER2-low, 16.7% had grade 3, 22.2% had T3 and 61.1% had N1 BCs.All patients received all planned vaccines. Grades 1 and 2 AEs that were at least possibly related to DCs (risk 10%) were fever (58.3%), chills (50%), pain (33.3%), headache (25%) fatigue (16.7%) and nausea (16.7%). No DC-associated grade ≥ 3 AEs were observed. None of the toxicities met the definition of DLT. Four pts had an SAE (i. bronchitis, ii. febrile neutropenia, iii. pulmonary embolism and iv. bilateral lower extremity weakness at weeks 13, 18, 22 and 14; respectively). One immune-related AEs was observed, hypothyroidism (1 pt). PCR was observed in 77.8% of the pts and it was observed at higher frequency among pts with T2 vs T3 tumors (92.3 vs 50%). Conclusion : Intratumoral DCs in combination with standard neoadjuvant chemotherapy and pembrolizumab showed preliminary efficacy and were well tolerated in pts with high-risk HR 20%, HER2-negative BC. This trial is ongoing to further assess the event-free survival of this novel treatment; correlative analyses will follow. The study is open at H. Lee Moffitt Cancer Center. Clinical trial information : NCT05504707 Funding : Shulas’ foundation.

R. S. Lee¹, Q. Jin², B. Binboga Kurt³, B. Koca¹, J. Gomez Tejeda Zanudo¹, A. Patel¹, A. Barkell⁴, J. Baginska¹, O. M. Cunningham¹, C. E. Stever¹, T. S. Parker¹, T. Rahman¹, M. Luo¹, I. G. Martino¹, B. M. Drummey¹, S. A. Virani¹, K. Santos¹, J. Bsat¹, C. Snow¹, N. Tung⁵, S. Lo⁶, M. G. Faggen⁷, N. Sinclair¹, N. Ahmad⁸, M. Constantinou⁹, S. Sinclair¹⁰, J. L. Meisel¹¹, S. A. Kirschner⁴, T. A. King¹², R. Salgado¹³, E. A. Mittendorf¹², N. U. Lin¹, N. Tayob², E. P. Winer¹⁴, E. C. de Bruin⁴, S. M. Tolaney¹, A. Moeini¹⁵, A. C. Garrido-Castro¹; ¹Medical Oncology, Dana-Farber Cancer Institute, Boston, MA, ²Data Science, Dana-Farber Cancer Institute, Boston, MA, ³Pathology, Mass General Brigham, Boston, MA, ⁴Translational Medicine, Oncology R&D, AstraZeneca, Cambridge, UNITED KINGDOM, ⁵Medicine, Division of Medical Oncology, Beth Israel Deaconess Medical Center, Boston, MA, ⁶Medical Oncology, Stamford Health, Stamford, CT, ⁷Medicine/Oncology, Dana-Farber Brigham Cancer Center at South Shore Health, South Weymouth, MA, ⁸Phelps Cancer Center, Berkshire Health System, Pittsfield, MA, ⁹Rhode Island Hospital Cancer Institute, Brown University Health, Providence, RI, ¹⁰Medical Oncology, Northern Light Health, Brewer, ME, ¹¹Hematology and Medical Oncology, Winship Cancer Institute at Emory University, Atlanta, GA, ¹²Surgery, Brigham and Women’s Hospital, Boston, MA, ¹³Pathology, ZAS Hospitals, Antwerp, BELGIUM, ¹⁴Medical Oncology, Yale Cancer Center, New Haven, CT, ¹⁵Translational Medicine, Oncology R&D, AstraZeneca, Barcelona, SPAIN.