621 Background: Neoadjuvant outcomes are variable in human epidermal growth factor receptor 2-positive (HER2+) early breast cancer (EBC), highlighting the unmet need for predictive biomarkers. The BluePrint (BP) assay classifies tumors by molecular intrinsic subtype, whereas the ImPrint hormone receptor-positive (HR+) and ImPrint triple-negative genomic signatures predicted immune sensitivity in the ISPY2 trial. Because HER2+ EBCs are genomically heterogeneous, and because anti-HER2 has an immunologic mechanism of action, we hypothesized that BP and ImPrint could be used to predict outcomes (pathologic complete response, pCR) following neoadjuvant chemotherapy + anti-HER2 (trastuzumab + pertuzumab, HP). Methods: We evaluated n=252 HER2+ EBCs receiving neoadjuvant chemotherapy + HP in the prospective, observational FLEX study. Tumors were classified by BP subtype (HER2, Luminal A/B, or Basal) and ImPrint (+ or -). Fisher’s exact test was used to compare pCR rates across BP/ImPrint results. Multivariable logistic regression was used to evaluate independent associations with pCR, adjusting for nodal status and tumor size. Results: 72% of HER2+ EBCs were HR+ and 28% were HR–. In HR+/HER2+ EBC, 52% were classified by BP as non-HER2 (Luminal A 6%, Luminal B 44%, Basal 2%), whereas in HR-/HER2+ EBC, 14% were classified genomically as non-HER2, all of which were Basal. In the HR+/HER2+ subgroup, pCR rates differed significantly by BP subtype and ImPrint, with the highest pCR observed in BP-HER2/ImPrint+ tumors (Table). Similar significant differences were observed in HR-/HER2+ EBC. In a multivariable model of the HR+/HER2+ subgroup, BP and ImPrint independently predicted pCR after controlling for node status and tumor size (BP-HER2 odds ratio/OR: 8.23 [95% CI: 3.47–21.44]; p<0.001; ImPrint+ OR: 4.48 [95% CI: 1.48–14.89]; p<0.05). Conclusions: An integrated genomic approach that combines BP and ImPrint maximizes prediction of pCR outcome in this dataset, particularly for the HR+/HER2+ subgroup. Further evaluation is warranted, as these data could be useful for guiding clinical decision-making, including selection of chemotherapy backbone and adjuvant therapy. Importantly, this approach may help identify patients who are overtreated or undertreated with current strategies, including those with exceptional treatment sensitivity who may be candidates for chemotherapy de-escalation. Clinical trial information: NCT03053193 . pCR rates by genomic subtypes within HR+/HER2– EBC (Fisher’s exact test). Biomarker Subtype* N pCR rate (%) p-value BluePrint HER2 87 55 < 0.001 Luminal B 80 20 ImPrint Positive 30 60 0.006 Negative 151 32 BluePrint + ImPrint HER2, ImPrint+ 14 79 < 0.001 HER2, ImPrint- 73 51 Luminal B, ImPrint+ 14 43 Luminal B, ImPrint- 66 15 *Excluded Luminal A and Basal due to small N.

Early lymph node (LN) metastasis often precedes systemic metastasis and corresponds with significantly inferior survival for patients diagnosed with early-stage breast cancer (EBC). To understand the biological pathways involved in early LN metastasis, differential gene expression (DGE) analysis compared large tumors without evidence of LN metastasis (pT2-3pN0) to small tumors with LN metastasis (pT1pN+). This study included 2,349 patients with EBC who underwent MammaPrint and BluePrint testing as part of the FLEX (NCT03053193). DGE was performed between pT2-3pN0/pT1pN + and across their MP/BP subtypes. Immune deconvolution was assessed using gene-signature-based methods, complemented by conventional tumor-infiltrating lymphocyte (TIL) analyses on a representative subset of patients. Greater DGE was observed within the MammaPrint High Risk and BluePrint Luminal B subgroups compared to pathological stages. MammaPrint High Risk tumors saw 73 differentially expressed genes (DEGs), while 34 were found for Luminal B tumors. Gene set enrichment analysis (GSEA) of MammaPrint High Risk/Luminal B tumors showed upregulated proliferation pathways and downregulated epithelial-to-mesenchymal transition (EMT) and immune profiles in pT2-3pN0 vs. pT1pN+, respectively. Immune deconvolution analyses showed a higher abundance of T gamma delta cells and CD4 + Th1 cells and a lower abundance of T regulatory cells, M2 macrophages, and cancer-associated fibroblasts within pT2-3pN0 tumors. Conventional histological assessment revealed no significant differences in TILs. This study lays the groundwork for exploring mechanisms of LN metastasis in EBC and their relation to MammaPrint High Risk and Luminal B subtypes. These data support previous studies’ association of LN metastasis with EMT and immune dysregulation.

We compared clinicopathologic features, MammaPrint and BluePrint molecular subtype, and outcomes by race among females with hormone receptor-positive (HR+), HER2- early-stage breast cancer (EBC). Of 1018 participants with HR+ HER2- EBC enrolled from two registries (BEST and FLEX), 509 White females were propensity score matched 1:1 to 509 Black females based on age and/or menopausal status. MammaPrint classified tumors as High-Risk or Low-Risk; and together with BluePrint, classified tumors as Luminal A-Type, Luminal B-Type, or Basal-Type. Recurrence-free survival (RFS) was analyzed by race and molecular subtype. Cox proportional hazards models assessed association of clinicopathologic features with outcomes. Basal-Type tumors were more prevalent among Black vs White participants (11.0% vs 4.8%, p < 0.001). Independent of race, participants with Basal-Type tumors had lower 3-year RFS (83.7%) compared to Luminal B-Type (93.7%) and Luminal A-Type (96.5%, p < 0.0001). Multivariate analysis revealed that participants with High-Risk, Luminal B- and Basal-Type tumors had significantly worse 3-year outcomes compared to Low-Risk Luminal A-Type, after controlling for race and potential confounders. Genomic classification identified higher proportions of High-Risk HR+ HER2- EBC among Black participants. Molecular subtype was independently prognostic of 3-year survival, supporting the prognostic and potentially predictive importance of genomic testing to reduce racial survival disparities among Black females with EBC.

561 Background: Homologous recombination deficiency (HRD) is a biomarker for impaired DNA repair proficiency and genomic instability, associated with sensitivity to DNA-damaging agents. MammaPrint (MP), a 70-gene assay, stratifies early breast cancer (EBC) by UltraLow, Low, High Risk 1 (H1) or High Risk 2 (H2) of recurrence. Compared to patients (pts) with H1 tumors, pts with H2 tumors derive greater benefit from anthracycline-based chemotherapy. Here we analyze HRD signatures to assess whether MP gene expression profiling may detect pathway activity in H2 tumors indicating DNA repair deficiency. Methods: We analyzed tumors from pts enrolled in the FLEX study who consented to full transcriptome and clinical data collection, with available MP and BluePrint (BP; Luminal or Basal) results (N=1298). HRD status assessed using a 228-gene signature was further refined into a 26-gene panel (Jacobson et al., 2023). The R-package Limma was used to preprocess gene expression data. For each gene, the median expression value of probes mapping to the gene was calculated. Differences in HRD scores were compared across MP and BP groups using a t-test. Results: Pts with HR+HER2- EBC HRD scores differed significantly by both MP and BP. Using the 228-gene HRD signature, H2 tumors were associated with higher HRD compared with H1 cancers (p<0.001). Luminal H2 tumors were also associated with higher HRD scores than Luminal H1 tumors (p<0.001). Basal tumors were associated with higher HRD compared with Luminal tumors within both H1 and H2 (Luminal H1 vs Basal H1 and Luminal H2 vs Basal H2; both p<0.001), and Basal H2 tumors had higher HRD scores than Basal H1 tumors (p<0.001). Similar patterns were observed within the 26-gene HRD panel. H2 tumors were associated with higher scores than MP H1 tumors (p<0.001), and Luminal H2 tumors were associated with higher scores than Luminal H1 tumors (p<0.001). Basal tumors were associated with higher HRD than Luminal tumors within both MP risk groups (Luminal H1 vs Basal H1, p=0.002; Luminal H2 vs Basal H2, p<0.001). Basal H2 tumors demonstrated numerically higher scores than Basal H1 tumors using the 26-gene panel, but did not reach statistical significance (p=0.062). Conclusions: In this real-world analysis, MP and BP identified differences in underlying HRD biology, with H2 tumors enriched for HRD and Luminal H1 tumors demonstrating relative homologous recombination proficiency. These findings may provide a biological rationale for the anthracycline chemotherapy and immunotherapy benefit observed in pts with H2 tumors and further evidence for MP as a tool for refined therapy selection. Additional research is warranted to examine MP as a biomarker for HRD-sensitive therapies such as PARP inhibitors and carboplatin. These results further highlight biological diversity within genomically high-risk tumors, uniquely detected by MP and BP. Clinical trial information: NCT03053193 .

PURPOSE Randomized trials have not demonstrated clear benefit from anthracyclines for patients with hormone receptor-positive, human epidermal growth factor receptor 2-negative (HER2-) early breast cancer (EBC). The MammaPrint 70-gene signature, specifically the High-Risk 1 (H1) and High-Risk 2 (H2) subgroups, was evaluated in predicting benefit from anthracycline- and taxane-based chemotherapy (AC-T) compared with taxane/cyclophosphamide (TC) in hormone receptor-positive HER2- EBC.MATERIALS AND METHODS This analysis used prospectively collected real-world data from patients enrolled in the FLEX study with stage I-III hormone receptor-positive HER2- EBC, MammaPrint High-Risk, and BluePrint Luminal B tumors who received adjuvant AC-T or TC. Inverse probability of treatment weighting (IPTW) was performed to balance clinical characteristics between treatment groups. The association between the chemotherapy regimen and 3-year invasive disease-free survival (IDFS) was assessed using Kaplan-Meier estimates and Cox proportional hazards models.RESULTS Among patients with H1 cancers (n = 1,106), 3-year IDFS did not differ between AC-T and TC (95.9% v 95.9%; adjusted hazard ratio [HRs], 1.14 [95% CI, 0.64 to 2.06]; P = .70). By contrast, patients with H2 cancers (n = 153) demonstrated significantly improved 3-year IDFS with AC-T versus TC (100% v 94.8%; adjusted HRs, 0.10 [95% CI, 0.01 to 0.75]; absolute benefit 5.2%; P = .025). A significant treatment-by-MammaPrint interaction demonstrated that AC-T benefit increased with higher genomic risk (adjusted HRs, 0.09 [95% CI, 0.01 to 0.85]; P = .036). Clinical and pathologic variables did not predict anthracycline benefit.CONCLUSION Patients with MammaPrint H2 tumors had a clinically meaningful benefit from anthracycline-containing chemotherapy, whereas patients with H1 tumors did not derive more benefit from AC-T than from TC despite overlapping high-risk features. These findings support the use of MammaPrint to inform chemotherapy regimen selection and warrant further validation.

Background: Breast cancer (BC) is the most common malignancy among women worldwide. Comprehensive genomic profiling has been widely used to identify molecular biomarkers and signatures for developing novel therapeutic strategies. BC is a complex disease; besides the genomic and molecular characterization of tumors, racial/ethnic classification and body mass index (BMI) are key determinants of treatment outcomes. To explore gene-to-phenotype relationships, we obtained transcriptomic and clinical data from 978 patient samples comprising all four BluePrint® (BP) subtypes: Luminal A, Luminal B, HER-2, and Basal, representing ∼25% of each genomic subtype. We investigated the association of biomarkers of inflammation, apoptosis, oxidative stress, autophagy, and ER stress in patients with BC and correlated them with BMI, menopausal status, and ethnicity. Method: The FLEX registry (NCT03053193) enrolled 14000 patients with stage I-III BC across 90 institutions in the United States and stratified risk groups based on the 70-gene signature (MammaPrint®, (MP)) and molecular subtype based on the 80-gene signature (BP). We obtained transcriptomic (110 genes) and clinical data from 978 patient samples, representing ∼25% of each genomic subtype. Principal component analysis (PCA) was conducted using the ‘PCAtools’ package in R to reduce the gene expression data's dimensionality and identify major variation patterns among the samples. Analysis of variance (ANOVA) was performed between groups using the first five principal components (PCs). The family-wise error rate was set at 0.05, using the modified Holm-Bonferroni approach. Results and Conclusion: Of the 978 patients (Caucasians:81.7%; African Americans:2.58%; other races: 5.7%), the risk or genetic profile was as follows: MP: Ultra-Low = 76 (8%), Low = 176 (18%), High1 = 315 (32%), and High-2 = 411 (42%); BP: Luminal A 250 (26%), Luminal B = 250 (26%), HER2 = 228 (23%), and Basal = 250 (26%). The mean age was 58.84±13.04, the average BMI was 30.04±7.34, and the majority were postmenopausal (n = 717, 73%). The first five PCs accounted for 63.11% of the total variance of the dataset. These five PC loadings (variable contributes) were enriched for genes associated with various cellular mechanisms, including glucose metabolism, cellular growth, resistance to cell death, angiogenesis induction, oxidative stress handling, and epigenetic modifications. Interestingly, the pairs plot comparing PC1 through PC5 showed a certain degree of clustering and PC4 (total variance of 4.59%) indicated distinct clustering for each molecular subtype. The biplot for PC1 and PC4 showed a certain degree of separation when intrinsic molecular subtypes and risk of recurrence categories were used. SOD2, KLK5, KLK7, and IL8 showed a strong positive correlation with PC4, whereas GLI1 showed a strong negative correlation with PC4. The loading plot also showed a positive loading for SOD2, indicating a significant role of SOD2 in differentiating distinct molecular subtypes. Comparisons of PCs with clinical variables showed that PC4 was significantly correlated with all baseline clinical variables, including age (p < 0.001), BMI (p = 0.001), race (p < 0.001), menopausal status (p = 0.001), molecular subtype (p < 0.001), and risk of recurrence (p < 0.001). Additionally, PC4 was also found to be significantly correlated with tumor size (p < 0.001), lymph node status (p < 0.001), and the presence of distant metastasis (p < 0.001). In conclusion, our findings demonstrate that genes in PC4 play a crucial role in discriminating known molecular profiles when different risk categories, as it is significantly associated with baseline characteristics, and risk of recurrence. Further analysis should be conducted to explore genes that contribute to PC4, such as SOD2. These results highlight the potential of SOD2 as a prognostic biomarker and a therapeutic target in breast cancer. Citation Format: Sharda P. Singh, Chathurika Samudani Dhanasekara, Michael W. Melkus, Reshad S. Ghafouri, Soroush Shahrokh, Flavia Sardela de Miranda, Maria F. Mahecha, Adam Brufsky, Joyce O'Shaugnessy, VK Gadi, Cathy Graham, Mehran Habibi, William Audeh, Sahra Uygun, Lavanya Samraj, Rakhshanda Layeequr Rahman. Correlation between risk groups and molecular subtype of breast cancer with metabolic syndrome and ethnicity [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2024; 2024 Dec 10-13; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(12 Suppl):Abstract nr P3-02-09.

Background: For patients with hormone receptor-positive (HR+), HER2-negative early-stage breast cancer (EBC), pathological Complete Response (pCR) rates to neoadjuvant chemotherapy (NCT) are low. However, NBRST and I-SPY2 trials report that MammaPrint risk of distant recurrence and BluePrint molecular subtyping signatures can more accurately identify which HR+HER2- EBC patients are likely to achieve pCR. Here, MammaPrint and BluePrint as biomarkers for predicting pCR and residual cancer burden (RCB) class, was evaluated among patients with HR+HER2- tumors enrolled in FLEX. Methods: Patients with HR+HER2- tumors treated with NCT with available pCR and/or RCB data enrolled in the ongoing prospective, observational FLEX Trial (NCT03053193), were included in this study (N = 457). Tumors were characterized as MammaPrint UltraLow (UL), Low (LR), High 1 (H1) or High 2 (H2) Risk. BluePrint subtypes were classified as either Luminal A, Luminal B, HER2, or Basal. pCR and Pathological Response (PR= pCR + minimal residual cancer burden [RCB-I]) were the study endpoints. Differences in clinical characteristics and response rates were evaluated by Chi-Squared test or two-sided proportional Z-test. The association between MammaPrint, BluePrint, and PR was assessed using multivariate logistic regression and was adjusted for menopausal status, T stage, and N stage. Results: MammaPrint classified 2% as UL, 8% as LR, 56% as H1, and 34% as H2. Nearly 60% of patients in this study had clinical lymph node (LN) positive disease. Higher tumor stage and grade were significantly more likely to be associated with higher MammaPrint Risk. Age, menopausal status, race, and LN status were comparable among MammaPrint groups. All UL and LR tumors were Luminal A, most H1 tumors were Luminal B (98%), while 45% of H2 tumors were Luminal B and 54% were Basal (1% were HER2). pCR rates were significantly higher in H2 Basal (39.8%), followed by H2 Luminal B (15.7%), H1 Luminal B (7.4%), and lowest in UL and LR, Luminal A (4.5%) (p<0.001). Among H2 Luminal B tumors, LN negative tumors had a pCR rate of 21.0% whereas LN positive tumors had a pCR rate of 12.5%. Rates of RCB-I were significantly higher among H2 tumors (26.9%) than H1 tumors (7.3%), or UL and LR tumors (0%) (p=0.026). Among H2 tumors, PR rates measured higher than rates of pCR alone in Basal tumors (43.4% vs 39.8%) and more prominently in Luminal B tumors (21.4% vs. 15.7%). Multivariate analysis revealed that MammaPrint H2 was significantly associated with likelihood of PR (OR=12.6, 95% CI 2.43 – 23.28, p=0.016), whereas menopausal status, T stage, and N stage were not associated with PR. Conclusion: MammaPrint and BluePrint predict chemosensitivity to NCT in HR+HER2- EBC. Patients with MammaPrint H2 tumors, including Luminal B and Basal subtypes, are more likely to achieve pCR or PR to NCT, compared to UL, LR, and H1 groups. Additionally, pCR rates trended lower among LN positive disease, possibly due to the increased difficulty of clearing disease in LNs with systemic therapy. We observed very high rates of RCB-I in H2 Luminal B tumors. These findings should be confirmed in other studies. The very high rates of PR observed in this study suggest that MammaPrint H2 identifies a subset of patients with HR+HER2- disease that benefit from NCT, which may enable surgical downstaging efforts and translate to improvements in outcomes. Citation Format: Joyce O’Shaughnessy, Priyanka Sharma, Cynthia R C Osborne, Gregory Vidal, Pond Remsen Kelemen, Suzanne Hoekstra, Harshini Ramaswamy, Nicole Stivers, Andrea Menicucci, William Audeh, FLEX Investigators’ Group. MammaPrint® and BluePrint® predict pathological response to neoadjuvant chemotherapy in patients with HR+HER2- early-stage breast cancer enrolled in FLEX [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2024; 2024 Dec 10-13; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(12 Suppl):Abstract nr PS4-04.

Patients with hormone receptor-positive (HR + )/HER2- breast cancer may benefit from neoadjuvant immune checkpoint inhibitor (ICI) plus chemotherapy. The effect of chemotherapy or ICI run-in before combination therapy in this population is unexplored. In this randomized pilot trial, patients with HR + /HER2- breast cancer received two weeks of neoadjuvant nab-paclitaxel or pembrolizumab, with baseline and post-run-in tumor biopsy, followed by combined nab-paclitaxel/pembrolizumab. The primary endpoint was PD-L1 expression change between biopsies. Tumor whole exome/RNA sequencing were performed. Of 29 patients, 72% were node-positive. Residual cancer burden (RCB) 0-1 rate was 28% (inclusive of patients receiving additional neoadjuvant adriamycin/cyclophosphamide). No significant change in PD-L1 expression occurred following nab-paclitaxel or pembrolizumab run-in, thus the primary endpoint was not met. Other secondary outcome measures included overall response rate of 80% to the neoadjuvant regimen, and 3-year event-free survival of 86% (95% CI 69-100%); there were no unexpected safety signals. In exploratory biomarker analyses, higher baseline PD-L1 expression and inflammatory gene signatures were associated with favorable response (RCB 0-1); higher expression of estrogen response genes, with unfavorable response (RCB 2-3). Clinical Trial Number: NCT02999477.

Clinical T3 (cT3) breast cancer (BC) presents a challenge for achieving cosmetically acceptable breast conservation, and neoadjuvant chemotherapy (NAC) is commonly used for cytoreduction in these high-risk cancers. MammaPrint® risk-of-recurrence and BluePrint® molecular subtyping genomic signatures have demonstrated high accuracy in predicting chemotherapy benefits. Here, we examined the utility of MammaPrint/BluePrint for predicting pathological Complete Response (pCR) rates to NAC among 404 patients diagnosed with cT3 early-stage BC. The association of genomic subtype and clinical features with the likelihood of pCR was evaluated by multivariate logistic regression. Differences in pCR rates between genomic risk categories were evaluated by a two-sided proportional z-test and stratified by nodal status. MammaPrint/BluePrint subtyping was associated with significantly higher odds ratios (ORs) for pCR in MammaPrint High-Risk/BluePrint Basal-Type (OR = 3.06, 95% CI: 1.15-8.19, p = 0.025) and HER2-Type (OR = 6.27, 95% CI: 2.19-19.38, p = 0.001) compared to BluePrint Luminal-Type. Of the 209 patients with hormone receptor-positive, HER2-negative disease, 6.7% achieved pCR, and MammaPrint High-Risk was associated with a significantly higher pCR rate (9.3%) compared to MammaPrint Low-Risk cancers (0%), regardless of nodal involvement (p = 0.036). These data show that for patients with MammaPrint Low-Risk, cT3 tumors are less likely to have clinically impactful cytoreduction from NAC, regardless of nodal involvement.

INTRODUCTION:Genomic tests informing systemic therapy recommendations for breast cancer (BC) were developed and validated on primary tumor (PT) tissue. Agreement of genomic tests between the PT and matched lymph node metastases (LNm) in patients with BC is currently unclear, which limits their use in nodal tissue. METHODS:Prespecified study within the international phase III TAXIS trial (NCT03513614). Tissue from PT and matched LNm were assessed with the primary objective of comparing genomic recurrence risk and subtypes. Clinical risk was assessed using the Adjuvant!Online tool and clinical subtypes were determined based on hormone receptor and Her2 testing. Agreement was assessed using Cohen's Kappa (Κ). RESULTS:Eighty-nine patients with stage II/III BC from 26 European centers were included. Median age was 63 years (range 50-72). Agreement in genomic risk between the PT and LNm was found in 84.3 % (Κ=0.64). However, 15.6 % exhibited genomic high risk in the LNm and low risk in the PT, while 16.0 % demonstrated low risk in the LNm and high risk in the PT. Genomic subtypes of PT and LNm showed disagreement in 17.1 % of patients with luminal BC (Κ=0.70). 96.5 % of patients were clinically categorized as high risk, whereas genomically, 25.6 % of them were classified as low risk in the LNm (Κ=0.17) and 31.7 % in the PT (Κ=0.13). Genomic subtyping reclassified 35.5 % of the clinically luminal tumor subtypes. CONCLUSION:Disagreement in genomically estimated risk exists in over 15 % of patients, potentially leading to over- or undertreatment. Consequently, the applicability of genomic tests in LNm remains controversial.

OBJECTIVES:Breast carcinomas overexpressing human epidermal growth factor receptor 2 (HER2) are typically associated with higher tumor grade and faster progression. HER2 positivity is rare in low-grade breast carcinomas with unclear biological implications. We aimed to characterize their clinicopathologic and molecular profiles in this study. METHODS:There were 2 cohorts of Nottingham grade 1, HER2-positive invasive breast carcinomas examined: (1) an institutional series (n = 14) and (2) tumors from patients (n = 59) enrolled in the FLEX multicenter clinical registry with MammaPrint and BluePrint profiling. RESULTS:Most (79%) in the case series were both estrogen receptor (ER) and progesterone receptor (PR)-positive. Over half were pathologic or clinical T1N0 tumors. In the 9 cases with adequate material for next-generation sequencing, the majority (66%) demonstrated ERBB2 copy number variations. Most (66%) received HER2-targeted therapy. No recurrences were observed, with a median follow-up time of 43 months. In the FLEX cohort, most tumors were ER-positive (86%) and PR-positive (68%), and over half were clinical T1. Most (70%) were of the luminal phenotype, and over half (54%) were low-risk on MammaPrint. CONCLUSIONS:Low-grade HER2-positive breast carcinomas constitute mostly low-stage, luminal-type, and apparently low-risk tumors, warranting investigation into whether therapy de-escalation could achieve favorable outcomes with less toxicity in this population.

Background: The advent of subtype-specific treatments, particularly hormone therapies and HER2-targeted therapies, has significantly improved survival rates and quality of life for breast cancer (BC) patients. Over the last decade genomic signatures have enabled improved classification of BC into distinct molecular subtypes, providing prognostic and/or predictive information about the metastatic potential of the tumor beyond those of clinicopathologic features. Despite marked progress, BC remains the most frequent cause of cancer death among women globally, accounting for almost 15.5% of all new female BC cases in the US. These poor clinical outcomes warrant further understanding of tumor heterogeneity and identifying genomic signatures, particularly variation within the subset of ER positive early-stage breast cancer (EBC). Pairing the full genome expression data with comprehensive clinical information enables further tumor stratification and a deeper understanding of tumor biology driving EBC. The ongoing, multi-center FLEX study (NCT03053193) seeks to enroll 30,000 patients to create a large-scale, diverse, population-based registry of full genome expression data matched with clinical data to investigate new gene expression signatures of prognostic and/or predictive value in a real-world setting. Efforts are focused on increasing clinical trial enrollment of racial/ethnic minorities and other historically underrepresented groups in clinical trials in the US to promote efficacy in outcomes and health equity. Additional objectives include supporting investigator-initiated sub-studies to address yet unresolved clinically relevant questions in EBC over 5-10 years of follow-up. Methods: FLEX is a large, multi-center, prospective, observational trial that enrolls patients (male or female) who are ≥ 18 years old with histologically proven stage I-III breast cancer. All patients with up to 3 positive lymph nodes who receive standard of care MammaPrint (70-gene signature risk of recurrence), with or without BluePrint (80-gene signature molecular subtype) on a primary breast tumor and consent to clinically annotated full transcriptome data collection are eligible for enrollment. FLEX fosters collaboration across 95 sites in the US, 2 sites in Canada, 1 site in Greece, and Israel. This initiative encourages investigator-initiated sub-studies, promoting diverse research perspectives and potentially enhancing the scope and robustness of the overall study. Within 7 years of trial initiation, FLEX total enrollment amounts to 16,980 EBC patients. To address racial/ ethnic disparities in clinical trials, a concerted effort has led to the inclusion of 1,377 Black, 530 Latin American (LA)/Hispanic, 353 AAPI, out of 14,330 EBC patients with self-reported race and ethnicity, making FLEX the most diverse study on EBC patients to date. Such diversity in FLEX sets a valuable precedent for future research aiming to improve healthcare outcomes for all groups. Currently FLEX supports 12 in-progress investigator initiated sub-studies in 2024, with over 45 abstracts accepted at congresses internationally (2018-2024), including 11 presentations and 2 poster spot-light sessions that address the underlying differences in tumor biology and clinical outcomes in Black, LA, and AAPI populations. Overall, as FLEX continues to grow, the study strives to leverage full transcriptome data to enhance precision medicine in EBC. By identifying molecular subtypes and predictive biomarkers, the trial intends to equip clinicians with enhanced tools for tailoring treatment strategies more effectively in EBC. The FLEX trial represents a pioneering effort in integrating genomic data and clinical information on a large scale to improve outcomes and reduce disparities in EBC. Its emphasis on diversity, comprehensive data collection, and collaborative research pursuits places it at the forefront of precision medicine in EBC. Citation Format: Robert Maganini, Ellis Levine, Kent Hoskins, Sarah Thayer, Alfredo Santillan, Sung Ho-Lee, Eduardo Dias, Regina Hampton, Eric Brown, Maxwell Brown, Joyce O' Shaughnessy, Nicole Stivers, Harshini Ramaswamy, Katie Quinn, Isha Kapoor, William Audeh. FLEX: A Real-World Evidence, Full Transcriptome Study in 30,000 Patients with Early-Stage Breast Cancer [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2024; 2024 Dec 10-13; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(12 Suppl):Abstract nr P2-08-26.

Background: Gene expression assays play a key role in personalizing adjuvant chemotherapy (CT) treatment decisions for patients (pts) with hormone receptor (HR)-positive, HER2-negative (HR+HER2-) early stage breast cancer (EBC). The 70-gene signature, MammaPrint, determines distant recurrence risk in EBC and has demonstrated its ability to guide CT de-escalation in pts with genomically Low Risk tumors based on the MINDACT trial. In the FLEX Registry of Real World Data (RWD), we evaluated MammaPrint as a continuous variable to predict adjuvant CT benefit in HR+HER2- EBC. Methods: This study included 1002 HR+HER2- EBC pts treated from 2017 - 2020 with endocrine therapy (ET) only or ET plus chemotherapy (ET+CT), with 5-year (yr) median follow up in the prospective, observational FLEX RWD study (NCT03053193). ET treated pts were propensity score matched to ET+CT treated pts based on menopausal status, tumor stage, and lymph node (LN) stage (n = 501 each). The MammaPrint Index (MPI) categorizes 4 risk groups: UltraLow (+1.000 to +0.356), Low (+0.355 to +0.001), High 1 (0.000 to -0.569), and High 2 (-0.570 to -1.000). The primary endpoint was Distant Recurrence Free Interval (DRFI), defined as time from diagnosis to distant recurrence or breast cancer specific death per STEEP 2.0 criteria. Kaplan Meier analysis estimated 5-yr DRFI risk as a continuous function of the MPI for each treatment group separately, with predicted 95% CIs. Cox proportional hazards model was used to test for interaction between CT treatment and clinical variables or MPI risk. Results: Most pts were postmenopausal (73%) and LN-negative (76%), with tumor stages T1 (40.7%), T2 (46.9%), or T3 (10.8%). Tumor grades were Grade 1 (27%), Grade 2 (54.4%) or Grade 3 (19.0%). Clinical features were comparable between treatment groups, except for Grade 3, which was significantly higher in the ET+CT pts. The MammaPrint Index exhibited strong predictive accuracy for 5-yr risk of a DRFI event in both ET only (R2 = 0.99, p < 0.001) and ET+CT pts (R2 = 0.90, p < 0.001). In ET only pts, 5-yr risk of a DRFI event significantly increased with higher MPI risk group, ranging from 0.6% to 2.2% (average [avg] 1.0%) in UltraLow, 2.2% to 4.5% (avg 3.2%) in Low, 5.6% to 14.6% (avg 10.0%) in High 1, reaching highest risk estimates of 14.8% to 24.8% (avg 19.1%) in High 2. In the ET+CT treated pts, 5-yr risk of a DRFI event ranged from 0.1% to 1.0% (avg 0.4%) in UltraLow, 1.0% to 2.1% (avg 1.5%) in Low, 2.6% to 6.4% (avg 4.4%) in High 1, and 6.5% to 10.6% (avg 8.2%) in High 2. Significant DRFI risk differences were observed between treatment groups as MPI risk increased. CT benefit ranged from 3.1% to 8.2% (avg 5.6%) in High 1, and was highest in High 2, ranging from 8.3% to 14.2% (avg 10.9%). Minimal CT benefit was observed for Low (1.7%) and UltraLow (<1.0%) Risk groups. In a subgroup of Clinical Low Risk, MammaPrint High Risk pts (n = 209), the ET+CT group had lower risk of a DRFI event (1.8%) than the ET only group (4.5%). The multivariate Cox model demonstrated CT benefit was dependent on increasing MPI risk (HR = 0.15; 95% CI 0.02-0.97, p = 0.047). CT benefit was also significantly associated with premenopausal status, but not age, T stage or LN stage. The CT benefit was not associated with Grade 3 (HR = 0.99, 95% CI 0.10-9.76, p = 0.695). Conclusion: In this Real World Evidence prospective, propensity score matched study, increasing CT benefit was observed with increasing MPI risk (High Risk). Patients with increasing MPI risk had significantly lower 5-yr risk of a DRFI event when treated with ET+CT than ET alone. Patients with MammaPrint indices within Low and UltraLow Risk ranges did not derive significant CT benefit, consistent with findings from MINDACT. These RWD confirm MammaPrint’s comprehensive utility, as prognostic of recurrence risk and predictive of adjuvant CT benefit in HR+HER2- EBC. Citation Format: Adam Brufsky, Kent Hoskins, Henry Conter, Pond Kelemen, Mehran Habibi, Laila Samian, Robert Maganini, Rakshanda Rahman, Laura Lee, Eduardo Dias, Regina Hampton, Beth Seiling, Cynthia Osborne, Eric Brown, Jailan Elayoubi, Priyanka Sharma, Jayanthi Ramadurai, Laurie Matt-Amral, Alfredo Santillan, Sasha Strain, Philip Albaneze, Harshini Ramaswamy, Nicole Stivers, Andrea Menicucci, William Audeh, Pat Whitworth, Nathalie Johnson, Joyce O’Shaughnessy. Prediction of Chemotherapy Benefit by MammaPrint® in HR+HER2- Early Stage Breast Cancer Revealed by the FLEX Registry of Real World Data [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2024; 2024 Dec 10-13; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(12 Suppl):Abstract nr P2-08-12.

11175 Background: Older women (≥70) are less likely to receive chemotherapy (CT) due to quality-of-life concerns. Additionally, older patients are underrepresented in studies assessing the utility of genomic profiling in guiding CT decisions, thus guidelines around neo/adjuvant CT for this population are less clear. To identify the utility of the MammaPrint (MP) 70-gene and BluePrint (BP) 80-gene assays in informing treatment decisions in an elderly population, we examined the relationship of age ( ≥70 vs.<70) and treatment outcomes stratified by MP/BP subtypes in pts with HR+ HER2- EBC. Methods: The prospective, observational FLEX Study (NCT03053193) includes stage I-III pts with early-stage breast cancer (EBC) who received MP with or without BP testing and consented to full transcriptome and clinical data collection with therapy data available. Differences in the distribution of clinical characteristics between age groups were assessed by Chi-squared, Fisher’s exact, or Wilcoxon-Mann-Whitney tests. The endpoint recurrence-free interval (RFI) was defined as time to local, regional, or distant recurrence or breast cancer related death Kaplan-Meier survival analysis and log-rank tests were used to assess differences in endpoints between treatment groups. Results: A total of 4,519 HR+, HER2- EBC pts were included, with 1,047 ≥70 (23.2%) and 3,472 <70 (76.8%). Patients ≥70 were significantly less likely to present with high grade tumors and lymph node involvement than those <70 (12.8% vs 16.2% grade 3, p=.022; 20.6% vs 24.2% node positive, p=.017, respectively). The MP risk group distribution showed a significantly higher proportion of low genomic risk (Ultralow or low risk) tumors in the ≥70 vs. <70 group (Ultralow (UL) 14.7% vs 14.9% , Low 41.2% vs 38.7% , High 1 (H1) 37.3% vs 36.9% , and High 2 (H2) 6.8% vs 9.5% , p 0.048, respectively). Patients ≥70 with MP High Risk tumors were less likely to receive CT compared to those <70 (H1 55.8% vs 73%, p<0.001; H2 72.6% vs 82.2%, p=0.07, respectively). When evaluating 3-year RFI, the ≥70 pts with MP High Risk cancer trended towards better outcomes with CT than those receiving endocrine therapy only, especially in H2 cancers (H1 97% vs 94% , p=.137, H2 90% vs 79% , p=.078, respectively). Conclusions: This study underscores the potential CT benefits in MP H2 HR+ HER2- EBC pts ≥ 70 who may forgo treatment due to overall health and quality of life concerns. Notably, in MP H2 pts, the absolute improvement in 3-year RFI of 11% with neo/adjuvant CT in women ≥ 70 suggests that for many pts, the benefit outweighs the risks. Of note, this H2 CT benefit is similar to that observed in a group of 1000 pts with a median age of 59 recently reported (Brufsky, et al. SABCS 2024, P2-08-12). Patient centered discussions on performance status, comorbidities, and genomic profiling of HR+ HER2-EBC as well as the potential benefit from CT should guide personalized treatment. Clinical trial information: NCT03053193 .