Abstract P1-04-05: Independent validation of the HER2DX genomic test in HER2-positive breast cancer treated with neoadjuvant paclitaxel, trastuzumab and pertuzumab (THP): a correlative analysis from the DAPHNe phase II clinical trial

Background: HER2DX is a 27-gene prognostic (risk-score) and predictive (pathological complete response [pCR]-score) assay in early-stage HER2+ breast cancer (BC) based on clinical data and the expression of 4 gene signatures (immune, proliferation, luminal differentiation, and HER2 amplicon). Here we aim to evaluate, for the first time, the ability of HER2DX to predict pCR following neoadjuvant THP in HER2+ BC.

Methods: Standardized HER2DX was evaluated centrally on baseline pre-treatment FFPE tumor biopsies from the DAPHNe phase II trial (Waks et al. NPJ Breast 2022; NCT03716180), in which patients (pts) with newly diagnosed stage II-III HER2+ BC were treated with neoadjuvant weekly paclitaxel × 12 and HP every 3 weeks × 4. Primary aim was to test the ability of HER2DX pCR-score to predict pCR (ypT0/isN0). Secondary objectives were to test the ability of HER2DX pCR-score to predict pCR independent of clinical-pathological variables and PAM50 subtype (HER2-enriched vs not) and to evaluate the association of HER2DX pCR-score with HER2DX risk-score. Five patients who received additional neoadjuvant chemotherapy after THP were excluded from this analysis. Logistic regression and receiver-operator curve (ROC) analysis were assessed. Statistical analyses were performed in R code 4.0.5.

Results: HER2DX was evaluated in 80 of 97 pts (82.5%) enrolled in the DAPHNe trial who received study treatment. Clinical T2-4 disease represented 81.3% of cases (n=65), clinical node-negative disease (cN0) represented 65.0% of cases (n=52), and 70.0% of tumors (n=56) were hormone receptor-positive. The overall pCR rate was 60.0% (95% confidence interval [CI] 49.3-70.7): 87.0% (95% CI 79.6-94.4) in hormone receptor-negative disease and 48.2% (95% CI 37.2-59.1) in hormone receptor-positive disease. The proportion of HER2DX low-, medium- and high-pCR groups was 38.8%, 27.5% and 33.7%, respectively. HER2DX pCR-score (as a continuous variable from 0 to 100) was significantly associated with pCR (odds ratio [OR]=1.05, p< 0.0001). In the overall population, the pCR rates in HER2DX pCR-high, pCR-med and pCR-low groups were 92.6%, 63.6% and 29.0% (pCR-high vs pCR-low OR=30.6, p< 0.0001), respectively. The AUC ROC of HER2DX pCR score (as a continuous variable) and pCR status was 0.835. In the ER-negative population, the pCR rates in HER2DX pCR-high, pCR-med and pCR-low groups were 94.7%, 66.7%, and 0%, respectively (Table 1). HER2DX pCR-score was significantly associated with pCR independent of hormone receptor status, HER2 immunohistochemistry (IHC) score, clinical stage, and PAM50 HER2-enriched subtype. The correlation between HER2DX pCR-score and HER2DX risk-score was weak (Pearson coefficient=-0.12), as previously described (Prat et al. EBiomedicine 2022). 51.3% of patients were categorized as HER2DX low-risk.

Conclusion: The 27-gene HER2DX genomic test predicts pCR following neoadjuvant THP in newly diagnosed stage II-III HER2+ BC. Patients with HER2DX pCR-low score and HER2DX high-risk score, representing 22.5% of pts, warrant further attention in order to optimize therapeutic strategies in this subset. The combination of HER2DX pCR-score and risk-score might guide therapeutic decisions by identifying patients who are ideal candidates for de-escalated or escalated systemic and locoregional treatments.

Table 1

Citation Format: Adrienne Waks, Esther R. Ogayo, Laia Paré, Mercedes Marín-Aguilera, Fara Brasó-Maristany, Patricia Galván, Oleguer Castillo, Olga Martínez-Sáez, Ana Vivancos, Patricia Villagrasa, Paolo Tarantino, Neelam Desai, Jennifer Guerriero, Otto Metzger, Nadine Tung, Ian Krop, Joel S Parker, Charles M. Perou, Aleix Prat, Eric Winer, Sara Tolaney, Elizabeth A. Mittendorf. Independent validation of the HER2DX genomic test in HER2-positive breast cancer treated with neoadjuvant paclitaxel, trastuzumab and pertuzumab (THP): a correlative analysis from the DAPHNe phase II clinical trial [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P1-04-05.

Abstract P5-02-09: Quantitative analysis of fiber-level collagen features in H&E whole-slide images predicts neoadjuvant therapy response in patients with HER2+ breast cancer

Background: Neoadjuvant treatment (NAT) combining chemotherapy and HER2-targeted agents is frequently administered to HER2-positive (HER2+) breast cancer (BC) patients, with some experiencing a pathological complete response (pCR) and others having residual disease measured by the residual cancer burden (RCB) score. Here, we use a physics-guided machine learning (ML)-based approach to extract fiber-level collagen features from hematoxylin and eosin (H&E)-stained whole slide images (WSIs) and identify collagen-related associations with treatment response in HER2+ patients receiving NAT.

Methods: Clinical data and specimens from stage II-III HER2+ BC patients enrolled on the De-escalation to Adjuvant Antibodies Post-pCR to Neoadjuvant THP (DAPHNe; NCT03716180) clinical trial and treated with neoadjuvant paclitaxel/trastuzumab/pertuzumab were analyzed. An ML-based model trained to identify regions of BC tissue as invasive carcinoma, ductal carcinoma in situ (DCIS), diffuse inflammatory infiltrate, stroma, necrosis, or normal tissue was deployed on WSIs of H&E-stained diagnostic core needle biopsies (N=89) to generate tissue overlays. Additional tissue areas were computed from the tissue model predictions using heatmap transformation, including tumor nests (continuous regions predicted as invasive cancer epithelium or DCIS), tumor nest borders (stromal region boundaries 10 μm from tumor nests), and bulk tumor borders (stromal region boundaries 300 μm from aggregated tumor nests). A separate ML-based model trained to identify fiber-level collagen features in WSIs of H&E-stained specimens was also deployed to generate collagen overlays. A fiber feature extraction pipeline was utilized to characterize properties of all identified collagen fibers in the WSI (on the order of hundreds of thousands per slide), including length, width, tortuosity, and angle. These fiber features were then assessed based on their position within the tumor (e.g. relative to the tumor nest border). Combinatorial features (e.g. angle of fibers with respect to tumor boundary) were then explored univariately for associations (N=609) with treatment response. Patients with pCR (RCB=0; N=53) were considered responders, while all other cases (RCBI-III; N=36) were designated non-responders. Due to the small size of the cohort analyzed here, raw p-values are reported.

Results: Using estrogen receptor status as a clinical covariate, a logistic regression-based univariate analysis of 609 collagen-associated features revealed six features to strongly associate with pCR (p< 0.05, AUC≥0.75; Table 1). Notable feature themes were identified: 1) fiber tortuosity in tumor nest borders and tumor borders, 2) angle of fibers in tumor border with respect to tumor boundary, and 3) distribution patterns of fiber width in tumor nest borders. The presence of fibers perpendicular to tumor boundary tangents was negatively associated with pCR, as was higher fiber tortuosity and thickness in tumor nest borders.

Conclusions: Improved prediction of response to NAT in patients with BC is needed to determine appropriate treatment strategies for each patient. Here, using ML-based models to identify tissue features and collagen fibers, we identify collagen-associated features, measured directly from WSIs of H&E-stained diagnostic BC biopsies, that negatively correlate with pCR. Additional development of this strategy, including the addition of cell identification models and known clinical information, is underway to further refine this novel predictive model.

Abstract PD18-01: Adjuvant Trastuzumab Emtansine Versus Paclitaxel plus Trastuzumab for Stage I HER2+ Breast Cancer: 5-year results and correlative analyses from ATEMPT (TBCRC033)

Background: The ATEMPT trial primary analysis found that one year of adjuvant trastuzumab emtansine (T-DM1) achieved a 3-year iDFS of 97.8% for patients with stage I HER2+ breast cancer, but was not associated with fewer clinically relevant toxicities (CRTs) compared with paclitaxel and trastuzumab (TH). In this end-of-study analysis, we report 5-year survival outcomes and correlative analyses from the trial. Methods: Patients with stage I centrally confirmed HER2+ breast cancer were randomly assigned 3:1 to adjuvant T-DM1 for one year or TH and received T-DM1 3.6 mg/kg IV every 3 weeks for 17 cycles or paclitaxel 80 mg/m2 IV with weekly trastuzumab IV followed by trastuzumab for 9 months. The co-primary objectives were to compare the incidence of CRTs between the 2 arms and to evaluate iDFS in patients receiving T-DM1. To investigate proteomic correlates of recurrence, spatial proteomic analyses were performed on samples from 13 patients experiencing iDFS events (cases) and 24 matched controls using the NanoString GeoMx Digital Spatial Profiler. The impact of HER2 heterogeneity on outcomes was investigated among 17 cases and 51 matched controls by fluorescence in-situ hybridization (FISH). HER2 genetic heterogeneity was assessed by scrutinizing the whole tumor area and defined as the occurrence of HER2 gene amplification in >5% but < 50% invasive tumor cells. The risk of recurrence was evaluated centrally with the HER2DX genomic assay from 225 primary tumor samples. Germline whole genome sequencing (WGS) was conducted among 55 patients experiencing T-DM1-induced thrombocytopenia and/or bleeding and 55 matched controls to identify genomic correlates for this side effect. Results: A total of 497 patients who initiated protocol therapy were included in this analysis (383 T-DM1 and 114 TH). After a median follow up 5.8 years, among patients receiving T-DM1 there were a total of 11 iDFS events, with 3 distant recurrences. The 5-year iDFS for T-DM1 was 97.0% (95% CI, 95.3-98.8%), the 5-year recurrence-free interval (RFI) was 98.6% (95% CI: 97.4-99.8%) and the 5-year overall survival (OS) for T-DM1 was 97.8 % (95% CI, 96.3-99.3%). Although the study was not powered to evaluate the efficacy of TH, among the 114 patients receiving TH, a total of 9 iDFS events were observed, including 2 distant events; the 5-year iDFS with TH was 91.3% (95% CI: 86.0-96.9%), 5-year RFI was 93.3% (95% CI: 88.6-98.2%) and 5-year OS was 97.9% (95% CI: 95.2-100%). A total of 56 samples were evaluable for heterogeneity analyses, among which 14% (n=8) harbored HER2 genetic heterogeneity. Spatial proteomic analyses found that NF1 (adjusted p=0.72 × 10-6) and CTLA-4 (adjusted p=0.15 × 10-3) were significantly upregulated in primary samples from cases, while cleaved caspase 9, CD25, GITR, ICOS, p53 and PD-L2 were significantly upregulated in controls (all with adjusted p< 0.05). Germline WGS found that the top gene associations with thrombocytopenia and thrombocytopenia or bleeding were ALMS1 (p=0,19 × 10-3) and APBA3 (p=0,23 × 10-3), respectively, although none reaching the threshold for genome wide significance. rs62143195 and rs114169776 were the top single nucleotide polymorphisms associated with thrombocytopenia and thrombocytopenia or bleeding, respectively. Data on the impact of HER2 heterogeneity and of HER2DX score on survival outcomes will be presented. Conclusion: With longer follow-up, adjuvant T-DM1 confirmed outstanding long-term outcomes among patients with stage I HER2+ breast cancer, demonstrating a 5-year RFI of 98.6%. Spatial proteomic analyses identified a potential association between NF1 and CTLA-4 expression with recurrence. Details on the impact of HER2 heterogeneity and HER2DX assay on prognosis will be presented.

Citation Format: Paolo Tarantino, Nabihah Tayob, Chau T Dang, Denise Yardley, Steven J. Isakoff, Vicente Valero, Meredith Faggen, Therese Mulvey, Ron Bose, Douglas Weckstein, Antonio C. Wolff, Katherine Reeder-Hayes, Hope Rugo, Bhuvaneswari Ramaswamy, Dan Zuckerman, Lowell Hart, Vijayakrishna K. Gadi, Michael Constantine, Kit Cheng, Audrey Merrill Garrett, Paul K. Marcom, Kathy S. Albain, Patricia DeFusco, Nadine Tung, Blair Ardman, Rita Nanda, Rachel C. Jankowitz, Mothaffar Rimawi, Vandana Abramson, Paula R. Pohlmann, Catherine Van Poznak, Andres Forero-Torres, Minetta C. Liu, Kathryn Ruddy, Yue Zheng, Romualdo Barroso-Sousa, Adrienne Waks, Michelle K. DeMeo, Molly K. DiLullo, Giuseppe Curigliano, Harold Burstein, Ann Partridge, Eric Winer, Giuseppe Viale, Winnie Hui, Elizabeth A. Mittendorf, Bryan P. Schneider, Aleix Prat, Ian Krop, Sara Tolaney. Adjuvant Trastuzumab Emtansine Versus Paclitaxel plus Trastuzumab for Stage I HER2+ Breast Cancer: 5-year results and correlative analyses from ATEMPT (TBCRC033) [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr PD18-01.

Abstract PD18-02: Adjuvant Paclitaxel and Trastuzumab Trial (APT) for Node-Negative, Human Epidermal Growth Factor Receptor 2–Positive (HER2+) Breast Cancer: final 10-year analysis

Background: The APT trial evaluated the activity of adjuvant paclitaxel and trastuzumab (TH) among patients with small, node negative HER2+ breast cancer. This regimen showed a 7-year invasive disease-free survival (iDFS) of 93%, a recurrence-free interval (RFI) of 97.5% with only four (1.0%) distant recurrences, and a 7-year overall survival (OS) of 95%. In this end-of-study analysis, we report the survival outcomes at 10 years and assess the role of HER2DX testing in predicting long-term outcomes with adjuvant TH.

Methods: APT was a single-arm multicenter investigator-initiated phase II study in which patients with HER2+ breast cancer with tumors ≤3 cm and negative nodes (one single micrometastatic node allowed) received IV weekly paclitaxel (80 mg/m2) with IV weekly trastuzumab for 12 weeks, followed by IV trastuzumab for 9 months. The primary endpoint was 3-year iDFS. Here we report 10-year iDFS, RFI, breast cancer–specific survival (BCSS) and OS. In an exploratory analysis, the risk of recurrence was evaluated with the HER2DX genomic assay.

Results: A total of 410 patients were enrolled from October 2007 to September 2010, of which 406 started the study treatment and were included in the intent to treat analysis. Median age at enrollment was 55 years (range, 24 to 85 years), and most patients (67%) had hormone receptor (HR)-positive disease. Fifty percent of patients had tumors 1.0 cm or smaller and only 9% of patients had tumors between 2 cm to 3 cm. Mean tumor size was 1.1 cm. After a median follow-up of 10.2 years (122 months), 36 iDFS events were observed, consistent with a 10-year iDFS of 89.7% (95% CI, 86.3%-93.1%). Ten-year iDFS was 90.2% (95% CI, 86.3%-94.3%) and 88.5% (95% CI, 82.4%-95.1%) for patients with HR-positive and HR-negative tumors at baseline, respectively. 10-year RFI was 96.8% (95% CI, 95.0%-98.7%), 10-year OS was 94.2% (95% CI, 91.6%-96.8%) and 10-year BCSS was 99.1% (95% CI, 98.1%-100.0%). Of the iDFS events observed in the trial, 6 were non-breast cancer related deaths and 9 were contralateral tumors, all but one locally found to be HER2-negative upon biopsy (Table 1). Among patients experiencing an iDFS event, 7 patients (1.7%) had distant recurrences, including 1 with a T2 tumor, 3 with a T1c tumor and 3 with a T1b tumor. At baseline, 6 of them had HR-positive disease, 1 had HR-negative disease, and 6 had high-grade disease. Upon biopsy of metastatic lesions, 5 of the 7 distant recurrences were locally found to be HER2+, 1 was HER2-negative and 1 had unknown HER2 status. HER2DX testing was conducted on available baseline archival tumor tissue and analyses of patients’ survival outcomes based on the HER2DX score will be presented.

Conclusion: After 10 years of follow-up, adjuvant TH confirmed excellent long-term outcomes for small, node-negative HER2+ breast cancer, with a 10-year RFI of 96.8% and a 10-year BCSS of 99.1%.

Table 1: iDFS events with adjuvant paclitaxel plus trastuzumab after 10.2 years of follow up

Citation Format: Sara Tolaney, Paolo Tarantino, Noah Graham, Nabihah Tayob, Chau T Dang, Denise Yardley, Beverly Moy, Paul K. Marcom, Kathy S. Albain, Hope Rugo, Matthew Ellis, Iuliana Shapira, Antonio C. Wolff, Lisa Carey, Romualdo Barroso-Sousa, Michelle K. DeMeo, Molly K. DiLullo, Ann Partridge, Adrienne Waks, Clifford Hudis, Ian Krop, Harold Burstein, Aleix Prat, Eric Winer. Adjuvant Paclitaxel and Trastuzumab Trial (APT) for Node-Negative, Human Epidermal Growth Factor Receptor 2–Positive (HER2+) Breast Cancer: final 10-year analysis [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr PD18-02.

Abstract 1773: Overcoming paclitaxel resistance in triple-negative breast cancer using a novel barcoding technology

Triple-negative breast cancer (TNBC) is an aggressive breast cancer subtype for which chemotherapy remains a part of standard treatment. Although pathologic complete response (pCR) after neoadjuvant chemotherapy, including paclitaxel (PTX), is associated with good outcomes, 50-60% of TNBC patients do not experience pCR and suffer poor long-term outcomes, often due to chemotherapy resistance. Identifying and analyzing the tumor cells responsible for chemotherapy resistance will lead to improved treatment strategies.

We developed a model of TNBC resistance to neoadjuvant PTX using Met1 murine mammary carcinoma cells. To study clonal dynamics in response to PTX we developed a barcoding method, SunCatcher. First, 31 single-cell derived clonal populations were generated from parental Met1 cells. Each clone was infected with a lentiviral vector containing a unique DNA barcode detectable by qPCR. All barcoded clones (BCs) were mixed in equivalent numbers to generate a BC pool that we confirmed captures the parental Met1 heterogeneity and tumor growth kinetics.

First, PTX responses were tested in vitro. The BC pool had an IC50 of ~100 nM, while individual BCs ranged in IC50 from 5 nM to 25 µM. We maintained the BC pool in 100 nM PTX for 39 days (termed long-term PTX; LTP), at which point it became clonal for BC25, suggesting that BC25 (IC50 of 100 nM) is uniquely PTX resistant, despite being the most proliferative and Zeb1hi/EpCAMlow clone in vitro. BC25 was not detected in the control-treated BC pool at 39 days.

Next, the BC pool was orthotopically injected into FVB/NJ mice and treated the mice with 20 mg/kg PTX on days 7, 11, and 15. At the day 18 experimental endpoint, tumor volume significantly decreased by 56% in response to PTX compared to control. BC25 composition increased from 7.7% (controls) to 15% (PTX treated) within the tumors at endpoint. We also injected BC25 and LTP cells alone and administered the same PTX regimen once tumors reached 50 mm3. Both BC25 and LTP tumors were unresponsive to PTX and had a longer latency period (35d) than BC pool (10d).

We performed a drug screen of 2313 compounds spanning FDA-approved cancer therapeutics on the LTP cells to identify compounds to target the PTX-resistant clone. HDAC inhibitors were the most potent class of hits and one, Panobinostat, killed LTP cells with an IC50 of 4.3 nM.

Utilizing SunCatcher, we identified a unique PTX-resistant TNBC subclone that represents residual disease associated with poor long-term outcome. Typical in vivo experiments would have reached ethical endpoint before BC25 had a chance to grow, given its long latency, therefore this PTX-resistant clone would not have been identified. This is an important finding because therapeutic resistance can emerge after a protracted period, longer than typical pre-clinical experiments. Further work will explore mechanisms of resistance and the potential for combination therapies to prevent recurrent disease.

Citation Format: Milos Spasic, Qiuchen Guo, Adam Maynard, Gregory Goreczny, Adrienne Waks, Sara Tolaney, Elizabeth Mittendorf, Sandra McAllister. Overcoming paclitaxel resistance in triple-negative breast cancer using a novel barcoding technology [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1773.

Abstract P1-04-13: Generation and validation of an estrogen receptor signaling (ERS) gene panel that inversely correlates with antigen presentation and T cell infiltration and activity in hormone receptor positive (HR+) breast cancer

Background: Tumor infiltrating lymphocytes (TILs) are observed in low numbers in HR+ breast cancer relative to other subtypes. For T cells (TC) to recognize and respond to a tumor, antigens must be presented on the tumor cell surface via human leukocyte antigen class one (HLA-I) molecules. Hence, the lack of immune infiltration into HR+ tumors could be explained by limited antigen or impaired antigen presentation. We hypothesized that ERS inversely correlates with antigen presentation and T cell infiltration in HR+ tumors. The objective of this study was to comprehensively examine the relationship between ERS, antigen presentation machinery (APM) and TC gene expression in HR+ breast cancer.Methods: Comprehensive gene panels for ERS, APM and TC expression were generated from literature review, GO terms, KEGG pathways, REACTOME, and computationally and manually curated gene lists. Genes expressed in both The Cancer Genome Atlas (TCGA) and Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) were used for subsequent analyses. Tumors were classified into 4 major subtypes (HR+/HER2-, HR+/HER2+, HR-/HER2+, HR-/HER2-) based on estrogen receptor (ER), progesterone receptor and HER2 expression as defined by immunohistochemistry. To statistically refine each gene panel, the genes were hierarchically clustered based on their pairwise Spearman correlation coefficients among HR+/HER2- samples in TCGA (n = 441) and METABRIC (n = 1028). Specifically, clusters were identified by linkage with 2 hallmark genes for each panel: ESR1 and FOXA1 for the ERS panel, HLA-A and NLRC5 for the APM panel, and CD8A and CD8B for the TC panel. Due to overlapping genes in the APM and TC panels, these two panels were combined for subsequent analyses. Final gene panels for ERS and APM/TC were generated from overlapping genes identified in corresponding TCGA and METABRIC clusters. Internal validity of the final gene panels was assessed through pathway enrichment analysis. The panels were then validated through correlation analysis in an independent single institution cohort (HR+ = 25, HER2+ [regardless of HR] = 25, TNBC = 23). Finally, intra and inter-panel correlation analysis results were compared between breast cancer subtypes in both TCGA and METABRIC datasets. Results: Among the 988 genes identified in our manually curated panels, 788 genes were recognized in both TCGA and METABRIC datasets. Statistical refinement resulted in a final 28-gene ERS panel and a final 135-gene combined APM/TC panel. Early and late-estrogen response pathways were enriched in the ERS panel, whereas interferon-gamma response and other innate and acquired immune-related pathways were enriched in the APM/TC panel. Strong inverse correlations between ERS and APM/TC panels were identified in both TCGA and METABRIC datasets. These findings were validated in the single institution cohort where we noted the strength of the correlations varied with the subtype of disease and extent of HR expression. Further analyses in all 4 breast cancer subtypes, in both TCGA and METABRIC datasets, revealed consistent positive correlations within the APM/TC panel across all subtypes. However, positive correlations within the ERS panel corresponded to the subtypes’ dependency on ER pathway, with a strong correlation in HR+ breast cancer and limited correlation in HR- tumors. Conclusions: Using an unbiased data-driven approach, ERS and APM/TC gene panels were generated. Among HR+ tumors, high levels of ERS gene expression significantly correlated with lower levels of APM/TC gene expression providing one potential mechanism for low TC infiltration in HR+ breast cancer. The prognostic and predictive values of these panels are currently being investigated.

Citation Format: Jonathan S Goldberg, Xiaoyong Cui, Kenichi Shimada, Sandra McAllister, Sara Tolaney, Adrienne Waks, Rinath Jeselsohn, Jennifer Guerriero, Judith Agudo, Elizabeth Mittendorf. Generation and validation of an estrogen receptor signaling (ERS) gene panel that inversely correlates with antigen presentation and T cell infiltration and activity in hormone receptor positive (HR+) breast cancer [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P1-04-13.

Abstract P5-13-15: High dimensional flow cytometric analysis or the peripheral immune profile and response to HER2-targeted antibody therapy

Background: DAPHNe was a prospective trial designed to assess adherence to de-escalated antibody doublet therapy in the adjuvant setting among HER2+ breast cancer patients experiencing a pathologic complete response (pCR) following neoadjuvant taxol (T), trastuzumab (H) and pertuzumab (P). Peripheral blood mononuclear cells (PBMC), were collected from all patients at baseline and after THP completion. The goal of this study was to determine if a patient’s peripheral blood immune profile at baseline, or the longitudinal change with treatment, could predict response to THP. Methods: Blood samples were subjected to high dimensional (28-30 parameter) flow cytometry with comprehensive T- and NK-cell panels. A fully automated computational analysis strategy was undertaken consisting of unsupervised clustering of the high dimensional data into groups of cells with similar immunophenotypic signatures. Clustering was performed using 2 algorithms: Fingerprint-based clustering (Fluster) and High Throughput Mapper (HiTMapper). Clusters were tested using the Wilcoxon rank-sum test for correlation with the clinical response. Responders were those with pCR (=residual cancer burden [RCB] 0) or RCB 1; non-responders were those with RCB 2/3 disease. P values were adjusted with the Benjamini-Hochberg method to control for FDR. In addition to P values, effect size was evaluated using the nonparametric Cliff’s Delta measure. An effect was determined to be large if the magnitude of Delta was >0.4 which corresponded to one cohort coming ahead 70% of the time. In addition, groups of clusters were evaluated using multivariate statistical modeling or dimensionality reduction to determine if there was an association with pCR. Results: Matched baseline and pre-op PBMC were available to perform the NK panel in 66 patients and the T cell panel in 40. In both groups 70% were responders and 30% were non-responders. No cluster produced by Fluster or HiTMapper differed significantly between responders and non-responders however, in the T cell panel, several clusters had a large effect size (table) suggesting the clusters are good at differentiating some, responders from non-responders. Both algorithms agreed that the median responder has more CD4 naïve and CD8 naïve cells than the median non-responder. While no individual cluster differed significantly between responders and non-responders, cross-validated logistic regression analyses showed that 2 clusters, activated CD4 central memory clusters, and activated CD4 naïve clusters, predicted responder status with AUC of 0.70 and 0.68 respectively. Numerous clusters showed robust and significant longitudinal changes between baseline and pre-op samples. Stratifying longitudinal changes by response status revealed no significant differences between responders and non-responders, however evaluation of effect size suggested a naïve CD4 cluster that increased in non-responders and decreased in responders. The latter could be explained as naïve T cells acquiring a memory phenotype in response to treatment in responders. Conclusion: High dimensional flow cytometry suggested a potential role for monitoring several T cell subsets to predict response in HER2+ patients receiving THP. Additional analyses to include cyTOF evaluation of PBMCs are ongoing to further characterize the peripheral immune profile of these HER2+ patients.

T cell clusters with high effect sizeMajor PhenotypeOther markersMethodp-valueeffect sizeCD4 CM.actCD38, CD226HiTMapper0.351-0.469CD4 Naive.act.2CD38, CD226HiTMAPPER0.4-0.413CD4 Naive.act.4CD226HiTMapper0.351-0.490CD4 Naive.act.5CD38HiTMapper0.351-0.524CD8 Naive 1CD226HiTMapper0.396-0.427CD3 Neg-Fluster0.3190.476CD3 Neg3CD45RA, CD185, CD197Fluster0.3190.476CD3 Neg4CD45RA, Eomes,tBETFluster0.4240.413CD4 Naive-Fluster0.319-0.517CD8 Naive1-Fluster0.319-0.469Unassigned 20 (CD4)CD45RA, CD27-, CD28-Fluster0.364-0.441

Citation Format: Esther R Ogayo, Adrienne Waks, Wade Rogers, Matei Ionita, Kenechukwu Adigwe, Jillian Alberti, Sapana Kadel, Jonni Moore, Tari King, Ian Krop, Sara Tolaney, Eric Winer, Jennifer Guerriero, Elizabeth Mittendorf. High dimensional flow cytometric analysis or the peripheral immune profile and response to HER2-targeted antibody therapy [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P5-13-15.

Aiming at a Tailored Cure for ERBB2-Positive Metastatic Breast Cancer: A Review

Importance

Metastatic breast cancer (MBC) has traditionally been considered incurable. Accordingly, current treatment algorithms are aimed at maintaining quality of life and improving overall survival, rather than at complete eradication of the disease. Attempts to achieve cure with high-dose chemotherapy were conducted in the 1990s, with no observed long-term benefit compared with conventional chemotherapy. Nonetheless, Erb-B2 receptor tyrosine kinase 2 (ERBB2, formerly HER2)-targeted biologic treatments, developed in the past 2 decades, are currently challenging this paradigm. Indeed, a fraction of patients with ERBB2-positive MBC achieve long-lasting responses to chemotherapy and ERBB2-blockade, resembling a cure. In this setting, the challenge of identifying the optimal curable population has emerged, including identifying populations in whom treatment escalation strategies may be beneficial, while avoiding overtreatment in patients with incurable disease.

Observations

A number of clinical and pathologic features allow physicians to identify patients with ERBB2-positive MBC who are more likely to experience a long-lasting response to chemotherapy and ERBB2-blockade. Long-term responders tend to be de novo metastatic, have a reduced disease burden, and tend to show deep responses to systemic treatment. In pathologic terms, features associated with long-term response are high ERBB2 expression, lack of detrimental genomic aberrations, and antitumor immune activation. This population of patients may potentially derive benefit from a tailored escalation of frontline treatment with novel anti-ERBB2 drugs, such as trastuzumab deruxtecan, tucatinib, or margetuximab. Additional recent therapeutic and diagnostic advancements could further aid in the path toward a cure for ERBB2-positive MBC.

Conclusions and Relevance

Careful implementation of novel diagnostic and treatment tools could potentially expand the population of patients with ERBB2-positive MBC experiencing long-lasting disease response. Trials are in preparation to confirm this paradigm, and hopefully lead to a new era of precision therapy for breast cancer.

Abstract P1-15-01: Withdrawn

This abstract was withdrawn by the authors.

Citation Format: Metzger Filho O, Janiszewska M, Guo H, Yardley D, Mayer I, Spring L, Arteaga C, Wrabel E, DeMeo M, Freedman R, Tolaney S, Waks A, Bardia A, Parsons H, Partridge A, Mayer E, King T, Polyak K, Viale G, Winer E, Krop I. Withdrawn [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P1-15-01.

Abstract PD9-02: Evolutionary analysis of 462 serial metastatic biopsies from 208 patients with estrogen receptor-positive (ER+) metastatic breast cancer (MBC) using whole exome sequencing (WES)

Background: While great strides have been made in the treatment of ER+ MBC, therapeutic resistance is nearly universal. The genomic evolution of ER+ breast cancer in the metastatic setting under the selective pressure of multiple lines of therapies is not well understood. To address this, we analyzed the clonal dynamics of serial metastatic samples (mets) to evaluate how tumors evolve and to identify acquired resistance mechanisms.

Methods: We performed WES on 462 clinically annotated samples from 208 patients (pts) with ER+ MBC, including 67 primary tumor biopsies, 229 metastatic biopsies and 160 blood samples (cfDNA). Pts with multiple mets included cases with temporally concordant metastatic tumor and blood samples (48 pts) and cases with serial mets obtained over the course of treatment in the metastatic setting (69 pts). Treatments given between the serial mets included CDK4/6 inhibitors (23 pts), and selective estrogen receptor degraders (19 pts), among others.

Results: In the temporally-concordant mets, we found that cfDNA mutations (muts) largely overlap with muts found in tumor biopsies, capturing >85% of clonal tumor muts. However, we observed a higher level of heterogeneity in cfDNA compared to biopsies (p.value< 1.05e-19, Welch test) and a subset of high-confidence muts that were only detected in cfDNA, including in clinically important genes such as ESR1, PIK3CA, KRAS, and ERBB2. Analysis of serial mets was used to elucidate the evolutionary dynamics within the metastatic setting under the selective pressure of treatment. The median duration between mets was 112 days and the median number of inter-biopsy unique treatments was two. Most tumors continued to evolve within the metastatic setting, with 50 out of 69 pts (72%) acquiring a meaningful sub-clone (50% increase in relative cancer cell fraction) and 31 out of 69 (45%) acquiring muts in known cancer genes, including a subset acquiring a plausible resistance alteration such as alterations that dysregulate ER (5 out of 69 pts, 7%; ESR1 mut, FOXA1 amplification (amp), NCOR1 bi-allelic deletion (del)), ERBB (4%; ERBB2 amp, ERBB3 mut), RAS (4%; KRAS mut, NRAS amp, NF1 del), FGF/FGFR (12%; FGFR2 mut, FGFR1/2 amp, FGF3 amp), and cell cycle (13%; RB1 del, CDK4 amp, AURKA amp, CDKN2A del). Finally, in pts who had multiple mets, we observed several cases of evolutionary convergence toward equivalent resistance mechanisms including convergent RB1 loss as a mechanism of resistance to a CDK4/6 inhibitor and convergent BRCA2 reversion following resistance to a PARP inhibitor.

Conclusions: This study demonstrates that ER+ MBC continues to evolve under the selective pressure of treatments in the metastatic setting. These findings elucidate the challenge of studying high complexity and heavily treated tumors, while also highlighting some commonalities in the evolutionary trajectories selected by these treatments. The multiplicity of clinically relevant genomic alterations acquired in these advanced stages highlights the need for serial biopsies and the potential to inform post-progression therapeutic choices through targeting the acquired dependencies in post-progression tumors.

Citation Format: Cohen O, Buendia-Buendia J, Wander S, Nayar U, Mao P, Waks A, Kim D, Freeman S, Adalsteinsson V, Helvie K, Livitz D, Rosebrock D, Leshchiner I, Dellostritto L, Garrido-Castro A, Jain E, Periyasamy S, Mackichan C, Lloyd M, Marini L, Krop I, Garraway L, Getz G, Winer E, Lin N, Wagle N. Evolutionary analysis of 462 serial metastatic biopsies from 208 patients with estrogen receptor-positive (ER+) metastatic breast cancer (MBC) using whole exome sequencing (WES) [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr PD9-02.

Abstract S1-01: Whole exome and transcriptome sequencing of resistant ER+ metastatic breast cancer

Background: While great strides have been made in the treatment of estrogen receptor-positive (ER+) metastatic breast cancer (MBC), therapeutic resistance invariably occurs. A better understanding of the underlying resistance mechanisms is critical to enable durable control of this disease.

Methods: We performed whole exome sequencing (WES) and transcriptome sequencing (RNA-seq) on metastatic tumor biopsies from 88 patients with ER+ MBC who had developed resistance to one or more ER-directed therapies. For 27 of these patients, we sequenced the treatment-naïve primary tumors for comparison to the resistant specimens. Tumors were analyzed for point mutations, insertions/deletions, copy number alterations, translocations, and gene expression. Detailed clinicopathologic data was collected for each patient and linked to the genomic information.

Results: WES of all metastatic samples demonstrated several recurrently altered genes whose incidence differed significantly from primary, treatment-naïve ER+ breast cancers sequenced in the TCGA study (TCGA). These include ESR1 mutations (n=17, 19.3%; 32.86 fold enrichment, q.value<7.5e-12), CCND1 amplification (n=52, 59.1%; 2.3 fold enrichment, q.value<0.0073), and MAP2K4 biallelic inactivation (n=14, 15.9%; 3.04 fold enrichment, q.value< 0.054).

Comparing to matched primary samples from the same patient, many alterations were found to be acquired in several cases, including for ESR1, ERBB2, PIK3CA, PTEN, RB1, AKT1, and others. Initial analysis of RNA-seq data from metastatic samples (n=59) allowed classification of individual resistance mechanisms into broader resistance modes based on the observed transcriptional state.

Conclusions: We present a genomic landscape of resistant ER+ MBC using WES and RNA-seq. Multiple genes were recurrently altered in these tumors at significantly higher rates than in ER+ primary breast cancer. When compared with matched primary tumors from the same patient, alterations in these and other genes were often found to be acquired after treatment, suggesting a role in resistance to ER-directed therapies and/or metastasis. Potential resistance mechanisms appear to fall into several categories; integrating RNA-seq data may enhance the ability to identify these categories even when genomic alterations are not identified. Multiple clinically relevant genomic and molecular alterations are identified in metastatic biopsies– with implications for choice of next therapy, clinical trial eligibility, and novel drug targets.

Citation Format: Cohen O, Kim D, Oh C, Waks A, Oliver N, Helvie K, Marini L, Rotem A, Lloyd M, Stover D, Adalsteinsson V, Freeman S, Ha G, Cibulskis C, Anderka K, Tamayo P, Johannessen C, Krop I, Garraway L, Winer E, Lin N, Wagle N. Whole exome and transcriptome sequencing of resistant ER+ metastatic breast cancer [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr S1-01.

Recognition of regions in brain sections

This paper addresses the problem of region identification in sequential brain sections and presents a recognition system that finds and tracks region boundaries in those sections. The characteristics of the areas of interest are unique in one sense because they are not stationary. Some regions are hardly discernible. In others, parts of the boundary are missing or so completely blurred that parts of the background may be considered as an extension of the region itself. Moreover, outliers are likely to exist in many cases. Due to the unique properties of brain regions, the emphasis is on robustification and efficiency. The region segmentation problem was expressed as a multi-hypothesis test seeking boundaries that maximize a performance criterion which is general in terms of blur and noise. Boundary candidates are restricted to an adaptive search area around a reference boundary which is usually the outcome of the algorithm from the previous section. The search for the maximum criterion uses a fast first order dynamic programing (DP) procedure, reducing the processing time. Outlier rejection techniques are integrated with the multi-hypothesis test to compensate for both outliers and noise. The result is the reference for the next section. Experimental results on boundary detection are presented. The algorithm is successful in tracing boundaries when the contrast is smaller than the noise power, and when parts of the outlines are missing.

Mechanisms of hypertension in obesity

We conclude that the following may explain the rise in blood pressure with obesity and the subsequent fall in blood pressure (Fig. 2): (1) An increase in calories, protein, or carbohydrate leads to an increase in plasma catecholamines, sympathetic nervous system activity, and insulin secretion. (2) These factors, in turn, lead to increased renal sodium retention and stimulation of the renin-aldosterone system which, in turn, leads to: (3) An increased cardiac output with an inability to appropriately adjust the peripheral resistance to maintain normotension with resultant hypertension. Conversely, the fall in blood pressure with weight reduction can be explained by (Fig. 3): (1) A decrease in calorie, carbohydrate, or protein intake which leads to: (2) A decrease in circulating plasma catecholamines, sympathetic nervous system activity, and insulin secretion which results in: (3) A natriuresis and decrease in the renin-aldosterone system, which causes a decrease in circulating blood volume and in cardiac output. This, in turn, lowers blood pressure towards normal. The unanswered question still remains: why do some obese patients become hypertensive and others remain normotensive? Perhaps there are weight-sensitive individuals and weight-resistant individuals just as there appear to be salt-sensitive and salt-resistant hypertensive patients. Perhaps the answer is genetic. These questions also remain to be answered.