Search

Testing peripheral blood for circulating tumor DNA (ctDNA) offers a minimally invasive opportunity to diagnose, characterize, and monitor the disease in individual cancer patients. ctDNA can reflect the actual tumor burden and specific genomic state of disease and thus might serve as a prognostic and predictive biomarker for immune checkpoint inhibitor (ICI) therapy. Recent studies in various cancer entities (e.g., melanoma, non–small cell lung cancer, colon cancer, and urothelial cancer) have shown that sequential ctDNA analyses allow for the identification of responders to ICI therapy, with a significant lead time to imaging. ctDNA assessment may also help distinguish pseudoprogression under ICI therapy from real progression. Developing dynamic changes in ctDNA concentrations as a potential surrogate endpoint of clinical efficacy in patients undergoing adjuvant immunotherapy is ongoing. Besides overall ctDNA burden, further ctDNA characterization can help uncover tumor-specific determinants (e.g., tumor mutational burden and microsatellite instability) of responses or resistance to immunotherapy. In future studies, standardized ctDNA assessments need to be included in interventional clinical trials across cancer entities to demonstrate the clinical utility of ctDNA as a biomarker for personalized cancer immunotherapy.

Background: Neoadjuvant endocrine therapy (NET) is an approach that allows real-time evaluation of drug efficacy as well as investigation of the biological and molecular changes that occur after estrogenic deprivation. This evaluation is mainly performed in surgical specimens after NET and it has been used as a research tool to obtain prognostic and predictive information using tumour response to decide adjuvant treatment. In this setting, there are not many validated biomarkers to predict response beyond Ki67 expression and Preoperative Prognostic Index (PEPI score) in NET. The aim of this study is to determine if the tumour cellularity size (TCS) in surgical specimen after NET correlates with PEPI score and Ki67 expression. Methods: Retrospective study of postmenopausal patients with estrogen receptor (ER) positive/HER2 negative resectable breast cancer, treated with an aromatase inhibitor for at least 4 months prior to surgery. Pathological characterization of tumour specimens: Evaluation of the percentage of residual tumour cellularity of formalin fixed paraffin embedded surgical specimens and immunohistochemistry characterization of ER and Ki67. Tumour cellularity size: calculated by combining the percentage of residual tumour cellularity and tumour pathological size. Results: N=104. Tumour characteristics at surgery and breakdown for the calculated PEPI score: table 1. Correlation between the percentage of Ki67 positive cells at surgery and TCS: (r=0.2503) p=0.04 (95% CI, 0.0014 to 0.4700). Correlation between TCS and PEPI score: (r=0.2582) p=0.05 (95% CI, -0.0131 to 0.4940). Conclusions: Tumour cellularity size is a promising biomarker to determine response and prognosis after NET. There is a need to find other biomarkers to predict response after NET. Table 1Pathology/Biomarker statusPEPI score RFS pointsNo. of patientspTNot available (NA)1pT1/T20102pT3/T431pNNA7Negative075Positive322KI67 levelNA10%-2.7%039>2.7%-7.3%131>7.3%-19.7%119>19.7%-53.1%213>53.1%31ER-status Allred scoreNA120-2303-8092PEPI scoreNA190281342533495561PEPI groupNot calculated19I (0 score)28II (1-3 score)42III (≥4 score)15 Citation Format: Joanna Ines Lopez Velazco, María Otaño, Lide Larburu, Ainhara Lahuerta, Kepa Elorriaga, Virginia Segur, Juan Carlos Irizabal, Ana Martínez, Lourdes Jáuregui, Maria M. Caffarel, Ander Urruticoechea. Tumour cellularity size as a biomarker to predict response after neoadjuvant endocrine therapy: Correlation analysis between Ki67 expression and PEPI score [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS5-39.

Background: Several biomarkers such as estrogen receptor (ER), progesterone receptor (PR), androgen receptor (AR) and Ki67 have been implicated in the pathogenesis and/or as prognostic biomarkers of breast cancer, and are utilized to determine treatment. Given the heterogeneity of response to endocrine therapy, however, predictive biomarkers are critical to better individualize patient care. Previous results from the Trans-aTTom study demonstrated that the Breast Cancer Index HOXB13/IL17BR [BCI (H/I)] biomarker significantly predicts extended endocrine benefit from 10 vs 5y of tamoxifen. In this correlative study, the predictive activity of BCI (H/I) was compared with ER, PR, AR and Ki67 protein expression in node positive patients treated in the aTTom trial. Methods: Patients with available tumor tissue and biomarker analyses were included. ER, PR, AR and Ki67 were centrally assessed by immunohistochemistry (IHC) utilizing tissue microarrays. BCI gene expression analysis by RT-PCR was performed blinded to clinical outcome. Multivariate Cox models adjusting for age, tumor size, tumor grade and HER2 status were used to assess the significance of the interaction between treatment and each biomarker as continuous variables. 17-year risk of recurrence, as a function of each continuous biomarker, was estimated from Cox models in each of the 2 treatment arms. Results: Analysis of 789 HR+, N+ patients showed a weak negative correlation between BCI (H/I) and ER, PR, and AR expression whereas Ki67 and the AR/ER ratio showed no correlation (ER, cor=−0.18; PR, cor=−0.25; AR, cor=−0.14; Ki67, cor=0.04; AR/ER ratio, cor=0.02). The interaction between BCI (H/I) and extended tamoxifen treatment was significant (p=0.014). In addition, analysis of risk of recurrence as a function of continuous BCI (H/I) demonstrated that the magnitude in the reduction in recurrence risk with extended tamoxifen correlated with increasing H/I levels. In contrast, interaction P values were nonsignificant (ER, p=0.829; PR, p=0.659; AR, p=0.783; Ki67, p=0.865; AR/ER ratio, p=0.835) and the magnitude of endocrine benefit did not correlate with expression levels of any of other biomarkers. Conclusion: Results from this post-hoc analysis of the Trans-aTTom study demonstrated that whereas BCI(H/I) is a significant predictive biomarker of endocrine response, analysis of ER, PR, AR, Ki67 and AR/ER expression showed no interaction with treatment, and lacked the ability to predict benefit of extended tamoxifen in HR+ early stage breast cancer. These results add to the growing body of evidence that BCI (H/I) is distinct in its ability to predict benefit from therapy and interrogates distinct tumor biology that is not captured by other traditional biomarkers. Citation Format: Dennis C Sgroi, Kai Treuner, Yi Zhang, Tammy Piper, Ranelle Salunga, Ikhlaaq Ahmed, Lucy Doos, Sarah Thornber, Karen J Taylor, Elena F Brachtel, Sarah Pirrie, Catherine A Schnabel, Daniel W Rea, John MS Bartlett. Correlative studies of the breast cancer index (HOXB13/IL17BR) and ER, PR, AR, AR/ER ratio and Ki67 for prediction of extended endocrine benefit: A trans-aTTom study [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr GS4-09.

INTRODUCTION Early breast cancer (EBC) which is hormone receptor positive, HER2 negative and 0-3 Lymph node positive, defines a Luminal subtype with an excellent outcome and the benefit of adjuvant chemotherapy in this subgroup of patients is increasingly questioned. Gene-expression profile tests have evolved to help stratify risk and to individualise treatment decisions. The 21-gene recurrence-score assay, “Oncotype DX” (ODX) is a genomic test used to quantify the likelihood of distant recurrence (Recurrence Score i.e. RS) and predict chemotherapy benefit. Although it is included in international guidelines as a tool to identify patients who could safely avoid adjuvant chemotherapy, its utility is restricted due to cost in countries with limited resources where it is not covered by insurance. The “NHS Predict” (NHSP) is a freely available online prognostic and predictive test based on cancer registry data. It utilises an algorithm based on clinical parameters and biomarker status to provide an estimate of survival as well as the potential benefit of third generation chemotherapy, hormonal and targeted therapy. The NHSP is based on “Clinical Risk (CR)” while the ODX is based on “Genomic Risk (GR)” assessment. The present study was undertaken to compare the results obtained by ODX with those obtained by NHSP so as to determine the concordance of the two tests and establish the utility of the NHSP in resource-constrained countries. PATIENTS AND METHODS Patients with early ER positive, HER2 negative, 0-3 nodes positive breast cancer, who underwent the ODX assay from January 2010 to March 2020 at a tertiary referral centre were entered into this study. Patients were stratified as per the ODX RS into Low (RS= 0-20 for ≤ 50 years, 0-25 for >50 years) and High (RS= >21 for ≤ 50 years and 26-100 for >50 years) risk for recurrence with absolute chemotherapy benefit (ACB) estimated to be ≤7% and >7% respectively at 9 years.Subsequently, the NHS Predict 2.2 tool (https://breast.predict.nhs.uk/tool ) was used to compute the absolute benefit of third generation chemotherapy for each of these patients. Relevant clinical data i.e. age, menopausal status, tumour size, grade, Ki-67, ER/PR/HER2 status, nodal status and method of detection was entered into the algorithm. The ACB at 10 years, as determined by NHSP, was stratified into Low risk≤ 7% and High risk >7% so as to correspond to the ACB derived from ODX. The results obtained by NHSP were compared to those of the ODX to determine the concordance between the two tests. RESULTS 104 patients were entered into the study. The mean age was 53 years (range 31-76 years), 39.4% were premenopausal and 60.5% post-menopausal. The average tumour size was 19.4mm (range 8-55mm) with 61.53% T1, 36.53% T2 and 1.92% with T3 disease.14.4% were grade 1, 66.3% grade 2, and 19.2% grade 3. 74% were lymph node negative and 14.4%, 9.6% and 0.96% had 1, 2 and 3 positive nodes respectively. On ODX evaluation, low RS was seen in 87 (83.65%), and high RS in 17(16.34%) patients. As per NHSP, 99 (95.19%) patients were classified into low CR and 5 (4.8%) into high CR. There was 82.7% (86/104) concordance between NHSP and ODX with discordance in 18 patients. Of the concordant group, 84 had low CR and GR and 2 had high CR and GR. In the discordant group 15 patients had low CR on NHSP but high GR on ODX and while 3 patients had a high CR on NHSP but ODX showed low GR. CONCLUSIONS In a group of patients with low clinical risk EBC there is high concordance between the NHSP and ODX tests. There is a subgroup of patients with discordant results in whom the ODX may identify patients who could potentially benefit with adjuvant chemotherapy. NHSP can aid decision making regarding adjuvant chemotherapy and identify patients who could potentially avoid chemotherapy in low risk early breast cancer. Citation Format: Vinay Deshmane, Anuja Raniwala. A comparison of oncotype DX and NHS predict to assess the benefit of adjuvant chemotherapy in patients with early breast cancer [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS6-38.

Invasive lobular carcinoma (ILC) accounts for 10-15% of primary breast cancer and is typically ER+ and ERBB2 non-amplified. There is preclinical evidence that somatic ERBB2 mutation may provide an alternative and tractable mechanism for upregulation of HER2 activity in tumors that do not express HER2 by current clinical criteria. Using large public datasets, we previously demonstrated that targetable ERBB2 mutations are enriched in ILC versus invasive ductal carcinoma (IDC) and are an independent prognostic factor in ILC (HR=3.7, 95% CI 1.2-11.0; p=0.021)*. We next hypothesized that a gene expression signature incorporating HER2 activity due to ERBB2 mutation and / or amplification would validate the prognostic signal we found in ILC. To derive a novel gene expression signature of HER2 activity that accounted for the effect of potentially targetable ERBB2 mutations in ERBB2 non-amplified tumors, we applied a weighted average difference method to gene expression data in cases from the METABRIC 2012 (N=1,980) and TCGA 2015 (N=817) cohorts. To compare our novel gene expression signature with an established signature of HER2 activity, we performed multivariate regression modeling of response to neratinib, a small molecule tyrosine kinase inhibitor of HER1, 2 and 4, using pharmacogenomic data accessed via the CellMinerCDB online portal. We show that our novel HER2 pathway signature score uniquely enriches for ERBB2 mutated tumors. Using a Cox regression model and stratifying gene expression scores into upper versus lower quartiles, we were able to validate the prognostic signal of ERBB2 mutations in ILC tumors (HR for 10-year OS in ILC=2.3, 95% CI 1.04-5.05; p=0.040). In contrast, no relationship was found between ERBB2 mutation status or novel HER2 pathway enrichment score and patient outcome in cases of IDC. We conclude that ERBB2 mutations that are enriched in ILC provide a robust biomarker of HER2 pathway activation and can be detected via gene expression signature. Novel clinical trials of HER2-targeted therapy in ERBB2 non-amplified primary ILC are warranted. *Reference: Kurozumi S et al, Cancer Research 2019, 80(4) suppl: SABCS 2019 Abstract P1-18-06 and medRxiv 2020.01.24.20018622 Citation Format: Mansour Alsaleem, Sasagu Kurozumi, Kartikeya Bhardwaj, Cintia Monteiro, Stacey EP Joosten, Andrew R Green, Takaaki Fujii, Ken Shirabe, Ian O Ellis, Emad A Rakha, Nigel P Mongan, David M Heery, Wilbert Zwart, Steffi Oesterreich, Simon J Johnston. Targetable ERBB2 mutation status is an independent marker of adverse prognosis in estrogen receptor positive, ERBB2 non-amplified primary lobular breast carcinoma: Validation using a novel gene signature of HER2 activation [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS6-11.

Background: Cyclin dependent 4/6 kinase inhibitors (CDK4/6i) and endocrine therapy (ET) have improved progression-free survival (PFS) and overall survival in hormone-receptor (HR)-positive metastatic breast cancer (MBC), but progression of disease is inevitable. Serum thymidine kinase-1 (TK1) is a secreted marker of proliferation that is prognostic in patients (pts) with HR-positive, HER2-negative MBC and may be predictive of ET and CDK 4/6i response. PROMISE (NCT0281902) is a prospective study enrolling women with HR-positive MBC starting palbociclib (Pb) + letrozole (L) (1st line) or Pb + fulvestrant (2nd line). We undertook a comprehensive “omic” assessment of blood, tumor, urine and the fecal microbiome in order to identify novel genomic variants and pathways associated with an early decline in TK1 (measured after 2 months) and PFS. Additionally, patient derived xenografts/organoids were generated at baseline and progression to test new therapeutic approaches to overcome resistance to CDK4/6i and ET. We report the initial association between the baseline genomic landscape and baseline TK1 levels. Methods: FFPE tumor biopsies were obtained for DNA/RNA sequencing (TempusTM) and blood samples for TK1 (Divitum® assay, Biovica) were collected pretreatment (pre-Pb) and after 2 cycles of Pb + ET (post-Pb2). Both whole-exome (exome capture) sequencing (WES) and RNA-Seq used the Integrated DNA Technologies xGen Exome Research Panel v1.0 capture kit. TK1+ disease was defined as > 200 Du/L and TK1- disease as below limit of detection up to 200 Du/L. We tested the association between genomic and transcriptomic characteristics with baseline TK1 data in pretreatment samples where both WES and RNA-seq and TK1 was available. The data were analyzed using bioinformatics pipelines for somatic and germline mutations and copy number alterations. The current analysis focuses on baseline 1st-line pre-Pb omics data in conjunction with baseline TK1 levels. The database was locked for analysis on 5/29/2020. Results: Thirty-three pts (median age: 59 yrs.) were evaluable, with paired samples for TK1 in 32. Six pts had TK1+ disease pre-Pb and post-Pb2. Twenty-two pts had TK1- disease pre-Pb and post-Pb2. Four pts had a decrease in TK1 after 2 cycles of treatment that altered the classification from TK1+ to TK1-. Both baseline RNA seq and serum TK1 (n=16) were available for 4 TK1+ and 12 TK1- pts. In this group, 476 genes were differentially regulated (398 upregulated; 78 downregulated). Pathway analysis demonstrated enrichment in complement and coagulation cascade pathway, PPAR signaling pathway, and metabolism-related pathways related to up-regulation of CYP and UGT gene families. Further testing for the association of WES data with baseline TK1+ (n=8) and TK1- (n=16) disease demonstrated somatic copy number variations on chromosomes 6, 11, 12 and 15. CDK4 copy number gains were observed in 3/8 TK1+ pts and 0/16 TK1- pts. We also observed that somatic mutations (LOH, copy number and/or SNV/INDELs) were more prevalent in the TK1+ compared to the TK1- pre-Pb group in several cancer-associated genes (FAS [p=0.06] PTEN, PIK3CB, NAB2, SOX9 and FAT1 [p=0.08], TP53, and MAP2K4 [p=0.22]). Conversely, we also noted that 6/7 pts with GATA3 mutations had TK1- disease (p=0.23). Conclusions: Using a comprehensive “omics” approach, our data suggest that a secreted biomarker of proliferation (TK1) obtained prior to initiating CDK4/6i and ET for the first line treatment of HR+ MBC is associated with established and novel genes and pathways associated with prognosis of pts receiving ET and CDK 4/6i. Analysis of on-treatment (after 2 cycles) tumor RNA seq and its association with change in TK1 as well as data from the 2nd-line cohort will be presented at the meeting. Citation Format: Ciara C O Sullivan, Krishna R Kalari, Vera J Suman, Peter T Vedell, Ann Moyer, Abraham D Eyman Casey, Jason Sinnwell, Xiaojia Tang, Kevin Thompson, Alvaro Moreno-Aspitia, Donald W Northfelt, Minetta C Liu, Tufia C Haddad, Saranya Chumsri, Prema Peethambaram, Kathryn J Ruddy, Karthik V Giridhar, Roberto A Leon-Ferre, Adrian Nordmark, Mattias Bergqvist, Brendan P McMenomy, Richard M Weinshilboum, Liewei Wang, Matthew P Goetz. Novel genomic variants and pathways associated with baseline serum thymidine kinase 1 levels in HR-positive HER2-negative MBC patients commencing palbociclib and letrozole [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS5-24.

Background and Rationale: Biomarker evaluation on breast tumor tissue is an important component of clinical research. There is a concern that tissues collected at different times (pre- vs intra-op) and with different techniques (core biopsy vs. surgical excision) produce different results that may confound comparisons of patient samples. Our objective is to address the difference between tumor tissue from the same patient collected pre- vs intra-op and by core biopsy (Core Bx) vs surgical excision, with an emphasis on ER/PgR/HER2/Ki-67 and biomarkers related to insulin metabolism. Design: Following a protocol approved by the IRB, patients with Core Bx proven invasive breast cancer (BC) >1.5 cm in size by imaging underwent a peri-operative Core Bx followed by surgical excision. No neoadjuvant therapy was administered. Formalin-fixed paraffin-embedded tumor sections of the diagnostic and peri-operative Core Bx and surgical excision were immunohistochemically (IHC) stained for ER, PgR, HER2, Ki-67, insulin receptor (IR), phospho-AKT (pAKT), and phospho-AMPK (pAMPK) using standardized protocols on the same platforms. A pathologist (MCC) scored all sections in blinded order to yield H-Scores (which combine percentage of stain-positive cells and staining intensity). The level of agreement for each assay between specimens was assessed by using pairwise models based on normal theory. Results: 18 women (mean age = 66.8 years, 16/18 postmenopausal) provided all specimens. Mean (± SD) invasive tumor size was 2.7 (± 1.2) cm. 1 BC was Nottingham Grade 1 (6%), 9 Grade 2 (50%), and 8 Grade 3 (44%). An in situ component was present in 4/18 (22%) cases (all non-extensive). Lymphovascular invasion was present in 6/18 (33%) cases. 10/18 were node-negative (N0, 56%), 2/18 had isolated tumor cells (pN0[i+], 22%), 5/18 had 1 to 3 nodes (pN1, 28%), 1/18 had >9 nodes involved (pN3, 6%).There was high concordance between diagnostic Core Bx, peri-op Core Bx, and excisions for the standard prognostic markers ER, PgR, and HER2. ER was concordant in all samples including 16/18 (89%) ER-positive and 2/19 (11%) ER-negative BCs. PgR was concordant in 16/18 (89%) of cases. In the 2 discordant cases, PgR was negative in 2 of 3 tissue samples with low expression in the 3rd sample (1 diagnostic Core Bx and 1 excision). HER2 was concordant in all samples in 16/18 (89%) HER2-negative cases and in 2/18 (11%) HER2-positive cases including 1/18 cases (6%) positive by IHC, and 1/18 cases (6%) positive by ISH. Table 1 summarizes the agreement between samples for Ki-67, IR, pAKT, and pAMPK. Ki-67 scores were statistically similar between diagnostic and peri-op Core Bx and excision samples. The IR, pAKT, and pAMPK H-Scores were statistically similar between diagnostic and peri-op Core Bx, but significantly different between Core Bxs versus excisional specimens. There was a systematic tendency towards lower IHC H-Scores in the excisional specimen for IR, pAKT, and pAMPK. Conclusion: Tissue from surgical excisions are susceptible to reduced IHC staining for metabolic markers such as IR, and phosphorylated kinases, when compared to core biopsies. When evaluating non-standard biomarkers for research, core biopsies should be used when possible. Funding: This study was funded by the Hold'Em For Life Charity and the Breast Cancer Research Foundation Table 1: Concordance Between IHC Scores for Ki-67, Insulin Receptor, pAKT, and pAMPKScore: Mean ± Standard DeviationPeri-Operative Core Bx versus Diagnostic Core BxExcision versus Diagnostic Core BxDiagnostic Core BxPeri-Operative Core BxExcisionDifferenceP-valueDifferenceP-valueKi-67 (% Positive)33.5 ± 28.235.9 ± 31.034.2 ± 29.1Mean ± SD: 2.4 ± 5.60.09Mean ± SD: 0.6 ± 6.30.68Range: -7.7 to 12.3Range: -11.5 to 12.2Insulin Receptor (H-Score**)136 ± 102124 ± 9986 ± 78Mean ± SD: -13 ± 530.32Mean ± SD: -51 ± 540.001*Range: -215 to 40Range: -150 to 5Phospho-AKT (H-Score**)102 ± 70100 ± 8641 ± 41Mean ± SD: -1 ± 750.95Mean ± SD: -61 ± 520.0001*Range: -170 to 170Range: -144 to 0Phospho-AMPK (H-Score**)201 ± 79185 ± 78157 ± 86Mean ± SD: -15 ± 760.41Mean ± SD: -44 ± 690.016*Range: -225 to 120Range: -215 to 70* Statistically significant difference in pairwise values based on t-test** H-Score is the percentage of stain-positive cells multiplied by the average intensity score (0 = absent, 1 = faint, 2 = moderate, 3 = strong) Citation Format: Martin C. Chang, Marguerite Ennis, Jaime Escallon, Karina Bukhanov, Ryan JO Dowling, Vuk Stambolic, Pamela J Goodwin. Concordance of immunohistochemical assays between peri-operative and post-operative breast tumor specimens: A prospective observational study of 18 cases [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS18-25.

Recently MHC-II protein expression levels in breast cancer tumor cells were shown to predict response to the PD-1 inhibitor pembrolizumab on a neoadjuvant clinical trial for ER+ and TNBC (ISPY-2). To better understand this biomarker and avoid post-market issues, we aimed to investigate possible pitfalls of MHC-II/HLA-DR quantification, prevalence of HLA-DR expression in breast cancer (BC) cells and its association with clinical data in a cohort of 372 BC patents. We assessed tumor and non-tumor HLA-DR expression by multiplex fluorescence immunohistochemistry on 8 TMAs containing 372 surgical BC samples, comprised of 239 ER+ and 123 TNBCs. Patient characteristics are shown in table 1. PanCK was used to define the tumor area and HLA-DR quantification was performed by training an object classifier on QuPath. Upon visual examination, 23 samples showed patchy panCK expression. We excluded normal breast epithelium and in situ components, which showed varying degrees of HLA-DR. 44.4% of breast cancers expressed HLA-DR, with 8.3% showing high tumor expression (HLA-DRHI; ≥20% of tumor cells). TNBC showed higher prevalence of HLA-DRHI cases (35% HLADR+, 14.6% HLA-DRHI, mean tumor HLA-DR expression (% of tumor cells) 10.3%, range 0-84%) than ER+ BC (51% HLADR+, 5.4% HLA-DRHI, mean 7.4%, range 0-66.5%). In some cases, HLA-DR and panCK expression were mutually exclusive. In other cases the lack of panCK expression highlighted the presence of stromal cells within the tumor area that were not evident on H&E. Tumoral MHC-II expression correlated with the presence of TILs for ER+ BC (r=0.18, p=0.0158) and TNBC (r=0.20, p= 0.0450), specifically CD3 (r=0.49 p Table 1: Patient characteristicsTNBCER+Total number of cases239123Median age48,563Treatmenttreatment naïve26%45%neoadjuvant chemotherapy74%4%pre-surgical letrozole-51%Positive lymph node58%14%StageI27%82%II20%27%III67%8%IV0%1% Citation Format: Paula Ines Gonzalez Ericsson, Justin M Balko, Violeta Sanchez, Melinda Sanders. Evaluation of tumor-specific MHC-II expression as a biomarker [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS4-19.

Purpose: To describe the pathological response rates (according to the MD Anderson RCB score) achieved by patients with 35 years old or younger, with HER2 negative breast cancer, who received neoadjuvant chemotherapy (NAC) and long term outcomes according to the response. Methods: A retrospective review of 526 medical records of women aged less than or equal to 35 years at diagnosis between 2009 and 2014 at INEN in Lima-Peru, who received NAC for invasive breast cancer, was performed. HER2 positive patients were excluded because of a lack of access to neoadjuvant anti-HER2 therapy during this period in our hospital. Descriptive statistics were used to describe baseline characteristics. OS was calculated from the date of diagnosis to death or last follow-up and presented as Kaplan-Meier curves. The comparison among patients who achieved pCR (RCB-0) and those with residual disease, was based on a log-rank test. Results: Seventy women (One with synchronous bilateral disease) were selected according to selection criteria, median age: 31.7 years old (IQR: 30.0-33.2). Hormone receptor-positive: 38 (53.5%). Clinical stage: IIA: 13(18.3%), IIB: 8(11.3%), IIIA: 25(35.2%), IIIB: 21 (29.6%) and IIIC: 4 (5.6%). Type of NAC: Anthracyclines 6 (8.6%), Anthracyclines and taxanes 55 (78.6%), Platin 5 (7.1%), Taxanes 4 (5.7%). Type of surgery: Conservative 16 (22.5%), radical 50 (70.4%), toilet 5 (7.0%). Type of pathological response RCB-0: 9(12.7%), RCB-I: 2(2.8%), RCB-II:29(40.8%), RCB-III: 24(33.8%), non-pCR: 7(9.9%). With a median follow-up of 46.5m, 5y OS: 100/100/79.3/45.8% according to RCB-0/I/II/III respectively (p=0.004). Conclusions: "Very young" women reach fewer pCR rates than reported for "young" and older women. Even so, RCB-0 can be considered as a biomarker of excellent prognosis and could be used as a surrogate biomarker to evaluate the efficacy of neoadjuvant therapy even in very young women who receive NAC. Citation Format: Fernando Valencia, Jorge Laboriano, Luis A. Chirinos, Henry L. Gomez. Pathological response after neoadjuvant chemotherapy and long-term outcomes among very young women with HER2 negative breast cancer [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS7-77.

Background:Many studies have explored the predictive factors for the efficacy of neoadjuvant therapy in HER2 + breast cancer. However, no reliable and widely used biomarker was found till now except several clinicalpathologic factors including HER2. HER2 oncogene can regulate the recruitment and activation of tumor-infiltrating immune cells (TILs) and trastuzumab therapeutic effects by inducing programmed death ligand 1 (PD-L1), suggesting that TILs and the expression of PD-L1 may be associated with the efficacy of trastuzumab. Several studies have verified certain predictive value of TILs and PD-L1 in HER2 + breast cancer patients, but controversy remains. Besides, most of them focus on the expression of PD-L1 or TILs before neoadjuvant therapy, but not the change in the paired tissues before and after neoadjuvant therapy. This study aimed to investigate the change of TILs and the expression of PD-L1 in the paired tissues before and after neoadjuvant therapy and the correlation to the efficacy of neoadjuvant trastuzumab plus chemotherapy and disease-free survival (DFS) in HER2+ breast cancer patients.Methods:HER2+ breast cancer cases receiving neoadjuvant therapy (n=115) were retrospectively collected between July 2013 and November 2018. The expression of PD-L1 was detected by immunohistochemistry using SP142 antibody and the percentage of positive membranous staining in tumor cells (TC-PD-L1) and TILs (TILs-PD-L1) was scored respectively. TIL percentile score for full sections were assessed by two pathologists independently. Results:In our study, 87 patients receiving neoadjuvant chemotherapy alone, and 68 patients receiving neoadjuvant trastuzumab plus chemotherapy. Among them, 39 patients achieved pCR and 116 patients were non-pCR. Univariate analysis confirmed that the pCR was positively correlated with high TILs and TILs-PD-L1 expression before neoadjuvant therapy (P< 0.05). Multivariate logistic regression analysis confirmed that pre-treatment TILs-PD-L1 expression but no other clinicalpathologic factors,was independent predictor of pCR in neoadjuvant therapy (P Citation Format: Huihui Li, Mao Shang, Yajing Chi, Sha Yin. The predictive value of tumor-infiltrating lymphocytes and PD-L1 expression on the efficacy of neoadjuvant trastuzumab plus chemotherapy in HER2-positive breast cancer [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS5-27.

Accurate diagnostic biomarker testing, including estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (Her2) expression profile, is crucial to appropriate management decisions in breast cancer. Receptor status testing is typically performed on the initial core needle biopsy (CNB) and surgical specimen (SS). The rate of concordance between CNB and SS is unclear as is the impact this has on clinical decision making. The current guidelines on retesting are vague, which results in individual institutions and providers determining retesting policies. Several studies worldwide have assessed the concordance of receptor testing with mixed conclusions. We aim to determine concordance between CNB and SS, and whether this leads to clinically relevant management changes. A retrospective analysis was performed on patients with invasive breast cancer with available CNB and SS pathology at our institution between January 2010- May 2020. Patients who were treated with primary surgical resection and neoadjuvant chemotherapy with residual disease were included. Concordance rates between CNB and SS were assessed for ER, PR, and Her2 IHC/ FISH amplification. ER and PR status were defined per NCCN guidelines as “positive,” “low-positive (ER)” or “negative.” Major discrepancy was defined as a change in label of “positive” or “negative”. Minor discrepancy was defined as change >10% without a “positive” or “negative” label change. Major discordance in Her2 was defined as change in label of “positive” or “negative” based on IHC or FISH amplification results. A minor discrepancy was a change in Her2 IHC (0-3) or FISH amplification without change in label of “positive” or “negative.” The clinical impact of discordant results was determined by investigator review. 748 patients met the eligibility criteria, and 64 of these patients received neoadjuvant therapy. For ER, there was 90.6% concordance, 2.2% major discordance, and 7.8% minor discordance between CNB and SS. For PR, there was 59.64% concordance, 11.9% major discordance, and 28.46% minor discordance. For Her2, there was 54.7% concordance, 1.6% major discordance, and 43.8% minor discordance. Of major discordance, ER (43.8%) led to the most change in management compared to Her2 and PR (12.5% and 2.2%, respectively). Retesting Her2 on SS did not change management when initial CNB was Her2 positive. For major discrepancies, patient demographics, tumor characteristics, treatment course, recurrence, and survival were reviewed. Although discordance was more common in PR and Her2 than ER biomarker profiles, major discordance leading to treatment changes were more common in ER and Her2. Retesting ER and Her2 on CNB and SS may be more clinically beneficial than retesting PR. Guidelines for retesting receptor profiles on CNB and SS are needed to best guide patient care management decisions that maximize clinical benefits while minimizing healthcare costs. Table: Concordance of ER, PR, and HER2 Expression ProfileConcordanceMajor DiscrepancyMinor Discrepancy# Patients with Change in Treatment with Major DiscrepanciesER90.6%2.2%7.8%7/16 (43.8%)PR59.6%11.9%28.5%2/90 (2.2%)Her254.7%1.6%43.8%5/12 (12.5%) Citation Format: Jessica Anne Slostad, Nicole Yun, Aimee Schad, Surbhi Warrior, Ruta Rao. Concordance of breast cancer biomarker testing in core-needle biopsy and surgical specimens: A single institution experience [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS2-34.

Background: In the phase 2 KEYNOTE-086 study (NCT02447003), pembrolizumab monotherapy had durable antitumor activity in a subset of patients with previously treated mTNBC (cohort A; n = 170) and in patients with previously untreated PD-L1-positive mTNBC (cohort B; n = 84). In this exploratory analysis of KEYNOTE-086, we evaluated the association between several biomarkers and response to pembrolizumab. Methods: Cohort A enrolled patients regardless of PD-L1 expression who had documented disease progression following ≥1 systemic therapy for metastatic disease. Cohort B enrolled patients with PD-L1-positive (combined positive score [CPS] ≥1) tumors who had not received prior systemic therapy for metastatic disease. Immunohistochemistry was used to measure PD-L1 CPS and CD8 density; H&E staining for percentage of stromal tumor infiltrating lymphocytes (sTILs); RNA-sequencing for 18-gene T-cell-inflamed gene expression profile (GEP), angiogenesis, and glycolysis signatures; and whole exome sequencing (paired tumor and germline) for TMB (TMB-H defined as ≥175 mut/exome), HRD-LOH (DNA damage), Signature 3, and APOBEC. Biomarkers were analyzed as continuous variables in cohorts A and B combined and individually. Area under the receiver operating characteristic curve was estimated between each biomarker and overall response rate (ORR). Wald test P-values were calculated using logistic regression adjusted for cohort and Eastern Cooperative Oncology Group performance status. Germline and somatic BRCA1/2 mutations were pooled; one-sided P-value was calculated using Fisher’s exact test. Spearman’s correlation was used for correlations. Results: Biomarker data were available in the following number of patients: 253 (99.6%; PD-L1), 204 (80.3%; CD8), 187 (73.6%; GEP), 171 (67.3%; TMB/HRD), 228 (89.8%; sTILs), 163 (64.2%; Signature 3/APOBEC), and 132 (52.0%; angiogenesis/glycolysis). When data from cohorts A and B were combined, PD-L1 CPS (median 2; IQR 0-10), CD8 (median 159; IQR 62-319), GEP (median -0.34; IQR -0.57 to -0.11), TMB (median 82 mut/exome; IQR 50-139), and sTILs (median 5; IQR 2-20) were significantly associated with ORR (Table). There were moderate correlations between PD-L1 and GEP (r = 0.532), PD-L1 and sTILs (r = 0.451), and GEP and sTILs (r = 0.490). No correlation was observed between TMB and PD-L1 (r = 0.038), GEP (r = -0.035), and sTILs (r = -0.031). When cohorts were combined, TMB was significantly associated with ORR, PFS, and OS after adjustment for PD-L1, GEP, CD8, or sTILs. HRD-LOH score, Signature 3, and APOBEC were not significantly associated with ORR (Table); P for BRCA1/2 was 0.2385. The angiogenesis signature was associated with lack of response while the glycolysis signature was associated with response to pembrolizumab (Table). Conclusions: In this exploratory biomarker analysis from KEYNOTE-086, higher levels of PD-L1, GEP, TMB, CD8 IHC, sTILs, and the glycolysis signature were associated with increased response to pembrolizumab monotherapy. These findings may help identify patients with mTNBC who are most likely to respond to pembrolizumab. Table. Association of Biomarkers as Continuous Variables With Pembrolizumab Objective ResponseBiomarkerCombined Cohorts AUCCombined Cohorts P*Combined Cohorts Multitest corrected PCohort A AUCCohort B AUCPD-L10.6740.040-0.5440.654GEP0.7480.003-0.8370.561TMB0.6270.007-0.5480.710CD8 IHC0.760.000020.000120.850.68sTILs0.6710.012-0.6320.641HRD0.3940.874-0.5220.316Signature 30.6830.072-0.6970.736APOBEC0.5280.537-0.6740.623Angiogenesis0.6770.0090.0450.6610.731Glycolysis0.6120.0090.0360.8590.459AUC, area under the curve.*One-sided P-values are shown for all biomarkers except for Signature 3 and APOBEC, for which 2-sided P-values are shown. Citation Format: Sherene Loi, Peter Schmid, Javier Cortes, David W. Cescon, Eric P. Winer, Deborah L. Toppmeyer, Hope S. Rugo, Michelino De Laurentiis, Rita Nanda, Hiroji Iwata, Ahmad Awada, Antoinette R. Tan, Roberto Salgado, Vassiliki Karantza, Petar Jelinic, Anran Wang, Lingkang Huang, Razvan Cristescu, Lakshman Annamalai, Jennifer Yearley, Jennifer Yearley, Sylvia Adams. Association between biomarkers and response to pembrolizumab in patients with metastatic triple-negative breast cancer (mTNBC): Exploratory analysis from KEYNOTE-086 [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PD14-07.

Background: Obesity is associated both with a higher risk of developing breast cancer, particularly in postmenopausal women, and with worse disease outcome for women of all ages. The molecular mechanisms that link obesity-breast cancer are complex and unclear. We hypothesized that overexpression of Aurora A was associated with poor survival in obesity breast cancer and related axis mechanism were involved.Methods: Fifty hundred and eighteen primary breast cancer specimens were collected from First Affiliated Hospital of China Medical University between January 2011 and November 2016. Our independent variable was BMI at baseline, categorized as overweight (BMI ≥ 25 kg/m2, as Obesity cohort), and normal (18.5 ≤ BMI < 25 kg/m2, as Non-obesity cohort). The Immunohistochemical (IHC) staining was performed with Aurora A, Survivin, MMP11, Cyclin B1 and Cathepsin L. Kaplan-Meier curve was used to analyze overall survival in our cohorts and TCGA-BRCA data (GSE3494). Log-rank test was used to calculate P value. PPI network analysis and MCODE model was used to analysis Aurora- altered signal pathway from GSE78958.Results: We validated Aurora A over expression using an independent cohort with a total of 517 breast carcinoma tissues (319 Non-Obesity patients and 198 Obesity Patients). The expression level of Aurora A-positive was significant higher in obesity breast carcinoma compared with non-obesity cancer carcinoma (χ2=9.79, P=0.002). In the immunohistochemistry assay, the results confirmed that the expression level of Aurora A-positive was significantly associated with hormone receptor status (68.4% vs 77.9%, P=0.015) and HER2 status (28.7% vs 17.9%, P=0.003). Aurora A-positive tumors had larger tumor size, though not statistically significant (78.1% vs. 71.1%, P=0.070). High Aurora A expression was remarkably and significantly associated with OS (8-year OS ratio: 69.5% vs 81.1%, OR=1.76, 95% CI:1.03~3.02, P=0.041) in Obesity Cohort. Interesting, higher expression of Aurora A was not associated with a shorter overall survival time among the Non-obesity breast cancer (8-year OS ratio:81.4% vs 85.8%, OR=1.40, 95% CI:0.79~2.45, P=0.229). As for RFS, the expression levels of Aurora A expression genes have no significance with RFS statistically in Non-Obesity patients. The in Aurora A low expression group was in high Aurora A expression in Non-Obesity Cohort ( 5-year RFS ratio:82.2% vs 83.2%, OR=1.29, 95% CI: 0.77~2.16, P=0.307). While 5-year RFS ratio in high Aurora A expression group of Aurora A were 71.3% was a little shorter than the 5-year RFS ratio in low Aurora A expression group (OR=1.58, 95% CI:0.96~2.57, P=0.072). Aurora A and Lymph node metastases were significantly poor prognostic factors for OS, and borderline significance was noted for high BMI. Kaplan-Meier survival analysis from TCGA database confirmed that the high Aurora A expression group had worse prognosis (HR=1.47, 95%CI:1.14-1.90, P=0.003). The KEGG pathway enrichment results were consistent with GO biological process term analysis, in which CCNB1 was enriched for upregulated Aurora A. In IHC correlation analyasis, Aurora A level on tumor cytoplasm had broad connections with Cyclin B1 (correlation coefficient = 0.227, P=0.001).Conclusions: Our finding demonstrate here for the first time that high expression of Aurora A was notably correlated with early recurrence and poor overall survival in obesity patients with early breast cancer. The Aurora A-Cyclin B1 axis could be a potential promising therapeutic target for cancer intervention and therapy. Keywords: Early breast cancer, obesity, Aurora A, Overall survival, Cyclin B1 Citation Format: Junnnan Xu, Tao Sun. Identification of Aurora kinase A as a biomarker for prognosis in obesity patients with early breast cancer [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS6-52.

Cellular dedifferentiation is a key mechanism driving cancer progression. Acquisition of mesenchymal features has been associated with drug resistance, poor prognosis, and disease relapse in many tumor types. Therefore, successful targeting of tumors harboring these characteristics is a priority in oncology practice. The SWItch/Sucrose non-fermentable (SWI/SNF) chromatin remodeling complex has also emerged as a critical player in tumor progression, leading to the identification of several SWI/SNF complex genes as potential disease biomarkers and targets of anticancer therapies. AT-rich interaction domain-containing protein 1A (ARID1A) is a component of SWI/SNF, and mutations in ARID1A represent one of the most frequent molecular alterations in human cancers. ARID1A mutations occur in approximately 10% of pancreatic ductal adenocarcinomas (PDAC), but whether these mutations confer a therapeutic opportunity remains unclear. Here, we demonstrate that loss of ARID1A promotes an epithelial–mesenchymal transition (EMT) phenotype and sensitizes PDAC cells to a clinical inhibitor of HSP90, NVP-AUY922, both in vitro and in vivo. Although loss of ARID1A alone did not significantly affect proliferative potential or rate of apoptosis, ARID1A-deficient cells were sensitized to HSP90 inhibition, potentially by promoting the degradation of intermediate filaments driving EMT, resulting in cell death. Our results describe a mechanistic link between ARID1A defects and a quasi-mesenchymal phenotype, suggesting that deleterious mutations in ARID1A associated with protein loss exhibit potential as a biomarker for patients with PDAC who may benefit by HSP90-targeting drugs treatment. Significance: This study identifies ARID1A loss as a promising biomarker for the identification of PDAC tumors that are potentially responsive to treatment with proteotoxic agents.

Noninvasive early indicators of treatment response are crucial to the successful delivery of precision medicine in children with cancer. Neuroblastoma is a common solid tumor of young children that arises from anomalies in neural crest development. Therapeutic approaches aiming to destabilize MYCN protein, such as small-molecule inhibitors of Aurora A and mTOR, are currently being evaluated in early phase clinical trials in children with high-risk MYCN-driven disease, with limited ability to evaluate conventional pharmacodynamic biomarkers of response. T1 mapping is an MRI scan that measures the proton spin-lattice relaxation time T1. Using a multiparametric MRI-pathologic cross-correlative approach and computational pathology methodologies including a machine learning–based algorithm for the automatic detection and classification of neuroblasts, we show here that T1 mapping is sensitive to the rich histopathologic heterogeneity of neuroblastoma in the Th-MYCN transgenic model. Regions with high native T1 corresponded to regions dense in proliferative undifferentiated neuroblasts, whereas regions characterized by low T1 were rich in apoptotic or differentiating neuroblasts. Reductions in tumor-native T1 represented a sensitive biomarker of response to treatment-induced apoptosis with two MYCN-targeted small-molecule inhibitors, Aurora A kinase inhibitor alisertib (MLN8237) and mTOR inhibitor vistusertib (AZD2014). Overall, we demonstrate the potential of T1 mapping, a scan readily available on most clinical MRI scanners, to assess response to therapy and guide clinical trials for children with neuroblastoma. The study reinforces the potential role of MRI-based functional imaging in delivering precision medicine to children with neuroblastoma. Significance: This study shows that MRI-based functional imaging can detect apoptotic responses to MYCN-targeted small-molecule inhibitors in a genetically engineered murine model of MYCN-driven neuroblastoma.

Neuroblastoma is a childhood cancer with heterogeneous clinical outcomes. To comprehensively assess the impact of telomere maintenance mechanism (TMM) on clinical outcomes in high-risk neuroblastoma, we integrated the C-circle assay [a marker for alternative lengthening of telomeres (ALT)], TERT mRNA expression by RNA-sequencing, whole-genome/exome sequencing, and clinical covariates in 134 neuroblastoma patient samples at diagnosis. In addition, we assessed TMM in neuroblastoma cell lines (n = 104) and patient-derived xenografts (n = 28). ALT was identified in 23.4% of high-risk neuroblastoma tumors and genomic alterations in ATRX were detected in 60% of ALT tumors; 40% of ALT tumors lacked genomic alterations in known ALT-associated genes. Patients with high-risk neuroblastoma were classified into three subgroups (TERT-high, ALT+, and TERT-low/non-ALT) based on presence of C-circles and TERT mRNA expression (above or below median TERT expression). Event-free survival was similar among TERT-high, ALT+, or TERT-low/non-ALT patients. However, overall survival (OS) for TERT-low/non-ALT patients was significantly higher relative to TERT-high or ALT patients (log-rank test; P < 0.01) independent of current clinical and molecular prognostic markers. Consistent with the observed higher OS in patients with TERT-low/non-ALT tumors, continuous shortening of telomeres and decreasing viability occurred in low TERT–expressing, non-ALT patient-derived high-risk neuroblastoma cell lines. These findings demonstrate that assaying TMM with TERT mRNA expression and C-circles provides precise stratification of high-risk neuroblastoma into three subgroups with substantially different OS: a previously undescribed TERT-low/non-ALT cohort with superior OS (even after relapse) and two cohorts of patients with poor survival that have distinct molecular therapeutic targets. Significance: These findings assess telomere maintenance mechanisms with TERT mRNA and the ALT DNA biomarker C-circles to stratify neuroblastoma into three groups, with distinct overall survival independent of currently used clinical risk classifiers.

Purpose: Diagnosis of bone metastasis (BM) in breast cancer relies on structural changes of bone. We investigated whether circulating osteocalcin-positivecells (cOC) could detect the incipient and/or progressive BM earlier than image-based diagnostics. Experimental design: Metastatic breast cancer patients with or without bone metastasis (designated BM+ or BM-) were enrolled and cOC were quantified at baseline by flow cytometry. The progression of BM was evaluated after 18 months. Murine BM models were established by intra-tibial or -cardiac injection of MDA-MB-231 or 4T1 breast cancer cells. Results: In the baseline analysis, cOC was significantly higher in BM+ than BM- group. In the 18-month follow-up study of BM+ patients (n=63), baseline cOC was significantly associated with BM disease progression. Among BM- patients (n=33), three patients developed new bone metastasis within 18 months and were found to have had higher cOC at baseline. The patients with higher cOC showed shortened BM progression-free survival, compared with those with lower cOC (cutoff=0.045%; P Conclusions: cOC is increased in the early phase of breast cancer BM and predict the disease progression. cOC can be a novel biomarker for early diagnosis and progression of BM. Citation Format: Sun Wook W Cho, Kyung-Hun Lee, Tae-Yong Kim, Serk In Park, Seock-Ah Im. Circulating osteocalcin-positive cells as a novel diagnostic biomarker for bone metastasis in breast cancer patients [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P2-19-02.

Background: Approximately 10%-20% of breast cancers (BCs) are classified as triple-negative breast cancer (TNBC). TNBC lacks targeted treatment options and has a poor prognosis. Bintrafusp alfa (M7824) is an innovative first-in-class bifunctional fusion protein composed of the extracellular domain of the TGF-βRII receptor (a TGF-β “trap”) fused to a human IgG1 mAb blocking PD-L1. We report interim results in patients with advanced TNBC treated with bintrafusp alfa in an expansion cohort of an ongoing, open-label, phase 1 trial (NCT02517398). Methods: Eligible patients had confirmed TNBC (ER: IHC Results: As of August 24, 2018, 33 patients with heavily pretreated TNBC (54.5% of patients had ≥4 prior regimens) were enrolled in this study of bintrafusp alfa. Patients received a median of 3 doses of bintrafusp alfa (range, 1-24). Confirmed responses as assessed by IRC and investigator occurred in 3 patients (1 complete response and 2 partial responses as assessed by IRC; ORR, 9.1% [95% CI, 1.9%-24.3%]); disease control was achieved in a total of 5 patients (15.2% [95% CI, 5.1%-31.9%]) by both IRC and investigator read. The median progression-free survival per IRC was 1.3 months (95% CI, 1.2-1.4 months), and the median overall survival was 7.8 months (95% CI, 2.1-12.8 months). Importantly, a biomarker related to TGF-β biology was identified via RNAseq analysis of tumor samples with 32.0-fold higher expression in patients who experienced disease control (response or stable disease) compared with those who had progressive disease (biomarker results for this expansion cohort are presented in a separate abstract). PD-L1 expression was not associated with response to treatment. In the entire cohort, 6 patients (18.2%) experienced ≥1 grade 3 treatment-related adverse event (TRAE), including anemia (n=3), asthenia, decreased appetite, generalized rash, hypophysitis, and increased transaminases (all n=1). No grade 4 TRAEs were reported. Three patients discontinued treatment due to TRAEs, including 1 death that was assessed by the investigator as related to treatment. Dyspnea, hemolysis, and thrombocytopenia were reported as grade 5 TRAEs in this patient, who had extensive disease at trial entry and was noted to have multiple pulmonary emboli, progressive disease, and expanding pleural effusion after 3 doses. No autoantibodies mediating hemolysis or thrombocytopenia were identified on workup. Conclusions: In summary, bintrafusp alfa was well tolerated, with a safety profile consistent with expectations in this heavily pretreated, advanced TNBC cohort. We identified high expression of a potential predictive biomarker of response in patients with advanced TNBC treated with bintrafusp alfa. Detailed biomarker data are presented in the separate abstract and warrant further investigation of bintrafusp alfa in TNBC. Citation Format: Alexander Spira, Ahmad Awada, Nicolas Isambert, David Lorente Estellés, John Nemunaitis, Nicolas Penel, Laureen S Ojalvo, Christoph Helwig, Christian Borel. Bintrafusp alfa (M7824), a bifunctional fusion protein targeting transforming growth factor-β and programmed death ligand 1, in advanced triple-negative breast cancer: Preliminary results from a phase 1 cohort [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P3-09-06.

Introduction: A number of different drugs have been developed to target the PI3 kinase/AKT/mTOR pathway. Most randomized trials in breast cancer have been performed in patients unselected for pathway activation although aberrations of components of this pathway are common. Here we explore the predictive value of PIK3CA and AKT mutations in randomized trials of drugs targeting this signaling route. Methods: We searched PubMed for randomized trials of drugs targeting the PI3 kinase/AKT/mTOR pathway in metastatic breast cancer. The search was supplemented by a manual search of recent ASCO and ESMO meetings. Eligible studies needed to report efficacy data based on presence or absence of PIK3CA or AKT status measured either in tumor specimens or in circulating DNA. Hazard ratios (HR) for progression-free survival (PFS) in PIK3CA or AKT mutant and wild-type groups were extracted and pooled in a meta-analysis using generic inverse variance and random effects modeling. Sensitivity analyses were performed based on the mechanism of action of the experimental drug. Results: Of the 15 studies identified, 10 studies comprising of 3615 patients were eligible for analysis. Studies included pan-PI3kinase inhibitors (n=5), alpha-specific PI3kinase inhibitors (n=1), AKT inhibitors (n=2) and mTOR inhibitors (n=2). Approximately 38% of patients had either PIK3CA or AKT mutations. The addition of drugs targeting the PI3 kinase/AKT/mTOR pathway was associated with a significantly longer improvement in PFS in patients with PIK3CA or AKT mutations than in those with wild-type status (HR 0.61, 95% CI 0.50-0.75 versus HR 0.83, 95% CI 0.67-1.02, p for difference = 0.04). Similar results were observed in sensitivity analyses (see Table). Conclusion: Benefit from drugs targeting the PI3 kinase/AKT/mTOR pathway in breast cancer appears to occur exclusively in those with PIK3CA or AKT mutations. Future trials of these drugs should be designed only in biomarker selected patients. TableGroupPIK3CA or AKT mutationPIK3CA and AKT wildtypep for differenceHR95% CIHR95% CIExcluding AKT inhibitors0.640.52-0.790.810.63-1.030.16Excluding pan-PI3 kinase inhibitors0.580.47-0.710.710.51-1.150.24Excluding AKT and alpha-specific PI3 kinase inhibitors0.690.48-1.000.980.83-1.150.09Excluding mTOR inhibitors0.620.49-0.800.920.80-1.060.008 Citation Format: Fahad A Almugbel, Ramy Saleh, Alberto Ocana, Atanasio Pandiella, Eitan Amir. PIK3CA status as a predictor of benefit from drugs targeting the PI3K/AKT/mTOR pathway: A meta-analysis of randomized trials [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P4-10-11.

Introduction: The PREDICT DNA trial is the first prospective, multi-center study aimed at validating cell-free plasma derived circulating tumor DNA (ctDNA) as a biomarker for treatment response and recurrence in early stage, triple-negative or HER2-positive (any hormone receptor status) breast cancer. Its primary aim is to determine the negative predictive value (NPV) of the absence of ctDNA after neoadjuvant therapy (NAT) for the achievement of pathologic complete response (pCR). This study has met its accrual goals and results of the overall trial are anticipated within the next year. The initial PREDICT DNA study design stipulated that tumor specific mutations (TSMs) to be tracked in blood would be identified by next gen sequencing (NGS) of tumor biopsy tissue. A disadvantage of this design is the dependence on adequate biopsy tissue. Recently, the advent of Safe-SeqS technology has enabled robust detection of rare variants using NGS with a sensitivity of approximately 0.05% mutant allele fraction. We employed these new NGS methods to pilot a novel tissue-independent approach to ctDNA detection and monitoring. Objective: The primary objective of this pilot study was to determine whether ultrasensitive NGS using a targeted cancer mutation panel can identify TSMs in ctDNA of early-stage breast cancer patients without the use of biopsy tissue. Methods: The PREDICT DNA trial enrolled 228 women from 22 sites with stage II/III breast cancer for whom standard neoadjuvant therapy was planned. Of these, 58 patients had matched pre-and post-NAT samples available for analysis at the time of this pilot. All pre-NAT samples were analyzed for the presence of TSMs using Sysmex Inostics’ SafeSEQ. Patients with detectable ctDNA before NAT were also evaluated for residual ctDNA after completion of NAT but prior to surgery. Five samples were also tested by digital PCR (BEAMing) for cross-platform comparison. Results: TSMs in ctDNA were identified in 29 of 58 patients (50%) prior to NAT. Of pre-NAT ctDNA(+) patients, TSMs were detected in TP53 (90%) and PIK3CA (10%); three patients (10%) were found to have 2 TSMs. Concordance between SafeSEQ and BEAMing was 100% in five samples tested [3 ctDNA(+), 2 ctDNA(-)]. Of 29 ctDNA(+) patients, 24 (83%) demonstrated reduction or elimination of detectable ctDNA following neoadjuvant therapy, with 16 (55%) converting to ctDNA(-). Conclusion: Identification of TSMs in the plasma of early-stage breast cancer patients without the need for biopsy tissue is feasible using a SafeSEQ cancer mutation panel. Further measures to improve the sensitivity of pre-treatment TSM analysis, such as increased plasma volume input and comprehensive TP53 mutational analysis are currently under investigation. Correlations between clinicopathologic factors with ctDNA detection and burden, as well as the NPV of post-NAT ctDNA for pCR and residual cancer burden, will be reported at the time of abstract presentation. NH and HP contributed equally to this work. Citation Format: Natasha Hunter, Heather Parsons, Alexander Sherry, Daniel Shinn, Dong Ho Shin, Alex Cole, Giovanni Cragnotti, Taylor Groginski, Margaret Leathers, Andrea L Richardson, Pedram Argani, Antonio Wolff, Leslie Cope, Dan Edelstein, Frank Holtrup, Hilary Sloane, Bapsi Chakravarthy, Vered Stearns, Ben H Park. TBCRC 040: Pathologic response evaluation and detection in circulating tumor DNA (PREDICT DNA): Initial results piloting a tissue-biopsy independent method of identifying and monitoring tumor-specific mutations in early stage breast cancer [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P6-10-05.

Background: Radiotherapy (RT) plays an important role in the multimodal treatment of breast cancer (BC). Despite improvements in the accuracy of delivering radiation to specific biological target volumes, the clinical response of BC to RT is still affected by intrinsic/acquired radioresistance. These resistant cancer cells can contribute to the development of recurrent disease and poor patient outcomes. Clinical signs of RT response are often not apparent for several weeks post-treatment; patients who fail to respond will therefore initially go undetected. There are currently no clinically validated biomarkers that can predict which patients will respond to RT or assess response during treatment. Our study aims to address this major clinical need through the identification and validation of biomarkers of radiation responsiveness. Methods: The effects of different radiation doses (2 - 10Gy) at a range of time points (1 - 24h) were investigated by analysing the protein secretion profiles from 3 BC cell lines: MCF-7 (ER+), ZR-751 (ER+) and MDA-MB-231 (ER-). Conditioned media was collected from each dose/time point and proteins isolated for mass spec analysis. For comparison, radioresistant models were derived from each of the 3 cell lines and were characterized by proliferation and colony formation assays, invasion and migration assays, whole-genome transcriptomic sequencing (WGTS) analysis and western blotting. To assess intrinsic response to RT a panel of 16 BC cell lines were evaluated by colony formation assays following a 2Gy dose of radiation. WGTS of a patient cohort of 230 (138 ER+ve, 92 ER-ve) post-menopausal women with BC, treated with breast conserving surgery and adjuvant RT but no systemic adjuvant therapy, with a median follow-up of 14 years, is currently underway. Results: 9 biomarkers emerged whose secretion was significantly increased with radiation. These were evaluated by western blotting of conditioned media 24h after a 2Gy dose of RT in matched radio-sensitive and resistant cell lines; this confirmed significantly higher levels of radiation induced secretion from sensitive cells compared to resistant. Radioresistant cell lines were characterised by epithelial-to-mesenchymal-transition, enhanced invasion/migration, loss of ER and PgR and increased EGFR and PI3K signaling activity. Initial mechanistic investigations suggest that biomarker release in response to radiation occurs via microvesicles. A blood-based assay to test the level of these secreted biomarkers is currently under development. Pre-treatment levels of the 9 biomarkers were also found to be associated with prediction of intrinsic response to RT at both gene and protein level. A gene expression signature comprising the 9 candidates is strongly associated with the intrinsic response to RT across the 16 BC cell lines studied, with higher expression found in those more sensitive to RT compared with those less sensitive or resistant. Validation of the predictive power of these biomarkers in terms of recurrence-free and overall BC specific survival is currently being assessed at gene and protein level in the patient cohort. Conclusions: We have identified 9 biomarkers which are released from BC cells sensitive to radiation 24h after a 2Gy dose (in line with current clinical standards) but not from radio-resistant derivatives.A blood based assay is currently under development which has the potential to monitor response to RT in the neoadjuvant and palliative settings.Intracellular levels of the 9 biomarkers are strongly associated with intrinsic response to RT and may hold predictive potential.These biomarkers may have the potential to improve patient care by identifying patients less likely to benefit from RT, paving the way for personalization of treatment, including altered dosing schedules and the future use of emerging radio-sensitizers. Citation Format: James Meehan, Mark Gray, Carlos Martinez-Perez, Charlene Kay, J Michael Dixon, Jimi Wills, Carol Ward, Alex von Kriegsheim, Niall Quinn, Olga Oikonomidou, David Cameron, Simon P Langdon, David Argyle, Ian H Kunkler, Arran K Turnbull. Development and validation of novel biomarkers of response to radiotherapy in breast cancer [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P6-10-18.

Background: In the Phase III MONALEESA-3 study (NCT02422615),the cyclin-dependent kinase (CDK) 4/6 inhibitor ribociclib(RIB) + fulvestrant significantly prolonged progression-free survival (PFS) vs placebo (PBO) + fulvestrant in postmenopausal patients (pts) with hormone receptor–positive, human epidermal growth factor receptor 2–negative (HR+/HER2−) advanced breast cancer (ABC; median, 20.5 vs 12.8 mo; hazard ratio 0.593 [95% CI, 0.480-0.732]; P< .001; Slamon D et al. J Clin Oncol. 2018). Previous MONALEESA-3 biomarker analyses, including those based on immunohistochemistry and next-generation sequencing of baseline circulating tumor DNA (ctDNA), showed consistent PFS benefit with ribociclib (Neven P et al. ESMO 2018; Neven P et al. SABCS 2018). We present analyses of gene expression in baseline tumor samples from the MONALEESA-3 trial and potential correlations with PFS. Methods: Postmenopausal pts with ≤1 prior line of ET for advanced disease were enrolled and randomized 2:1to receive either RIB or PBO plus fulvestrant. Gene expression data were generated for 531 patients using tumor samples collected before treatment using the NanoString 800-gene nCounter® GX Customized Panel. Correlations between PFS and mRNA expression levels were assessed for various genes, including those relevant to breast cancer, the CDK4/6 pathway, the mitogen-activated protein kinase (MAPK) pathway, receptor tyrosine kinases (RTKs), estrogen receptor (ER) signaling and response to estrogen, and proliferation; specific genes are listed in the Table. Pts were classified into low and high mRNA expression subgroups, with median values of gene expression levels or pathway scores as the cutoff. The Kaplan-Meier method was used to calculate median PFS with 95% CIs in each treatment arm by biomarker subgroup. Cox proportional hazards models were used to estimate hazard ratios and 95% CIs. Results: PFS hazard ratios for all biomarker subgroups favored RIB + fulvestrant vs PBO + fulvestrant (Table). There appeared to be a trend toward greater PFS benefit with RIB in patients with lower expression of BCL2 or genes involved in ER signaling. Similar RIB PFS benefit was observed regardless of the expression level of the CCNE1 gene, RTK genes, genes involved in tumor cell proliferation, CDK4/6 pathway genes, or MAPK pathway genes. Conclusions: The PFS benefit observed with RIB vs PBO across gene expression subgroups in this analysis was generally consistent with that seen in the primary analysis of MONALEESA-3. Table. Progression-Free Survival by Gene Expression SubgroupGene/PathwayLow ExpressionHigh ExpressionPBORIBPBORIBCCNE1an9017682183Median PFS, mo14.5220.0412.7816.62HR (95% CI)0.62 (0.44-0.87)0.66 (0.46-0.94)BCL2an7918793172Median PFS, mo9.4317.9418.4620.04HR (95% CI)0.47 (0.33-0.66)0.8 (0.55-1.14)Proliferationbn9017682183Median PFS, mo12.6819.0912.9120.04HR (95% CI)0.58 (0.41-0.81)0.72 (0.5-1.03)ER signalingcn8018692173Median PFS, mo9.4619.2914.8219.12HR (95% CI)0.53 (0.37-0.75)0.78 (0.55-1.11)Estrogen responsivedn8518187178Median PFS, mo9.116.3619.3823.46HR (95% CI)0.53 (0.38-0.74)0.76 (0.52-1.11)RTK genesen8717985180Median PFS, mo12.29NA16.4916.62HR (95% CI)0.55 (0.38-0.79)0.73 (0.51-1.03)CDK4/6 pathwayfn9017682183Median PFS, mo16.5622.0810.8416.49HR (95% CI)0.63 (0.43-0.9)0.67 (0.48-0.94)MAPK pathwaygn9117581184Median PFS, mo12.2922.0812.9117.94HR (95% CI)0.66 (0.47-0.94)0.65 (0.45-0.92)aBreast cancer–related gene; b MKI67, TYMS, MYC, TOP2A; c GATA3, FOXA1, AR, ESR1; d CCND1, PGR, NAT1, FOXA1, RARA, BCL2, BAG1, GATA3, MYC, STC2, XBP1; e AREG, EGFR, ERBB2, ERBB3, ERBB4, FGFR1, FGFR2, FGFR3, FGFR4, FYN, IGF1, KDR, KIT, PDGFA, PDGFRA, PDGFRB, TYRO3; f CCNA2, CCND1, CCND2, CCND3, CCNE1, CDK2, CDK4, CDK6, E2F1, E2F3, TFDP1, CCNB1, CDKN1A, CDKN1B, CDKN2A, CDKN2B, CDKN2C, CDKN3, RB1, RBBP8, RBL1, RBL2, TP53; g BRAF, HRAS, KRAS, NRAS, RAF1, MAP3K8, NF1. Citation Format: Stephen Chia, Faye Su, Patrick Neven, Seock-Ah Im, Katarina Petrakova, Giulia Val Bianchi, Wei He, Karen Rodriguez-Lorenc, Tetiana Taran, Naveen Babbar, Dennis Slamon. Gene expression analysis and association with treatment response in postmenopausal patients with hormone receptor-positive, HER2-negative advanced breast cancer in the MONALEESA-3 study [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr PD2-08.

Background Few currently available biomarker panels can predict the risk of long-term distant recurrence (DR) in endocrine-responsive breast cancer (ERBC) patients based on lymph-node (LN) status. A new model, DGM-CM6 (Distant Genetic Model-Clinical variable Model 6), has been developed for the prediction of DR risk in our previous studies. To interrogate the prognostic value of our DGM-CM6 model for ERBC patients stratified by LN status, we retrospectively studied 752 operable breast cancer patients treated in a cancer center from 2005 to 2014. Patients and Methods Of 752 tumors, a total of 499 ERBC patients was identified from both IHC method (n=490) and PAM50 method (n=404). ERBC patients were classified by high and low-risk groups using the cutoff ( Results Significant differences were observed between DGM-CM6 low- and high-risk patients with 10-year DRFS in LN- (ERBC identified by IHC: P = 0.011; by PAM50: P< 0.0001) and LN+ (ERBC identified by IHC: P = 0.015; by PAM50: P = 0.019) by log-rank test. Compared to risk prediction by intrinsic subtype, DGM-CM6 performed better in LN- patients and equally good in LN+ patients. Further multivariate analysis confirmed the independent strength of DGM-CM6 model in the prediction of high- vs. low- risk groups in DRFS (P < 0.0001, HR: 6.76; 95% CI, 1.8-25.42) and OS (P = 0.01, HR: 6.06; 95% CI:1.55-23.47), respectively. Conclusion DGM-CM6 model can be used to predict low- and high-risk of 10-year distant recurrence in both LN- and LN+ ERBC patients. This model needs a large scale of clinical trials to validate its clinical utility. Table 1. Multivariate cox regression analyses for the prognosis of predicted risk groups after adjusted for other clinical variables in the distant recurrence (DR, p=0.005) and overall survival (OS, p=0.01) respectively. Results shows that there is no interaction between chemotherapy and risk groups (DR: p=0.163; OS: p=0.195). Abbreviations: DR, distant recurrence; OS: overall survival.GroupsDROSHR [95% CI]pvHR [95% CI]pvRisk: high vs.low6.76 [1.8;25.42]0.005 ***6.06 [1.55;23.74]0.01 ***Age: 50yrs0.86 [0.51;1.44]0.5630.68 [0.38;1.19]0.175LN: Pos. vs. Neg.1.64 [0.88;3.05]0.1161.6 [0.79;3.22]0.189Stage: II vs. I1.35 [0.76;2.4]0.31.45 [0.76;2.77]0.264Stage: III vs. I2.08 [0.59;7.3]0.2531.59 [0.34;7.46]0.557Grade: 2 vs. 12.14 [0.77;5.97]0.1461.81 [0.63;5.25]0.272Grade: 3 vs. 11.2 [0.4;3.62]0.7481.21 [0.38;3.86]0.752PAM50: Normal vs. LumA0.6 [0.13;2.64]0.4960.7 [0.16;3.16]0.645PAM50: LumB vs. LumA1.58 [0.84;2.98]0.1591.29 [0.65;2.59]0.469PAM50: Her2 vs. LumA0.81 [0.25;2.68]0.7311.01 [0.29;3.47]0.986RT: Yes vs. No0.88 [0.47;1.66]0.7010.71 [0.36;1.41]0.326CT: Yes vs.No0.36 [0.11;1.12]0.0770.3 [0.09;1]0.05 *Interaction: CT vs. Risk0.36 [0.09;1.51]0.1630.36 [0.08;1.68]0.195 Citation Format: Lei Lei, Skye Hung-Chun Cheng, Xiao-Jia Wang, Yin-Yuan Mo, Yun-Yun Zhou. A new model to predict 10-year distant recurrence risk for operable endocrine-responsive breast cancer population [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P6-10-14.

Background: Ductal carcinoma in situ (DCIS) lesions are non-obligate precursors to invasive breast cancer (IBC). With the ultimate goal of preventing the development of invasive disease, DCIS is typically treated by breast-conserving surgery (BCS). Adjuvant radiotherapy (RT) is given for high-grade disease to reduce the risk of in-breast tumour recurrence (IBTR). The use of endocrine therapy (ET) for DCIS varies, as studies show a modest benefit but no survival improvement; in the UK, guidelines recommend ET for DCIS in some scenarios but only instead of, rather than in addition to, RT. This project sought to characterise the biology of ER+ DCIS and identify a group of women who are likely to gain the most benefit from the addition of adjuvant ET. Patients: Cohort A - 77 women with ER+ (Allred 7/8) high-grade DCIS treated with BCS plus RT, 20 of whom developed IBTR.Cohort B - 70 women with ER+ (Allred 7/8) low/intermediate-grade DCIS treated with BCS alone, 12 of whom developed IBTR.Cohort C - 68 women with ER+ (Allred 7/8) DCIS treated with BCS plus ET. All patients were treated locally between 2000 and 2016 and the median follow-up is 6 years. Methods: We performed whole-genome transcriptomic QuantSeq sequencing of samples from cohort A. Sequencing of cohort B and C is currently underway. IL6ST levels were validated using immunohistochemistry and RNAScope. Results: In cohort A, only a subset (34/77) of tumours had gene expression profiles consistent with active ER signalling. Levels of IL6ST, a biomarker for ET response, could differentiate these two subgroups and this was validated at protein level using immunohistochemistry. The low ER signalling subgroup were associated with higher levels of EGFR, HER2 and MAPK signalling. 20/77 high-grade DCIS cases recurred within 10 years. 50% of these recurred as IBC (rather than DCIS) and these were associated with higher levels of IL6ST, had active ER signalling and higher levels of proliferation-associated and estrogen receptor target genes, known to be decreased by ET, in the primary DCIS lesion. Discussion: Our findings suggest that some high-grade ER+ DCIS patients have active ER signalling while in others ER signalling remains low despite highly expressing the ER protein. IL6ST, a biomarker of endocrine therapy response can be used to differentiate these two groups of ER+ DCIS. DCIS lesions which recurred as IBC had active ER signalling and also higher levels of proliferation genes known to be decreased by ET compared with DCIS which recurred as further DCIS. These findings suggest that IL6ST may have a role in identifying a subset of ER+ DCIS which are at a higher risk of developing advanced disease and are also more likely to benefit from the addition of adjuvant ET, with a better risk-to-benefit ratio than observed in previous studies that considered a less targeted use of this treatment strategy, thus potentially reducing the risk of IBC recurrence. These findings will be validated in a cohort of low/intermediate-grade DCIS who received no adjuvant RT (cohort B) and a cohort of patients who received adjuvant ET as part of their treatment (cohort C). Citation Format: Carlos Martinez-Perez, Charlene Kay, Rebecca Swan, Gregory E Ekatah, Laura M Arthur, James Meehan, Mark Gray, Andrew H Sims, Olga Oikonomidou, Arran K Turnbull, J Michael Dixon. IL6ST, a biomarker of endocrine therapy response, has potential in identifying a subgroup of women with ER+ DCIS who are more likely to benefit from adjuvant endocrine therapy [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P6-16-04.

Background: A variety of investigational HER2-inhibitor agents/combinations have been tested in I-SPY 2, including neratinib (N), TDM1 combined with pertuzumab (P) (TDM1/P), and trastuzumab (H) combined with pertuzumab (H/P; prior to this combination becoming standard of care), all with trastuzumab as control (Ctr). All three experimental arms graduated, showing improved efficacy over control in one or more receptor subsets (HR+HER2+, HR-HER2+, or/and HER2+). Here we assess 10 biomarkers in the HER2, ER/PR, and proliferation pathways on multiple levels of resolution (expression, protein, phospho-protein) as predictors of response in these four arms, hypothesizing that highly HER2-activated, proliferative tumors may be more sensitive to HER2-inhibition than those that are more luminal and quiescent. Methods: 192 HER2+ patients were considered in this analysis: (31 Ctr, 65 N, 52 TDM1/P, and 44 H/P). 10 biomarkers relating to HER2, ER, or proliferation were evaluated from pre-treatment biopsies: HER2 IHC (n=145), 3 expression signatures (n=192), BluePrint subtype (n=192), and 5 protein/phospho-protein analytes by RPPA (n=175). Each biomarker was tested for association with pCR in the whole population and within each arm using a logistic model. This analysis was adjusted for HR status and treatment arm as covariates, and performed within receptor subtypes. This analysis does not adjust for multiplicities of other biomarkers. Results: In the population as a whole, HER2 and HER2-signaling biomarkers, evaluated at multiple levels of resolution: IHC, total-/phospho-protein by RPPA, and mRNA (HER2 amplicon module) - are highly correlated (rho=0.8 [0.65-0.92]). Higher HER2 levels and activity are associated with response: HER2 IHC 3+ status (LR p=0.00032), total quantitative ERBB2 protein by RPPA (LR p=5.4E-09), ERBB2 activation levels (LR p=6.58E-06 (pERBB2 (Y1248)) and 9.95E-06 (pEGFR Y1173)), and the ERBB2 amplicon expression signature (LR p=2.38E-08). In contrast, higher average ER/PR expression associates with non-response to HER2-targeted therapy (LR p=4.28E-08). Both HER2 and ER/PR signaling phenotypes are captured by BluePrint subtyping; and consistent with the individual pathway markers, tumors classified Luminal-type had a lower pCR rate relative to those classified as Her2-type (or Basal-type) (LR p=4.84E-11). These associations all retain significance in a model adjusting for HR status and treatment arm, and in the HR+HER2+ subset. In addition, we quantitatively assessed proliferation markers at the total protein (RPPA: Ki67), phospho-protein (pAURKA) and mRNA (proliferation signature Module11_Prolif) levels. All three proliferation biomarkers predict response overall; but this association is strongest within the HR+HER2+ subset (LR p=0.0012 (Module11_prolif), 0.0036 (pAURK), and 0.045 (Ki67)). None of the biomarkers tested were associated with response in the HR-HER2+ subset. Numbers are small within individual arms. Within the HR+HER2+ subtype, higher HER2 and lower ER/PR is observed in responders in all experimental arms; but the proliferation markers Module11_Prolif (LR p=0.0031) and Ki67 total protein (LR p=0.0029) are associated with response to TDM1/P but not H/P or N. Conclusion: High HER2 signaling at the expression, protein, and phospho-protein levels, and low ER signaling, all predict response to HER2-inhibition across treatment arms. Proliferation markers may be useful for prioritizing therapies in the HR+HER2+ subset. Citation Format: Denise M Wolf, Christina Yau, Julia Wulfkuhle, Lamorna Brown-Swigart, Smita M. Asare, Gillian L. Hirst, Laura Sit, Jane Perlmutter, I-SPY 2 TRIAL Consortium, Minetta Liu, John Park, Angela DeMichele, Douglas Yee, Don Berry, Laura Esserman, Emanuel Petricoin, Laura van t' Veer. HER2 signaling, ER, and proliferation biomarkers predict response to multiple HER2-targeted agents/combinations plus standard neoadjuvant therapy in the I-SPY 2 trial [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P4-10-02.

Background: In ER+ HER2- metastatic breast cancer (MBC), activating ESR1 mutations (ESR1mut) confer resistance to aromatase inhibitors (AI) when used as single agent. The impact of ESR1mut, when detected at baseline (prior to the initiation of therapy), on the efficacy of first line combined AI- CDK4/6 inhibitor therapy remains however unknown. Methods: The PADA-1 phase 3 trial (NCT03079011) aims at evaluating the utility of monitoring the onset of ESR1mut in cell-free DNA (cfDNA) from patients (pts) receiving AI-palbociclib, as a biomarker to trigger a switch from AI to fulvestrant. Main inclusion criteria were pts with no prior therapy for MBC and with no overt resistance to AI (i.e. no adjuvant AI or adjuvant AI completed >12 months prior to the metastatic relapse). The present exploratory analysis evaluated the biological and clinical outcome of patients who tested positive for ESR1mut at baseline (before any treatment). Results: From 04/2017 to 01/2019, N=1017 ER+ HER2- MBC pts were included and had their cfDNA tested for ESR1mut at inclusion and during therapy (at 1 month and then every 2 months). N=33/1017 pts had a detectable circulating ESR1mut at inclusion (3.2%, 95%CI [2.2;4.5]), with a median allelic frequency (AF) of 2.5% (range: 0.1- 46%). First, we evaluated the kinetics of ESR1mut in the 33 ESR1mut-positive pts at inclusion. N=1 pt died after 1 month on treatment. In N=25/32 pts (78%), ESR1mut became undetectable (AF Conclusions: ESR1mut are rarely detected in the cfDNA of ER+ HER2- MBC patients with no overt resistance to AI. The quick ‘clearance’ of ESR1mut observed in most patients following treatment initiation and the observed 17.5 months-long median PFS both suggest that the AI palbociclib combination retain clinical activity in ESR1mut-positive MBC. However, in most patients, these mutations were eventually detected again later during therapy and ESR1mut- positivity was associated with a significantly shorter PFS, suggesting that ESR1mut positivity at baseline could accelerate the onset of resistance to AI-palbociclib. Funding: Pfizer Citation Format: François-Clément Bidard, Céline Callens, Florence Dalenc, Barbara Pistilli, Thibault de la Motte Rouge, Florian Clatot, Véronique D'Hondt, Luis Teixeira, Sibille Everhard, Hélène Vegas, Jérôme Lemonnier, Ivan Bieche, Anne Pradines, Jean-François Paitel, Dominique Spaeth, Jean-Luc Canon, Isabelle Moullet, Jean-Yves Pierga, Frédérique Berger, Thomas Bachelot, Suzette Delaloge. Clinical and biological efficacy of first line AI and palbociclib in ER+ HER2- MBC with detectable circulating ESR1 mutation prior to treatment initiation [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P5-11-22.

Introduction: CDK 4/6 inhibitors (CDK4/6i) are FDA approved for treatment of hormone receptor (HR) positive, HER2 negative metastatic breast cancer (MBC). Recent work demonstrated that a rise in MCV was correlated with longer median progression free survival (Anampa et al, Haemoatologica 2018). We performed a single-institution retrospective analysis to evaluate whether palbociclib induced early changes in MCV serve as a pharmacodynamic biomarker to predict response to CDK4/6i. Methods: We identified a retrospective cohort of 81 patients with HR+, HER2- MBC treated with CDK4/6i as first line metastatic treatment at Mayo Clinic, Rochester. Hematologic indices were abstracted pretreatment and at the start of each cycle through the start of cycle 4 (C4). Wilcoxon signed test was used to examine the changes in MCV (ΔMCV) comparing pretreatment up to C4 and ΔMCV between cycles. Optimal cut points for ΔMCV were determined using Cutoff finder (Budczies et al., PLOS One 2012). Cox proportional hazard regression analysis was performed to evaluate associations of MCV and time to treatment failure (TTF), which was defined as the time from the start of the analyzed cycle to treatment discontinuation for any reason(in months). As eight comparisons were made, a Bonferroni correction established p-value Results: 60/81 patients had pretreatment and at least one subsequent lab data point that included MCV. The pretreatment mean hemoglobin (Hgb) was 12.8 [standard deviation (SD) 1.37 g/dL] and 6/60 (10.2%) patients with Hgb 100 fl) was observed. At C4, the mean Hgb was 12.1 g/dL (SD 1.11) with 4/49 (8.2%) with Hgb Discussion: We observed an increase in MCV in all patients receiving CDK4/6i by start of C4. The rise in MCV during CDK4/6i treatment may reflect drug induced cell-cycle arrest in hematopoietic cells. Evaluation in larger studies is needed to validate dynamic changes in MCV as a predictive biomarker of response to CDK4/6i therapy. Citation Format: Grace Mei Yee Choong, Roberto A Leon-Ferre, Ciara C O'Sullivan, Kathyrn J Ruddy, Tufia C Haddad, Timothy J Hobday, Prema P Peethambaram, Charles L Loprinizi, Minetta C Liu, Vera J Suman, Matthew P Goetz, Karthik V Giridhar. Evaluation of mean corpuscular volume (MCV) as a pharmacodynamic predictive biomarker in patients receiving CDK4/6 inhibitors for metastatic breast cancer (MBC) [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P1-19-42.

Background: Estrogen (ER) and progesterone (PR) receptor staining with immunohistochemistry (IHC) forms the basis for endocrine therapy (ET) eligibility. However, ER protein level does not necessarily reflect ER functionality and is insufficient as a predictor of response or as a pharmacodynamic (PD) biomarker to assess the impact of ET on ER activity. Furthermore, assessing PD modulation based on a single ER target gene, PR, lacks robustness and cannot be utilized in PR- tumors. We established a transcriptional signature that more comprehensively reflects ER pathway activity. We showcase, in vitro, in vivo, and in patients enrolled in a Phase I trial of GDC-9545 -a novel, potent, orally bioavailable, selective estrogen receptor antagonist / degrader- that this signature may have utility as a predictive and PD biomarker. Methods: A signature of 21 estradiol (E2)-induced and 17 E2-repressed genes was experimentally derived by transcriptional profiling of 8 E2-stimulated ER+ breast cancer (BC) cell lines in vitro, and refined by leveraging TCGA RNA-seq data. The effect of ER inhibition by GDC-9545 on cellular proliferation was determined in an 8 day in vitro viability assay, and correlated with pre-treatment ER pathway activity as determined by the signature. The signature was evaluated in an in vivo efficacy study with GDC-9545 (0.1, 1 or 10 mg/kg) in PDX breast model HCI-013, measured using a Fluidigm® panel. Paired pre- and on-treatment tumor biopsies from 7 patients enrolled in the Phase I dose escalation study of GDC-9545 (10, 30 or 90 mg) in ER+/HER2- advanced or metastatic BC were collected to retrospectively assess ER pathway modulation. ER, PR and Ki67 protein levels were analyzed by IHC, while gene expression analysis from FFPE tissue was performed using Illumina’s RNA Access protocol and HiSeq. The ER pathway activity score was defined as the difference in average expression of the E2-induced versus E2-repressed genes, relative to ER pathway activity in a reference collection of 139 procured ER+/HER2- breast tumors. Results: A panel of 14 ER+/HER2- BC cell lines exhibited a spectrum of in vitro responses to GDC-9545 that strongly correlated with the pre-treatment ER pathway activity score. GDC-9545 had little anti-proliferative effect in cell lines with lowest scores, while having a considerably greater impact on cell lines exhibiting higher scores. Besides potential applicability as a predictive biomarker, we explored the utility of the ER activity score in vivo as a PD biomarker. We observed enhanced transcriptional pathway suppression with increased dose of GDC-9545, and a concomitant greater impact on Ki67 expression, in the in vivo HCI-013 PDX breast model. The ER signature was further evaluated in pre- and on-treatment tumor biopsy pairs from 7 patients enrolled in the GDC-9545 Phase I trial. ER activity scores of tumors from these patients were benchmarked against a cohort of 139 procured ER+/HER2- breast tumors, revealing a range in ER pathway activity prior to treatment with GDC-9545. Two tumors exhibiting profound Ki67 responses (post-treatment IHC Citation Format: Anneleen Daemen, Jill M Spoerke, Wei Zhou, Jane Guan, Ellen Ingalla, Amy Young, Marc Hafner, Junko Aimi, Ching-Wei Chang, Jennifer M Giltnane, Mary Gates, Ingrid A Mayer, Analia Azaro, Eric P Winer, Sherene Loi, Komal Jhaveri, Jennifer Lauchle, Steven Gendreau, Eric W Humke, Ciara Metcalfe. ER pathway activity signature as a biomarker for endocrine agent GDC-9545 [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P2-11-05.

Background Increasing data support the role of BCSCs in recurrence and resistance to chemotherapy. However, detecting these cells specifically and targeting them therapeutically remain challenging. We set to identify miRNAs involved in chemoresistance of BCSCs in vitro and the results were assessed in the sera of healthy donors and breast cancer patients treated with neoadjuvant chemotherapy (NAC). Methods In vitro experiments were conducted in the MCF7 cell line grown as mammospheres (MS) in order to enrich a BCSC phenotype. After treatment with 5FU or Paclitaxel (Pac), microRNA profile of chemoresistant cells was analyzed by microarray. The results were compared to miRNAs found in immortalized non-tumorigenic MCF10A cells in order to exclude miRNAs related to normal stem cells. A signature of 6 miRNAs was identified. miR-363-3p, appearing the most relevant, was chosen for further assessment. By RT-qPCR, miR363-3p levels were 20- and 100-fold higher in ALDH+ sorted cells compared to ALDH- MCF7 and MDA-MB-231 cells, respectively. Moreover, overexpression of miR363-3p in MCF7 cells correlated with an increased number of ALDH+ cells and 1.5 more MS formed. In contrast, downregulation of miR363-3p levels induced a decrease in MS size and 2-fold reduction in number. Consistently, miR-363-3p downregulation decreased tumor growth and metastasis of MCF7 cells in an intraductal human-in-mice transplantation model. miRNA was quantified by RT-qPCR in tissue and sera collected prospectively from 40 breast cancer patients before and after NAC (anthracycline-taxane) and also from 25 healthy donors. Results Fifty percent of the patients had luminal A and B tumors (n=20), 32,5% (n=13) had triple negative breast cancer (TNBC), and 17.5% (n=7) had HER2 positive BC. Excluding one patient, who refused the surgical resection, 40% achieved pathological complete response (pCR). Six patients presented disease recurrence. In patients’ tumor before and after NAC, a higher level of miR363-3p was observed compared with benign tissue. In sera of the BC patients, miR363-3p levels were significantly higher than that of healthy donors. After NAC, the levels of miR363-3p remained high among patients who relapsed, whereas they were equivalent to the healthy donors in patients who remained in remission. No correlation between Ki-67, grade 1 or 2 and miR363-3p levels was observed. However, in grade 3 BC, low level of miR363-3p before and after NAC was correlated with pCR and remission. The level of miR363-3p correlated with remission and pCR in patients with TNBC and HER2 BC. Conclusions Assessment in the sera of patients and healthy donors confirms our previous data supporting the role of miR363-3p as predictive factor of resistance to neoadjuvant chemotherapy and disease recurrence. Further investigations are warranted to confirm the role of miR363-3p as a biomarker and potential therapeutic target. Citation Format: Stephanie Renaud, Athina Stravodimou, Maryse Fiche, Ioannis Xenarios, Scabia Valentina, Valerian Dormoy, Marie Galmiche, Cathrin Brisken, Jean-Francois Delaloye, Assia Treboux, Jean-Yves Meuwly, Nicolas Mermod, Khalil Zaman. miR363-3p mediates maintenance of breast cancer stem cells (BCSCs) and predicts resistance to neoadjuvant chemotherapy and disease recurrence [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P6-10-22.

Background: The combination of everolimus and an endocrine agent has shown a progression-free survival (PFS) benefit but also significant toxicity in locally advanced and/or metastatic breast cancer (MBC) patients. Disappointingly, all translational research efforts conducted so far on primary/metastatic tissue or baseline ctDNA samples have failed to identify clinically useful biomarkers of benefit with the possible exception of p4EBP1 which requires fresh metastatic biopsies (Bachelot at al, ASCO 2019). Methods: Forty-seven patients were included in the PEARL trial in six Belgian centres. All patients underwent one baseline and one “early” 18FDG-PET/CT on day 14 (blinded to the treating oncologist) after treatment initiation. Eligible patients had prior progression on NSAI and at least one measurable lesion on baseline 18FDG-PET/CT (fulfilling PERCIST). Twenty seven of the 47 patients were treated in a pilot phase aiming to characterize 18FDG-PET/CT metabolic responses 14 and 28 days after treatment initiation. Day 14 18FDG-PET/CT was selected for the second phase in which 20 more patients were included to answer the primary endpoint of the trial, namely the potential of an early FDG-PET/CT to select patients who will not derive a long-term benefit from everolimus-exemestane combination. We considered as “metabolic responders” those patients in whom at least a 25% homogenous decrease in SUVmax was seen in all their target lesions; the so called “consistent” patient-based response, here correlated to PFS. Results: Seventy-four percent of the patients had visceral metastasis, 26% were pretreated with CDK 4/6 inhibitors and 58% with at least one line of chemotherapy (28% ≥ 2 lines) for the treatment of MBC. Metabolic non-response rate on day 14 was 64%. Two patients were considered non-evaluable because they stopped everolimus 3 and 8 days before early 18FDG-PET/CT respectively. The median PFS in non-responding patients was 2.3 months compared to 6 months in those who showed a consistent metabolic response (HR-1.96, 95%CI: 0.98-3.95, p=0.05). We’ll be able to assess whether day 14 ctDNA measurements add predictive value to imaging. Conclusion: Absence of 18FDG-PET/CT metabolic response after 14 days of everolimus-exemestane combination is associated with a low probability to derive benefit from this treatment. The advanced stage of the patients included in this trial (75% visceral metastasis) and their extensive prior treatments can explain the poorer outcome seen with exemestane-everolimus compared to the Bolero2 trial. Citation Format: Andrea Gombos, Martine Piccart-Gebhart, Lieveke Ameye, Peter Vuylsteke, Patrick Neven, Vincent Richard, Francois P Duhoux, Marianne Paesmans, Michail Ignatiadis, Patrick Flamen, Géraldine Gebhart. Pearl or the use of FDG positron emission tomography imaging as an early dynamic biomarker of everolimus efficacy in advanced luminal breast cancer [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P1-19-22.

Fundamental to the effective clinical use of new therapeutic agents for cancer is the identification of genetic biomarkers of sensitivity as well as the identification of additional patient populations who would benefit from treatment. Here, through an unbiased functional genomics approach in a model of aggressive disease using cancer cell line spheroid cultures, we have uncovered a role for the histone acetyltransferase CREBBP as a tumour suppressor in aggressive hard to treat triple-negative breast cancers (TNBCs). We find that CREBBP protein expression is altered in a substantial proportion of TNBCs, as well as other solid tumours, including lung and lymphoma where CREBBP inactivating mutations are common. We found CREBBP alterations are associated with higher genomic heterogeneity and poorer patient survival in TNBCs and resulted in upregulation of the FOXM1 transcriptional network. Strikingly, treatment with clinical CDK4/6 inhibitors (CDK4/6i), that indirectly target FOXM1 activity, selectively impaired growth in spheroids, cell line xenografts and patient derived models from multiple tumour types with CREBBP alterations. Here we provide the pre-clinical and mechanistic rationale for assessing CREBBP alterations as a biomarker of single-agent CDK4/6 inhibitor response in a new patient population. This represents a novel synthetic lethal approach to targeting triple negative breast cancers with limited therapeutic options. Citation Format: Rachael Natrajan. CREBBP alterations lead to CDK4/6 inhibitor sensitivity in triple negative breast cancer [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P3-10-17.

Rationale/Objectives: Tumor immune infiltration in HER2+ breast cancer is associated with improved outcomes and better responses to treatment with drugs such as trastuzumab. Breast MRI is a non-invasive technique that is part of the standard evaluation for localized breast cancer. Breast MRI dynamic-contrast enhanced (DCE) and diffusion-weighted imaging (DWI) sequences may detect immune responses in the tumor microenvironment. Apparent diffusion coefficient (ADC), which is derived from DWI, has been associated with tumor cellularity. In this pilot study, we hypothesized that patients with low ADC would be more immune responsive (i.e. have greater TIL infiltration) to trastuzumab. We also performed an exploratory analysis looking at differences in the expression of immune-relevant genes between “low” and “high” ADC patients. Methods: Patients with localized HER2+ breast cancers 10 mm or larger and planned for curative-intent therapy were eligible to enroll. After enrollment, patients received a single run-in dose of trastuzumab with paired breast MRI exams before and after treatment that included DCE and DWI sequences. Post-trastuzumab tissue was obtained from the patient’s surgical specimen or from a second biopsy (if planned for additional neoadjuvant treatment). MRI images were analyzed to measure the functional tumor volume and peak signal enhancement ratio from DCE and mean apparent diffusion coefficient (ADC) from DWI. TIL content was assessed by a breast pathologist in both the pre- and post-trastuzumab tissue specimens. The pre- and post-treatment samples were also analyzed using an immune-focused RNA-based multiplex gene expression assay. Results: Twelve women have enrolled to date with a median age of 53 years (range 37-70 years old). Five out of the twelve patients (42%) had estrogen-receptor negative tumors. For the 8 patients with pre- and post-treatment tissue currently available for analysis, the average pre-treatment TIL percentage was 22% (range 1-70%), the average post-treatment TIL percentage was 42% (range 1-90%), and the average absolute increase between pre- and post-treatment TILs was 20% (range 0%-58%). Pre- and post-treatment ADC showed a strong negative correlation with the increase in TIL content after trastuzumab treatment (r = -0.71 and r = -0.93, respectively). At submission, preliminary gene expression analysis of pre- and post-treatment tissue is available in 6 patients. The patients were binned into “high” and “low” ADC groups with the median pre-treatment ADC as the cut-off. The magnitude of change in gene expression after trastuzumab treatment (small vs large change) clustered by high or low pre-treatment ADC; this suggests that pre-treatment ADC may differentiate between two distinct immune phenotypes. Further analysis showed that the low ADC group had relatively higher pre-treatment gene expression of CD8, CD4, and CD19 (T and B cell markers) when compared to the high ADC group. The low ADC group also had a decrease in FoxP3 expression (regulatory T cell marker) versus the high ADC group, which had an increase in FoxP3 expression. The high ADC group also had a notable increase in CCR4 expression (associated with a TH2 type immune response), while the low ADC group did not have an increase. The combination of an increase in FoxP3 expression and an increase in CCR4 expression suggests that the high ADC group may have a more immune-inhibited microenvironment in response to trastuzumab. Conclusions: These initial analyses suggest that the pre-treatment ADC calculated from DWI may identify HER2+ breast cancers that will have an anti-tumor immune response to trastuzumab therapy. Final data from the full planned cohort will be presented at the meeting and include discussion of the DCE metrics. Based on this preliminary data, further study of breast MRI as an immune biomarker is warranted. Citation Format: Laura Carpin Kennedy, Grace Durenberger, Sasha E. Stanton, Shaveta Vinayak, Suzanne Dintzis, Savannah Partridge, VK Gadi. Breast MRI as a radiomic biomarker of immune response in HER2+ breast cancer [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P6-10-17.

Background: Cell-free DNA (cfDNA) molecules circulating in blood are emerging as important non-invasive biomarkers for monitoring physiological and pathological processes. Tumor-derived circulating cell-free DNA (cfDNA) is present in the plasma of individuals with cancer. Assays aimed at detecting common cancer mutations in cfDNA are being developed for the detection of multiple cancer types. In breast cancer, however, such assays have failed to detect the disease at a sensitivity relevant for clinical use, in part due to the absence of multiple common mutations that can be co-detected in plasma. Unlike individual mutations that exist in only a subset of tumors, unique DNA methylation patterns are universally present in cells of a common type and therefore may be ideal biomarkers. Here we describe the detection and quantification of breast derived cfDNA, using a breast-specific DNA methylation signature. Methods: In order to identify regions of the genome with unique methylation patterns in breast epithelium, we compared genome-wide methylation profiles of normal breast tissue and breast cancer to genome-wide methylation profiles of other normal and cancerous tissues and aimed to identify CpG sites that are uniquely hypermethylated or hypomethylated in the breast. By quantifying the proportion of molecules in a sample that harbored 3 chosen breast-specific methylation sites, we could obtain an estimate of the proportion of the sample derived from breast epithelial cells. We have collected plasma prospectively from healthy women, metastatic breast cancer patients and patients with localized (stage II-III) breast cancer who were assigned to preoperative chemotherapy at our cancer institute. In the latter cohort, samples were collected before and throughout treatment until surgery. The trial was approved by the Institutional Ethics Committee and all patients gave written informed consent. Results: All three markers clearly distinguished between the plasma of healthy donors (n=64) and patients with metastatic disease (n=14, p-value Conclusion: A breast-unique epigenetic signature in circulating cell-free DNA can be used as a universal biomarker for breast cancer, to allow sensitive detection and monitoring of localized breast cancer treatment outcomes. Further research will aid to evaluate this promising and affordable method in additional clinical scenarios. Citation Format: Albert Grinshpun, Aviad Zick, Joshua Moss, Einat Carmon, Myriam Maoz, Bracha Ohana, Ofri Abraham, Lili Germansky, Benjamin Glaser, Ruth Shemer, Yuval Dor, Beatrice Uziely. Non-personalized detection of circulating breast-derived DNA for monitoring of preoperative breast cancer treatment [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P1-10-10.

Background: Triple-negative breast cancer (TNBC) is an aggressive subtype of breast cancer. Identification of biomarkers that are predictive of response to investigational therapies will help identify patients that are more likely to clinically benefit from treatment. Bintrafusp alfa (M7824) is an innovative first-in-class bifunctional fusion protein composed of the extracellular domain of the TGF-βRII receptor (a TGF-β “trap”) fused to a human IgG1 mAb blocking PD-L1. In this expansion cohort of a global, phase 1, open-label trial (NCT02517398), an extensive tumor biomarker analysis was conducted to identify potential markers correlating with efficacy in patients with advanced TNBC treated with bintrafusp alfa. Methods: Patients received infusions of bintrafusp alfa 1200 mg every 2 weeks until confirmed progression, unacceptable toxicity, or trial withdrawal. Tumor archival material was collected from each patient for biomarker analysis. All biomarker analyses were exploratory. PD-L1 expression and immune phenotype were determined by immunohistochemistry. Tumor samples were also processed for RNA sequencing (RNAseq) and resulting data were analyzed to identify potential associations between gene expression and efficacy of bintrafusp alfa treatment. Results: As of August 24, 2018, 33 patients with heavily pretreated TNBC were treated with bintrafusp alfa (54.5% of patients received ≥4 prior regimens). The objective response rate was 9.1% (n=3) and the disease control rate was 15.2% in a total of 5 patients. PD-L1 expression levels in tumor cells and in the tumor microenvironment were generally low in this cohort with the majority of patients having 0% PD-L1+ tumor cells and/or ≤10% PD-L1+ immune cells in their tumor microenvironment. Moreover, several patients had an immune desert phenotype. PD-L1 expression was not associated with response to treatment. Importantly, exploratory analysis of RNAseq data (n=26 passing quality control) identified differential expression of the high mobility group AT-hook 2 (HMGA2) gene with 32.0-fold higher expression (as computed by DESeq2; q=2.23e−13) in patients who experienced disease control (response or stable disease) compared to those who had progressive disease. HMGA2 is a transcriptional regulator whose expression is upregulated by TGF-β signaling. Furthermore, HMGA2 is known to be an important factor in mediating TGF-β-induced epithelial-mesenchymal transition. In The Cancer Genome Atlas dataset, approximately 12% of subjects annotated as TNBC have high expression of HMGA2. Clinical efficacy and safety data for this expansion cohort are presented in a separate abstract. Conclusions: We identified high expression of HMGA2 as a potential predictive biomarker of response in TNBC patients treated with bintrafusp alfa. HMGA2 assay development is on-going to confirm this finding in a future clinical trial. Citation Format: George Locke, Yue Zhang, Laureen S Ojalvo, Christoph Helwig, Alex Rolfe, Olaf Christensen, Isabelle Dussault. Identification of a tumor biomarker in advanced triple-negative breast cancer that predicts response to bintrafusp alfa (M7824), a bifunctional fusion protein targeting transforming growth factor-β and programmed death ligand 1 [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P3-09-13.

Background: Breast cancer arising in young women is more likely to have higher risk features than in older women and more likely to be associated with germline BRCA1 or BRCA2 mutations. We previously reported a higher distribution of luminal B-like tumors among young women. Here we present the clinical and pathologic features of invasive disease in a large prospective cohort of young women with breast cancer, and associations between surrogate molecular subtype and BRCA 1/2 mutation status. Methods: Among 1,297 young women (aged ≤40 years) with breast cancer enrolled from 2006-2016, tumor histopathological features were examined by central review, and biomarker status was extracted from pathology reports. Based on histologic grade and biomarker expression, invasive tumors were categorized as luminal A-like (LA=ER/PR+ and HER2-, histologic grade 1 or 2), luminal B-like (LB=ER/PR+ and HER2+, or ER/PR+, HER2-, and grade 3), HER2 enriched (E) (ER/PR- and HER2+) or triple negative (TN=ER-, PR- and HER2-). Tumor stage, clinical characteristics, and BRCA status were retrieved by medical record review (and patient survey for BRCA status if not in medical record). Results: The majority of women presented with early stage invasive disease (8% stage 0, 32% stage 1, 40% stage 2, 15% stage 3; 5% stage 4). Among invasive tumors with evaluable molecular phenotype (n=1,136), 57% were high grade and the distribution of subtypes was: 33% LA, 42% LB, 8% HER2E, and 16% TN (Table 1). Among different age groups (≤ 30, 31-35, and 36-40 years), there were no significant differences in molecular phenotype, tumor stage, tumor grade or histopathological features including presence of tumor necrosis, lymphocytic infiltration, and central fibrotic focus. 96% of women were tested for BRCA1 and BRCA2 germline mutations, and 127 (11%) tested positive: 65% were BRCA1+, and 35% were BRCA2+ (table 2). Among BRCA1+ patients, 2% of cancers were LA, 30% LB, 5% HER2E, and 62% TN; among BRCA2+, 29% were LA, 56% LB, 2% HER2E, and 13% TN. These proportions were relatively consistent across age groups. BRCA1+ patients’ tumors were proportionately more TN and BRCA2+ patients’ tumors more likely to be LB/HER2 negative (p Table 1. Subtype by age groupTotal (n=1136)≤30 yrs (n=144)31-35 yrs (n=314)36-40 yrs (n=678)p=0.42Luminal A-like377 (33%)48 (33%)99 (32%)230 (34%)Luminal B-like477 (42%)59 (41%)137 (44%)281 (41%)• ER/PR+, HER2-, grade 3245 (22%)27 (19%)68 (22%)150 (22%)• ER/PR+, HER2+232 (20%)32 (22%)69 (22%)131 (19%)HER2E95 (8%)13 (9%)34 (11%)48 (7%)Triple Negative187 (16%)24 (17%)44 (14%)119 (18%) Table 2. Subtype by BRCA statusTotal(n=1136)BRCA1+(n=82)BRCA2+(n=45)No mutation/not tested(n=1009)p Citation Format: Yaileen D Guzman-Arocho, Shoshana M. Rosenberg, Philip Poorvu, Kathryn J. Ruddy, Greg Kirkner, Craig Snow, Rulla M. Tamimi, Jeffrey Peppercorn, Lidia Schapira, Virginia F. Borges, Steven E. Come, Elena F. Brachtel, Ellen Warner, Ann H. Partridge, Laura C. Collins. Clinicopathological features and BRCA 1/2 status in a large prospective cohort of young women with breast cancer [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P4-07-02.

Background: The enumeration of circulating tumour cells (CTCs) using the FDA-cleared CellSearch system has repeatedly been proven to bear prognostic significance in patients with localized and metastatic breast cancer (MBC) (Bidard et al. Lancet Oncol 2014; Janni et al. CCR 2016). In addition, CTC count has been suggested to be a valuable alternative for physician-based decision making in choosing between hormone therapy and chemotherapy as frontline treatment for patients with estrogen receptor-positive, HER2-negative MBC (Bidard et al. SABCS 2018). Challenges for the implementation of CTCs as a biomarker or a source of liquid biopsy in clinical practice include that most isolation methods only detect a limited number of CTCs in a blood sample in a subset of patients, even in those with clinically proven metastatic disease, and the limited availability of comparative data on technical variability between different CTC detection assays. Methods: In this study, we compare CTC detection rates in paired 7.5 ml blood samples from patients with MBC starting a new line of treatment using two commercially available CTC detection assays. All patients are recruited at GZA Hospitals, campus Sint-Augustinus, Antwerp, Belgium. The first CTC detection system is the CellSearch system, currently the only FDA-cleared system for the enumeration of CTCs in patients with MBC. This system employs an immunomagnetic CTC enrichment based on the expression of Epithelial Cell Adhesion Molecule (EPCAM) followed by an immunofluorescent CTC detection based on the expression of cytokeratins (CK) in EPCAM+ enriched cells. With this system, only EPCAM/CK double positive cells can be qualified as CTCs. The second system is the Rarecytes system, an automated scanner for immunofluorescently stained cell smears of the entire buffy coat of blood samples. With this system, CTCs will be enriched and detected in a single-step measurement using a cocktail of anti-EPCAM and anti-CK fluorescently labeled antibodies, thereby allowing to detect also single-marker positive CTCs. Based on CTC detection prevalence data in patients with MBC that are available for the CellSearch system and a prespecified hypothesis that with both detection methods a discordant result (relative to a cut-off of >= or Expected results and conclusions: Upon submission of the abstract, 30 patients with MBC have been included in this trial after an accrual period of 4 months. Full results are expected mid 2020. Detailed comparison of CTC detection rates between both techniques will not only render analytical data on the performance of the Rarecytes system compared to the CellSearch system - the current gold standard in the field - but might also shed light on the frequency of EPCAM/CK single-positive CTCs in patients with MBC. Conflicts of interest: This study is partially funded by a grant from Rarecyte, Inc. Contact details for collaboration: Dieter Peeters (HistoGeneX NV, Antwerp, Belgium): dieter.peeters@histogenex.com; Andy Buys (Translational Cancer Research Unit, Antwerp, Belgium): andy.buys@gza.be. Citation Format: Dieter Peeters, Andy Buys, Hannah Ledegen, Kathlyn Moehlig, Steffi Oeyen, Dieter Rondas, Alexia De Beukelaer, Steven Van Laere, Liesbet Vervoort, Peter Vermeulen, Christopher Ung, Mark Kockx, Luc Dirix. A comparative study of two detection platforms of circulating tumour cells in patients with metastatic breast cancer [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr OT3-18-02.

Background: G-protein coupled estrogen receptor (GPER) is a heptahelical transmembrane receptor that belongs to the G-protein coupled receptor superfamily and promotes specific estrogen binding and transactivation of epidermal growth factor receptors. GPER functions independently of estrogen receptors, ERα and ERβ, and it is commonly expressed in ER-negative breast tumors. Further evidence of its autonomy from ER is derived from the fact that GPER varies directly with advanced breast cancer, a relationship that is directly opposed to that shared by ER, indicating that GPER and ER are independent measures of estrogen responsiveness. Since triple negative breast cancer (TNBC) remains a high-risk subtype with a relative paucity of treatment options and a poorer prognosis compared to hormone receptor (HR) positive breast cancer, we have begun to evaluate GPER expression in TNBC tumors and to assess its correlation with clinicopathologic variables. Methods: Subjects diagnosed with breast cancer between 1996-2018 were identified using existing biorepositories at the University of Iowa. Inclusion criteria were patients diagnosed with TNBC with archived, treatment-naïve breast tumor tissue and adequate clinical data from the electronic health record. Subjects were excluded if they did not meet both those criteria or they had stage IV disease at presentation. GPER expression was determined by immunohistochemistry (IHC) and a semiquantitative scoring system combining staining intensity and extent. Clinicopathologic variables included in the analysis were clinical stage, Elston-Ellis grade, BMI, and age. Comparisons between GPER staining scores and clinicopathologic characteristics were performed using chi-squared (Fisher’s exact where appropriate) and t-tests. Results: Of the 116 subjects we identified, 78 (67%) were GPER-positive. Most of the tumors were clinical stage T1 or T2 (85%) and node negative (79%) while 18% were N1. Breakdown by clinical stage is as follows: 2.6% stage 0, 42.2% stage I, 44% stage II, 6.8% stage III, and 4.3% unknown stage. Of the variables analyzed, clinical stage was found to be significant, with an inverse relationship observed between cancer stage and level of GPER expression (p=0.02). Conclusion: GPER is a potential therapeutic target for TNBC. The majority of TNBC tumors in our sample expressed GPER, a result which suggests that this subgroup of TNBC tumors may be responsive to estrogen-targeted therapy. In addition, correlating GPER expression with clinical outcome would evaluate its usefulness as a biomarker of risk for patients with very early stage TNBC tumors, a group in which personalized treatment options are presently not well defined. Citation Format: Sneha Phadke, Sarah Mott, Amani Bashir, Andrew Bellizzi, Murray Resnick, Ashlee Sturtevant, Edward Filardo. Distribution of G-protein coupled estrogen receptor in treatment-naïve triple negative breast cancer and association with clinicopathologic characteristics [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P2-11-16.

Agents which both reduce risk for development of breast cancer and relieve vasomotor symptoms are likely to have good uptake and adherence. We conducted a pilot study with 6 months of the tissue selective estrogen complex bazedoxifene (20 mg) and conjugated estrogen (0.45 mg) (Duavee®) to assess feasibility and effects on biomarkers. Risk biomarkers for postmenopausal breast cancer included fully automated mammographic volumetric density (Volpara®), benign breast tissue Ki-67, and serum levels of progesterone, IGF-1 and IGFBP3, bioavailable estradiol and testosterone. Exploratory biomarkers included selected estrogen and progesterone responsive gene expression in benign breast tissue. 28 peri- and post-menopausal women at increased risk for breast cancer were enrolled; 13 in Cohort A with baseline Ki-67 < 1% and 15 in Cohort B with baseline Ki-67 of 1-4%. All completed the study with > 85% drug adherence. An improvement in median hot flash score from 15 at baseline to 0 at 6 months, and menopause specific quality of life total, vasomotor and sexual domain scores were also observed (p< 0.001). Significant changes in risk biomarkers, uncorrected for multiple comparisons, were a decrease in mammographic fibroglandular volume (p=0.043); decreases in serum progesterone, bioavailable testosterone, and IGF-1 (p This study was supported in part by grants from the Breast Cancer Research Foundation (BCRF-16-049, BCRF-17-049, BCRF-18-049); and an NIH Clinical and Translational Science Award grant (UL1 TR000001, formerly UL1RR033179) awarded to the University of Kansas Medical Center, and an internal clinical pilot grant program of the KUMC Research Institute. Citation Format: Carol J Fabian, Lauren Nye, Teresa A Phillips, Onalisa Winblad, Carola M Zalles, Christy R Hagan, Merit L Goodman, Byron J Gajewski, Devin C Koestler, Prabhakar Chalise, Bruce F Kimler. Biomarker modulation by bazedoxifene and conjugated estrogen (Duavee®) in women at high risk for development of breast cancer [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr PD3-06.

Introduction: Interpretation of HER2 result in breast cancer patients with monosomy of CEP17 has been controversial due to the loss of CEP17 signals and the subsequent relatively increased HER2/CEP17 ratio. HER2 FISH group 2 (HER2< 4 and HER2/CEP17 ratio ≥ 2) was historically considered HER2 positive (HER2+) irrespective of the negative HER2 overexpression and anti-HER2 therapy response. The updated 2018 ASCO/CAP guidelines incorporated immunohistochemistry (IHC) results to facilitate HER2 reporting. However both IHC and FISH can be subjective to analytical variables, which can lead to challenge in interpretation cases in this group. This study aimed to assess if repeat IHC and FISH tests could assist final HER2 interpretation for group 2 cases. Methods: We retrospectively reviewed HER2 FISH cases performed at our institution from 2012 to 2017 and identified 23 of 3554 (0.6%) breast cancer cases with at least one time HER2 FISH result fell into group 2 category. A chart review for clinicopathological characteristics and biomarker status was performed in all 23 cases. The HER2 IHC and FISH tests were repeated for cases with available alternative tumor samples. Final HER2 interpretation was modified by following 2018 ASCO/CAP guidelines. Results: Patients’ age ranged from 26-67. These tumors were predominantly ductal (n = 17, 74%), high grade (n = 13, 57%), ER positive (n = 17, 74%), and equivocal HER2 (2+) by IHC (n = 15, 60%). These HER2 FISH group 2 cases were from 18 primary tumors and 5 metastatic/recurrent tumors. All 23 cases (100%) were considered HER2 amplified according to 2013 ASCO/CAP guidelines, only 2/23 (9%) can be interpreted as HER2+ in conjunction with their corresponding IHC result according to 2018 ASCO/CAP guidelines. The corresponding primary tumors for the 5 metastatic/recurrent group 2 tumors were 2 HER2+ and 3 HER2-. Repeat HER2 tests were performed in 15 of 18 group 2 primary tumors; the concordance rates between initial and repeat tests were 27% (4/15) for the HER2 FISH group category and 73% (11/15) for the final HER2 results. Three (20%) cases changed from HER2- (group 2 and IHC 2+) to HER2+ (group 1 [HER2/CEP17 ratio ≥ 2 and HER2 ≥ 4] and IHC 2+); one case changed from HER2+ (group 2 and IHC 3+) to HER2- (group 5 [HER2/CEP17 ratio < 2 and HER2< 4] and IHC 1+); and 11 cases remained as HER2- on repeat (4 for group 2 with IHC 1-2+ and 7 for group 5 with IHC 0-2+). Eight of these 15 patients received neoadjuvant chemo and Herceptin therapy, 3 (38%) achieved a complete pathologic response and 5 (62%) did not respond (Table 1). The 3 responders had ductal tumors and at least 40% of Ki67 in contrast the other 5 non-responders with either lobular histology or Ki67 < 40% (p < 0.05). Conclusion: Breast cancer with a group 2 HER2 FISH result represents heterogeneous populations of tumor cells being originated de novo or preferentially selected secondary to chemotherapy, and leading to challenges in HER2 interpretation. Repeat HER2 IHC and FISH test on alternative samples either different tumor block of the same biopsy or tumor from excision may assist in final classification to guide treatment decision. Table 1. Repeat HER test in breast cancer patients with initial HER2 FISH test in group 2CaseInitial test on biopsy (IHC/FISH)Repeat test on biopsy (IHC/FISH)Repeat test on excision (IHC/FISH)Final HER2 resultHistologyGradeERKi-67Response to NAC+H12+/Group 22+/Group 1N/APosDuctal3Low Pos40%pCR22+/Group 22+/Group 1N/APosDuctal3Neg90%pCR32+/Group 22+/Group 1N/APosLobular3Pos5%RCB-II43+/Group 2N/A1+/Group 5PosDuctal3Pos15%RCB II51+/Group 20/Group 5N/ANegDuctal3Pos73%pCR62+/Group 2N/A2+/Group 5NegLobular3Pos10%RCB II72+/Group 2N/A1+/Group 2NegDuctal2Pos38%RCB II81+/Group 2N/A2+/Group 2NegDuctal/Lobular3PosNPRCB IINAC+H: Neoadjuvant chemo and HerceptinpCR: complete pathologic responseRCB: residual tumor burden Citation Format: Hui Chen, Minhua Wang, Hongxia Sun, Jun Gu, Zhenya Tang, Guilin Tang, Steven Sfamenos, Melissa Robinson, Aysegul Sahin. Breast cancer with HER2 FISH test result in group 2: Should HER2 test be repeated? [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P1-18-20.

Background: Stromal Tumour Infiltrating Lymphocytes (sTILs) has been established as a predictive biomarker for response to neoadjuvant chemotherapy irrespective of subtype. While increased sTIL levels is associated with prolonged survival among patients with triple-negative and HER2 positive breast cancer increased sTILs has in Estrogen Receptor positive (ER+) breast cancer been suggested to be an adverse prognostic factor. It has been hypothesised that differences in the cellular composition of immune cells may explain the dissimilarity in prognostic impact according to breast cancer subtype. Here, we report data from a neoadjuvant phase II study, treating postmenopausal patients with primary ER+, HER2 negative, operable breast cancer with letrozole for four months by the Danish Breast Cancer Group (DBCG). We analysed the association of sTILs, and changes in sTILs during neoadjuvant endocrine treatment (NET), with pathological response and correlation with other clinicopathological variables such as Ki67, as potential biomarkers for risk-stratification of patients following NET. Method: 113 postmenopausal patients with ER+, HER2 negative breast cancer were treated with NET for four months. Endpoint was pathological response assessed by a modified Miller Payne scale and the Residual Cancer Burden (RCB) score. Pretherapeutic core biopsies and post therapeutic surgical specimens were assessed centrally for the percentage of sTILs on HE sections according to the International TILs working group guidelines. sTILs association to pathological response were assessed with univariate logistic regression. Odds Ratio (OR) was estimated with a 95 % confidence interval. Correlation of sTILs and Ki67 were tested with Pearson´s correlation coefficient. Changes in sTILs with a paired t-test. Level of significance was set to 5 %. Results: sTILs concentration increased with mean 6.8 % (p The correlation between pre-therapeutic sTILs and Ki67 was moderate (Pearson 0.4; p = 0.0004), however the association grew stronger post treatment (Pearson 0.5; p Conclusion: Increased sTILs activity during NET was associated with poor treatment response. An increase in sTILs during NET could be considered indictive for potential immunogenic tumours, suggesting that patients with increasing sTIL in residual disease after NET might derive benefit from the addition of immunotherapy. We found a significant correlation between sTILs and Ki67 levels. Ki67 has been established as a window of opportunity biomarker for antiproliferative response. Combining biomarkers of antiproliferative response such as Ki67 with biomarkers of immunogenetic significance might be important to determine optimal combination of adjuvant therapy. A window of opportunity study with an aromatase inhibitor could show if it is feasible to detect a rise in sTILs early enough to guide adjuvant treatment in the perioperative setting. Citation Format: Signe Korsgaard Skriver, Maj-Britt Jensen, Ann Soegaard Knoop, Bent Ejlertsen, Anne-Vibeke Laenkholm. Changes in tumour infiltrating lymphocytes during neoadjuvant endocrine therapy and possible clinical implications for guiding therapy [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P4-10-18.

Background: In the FAIRLANE trial (NCT02301988) in early TNBC, adding the oral AKT inhibitor IPAT to neoadjuvant PAC led to numerical increases in rates of pathologic complete response (pCR; primary endpoint) and complete response by MRI in unselected patients, with a numerically greater treatment effect in patients with PIK3CA/AKT1/PTEN-altered tumors [Oliveira et al., Ann Oncol 2019]. The IMpassion130 trial established the efficacy of first-line atezolizumab + nab-PAC in patients with PD-L1-positive advanced TNBC [Schmid et al., NEJM 2018]. Preliminary phase 1b results showed promising activity with a triplet combination of IPAT, atezolizumab, and taxane as first-line therapy for advanced TNBC across PIK3CA/AKT1/PTEN and PD-L1 biomarker subgroups [Schmid et al., AACR 2019]. Using samples from the FAIRLANE trial, PD-L1 expression was characterized to evaluate relationships between IPAT treatment effect and expression of PD-L1 and biomarkers relevant to IPAT. Methods: Pretreatment tumor samples were evaluated for genomic alterations using the FoundationOne® assay (Foundation Medicine). Pretreatment PD-L1 status was determined using the SP142 immunohistochemistry assay (VENTANA Medical Systems), with PD-L1 positivity defined as PD-L1 expression on ≥1% of tumor-infiltrating immune cells. Tumor-infiltrating lymphocytes (TILs) were quantified using the Salgado method [Salgado et al., Ann Oncol 2015]. Gene expression was measured by RNA sequencing. Results: Samples were evaluable for genomic alterations in 144 patients, TILs in 135 patients, gene expression in 111 patients, and PD-L1 expression in 99 patients. The prevalence of PD-L1 positivity was 39%, consistent with the reported prevalence in IMpassion130. Compared with PD-L1-negative tumor samples, PD-L1-positive tumor samples showed significantly higher levels of TILs (mean 37% vs 24%; p=0.004) and significantly higher expression of the genes encoding for PD-L1 (CD274: geometric mean 10.5 vs 3.9 counts per million; p Citation Format: Matthew J Wongchenko, Mafalda Oliveira, Cristina Saura, Paolo Nuciforo, Isabel Calvo, Jay Andersen, José L Passos Coelho, Miguel Gil Gil, Begoña Bermejo, Debra A Patt, Eva Ciruelos, Zhen Shi, Na Xu, Patricia Villagrasa, Aruna Mani, Steven J Isakoff. Expression of PD-L1 is independent of PIK3CA/AKT1/PTEN alterations in triple-negative breast cancer (TNBC) and is not associated with response to ipatasertib (IPAT) plus paclitaxel (PAC) [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P4-10-23.

[background] In general, TNBC has a high distant metastasis rate and is highly aggressive. Despite good response to chemotherapy, it has a poorer prognosis compared to other subtypes. There is a need to develop new treatments for the selection of appropriate treatments. At present, treatments using an immune check point blockade are attracting attention, and the evaluations as biomarkers of and, PD-1 / PD-L1 expression on immune cells for therapeutic effects and prognosis in TNBC are controversial. In order to select patients with TNBC for who immunotherapy is likely to be effective, new biomarkers need to be explored. [Purpose] Our aim is to investigate the possibility as a biomarker of PD-L1 expression on immune cells(PD-L1/IC) in TNBC. We also assess the relationship among PD-L1/IC, PD-L1 expression on tumor cells(PD-L1/TC) and TIL. [Methods] This study included 248 patients with primary TNBC who underwent resection without neoadjuvant chemotherapy at Kyushu University Hospital(Fukuoka, Japan) and related hospitals between January 2004 and December 2014. [Results] PD-L1/IC was positive in 129 cases (52.0%) of 248 TNBCs. Although there was no significant difference in RFS and OS between the positive and negative groups, PD-L1/IC-positive and TIL-High group showed the best prognosis, considering the combination whether PD-L1/IC was positive or negative, and TIL was high or low. PD-L1/IC was significantly correlated with PD-L1/TC(p [Conclusion] In TNBC, the relationship between PD-L1/IC and prognosis is still not clear, but in IMpassion130 trial, PD-L1/IC is a predictive marker for the effect of Atezolizumab. PD-L1/IC has potential as an immune biomarker in TNBCs. This study also found that PD-L1/IC was related to the therapeutic effect of anthracycline. PD-L1/IC may be a predictive marker of the efficacy of various treatments. Citation Format: Yurina Harada, Makoto Kubo, Hitomi Mori, Masaya Kai, Mai Yamada, Kanako Kurata, Hitomi Kawaji, Kazuhisa Kaneshiro, Saori Hayashi, Akiko Shimazaki, Takafumi Morisaki, Rin Yamaguchi, Nobuyuki Arima, Reiki Nishimura, Masayuki Okido, Masafumi Nakamura. PD-L1 expression of tumor infiltrating immune cells has potential as a biomarker in triple-negative breast cancers [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P1-10-21.

Lung cancer is the leading cause of cancer-related deaths worldwide. Cytologic examination is the current “gold standard” for lung cancer diagnosis, however, this has low sensitivity. Here, we identified a typical methylation signature of histone genes in lung cancer by whole-genome DNA methylation analysis, which was validated by The Cancer Genome Atlas (TCGA) lung cancer cohort (n = 907) and was further confirmed in 265 bronchoalveolar lavage fluid samples with specificity and sensitivity of 96.7% and 87.0%, respectively. More importantly, HIST1H4F was universally hypermethylated in all 17 tumor types from TCGA datasets (n = 7,344), which was further validated in nine different types of cancer (n = 243). These results demonstrate that HIST1H4F can function as a universal-cancer-only methylation (UCOM) marker, which may aid in understanding general tumorigenesis and improve screening for early cancer diagnosis. Significance: These findings identify a new biomarker for cancer detection and show that hypermethylation of histone-related genes seems to persist across cancers.

The genetic events occurring in recurrent nasopharyngeal carcinoma (rNPC) are poorly understood. Here, we performed whole-genome and whole-exome sequencing in 55 patients with rNPC and 44 primarily diagnosed NPC (pNPC), with 7 patients having paired rNPC and pNPC samples. Previously published pNPC exome data were integrated for analysis. rNPC and pNPC tissues had similar mutational burdens, however, the number of clonal mutations was increased in rNPC samples. TP53 and three NF-κB pathway components (TRAF3, CYLD, and NFKBIA) were significantly mutated in both pNPC and rNPC. Notably, mutations in TRAF3, CYLD, and NFKBIA were all clonal in rNPC, however, 55.6% to 57.9% of them were clonal in pNPC. In general, the number of clonal mutations in NF-κB pathway–associated genes was significantly higher in rNPC than in pNPC. The NF-κB mutational clonality was selected and/or enriched during NPC recurrence. The amount of NF-κB translocated to the nucleus in samples with clonal NF-κB mutants was significantly higher than that in samples with subclonal NF-κB mutants. Moreover, the nuclear abundance of NF-κB protein was significantly greater in pNPC samples with locoregional relapse than in those without relapse. Furthermore, high nuclear NF-κB levels were an independent negative prognostic marker for locoregional relapse-free survival in pNPC. Finally, inhibition of NF-κB enhanced both radiosensitivity and chemosensitivity in vitro and in vivo. In conclusion, NF-κB pathway activation by clonal mutations plays an important role in promoting the recurrence of NPC. Moreover, nuclear accumulation of NF-κB is a prominent biomarker for predicting locoregional relapse-free survival. Significance: This study uncovers genetic events that promote the progression and recurrence of nasopharyngeal carcinoma and has potential prognostic and therapeutic implications. See related commentary by Sehgal and Barbie, p. 5915

Genomic instability is a key feature of carcinogenesis, giving rise to tumour promoting aberrations. The inherent unpredictability of these processes leads to a high degree of intratumour heterogeneity, previously shown to be advantageous for cancer progression in a series of cancer types. We sought to explore the role of genomic instability and intratumour heterogeneity in the progression of ductal carcinoma in situ (DCIS) to invasive disease, in a cohort of 49 women with pure DCIS that developed subsequent ipsilateral invasive disease and 90 women with pure DCIS with no evidence of recurrence with a median follow up of 5.4 years. All samples were identified from within the Sloane project, a prospective national cohort study of DCIS within the UK National Health Service Breast Screening Programme. DNA extracted from bulk tissue samples was processed on the HumanCytoSNP array, and allele-specific copy number aberration (CNA) data was obtained from ASCAT. We measured genomic instability using Scores Of Chromosomal Instability Scarring (SCINS), as well as by calculating the Shannon diversity index for the different copy number states occurring in the tumour genome. We also attempted to deconvolute the clonal structure of the tumours by applying the Shannon diversity index on the raw copy number estimates provided by ASCAT. Non-integer copy number states were interpreted as indicative of the presence of subclonal CNAs. As a positive control, we compared ER+ and ER- DCIS in a larger cohort of 173 DCIS samples (129 ER+ and 44 ER-), using the same methods. We expect ER- tumours to be more genomically unstable, and consequentially more clonally diverse. Measured using SCINS, both copy-neutral loss of heterozygosity (≥ 4 Mbp) and allelic-imbalanced CNAs (≥ 8 Mbp) were shown to be higher in ER- tumours (Fisher’s exact test, p = 0.014 and p = 0.034, respectively). Overall genomic instability was also shown to be higher in ER- tumours using the Shannon diversity index (Fisher’s exact test, p = 0.0085). Our attempt at clonal structure deconvolution also indicates increased heterogeneity in ER- tumours (Fisher’s exact test, p = 0.027). However, no significant difference was identified in the genomic instability or heterogeneity of DCIS that developed subsequent invasive disease compared to that of indolent DCIS, using any of the aforementioned measures. We were therefore unable to identify evidence of a direct link between DCIS heterogeneity and progression to invasive cancer. Our methods did, however, demonstrate that DCIS in general is highly genomically diverse, with ER- tumours exhibiting increased diversity compared to ER+ ones. We aim to validate these findings using spatial sampling data from our cohort, as well as integrating mutation data to allow the use of established clonal heterogeneity deconvolution pipelines. Citation Format: Anargyros Megalios, Vandna Shah, Rana Shami, Salpie Nowinski, Jelmar Quist, Mathini Sridharan, Carolina Salinas de Souza, Karen Clements, Anita Grigoriadis, Sarah Pinder, Alastair Thompson, Elinor Sawyer. Heterogeneity of DCIS does not appear to be a biomarker for development of subsequent invasive cancer [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P4-06-05.

Posttranslational modifications of proteins have been implicated in pathogenesis of numerous diseases. Arginine deimination (also known as citrullination) has a principal role in progression of rheumatoid arthritis through generation of autoantibodies and exacerbation of the inflammatory response. Recently, multiple research groups provided solid evidence of citrullination being in control of cancer progression; however, there is no comprehensive overview of these findings. This article summarizes and critically reviews the influence of citrullination on different aspects of tumor biology, including (i) regulation of apoptosis and differentiation, (ii) promoting EMT and metastasis, and (iii) potential use of citrullinated antigens for immunotherapy. In addition, (iv) the role of citrullination as a cancer biomarker and (v) implication of neutrophil extracellular traps in tumorigenesis are discussed. In summary, current findings testify to the significance of arginine deimination in tumor biology and thus more basic and translational studies are needed to further explore this topic.

Background: Meta-analysis of clinical trials has shown that adjuvant bisphosphonates reduce bone metastases and improve survival in postmenopausal (PM) breast cancer and are now recommended for routine clinical use by international guidelines1. However, evaluation of menopause is imprecise and the biological rationale for lack of benefit in premenopausal women unclear. To address this, the biomarker, transcription factor MAF on 16q23 was tested retrospectively in the prospective randomized AZURE trial of standard adjuvant therapy +/- zoledronic acid (ZOL). Initial evaluation indicated that women with MAF negative tumors treated with ZOL had a lower rate of disease relapse irrespective of menopausal status.2 Here we present the long-term findings of this predictive biomarker on 10 year overall survival. Materials and methods: The biomarker analysis was completed on TMAs from primary tumors. Quadruplicate cores of breast tumor tissue were arrayed across replicate TMAs. MAF+ was detected using a validated (MAF/D16Z3) FISH test (Inbiomotion SL, Spain). A central laboratory (Targos, Germany) validated the assay for analytic and diagnostic performance, established acceptance criteria, included appropriate quality controls for each assay, and performed the analyses in a blinded fashion. A copy number cut-off ≥2.5 was preset for MAF+ for both prognostic and predictive testing. Interactions between MAF+ and effects of ZOL on Invasive disease free (IDFS), overall (OS) survival and time to bone metastases by menopausal status were evaluated using a Cox proportional hazards model. Results: 1769 of the 3360 AZURE pts donated primary tumor samples. Median follow-up was 117 (interquartile range 70.4-120) months. 865 pts (49%) had 2 FISH evaluable cores and were included in the analysis. These pts had similar disease and treatment characteristics to the overall study population as well as similar IDFS and OS at 10 years. 184 (21%) had MAF+ tumors and these tumors were more likely to be of higher grade, ER-ve and HER2+. In 680 pts with MAF- tumors, ZOL was associated with improved IDFS (HR=0.75; 95%CI:0.58-0.97, [P=0.02]), reduced relapse in bone (HR=0.65; 95%CI:0.45-0.94, [P=0.022] and, most importantly, better OS (HR=0.69; 95%CI:0.50-0.94, [P=0.019]). In the 185 patients with MAF+ tumors, there was a suggestion of worse outcome (IDFS HR=1.54; 95%CI:0.96-2.47 and OS HR 1.40; 95%CI:0.83-2.33), with a strong interaction between treatment effects and menopausal status. Outcomes in ZOL treated MAF+ pts who were non-PM appeared to be much worse (IDFS HR=2.31; 95%CI:1.18-4.42] and OS HR=2.28; 95%CI:1.07-4.82) due predominantly to an excess of extra-skeletal metastases in ZOL treated patients (HR=4.47; 95%CI:1.66-12.57). Conclusions: Adjuvant ZOL significantly improved disease outcomes in 79% of patients with MAF negative tumors, irrespective of menopausal status and other clinico-pathologic features. Conversely, more extra-skeletal metastases and breast cancer deaths were seen in women with MAF+ tumors who were not PM at the start of treatment. If validated in ongoing studies, the MAF FISH test could provide a clinically useful biomarker for selection of patients for adjuvant bisphosphonate treatment. 1EBCTCG, Lancet 2015; 2Coleman RE et al, Lancet Oncol 2017 Citation Format: Coleman R, Gregory W, Jean-Mairet J, Tercero JC, Torres-Martin J, Gomis R. Long term survival benefits of adjuvant zoledronic acid associated with maf status of primary tumor [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 P1-17-01.

Background: The epichaperome is a new cancer target required for tumor survival (Joshi et al. Nature Reviews Cancer 2018). PU-H71 is a synthetic, purine scaffold epichaperome inhibitor that binds to the ATP-binding site of HSP90 specifically when HSP90 is integrated into the epichaperome (Rodina et al. Nature 2016). It has demonstrated antitumor activity in multiple xenograft models. Furthermore, sequential administration of nab-paclitaxel and PU-H71 in TNBC xenograft models augmented epichaperome levels, and in turn resulted in super-synergistic drug action with ablation of xenografted tumors and cures in mice. Methods: This is an open label phase1b study of PU-H71 + nab-paclitaxel in pts with HER2- MBC. Pts received nab-paclitaxel at a standard dose of 260mg/m2 IV Q 3weeks. PU-H71 was administered IV 6 hrs (+/-1 hr) post nab-paclitaxel Q3weeks in 2 escalating dose levels (225mg/m2 and 300 mg/m2). All pts underwent FDG PET/CT every 6 weeks. Additionally, patients had the option to enroll on a separate diagnostic PU-PET protocol to measure epichaperome expression prior to initiating treatment on the phase 1b study, wherein they received a single dose of up to 11mci of 124I-PU-H71 IV and underwent imaging at 3-4hrs and 20-24 hrs. Primary objective was to establish the MTD/RP2D of this regimen. Secondary objectives were to assess PK of PU-H71 + nab-paclitaxel and clinical efficacy. Exploratory analysis included correlation of epichaperome expression at baseline using PU-PET with tumor response. Results: 12 patients (5 ER+/HER2- ; 7 TNBC) were enrolled (6 at 225mg/m2 of PU-H71 and 6 at 300mg/m2). Median Age: 54 yrs (range: 37-71). Median ECOG: 0. Median lines of therapy in the metastatic setting: 6 (range 1-11) including prior taxanes in 75% of pts. Most common toxicities included diarrhea G1 58%; G2 7%, G3 7%) that was easily managed with anti-diarrheal agents, G1 fatigue (25%), G1/2 peripheral neuropathy (17%), G1 hyperglycemia (67%), G1 increases in alk phos (58%), AST (50%) and ALT (42%). Hematological toxicities included G3 leukopenia (42%), G3/4 neutropenia (67%), G3 anemia (50%) and G2 thrombocytopenia (17%). There were no DLTs. 33% (4/12) had PR, 58% (7/12) achieved SD with only 1 PD at the time of first scan; 5 pts are currently ongoing including 2 TNBC pts with PR who have been on therapy > 7 months. PK data will be presented. 8/12 patients also underwent PU-PET at baseline. A higher tumor to muscle SUV ratio at 24 hrs on PU-PET predicted response and increased PU-H71 retention on PU-PET at 24 hrs correlated with a longer duration of response. Conclusion: The RP2D of PU-H71 was 300mg/m2 with 260mg/m2 of nab-paclitaxel administered IV every 3 weeks. The regimen is well tolerated with promising clinical activity in this heavily pre-treated cohort. Tumor epichaperome expression at baseline using PU-PET has the potential to serve as a predictive biomarker of response. A Phase 2 trial of this combination along with baseline PU-PET is currently planned. Citation Format: Jhaveri K, Dunphy M, Wang R, Comen E, Fornier M, Moynahan ME, Bromberg J, Ma W, Patil S, Taldone T, Rodina A, Sterlin V, Khoshi S, Lewis J, Norton L, Chiosis G, Modi S. Tumor epichaperome expression using 124I PU-H71 PET (PU-PET) as a biomarker of response for PU-H71 plus nab-paclitaxel in HER2 negative (HER2-) metastatic breast cancer (MBC) [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 P6-20-03.

Purpose: Neo-Adjuvant chemotherapy treatment is increasingly being used in breast cancer to preoperatively shrink tumour volumes and facilitate surgical plans. These datasets however, are still scarce, making it difficult to assess the relative value of multiple time point biopsies compared to diagnostic only sampling. This study aims to identify sequential samplings intrinsic value. Method: A total of 97 samples from a cohort of 50 neoadjuvant chemotherapy treated primary breast cancer patients (aged 29-76 at diagnosis, Allred status 47:53% +/-, Her2 status 80:20% +/-, mixed grade and menopausal status) taken pre- treatment, at 2 weeks on-treatment, mid chemotherapy and at resection were sequenced with Ion Ampliseq transcriptome yielding expression values for 12,635 genes. Differential expression analysis was performed across response groups (16 Responders, 34 Non-Responders) as defined by Pathological Complete Response and over treatment time to identify significantly differentially expressed genes, pathways and markers indicative of response status. Results: An on-treatment marker for response was identified (AAGAB), which resulted in a testing accuracy of 100% and a validation accuracy of 78% in the I-SPY 1 Trial. AAGAB was predictive of long term survival (p = 0.048 testing, p = 0.031 validation) in both chemotherapy cohorts at the same expression level as defined for treatment response. The single gene on-treatment biomarker, AAGAB proves more performant than established prognostic tests, Mammaprint (Edinburgh NEO trial, pre-treatment 61%, on-treatment 63%. I-SPY 1 trial, pre-treatment 60%, on-treatment 66%) and Pam50 RORS (neo trial pre-treatment 50%, on treatment 58%, Magbanua trial pre-treatment 56%, on-treatment 64%) Conclusion: Changes in gene expression of on-treatment chemotherapy breast cancer resulted in the identification of a novel gene marker that was as effective in predicting prognostic status as established prognostic tests. These results support the use of on-treatment testing in breast cancer to improve the accuracy of tumour response prediction. Citation Format: Bownes RJ, Turnbull AK, Cameron D, Sims AH, Oikonomidou O. On-treatment biomarkers can improve prediction of response to neoadjuvant chemotherapy in breast cancer [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 P3-11-13.

Background: LIV-1 is an estrogen-inducible gene that has been implicated in epidermal-to-mesenchymal transition (EMT) in preclinical models of progression and metastasis. Its expression is associated with node-positivity in breast cancer; and has been detected in a variety of cancer types, including estrogen receptor positive breast cancers. SGN-LIV1A is a novel antibody drug conjugate targeting LIV-1 that is currently being evaluated in the I-SPY 2 TRIAL. In this pilot study, we evaluated LIV-1 levels by IHC within HR/HER2/MammaPrint (MP) defined subtypes among patients screening for the I-SPY 2 TRIAL and its correlation to microarray assessed LIV-1 expression levels. Method: In a pilot study, LIV-1 IHC staining was performed by Quest Diagnostics on the pre-treatment samples of 38 patients screening for the I-SPY 2 TRIAL. Pre-treatment expression data generated on a custom Agilent 44K platform was also available. We summarized the LIV-1 H-Scores and percent (%)-positivity across the population and within HR/HER2/MP subtypes; and we assessed the Pearson correlation between LIV-1 H-Score and LIV-1 gene expression levels. In addition, we compared the pre-treatment LIV-1 expression levels within HR/HER2/MP subtypes across I-SPY 2 TRIAL patients from completed arms and their relevant controls (n=989) using ANOVA and post-hoc Tukey tests. Our statistics are descriptive rather than inferential; and does not take into account multiplicities of other biomarkers outside of this study. Results: Of the 38 patients evaluated, 37 have LIV-1 %-positivity > 0; and 18 (47%) have 100% LIV1 positivity. The median LIV-1 H-Score is 200; and 89% of patients (34/38) have moderate/high LIV-1 staining (with H-Score≥100). Of the 34 patients who proceeded onto the trial (and have known HR/HER2/MP status), 9 are triple negative, 19 are HR+HER2-, and 6 are HER2+. Due to our small sample size, we did not further subset the triple negative and HER2+ cases; but within the HR+HER2- patients, 10 are MP1 compared to 9 who are MP2 class. LIV1 H-Score appears highest within the HR+HER2-MP1 cases (median: 290), followed by the HER2+ (median: 216), then the HR+HER2-/MP2 (median: 155), and the TN (median: 120) subtype. LIV1 H-score is significantly correlated with LIV-1 mRNA expression levels (Rp=0.79, p Conclusion: Our result suggest that although LIV-1 expression differs by subtype, it is expressed at a moderate/high level in the majority of patients. The good correlation between IHC and array-based LIV-1 expression levels enables us to leverage the entire existing I-SPY 2 dataset and confirm the high rates of LIV-1 expression across the I-SPY 2 population. Further studies to evaluate LIV-1 expression as a biomarker of response to LIV-1 targeting therapies for the neoadjuvant treatment of breast cancer are warranted and ongoing in I-SPY 2. Citation Format: Yau C, Brown-Swigart L, Asare S, I-SPY 2 TRIAL Consortium, Esserman L, van' t Veer L, Beckwith H, Forero A, Rugo H. LIV-1 expression in primary breast cancers in the I-SPY 2 TRIAL [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 P3-10-14.