Carter, Jodi M.

Abstract Introduction: Stromal tumor-infiltrating lymphocytes (sTILs) have established prognostic and predictive significance in triple-negative breast cancer (TNBC). However, the roles of other immune cells in TNBC are less well-established. We performed high-plex quantitative spatial profiling in a cohort of early-stage TNBC to 1) apply spatial context to tumoral immune landscapes and 2) identify immune proteins associated with clinical outcomes, independently of TILs and other established prognostic clinicopathologic variables, in patients (pts) treated with or without adjuvant chemotherapy (CTX). Methods: The Mayo TNBC cohort comprises pts with centrally-verified, CTX-naive tumors resected from 1985-2012. Using a cohort-based TMA, with Nanostring GeoMX DSP, we quantitated 58 proteins within spatially-distinct intra-epithelial, cytokeratin-positive tumor segments and adjacent cytokeratin-negative/nuclei-positive stromal segments. Differentially-expressed (DE) proteins were identified using a negative binomial generalized linear model (SNR>2, p< 0.05) and a target DE protein set was dichotomized (80th percentile). After adjusting for prognostic clinicopathologic variables, proteins associated with recurrence-free survival (RFS, defined as time from surgery to either local, regional, and distant recurrence, or death by any cause) were identified by performing variable selection using the Akaike Information Criterion (AIC) obtained from fitting all possible Cox proportional hazards regression models (performed separately for intra-epithelial/stromal segments, and in groups +/- adjuvant CTX. Results: From the TNBC TMA, DSP data (N=250 tumors) included 169 pts who received adjuvant CTX+ and 81 who did not (CTX-). Overall, 85/250 developed recurrent disease. In the CTX+ group, intra-epithelial tumor segments from pts without recurrent disease were enriched in 10 immune proteins, including CD8, markers involved in antigen presentation/dendritic cells (CD11c, CD40, HLA-DR) or NK cells (CD56) (FC: 1.4-2.1, p<0.05); CD14 was increased in stroma (FC: 1.5, p<0.05). In contrast, in the CTX- group, both the intra-epithelial tumor and stromal segments from pts without recurrences were enriched in immune proteins (N= 12 and 15 respectively; FC 1.6-5.5, p< 0.05) most markedly CD40, IDO1 and HLA-DR (FC: 3.2-5.5, p< 0.05). Overall, CD3, CD4, CD27, CD44, and ICOS among others were enriched only in the CTX- group; CD14 and CD56 were enriched only in the CTX+ group. Based on these spatial data, biologic function and DSP data from another set of TNBC (FinXX trial), CD11c, CD14, CD27, CD40, CD56, and IDO1 were selected for RFS analysis. After applying our model selection criterion and adjusting for pt age at surgery, tumor size, lymph node status, and sTILs, intra-epithelial CD56 was independently associated with improved RFS in the CTX+ group (HR: 0.31[0.12, 0.81]). In the CTX- group, intra-epithelial CD11c was independently associated with improved RFS (0.10 [0.01, 0.81]). Conclusion: In this early-stage TNBC cohort, spatially-distinct tumor immune landscapes were associated with RFS but differed according to receipt of CTX after surgical resection. In the patients who received CTX, the intra-epithelial compartment, rather than stromal compartment, was immune-enriched in pts without recurrences. Among a targeted protein set, intra-epithelial CD56 remained associated with improved outcomes, independent of sTILs and other clinicopathologic features. In the CTX- group, spatial landscapes were more balanced, and intra-epithelial CD11c was independently associated with improved outcomes. These data provide insight into the spatial context of intrinsic immune landscapes in TNBC, and identify candidate prognostic immune biomarkers which may inform therapeutic strategies. Citation Format: Jodi M Carter, Saranya Chumsri, David W Hillman, David M Zahrieh, Yaohua Ma, Xue Wang, Jennifer M Kachergus, Judy C Boughey, Minetta C Liu, Krishna R Kalari, JC Villasboas, Roberto A Leon Ferre, Fergus J Couch, Matthew P Goetz, E. Aubrey Thompson. Intra-epithelial tumor immune landscapes are associated with clinical outcomes in early-stage triple-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 PS16-01.

Abstract Introduction: The immune microenvironment of high-risk HR+ breast cancer (BC) is poorly understood. BEAUTY is a prospective neoadjuvant chemotherapy (NAC) study of stage I-III BC patients treated with neoadjuvant weekly taxane followed by anthracycline-based chemotherapy. Among clinical luminal B BC (St. Gallen criteria) from BEAUTY, we used high-plex digital spatial profiling (DSP) to characterize the intra-epithelial tumor and stromal immune landscapes to 1) assess the impact of NAC on these landscapes and 2) identify immune biomarkers predictive of response to NAC. Methods: The tissue set included FFPE sections of resected tumors from 35 patients (median 51y; range: 21-71y) with clinically-defined luminal B BC (ER > 10%/grade 2/Ki67 ≥ 15% or ER > 10%/grade 3), and 16 paired pre-NAC biopsies. Nanostring GeoMX DSP was used to quantitate 58 immune and BC biomarker proteins in intra-epithelial, cytokeratin-positive tumor segments and adjacent stromal (cytokeratin-negative/SYTO13 (nuclear stain)-positive) segments. Data were normalized using the geometric mean of two negative controls (RbIgG and MsIgG1). A general linear model with negative binomial identified differentially-expressed (DE) proteins in pre/post-NAC tumors. Based on DE protein data and biologic function, a targeted protein subset (N=19) was evaluated in pre-NAC biopsies for associations between protein abundance and residual cancer burden (RCB) class (0-II vs. III) (Wilcoxon-rank sum test: p-value < 0.025 considered significant) or RCB ‘breast only’ index (Spearman correlation). Results: Comparing tumor segments in pre-NAC specimens, intra-epithelial segments were predictably enriched in cytokeratin, ER, PR and Ki-67, but also CD127 and NY-ESO-1; whereas stromal segments were enriched in proteins associated with T cell subsets (e.g. CD3, CD4, CD8), macrophages (CD68 and CD163), antigen presentation/dendritic cells (CD11c, HLA-DR) and immune checkpoint proteins (PD-L1, PD-L2, B7-H3, TIM-3, VISTA) among others. While NAC did not alter the spatial distribution of proteins (intra-epithelial vs. stromal segments), it markedly attenuated the immune landscape, with decreased abundance of functionally-diverse immune proteins (e.g. CD45, CD3, CD4, CD127, granzyme B, CTLA4; STING, B7-H3, CD11c, and CD68, log2FC: 0.27-1.52 p < 0.05), including low abundance proteins PD-1, PD-L1, PD-L2, CD20, and OX40L (log 2FC: 0.15-1.16, p < 0.05). Both PR and Ki-67 decreased in post-NAC tumors whereas ER was not significantly altered (p < 0.05). CD8, CD14, CD163, HLA-DR, IDO-1 and TIM-3 were not significantly altered by NAC. In the pre-NAC biopsies (which had subsequent residual tumor burdens of RCB class 0 or I (n=1 each), II (n=6), and III (n=8), and median RCB ‘breast only’ index of 3.39 (range: 0.00-38.02), 19 proteins were used for RCB analysis (CD3, CD4, CD8, CD14, CD34, CD44, CD45, CD68, CD127, CD163, CTLA4, granzyme B, STING, B7-H3, fibronectin, Ki-67, HLA-DR, SMA and TIM3). Among them, stromal CD127 was significantly higher in RCB class III than RCB class 0-II (p=0.021) and RCB ‘breast only’ index was positively correlated with intra-epithelial granzyme B (rho = 0.61; p=0.012), and negatively correlated with intra-epithelial Ki-67 (rho= -0.71; p=0.0022). Conclusion: In this series of clinical luminal B BC, NAC markedly attenuated the tumoral immune landscapes with a small set of “NAC-resistant” immune proteins. Among a targeted set, stromal CD127 was significantly higher in RCB III, and RCB breast-only index was positively correlated with intra-epithelial granzyme B. These data provide insight into the impact of NAC on HR+ BC, and identify potential immune biomarkers to predict response to neoadjuvant chemotherapy. Citation Format: Jodi M Carter, Judy C Boughey, Jun He, Vera J Suman, Xue Wang, Jennifer M Kachergus, Krishna R Kalari, Liewei Wang, Richard Weinshilboum, Ann M Moyer, Sarah A McLaughlin, Alvaro Moreno-Aspitia, Donald W Northfelt, Richard J Gray, James N Ingle, E. Aubrey Thompson, Matthew P Goetz. Neoadjuvant chemotherapy selectively alters spatially-defined immune landscapes in clinical luminal B HR+/HER2- breast cancers: Analysis of the breast cancer genome guided therapy study (BEAUTY) [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 PD7-05.

Abstract Background: Patient derived xenografts are a standard tool used to study cancer biology and evaluate drug response phenotypes. It is well known that PDX take rate is associated with more aggressive clinical features [higher grade, biologic subtype (e.g. TNBC vs luminal)]. Previous data from a small retrospective cohort of patients (pts) with no prior chemo (n=24) suggested that primary breast tumor PDX take rate was associated with reduced overall survival. (Derose Nat Med. 2011) However, whether take rate is prognostic in newly diagnosed pts receiving standard of care neoadjuvant chemotherapy (NAC) is unknown. From the prospective BEAUTY clinical trial of pts treated with NAC in which PDX was attempted from pre-NAC as well as after chemo (post-NAC), we recently reported no difference in take rate comparing pts with or without a pathologic complete response (pCR) (Yu Breast Canc Res 2018). Herein we report the association between PDX take rate and recurrence from tumor samples implanted pre-NAC as well as post-NAC (time of surgery). Methods: The BEAUTY study is a prospective NAC study which enrolled breast cancer pts (Stage I-III; n=140) treated with neoadjuvant weekly taxane +/-trastuzumab followed by anthracycline-based chemotherapy. Percutaneous tumor biopsies were obtained prior to NAC and tumor samples from residual disease at surgery were additionally used to establish PDXs. Tumor take rate was defined as percent of pts with the development of at least one stably transplantable (passed at least for four generations) xenograft that was pathologically confirmed as breast cancer. Time to breast cancer recurrence was defined as the time from surgery to documentation of a local, regional, or distant recurrence. Gray’s test was used to assess whether the cumulative incidence (CI) of a breast cancer recurrence differs with respect to either pre-NAC PDX take or post-NAC PDX take. Results: Of 140 pts enrolled in the BEAUTY study, tumor tissue for PDX from the pre-NAC tumor was available for implantation in 113. As previously published, PDX take rate from pre-NAC tumor was 27.4% (31/113), and varied according to tumor clinical subtype [51.3% (20/39) in triple negative breast cancer (TNBC), 26.5% (9/34) in HER2+, 5.0% (2/40) in Luminal]. With median follow up of 5.7 years (range: 3 months to 6.75 years), 17 pts developed local, regional or distant disease relapse (4 of whom had a PDX established pre-NAC). The cumulative incidence of breast cancer relapse after surgery was not found to differ according to pre-NAC PDX take (5 yr CI rate: 13.6% no take; 13.4% take; p=0.8911). In pts with TNBC, the group with the highest take rate, there was also no significant difference in incidence of BC recurrence between those without PDX take and those with PDX take (5 yr CI rate: 21.4% vs 16.2% p=0.7314). PDXs were established from residual tissue from surgery (post-NAC) in 6 of 34 pts (17.6%), [specifically, TNBC: 5/9; Her2+: 1/8; and Luminal: 0/17]. Nine of these 34 pts developed a local, regional or distant recurrence. There was a tendency for pts who had a PDX established from their residual disease to have a higher incidence of a breast event (p=0.1092). The 5 year cumulative incidence of a breast event was 19.6% for pts whose post-NAC PDX did not take and 50.0% for pts whose post-NAC PDX took. Discussion: In pts receiving NAC for breast cancer, we observed no significant difference between establishment of a pre-NAC PDX and breast cancer relapse. In contrast, post-NAC PDX take rate (from residual tumor obtained from the breast at surgery) was associated with a tendency to have a higher incidence of a breast cancer event. PDXs remain a valuable tool for the evaluation of tumor biology and development of new therapeutics, especially those models established from pts with chemotherapy resistance. Citation Format: Judy C. Boughey, Vera J. Suman, Jia Yu, Katelyn Santo, Jason P. Sinnwell, Jodi M. Carter, Krishna R. Kalari, Xiaojia Tang, Sarah A. McLaughlin, Alvaro Moreno Aspitia, John A. Copland, III, Donald W. Northfelt, Richard J. Gray, Katie N. Hunt, Amy Lynn Conners, Richard Weinshilboum, Liewei Wang, Matthew P. Goetz. Association between patient derived xenograft (PDX) take rate and breast cancer recurrence in the prospective breast cancer genome guided therapy study (BEAUTY) [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 PD7-04.

Abstract Background: In ER+ breast cancer (BC) models, Aurora A kinase (AURKA) activation is associated with expansion of CD44+/CD24low/- tumor initiating cells, down-regulation of ERα, and endocrine therapy resistance. Alisertib, a selective AURKA inhibitor, can restore ERα expression and endocrine sensitivity. Early phase studies evaluating alisertib alone or with fulvestrant for ER+ metastatic BC (MBC) demonstrated a favorable safety profile and promising antitumor activity [Haddad, Breast Cancer Res Treat. 2018]. A phase II trial was conducted to determine if the addition of fulvestrant to alisertib improved objective response rate (ORR) and to assess clinical activity of alisertib alone or with fulvestrant in patients (pts) with prior fulvestrant and CDK 4/6 inhibitor (CDK 4/6i). Methods: Pts were randomized 1:1 to Arm A, alisertib (50 mg PO BID Days 1-3, 8-10, 15-17 q 28 days) or Arm B, alisertib with fulvestrant (500 mg IM Day 1, 15 for Cycle 1 and q 28 days thereafter). Eligibility included postmenopausal women, history of ER+ BC, prior fulvestrant, ≤ 2 prior chemotherapy lines, and measurable disease. Stratification factors included prior CDK4/6i, ER level (<10%, ≥10%), and primary/secondary endocrine resistance. Pts on Arm A could cross over to Arm B at progression. With 45 pts per arm, a one sided alpha=0.15 sequential binomial test would have an 85% chance of detecting an increase of ≥ 20% in the ORR of ArmB when the true ORR for Arm A is ≤ 20%. ORR was defined as a partial response (PR) + complete response (CR) by RECIST v.1.1 criteria. Secondary endpoints include progression free survival (PFS), 24-week clinical benefit rate (CBR = CR + PR + absence of progression for> 6 cycles), overall survival, duration of response (DoR), and safety. Blood and tumor specimens were collected at baseline, end of Cycle 1, and progression. Results: Pts enrolled July 2017 - November 2019 with 118 pre-registered, 96 registered, and 90 evaluable for the primary endpoint (Arm A: 45, Arm B: 45). Median age was 60 (range 33, 85). Nearly all received prior fulvestrant (n=89, 98.9%), aromatase inhibitor (n=83, 92.2%), and CDK4/6i (n=88, 97.8%). Most had secondary endocrine resistance (n=71, 78.9%). Pre-registration biopsy for ER was positive in 84 pts (86.7%) and negative in 6 pts (13.3%). More pts on Arm B had prior everolimus (A: 35.6%, B: 57.8%) and prior chemotherapy (A: 44.4%, B: 55.6%) for MBC. The ORR for alisertib and fulvestrant was 20.0% (90% CI: 10.9-32.3%), not significantly greater than alisertib alone 17.8% (90% CI: 9.2-29.8%). The 24-week CBR for Arm A was 42.2% (90% CI: 29.7-55.6%; n=19, including 7 PR) and Arm B was 31.1% (90% CI: 19.9-44.3%; n=14, including 8 PR). As of July 1, 2020, the median DoR was not reached in either arm. The median PFS time was 5.6 months (95% CI: 3.9 - 9.3) for Arm A and 5.1 months (95%CI: 3.8 - 7.6) for Arm B. Seventeen pts (18.9%) remain on treatment (A: 12, B: 5) having received at least 11 cycles (range up to 32+ cycles). At least one dose reduction was required for pts (A: 19, B: 18), most commonly due to neutropenia. The most common severe (grade ≥3) adverse events included neutropenia (n=19, 42.2% in each arm), anemia (A: 15.6%, B: 8.8%), and fatigue (n=5, 11.1% Arm B only). Pts discontinued therapy due to disease progression equally in each arm (n=28, 62.2%), however more pts on Arm B (n=12) than Arm A (n=5) discontinued therapy due to toxicity, refusal or other reasons. Conclusion: While the addition of fulvestrant to alisertib did not improve ORR, promising clinical activity with alisertib monotherapy was observed overall and notably for pts with endocrine and CDK 4/6i-resistant MBC. More severe toxicities and treatment discontinuation were observed in pts receiving combination therapy. Correlative blood (CTC/cfDNA) and tissue (AURKA, ERα, and stemness biomarkers) studies are underway. Citation Format: Tufia Haddad, Antonino D'Assoro, Vera Suman, Jodi Carter, Brendan McMenomy, Erica Mayer, Meghan Karuturi, Aki Morikawa, Paul Marcom, Claudine Isaacs, Sun Young Oh, Amy Clark, Ingrid Mayer, Khandan Keyomarsi, Roberto Leon-Ferre, Karthik Giridhar, Ciara O'Sullivan, Prema Peethambaram, Timothy Hobday, Minetta Liu, James Ingle, Matthew Goetz. Randomized phaseII trial to evaluate alisertib alone or combined with fulvestrant for advanced, endocrine-resistant breast cancer (TBCRC 041) [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 PD2-05.

Abstract BACKGROUND: Brain metastatic disease occurs in 10-30% of metastatic breast cancer cases. The incidence of brain metastases is increasing, yet overall survival remains < 2 years. Treatment of brain metastases is limited with current clinical practice centered on radiation, chemotherapy and surgery. Although these treatments may prolong survival in the short term, targeting oncogenic alterations in brain metastases may deliver a more sustained clinical benefit. In this multicenter study, we comprehensively characterized DNA and RNA alterations that aberrantly drive specific oncogenic pathway activity pertinent to breast cancer brain metastases (BCBM). METHODS: RNA sequencing was performed on a cohort of patient-matched primary and resected brain metastatic tumours (45 patients; N=90 samples). Whole exome DNA sequencing (WXS) was performed for 18/45 patients (54 trios consisting of primary tumor, brain metastasis and matched normal tissue). An independent brain metastatic WXS cohort (N=21 patients) (PMID: 26410082) was also analysed resulting in a total of 39 patient samples. Recurrent somatic copy number alterations (SCNA), somatic single nucleotide variants (SNVs) and mutational signatures were identified from WXS data. Expressed gene fusions were detected computationally from RNA-Seq (N=45 cases) RESULTS: Of the 45 BCBM patients, median age at diagnosis was 51 years [25, 67], median overall survival 57 months (range 18-255) with median brain metastases free survival 34 months (range 5-216).Clinical molecular subtype of the primary tumour included 13 ER+/HER2- (29%), 16 HER2+ (35.5%) and 16 TNBC (35.5%). Regions of significant recurrent amplifications and deletions in BCBM (N=39 patients) were identified in 4q12, 10q11.21, 8p11.23, 8q23.3 and 17q12 (FDR < 0.10). Recurrent expressed gene fusions identified in known cancer driver genes were associated with chromatin modification, MAPK and HER signaling pathways. Mutational signature analysis of SNVs identified signatures associated with ageing, mismatch repair and homologous recombination deficiency (HRD) mutational processes. The relative contribution of the HRD signature was significantly increased in brain metastases compared to matched primary tumour (p < 0.05). Concordantly increased HRD in the brain metastatic transcriptome was confirmed in these patients by gene set variation analysis (GSVA) of homologous recombination pathway genes from KEGG database in RNA-Seq data. Moreover, in the extended BCBM RNA-Seq (N=45 patients) GSVA pathway scores were elevated in brain metastases relative to primary tumours, validating the functional significance of altered DNA repair defects in brain metastases CONCLUSIONS: Here, we report recurrent genetic drivers, supported by altered functional transcriptome, unique to brain metastasis that may have clinical implications for prognosis and treatment choice. Specifically, targeting defects in the homologous recombination repair mechanism may represent new therapeutic strategies and management opportunities for breast cancer brain metastases patients. Citation Format: Nicola S Cosgrove, Damir Varešlija, Stephen Keelan, Simon Furney, Jodi M Carter, Steven N Hart, Siddhartha Yadav, Arnold DK Hill, Steffi Oesterreich, Adrian V Lee, Fergus J Couch, Leonie Young. Homologous recombination deficiency represents a new therapeutic strategy for breast cancer brain metastases [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 PD13-01.

Abstract Background: Women with benign breast disease (BBD) experience an increased risk of breast cancer (BC). Histologic classification of BBD, as non-proliferative disease (NP), proliferative disease without atypia (PDWA) or atypical hyperplasia (AH), stratifies groups of patients into progressively higher categories of BC risk. However, this classification does not comprehensively assess the proliferative state of the epithelium throughout the biopsy. In addition, while AH is considered the most high-risk class of BBD, it is not always a highly proliferative lesion; atypical ductal lesions may reflect focal cytologic and architectural changes. We evaluated the association of an alternative classification of BBD severity and BC risk based on subjective grading of: 1) the maximal degree of epithelial proliferation and 2) multifocality of epithelial proliferation. Methods: Pathologists reviewed biopsies from participants aged 18 to 85 years in the Mayo BBD cohort (2002-2013), masked to BC outcomes, ascertained via questionnaires, tumor registry data and medical record review. Biopsies were classified as NP, PDWA or AH and semi-quantitatively scored for: 1) maximal degree of epithelial proliferation within a focus (DP) (0-3; none to severe) and 2) multifocality of proliferation (MP) (0-3; none to multiple foci). DP and MP scores were also summed to give a DP+MP score (0-6). Associations of DP and MP with BC risk were examined using Cox proportional hazards regression analyses, adjusting for age at BBD biopsy. Women were followed from date of initial biopsy to date of BC, death or last follow-up. Results: Of the 1529 assessable biopsies, 544 (35.6%) were classified as NP, 708 (46.3%) as PDWA and 277 (18.1%) as AH. Both DP and MP scores had significant positive correlation with increasing BBD severity (DP: r=0.51, p< 0.001; MP: r=0.52, p< 0.001). Mean (SD) DP scores were 0.6 (0.6) for NP, 1.6 (0.9) for PDWA, and 1.8 (0.7) for AH (ANOVA p<0.001). Mean (SD) for MP scores were 0.6 (0.6) for NP, 1.4 (0.8) for PDWA, and 1.8 (0.8) for AH (ANOVA p<0.001). Mean (SD) for DP+MP scores were 1.2 (1.2) for NP, 2.9 (1.5) for PDWA, and 3.6 (1.2) for AH (ANOVA p<0.001). With median follow-up of 8.8 years for controls and 5.3 years for cases, 10.6% of the women in the cohort developed BC. Compared to those with DP scores of 0, women with DP scores of 3 had significantly increased BC risk (HR 1.42, 95% CI: 1.16, 1.74, p=0.003). MP was associated with a non-significant increase in BC risk for scores of 3 versus 0 (HR: 1.20, 95% CI: 0.97,1.49, p=0.11). DP+MP scores of 6 conferred the highest BC risk (HR (score 6 vs. 0): 1.62, 95% CI 1.18,2.21, p=0.02). Results did not substantively differ after adjusting for BBD severity as NP, PDWA or AH. Conclusions: In this preliminary analysis within the Mayo BBD cohort, both proliferative degree (DP) and multifocality (MP) scores were correlated with histologic severity of BBD. DP and DP+MP scores were each associated with increased BC risk. We conclude that improved characterization of epithelial proliferation in BBD biopsies may enable refined prediction of individual BC risk. Citation Format: Jodi M Carter, Matthew R Jensen, Robert A Vierkant, Stacey J Winham, Tanya L Hoskin, Marlene Frost, Karthik Ghosh, Derek C Radisky, Amy C Degnim, Mark E Sherman. Epithelial proliferation score as an independent breast cancer risk predictor in benign breast disease [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 PD10-10.

Abstract BACKGROUND Neuromuscular choristoma (NMC) is a peripheral nerve malformation frequently associated with a fibromatosis (NMC-DTF) that mimics sporadic desmoid-type fibromatosis (DTF). Sporadic DTF is often managed conservatively but its clinical behavior varies. CTNNB1 mutational subtypes in sporadic DTF have prognostic value. We have previously identified CTNNB1 mutations in NMC, and 3 paired NMC-DTF but the clinical behavior of NMC-DTF is poorly understood. OBJECTIVE To evaluate patients with NMC-DTF to determine (1) CTNNB1 mutational subtypes in NMC-DTF, and (2) associated clinical behavior and response to treatment. METHODS Retrospective review of clinical, imaging, and pathologic features of patients with NMC and NMC-DTF, and molecular testing for CTNNB1 mutations. RESULTS Among 7 patients with NMC of the sciatic nerve (median age: 18 yr), NMC-DTF (mean size 10.7 cm) developed shortly following NMC biopsy (N = 5) or spontaneously (N = 2): 6 NMC-DTF had CTNNB1 p.S45X mutations and 1 NMC-DTF had a p.T41A mutation. All patients with CTNNB1 -p.S45-mutated NMC-DTF developed local progression after wide local excision or active surveillance, including one distal metachronous NMC-DTF. No patient had spontaneous disease stabilization. Following adjuvant radiation or systemic therapy, disease stabilization was achieved in 4 (of 6) patients. One patient progressed on sorafenib treatment. CONCLUSION NMC-DTF frequently contain CTNNB1 p.S45 mutations, behave aggressively, and require adjuvant therapies for disease stabilization. We now use imaging alone to diagnose NMC, and routinely surveille the NMC-affected nerve segment to identify early NMC-DTF. In contrast to sporadic DTF, earlier adoption of systemic therapeutic strategies may be required for optimal disease management of NMC-DTF.

Context.— Guidelines for HER2 testing in breast cancer have changed over time, from the US Food and Drug Administration guideline to the American Society of Clinical Oncology (ASCO)/College of American Pathologists (CAP) guidelines published in 2007, 2013, and 2018. Objective.— To investigate the change in assignment of HER2 status in breast cancers with equivocal (2+) immunohistochemistry (IHC) results by fluorescence in situ hybridization (FISH) following implementation of the ASCO/CAP 2018 guideline. Design.— The study included 3556 invasive breast cancers that were HER2 equivocal (2+) by IHC and were submitted to our FISH laboratory after July 2018. Reflex testing (with repeat IHC staining) was performed on certain categories of FISH results known as groups 2, 3, and 4. Concomitant review of IHC and FISH was performed on these reflex cases per 2018 guideline recommendations. The FISH data were analyzed to compare US Food and Drug Administration and ASCO/CAP 2007, 2013, and 2018 interpretations. Results.— Of 3548 invasive breast cancers with complete data available, the percentage agreement for FISH according to different guidelines was highest for ASCO/CAP 2018 versus US Food and Drug Administration (96.5%), followed by ASCO/CAP 2018 versus 2007 (93.8%), and lowest with ASCO/CAP 2018 versus 2013 (83.7%). Per the 2018 guideline, reflex IHC testing was performed on 633 breast cancers (17.8%); the majority of reflex testing results were negative (541 of 633; 85.5%). The overall distribution of HER2 FISH results (per the 2018 guideline) was 88.5% negative and 11.5% positive. Conclusions.— By eliminating the equivocal FISH category, the 2018 ASCO/CAP guideline significantly reduced the HER2 FISH–positive rate in tumors with equivocal (2+) IHC results.

Abstract Background: Biopsy diagnosis of benign breast disease (BBD) based on the most severe lesion in a sample predicts future breast cancer risk and has implications for screening and management. Lobules are the functional unit of the breast and the structures from which BBD arises. We developed and preliminarily validated an automated computational pathology algorithm to discriminate normal from BBD breast lobules as a step toward automated comprehensive characterization of benign biopsies. Methods: 152 BBD biopsies (27 training, 125 validation) from the Mayo Clinic were examined using scanned digital images of H&E stained sections. For each image, a pathologist annotated up to 10 representative lobules as normal or BBD using standard pathology criteria. A deep learning algorithm to quantify lobular features was developed using 129 lobules from 27 subjects. Nine features were identified that discriminate normal vs BBD lobules, expressing lobule size, acini size and number, acinar lumen size, proportion of lobular stroma, and capillaries. Here, we validate their performance to discriminate normal vs BBD lobules in a set of 1250 lobules from 125 subjects using area under the ROC curve (AUC) analysis. Random forest analysis was used for multivariable modeling; model performance was assessed with a tenfold cross-validation approach. Results: Median subject age was 52 years. Among the 125 validation subjects, BBD findings were nonproliferative in 39%, proliferative in 45%, and atypical hyperplasia in 16%. Sections included 552 (44%) normal lobules and 698 BBD (56%) lobules, with representation of both lobule types on each section. In univariate analyses, four individual features showed good discrimination between normal vs. BBD lobules, yielding the following AUCs: lobule size= 0.74, mean acini size= 0.75, epithelial area= 0.75, and number of acini with large lumens= 0.76. Lobule size and epithelial area were highly correlated (Spearman rank correlation r = 0.95), but both of these features showed lower correlation with mean acini size and number of acini with large lumens (each r < 0.5). With random forest modeling, number of acini with large lumens was the strongest discriminating feature, followed by mean acini size, lobule size, and epithelial area. With ten-fold cross validation of the multivariable random forest model, the overall AUC was 0.82 (95% CI: 0.79-0.85). Conclusion: Our validation showed that automated quantitative computational pathology assessment of breast lobules can discriminate normal versus BBD on a per lobule basis. This finding supports the feasibility of developing automated algorithms to classify every lesion in breast biopsies, expanding beyond visual assessment. Further studies using deep learning may reveal novel pathology features for classifying BBD biopsies with the potential to strengthen estimation of breast cancer risk. Citation Format: Amy C. Degnim, Thomas de Bel, Mark E. Sherman, Derek C. Radisky, Stacey J. Winham, Tanya L. Hoskin, Melody L. Stallings Mann, Marlene Frost, Robert A. Vierkant, Brendan T. Broderick, Ethan P. Heinzen, Rushin Brahmbhatt, Muhammad Arshad, Celine M. Vachon, Jodi M. Carter, Lori A. Denison, Daniel W. Visscher, Jeroen van der Laak. Discrimination of benign breast disease from normal lobules using an automated computational pathology algorithm [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 2113.