Search

Background The status of CDH1, ERBB2, and ESR1 is important for the diagnostic and treatment workup for patients with breast cancer. Most alterations in these genes occur in the form of short variants (eg missense and indel alterations) or copy number alterations (eg amplifications of ERBB2 or deletions in CDH1). The prevalence and characteristics of rearrangement events have been understudied. Materials and Methods: Comprehensive genomic profiling using a hybrid-capture based approach was performed on 44,842 breast carcinomas during the course of routine clinical care (FoundationOne® or FoundationOne®CDx) examining all classes of alterations in up to 395 genes, including CDH1, ESR1, and ERBB2. All rearrangements were included in the analysis (known/likely pathogenic and variants of uncertain significance). Estrogen receptor status was extracted from pathology reports for a subset of samples. Results: Rearrangements in CDH1, ESR1, and ERBB2 were observed in 0.26% (115/44842), 0.34% (153/44842), and 1.33% (598/44842) of breast cancer samples, respectively. As expected, CDH1 rearrangements were most common in invasive lobular carcinoma (ILC) (0.64%; 16/2516) though events were observed in samples originally submitted as invastive ductal carcinoma (IDC) (0.16%; 26/15,836), suggesting possible misdiagnosis. CDH1 rearrangements were predominantly loss of function consisting of large deletions, inversions, and truncation events. ESR1 rearrangements were observed at the highest frequency in ER+/HER2- tumors (0.58%) and were never seen in ER-/HER2- and ER-/HER2+ tumors. ESR1 rearrangements were observed with a variety of partners, with recurrent events with CCDC170, SYNE1, RMND1, PLEKHG1, ARMT1, MTHFD1L, and ZBTB2. Consistent with a possible role in therapy resistance, ESR1 rearrangements were enriched in metastatic samples relative to those biopsied from the breast (OR = 2.25; p = 8E-05). ERBB2 rearrangement events were commonly observed in HER2 amplified tumors (13.4%) and rarely in other subtypes (0.12% in ER+/HER2- and 0.28% in ER-/HER2-). Most of the events were intra-chromosomal and typically represented non-fusion duplication fragments that may have generated the ERBB2 amplification. Fusions were much rarer. Out of the 598 rearrangement events, only 18 were predicted to create in-frame and in-strand fusion products retaining the HER2 kinase domain following manual review of the events. Most fusion events were unique, though a fusion with IKZF3 was seen recurrently (n=2). Conclusions Rearrangement events in diagnostically important breast cancer genes (CDH1, ESR1, and ERBB2) were commonly observed in breast cancer with subtype-specific enrichment. Since these alterations have implications in disease diagnosis and therapy response (eg endocrine therapy resistance), comprehensive genomic profiling can provide value in breast cancer care. Citation Format: Ethan Sokol, Dexter Jin, Jeffrey S. Ross, ADRIAN V. LEE, Steffi Oesterreich. Rearrangements in CDH1, ESR1, and ERBB2 are commonly observed in breast cancer and may influence diagnosis and treatment [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P5-14-11.

Background: Malignant phyllodes tumors (MPT) are rare fibroepithelial breast tumors that disproportionately affect young women. Despite high local and distant recurrence rates, treatment is almost exclusively surgical. With no known effective chemotherapy or targeted systemic therapy options, metastatic progression portends dismal prognosis. Current national guidelines do not recommend biomarker testing in the non-metastatic setting. Management of metastatic disease follows principles of soft tissue sarcoma recommending Next-Generation Sequencing on an individual basis for patients who may qualify for clinical trials or are refractory to standard therapies. In the largest genomic profiling effort of MPT to date, we sought to describe the genomic landscape of MPTs through comprehensive genomic profiling (CGP) and immunotherapeutic biomarker analysis. Methods: Cases of sequenced MPT were identified from a CLIA-certified, CAP-accredited laboratory database (Foundation Medicine, Cambridge, MA). Cases were categorized as localized/locally recurrent or metastatic. All cases underwent CGP by DNA extraction from FFPE samples using a hybrid capture, adaptor ligation-based next generation sequencing assay, and all classes of genomic alterations, microsatellite instability (MSI), and tumor mutational burden (TMB) were evaluated. PD-L1 expression by IHC was measured by the DAKO 22C3 assay (scored CPS) or the Ventana SP142 assay (scored as IC). Patient characteristics and results of CGP of MPT were summarized by either N (%) or median (range). Factors from MPT were compared to all cases of breast carcinoma in the lab database. Fisher’s Exact Tests were used to test for differences between groups and analysis of variance was used to test for differences for continuous variables. Results: Of 135 MPT cases identified; 94 (69.6%) were localized/locally recurrent (breast, chest wall, soft tissue, skin, or lymph node) and 41 (30.4%) were metastatic (73% lung, followed by bone, liver, other intraabdominal). All patients were female with a median age of 54 years (range 14-86). The median TMB across samples was 2.5 mut/Mb and only 3 were TMB-high tumors (>10mut/Mb). Over 1/3 (36.8%) of samples were PD-L1+ via Ventana SP142 assay (≥1) and 21.4% were PD-L1+ via Dako 22C3 assay (CPS ≥10). All evaluable cases (N=132) were microsatellite stable. The ten most commonly altered genes included TERT-promoter (69.7%), CDKN2A (45.9%), TP53 (37.8%), NF1 (35.6%), CDKN2B (33.3%), MED12 (28.9%), MTAP (27.7%), KMT2D (22.2%), PIK3CA (20.0%), PTEN (18.5%), and RB1 (18.5%); notably distinct in frequency from the breast carcinoma cohort (Table 1). Alterations in genes affecting cell cycle regulation (i.e. CDKN2A/B and TP53) were mutually exclusive from one another (p< 0.001). Several tumors harbored genomic alterations with FDA-approved indications in other tumor types were found including: NF1, PIK3CA, EGFR Exon 19/20 insertions, and BRAF V600E mutations. There were no statistically significant differences between alterations in localized/locally recurrent versus metastatic specimens. Conclusions: In the largest genomic evaluation of localized/locally recurrent or metastatic MPTs to date, multiple clinically actionable mutations for further exploration were found. TERT-promoter was altered in the majority of cases. PDL-1 was expressed in over 1/3 of cases suggesting that clinical investigation of immunotherapeutic strategies is warranted. Routine sequencing of metastatic MPT may provide additional information to guide treatment decisions and clinical trial enrollment. Table 1. Frequency of Genomic Alterations identified in 135 cases of Malignant Phyllodes Tumors. Citation Format: Laura H. Rosenberger, Richard F. Riedel, Natalie A. Danziger, Ethan Sokol, Jeffrey S. Ross, Sarah L. Sammons. Genomic Evaluation of Malignant Phyllodes Tumors Reveals Multiple Targetable Opportunities [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P2-23-03.

Background: Chemotherapy remains the mainstay of treatment for early-stage triple negative breast cancer (TNBC), yet targetable drivers of interest are under investigation. A subset of TNBCs express the androgen receptor (AR) and exhibit androgen-dependent growth. The AR-antagonist enzalutamide (ENZA) has shown activity in patients with metastatic AR+ TNBC. In this study, the feasibility of adjuvant ENZA in early-stage, AR+ TNBC was assessed (NCT02750358). As reported previously, this study met its primary endpoint of feasibility (Traina et al., ASCO 2019). Here we report secondary survival endpoints. Methods: In this single-arm, open-label, multi-center trial, patients with stage I-III, AR≥1% TNBC (ER/PR Citation Format: Elaine M Walsh, Ayca Gucalp, Sujata Patil, Marcia Edelweiss, Dara S Ross, Pedram Razavi, Shanu Modi, Neil M Iyengar, Rachel Sanford, Tiffany Troso-Sandoval, Mila Gorsky, Jackie Bromberg, Pamela Drullinsky, Diana Lake, Serena Wong, Patricia DeFusco, Nicholas Lamparella, Ranja Gupta, Tasmila Tabassum, Leigh Ann Boyle, Artavazd Arumov, Tiffany A Traina. Adjuvant enzalutamide for the treatment of early-stage androgen-receptor positive, triple negative breast cancer: A feasibility study [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P1-14-03.

Background: Several recent studies that compared small cohorts of metastatic and primary lesions, suggested substantial heterogeneity in tumor infiltrating lymphocyte count, immune gene expression and PD-L1 protein expression across different metastatic sites and between primary breast cancers and metastasis. Understanding the frequency of PD-L1 positivity rates across different tissue sites can indicate differences in the immune microenvironment and may also guide biopsy site selection. We compared PD-L1 positivity on immune cells and tumor cells in primary and metastatic triple negative breast cancer tumors (TNBC).Methods: A retrospective data analysis of the Foundation Medicine PD-L1 IHC database was conducted on 340 cases of TNBC. PD-L1 positivity was determined by IHC using SP142CDx. Results are reported as percent of PD-L1 stained immune cells (IC) in the tumor area. A tumor was considered PD-L1 positive if ≥ 1% IC stained positive with PD-L1. As an exploratory analysis, PD-L1 positivity of tumor cells (TC) was also assessed. PD-L1 percent positive staining results are reported as means with 95% CI. The proportion of PD-L1 positive and negative IC and TC in primary tumors vs metastatic sites was compared using Chi-Square test. Prism 8 was used for all data analysis.Results: All patients were female, with median age 56 years (range 26-89); 179 samples were from primary tumors and 161 from metastatic lesions, representing 15 different tissue sites. Overall, PD-L1 expression on immune cells was statistically significantly more frequent in primary tumors compared to metastatic sites (63.7% [n=114] vs 42.9% [n=69]), p Table 1: Sample Characteristics and % PD-L1 positivity on immune cellsSample typeTotal N (%)N PD-L1 positive (%, 95% CI)Primary Tumor179 (52.6)114 (63.7%, 56.2% - 70.7%)Metastatic Lesion161 (47.4)69 (42.9%, 35.1% - 50.9%)Sites of MetastasesN (% of metastatic samples)N PD-L1 positive (%, 95% CI)Lung16 (10.0)11 (68.8%, 41.3% - 90.0%)Soft Tissues23 (14.3)15 (65.2%, 42.7% - 83.6%)Lymph Nodes45 (28.0)23 (51.1%, 35.8% - 66.3%)Skin21 (13.0)5 (23.8%, 8.22% - 47.2%)Liver23 (14.3)4 (17.4%, 5.00% - 38.8%)Bone12 (7.5)2 (16.7%, 2.10% - 48.4%)Brain9 (5.6)5Mediastinum4 (2.5)1Pleura2 (1.2)0Muscle1 ( Table 2: Comparison of PD-L1 positivity in primary versus metastatic sitesTissuePDL1+ Immune CellPDL1- Immune CellP valuePDL1+ Tumor CellPDL- Tumor CellP valuePrimary114650.0001151640.1313Metastasis69927154 Citation Format: Mariya Rozenblit, Richard Huang, Natalie Danziger, Brian Alexander, Shakti Ramkissoon, Kim Blenman, Jeffrey Ross, David Rimm, Lajos Pusztai. Comparison of PD-L1 protein expression between primary tumors and metastatic lesions in triple negative breast cancers [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 PS5-08.

Introduction: De-regulation of the class I phosphoinositide 3’-kinase (PI3K) pathway has long been known to contribute to the development and progression of many tumors. However, the FDA only recently approved the first selective inhibitor of PIK3CA, the p110-alpha catalytic subunit of PI3K, specifically for use in treatment of a subset of PIK3CA-mutant breast carcinomas. Given the emergence of this therapeutic class of agents, we sought to identify other subsets of breast tumors that may be driven largely by PIK3CA mutations and which therefore could be candidates for these therapies. In breast fibroepithelial neoplasms, mutations in PIK3CA have been occasionally reported in borderline and malignant phyllodes tumors. In contrast, mutations in MED12 are common and recurrent across the entire spectrum of benign and malignant breast fibroepithelial tumors, and are enriched in borderline/malignant phyllodes cases that still have benign fibroadenoma-like areas. In the current study, we sought to define the histologic and molecular features of PIK3CA-mutant phyllodes tumors. Methods: From 2014 to 2019, we analyzed clinical tumor samples using comprehensive genomic profiling by a hybrid capture-based DNA sequencing platform. We searched our case archive to find breast phyllodes tumors with known or likely pathogenic alterations in PIK3CA and other known tumor-related genes. All cases were clinically advanced. We reviewed pathology reports, histopathology, and patient clinical data. Results: We identified 12 (16%) of 76 breast phyllodes tumors in our case archive as PIK3CA-mutant. Median patient age for PIK3CA-mutant phyllodes tumors was 56 years. Cases consisted of 6 primary tumors, 2 local recurrences, and 4 lung metastases. Primary tumor size measured from 38 to 220 mm (median 100 mm; mean 114 mm). Digital slides were available for histology review in 9 cases. Cases showed uniformly malignant histology, with no benign or fibroadenoma-like regions. 3 cases showed malignant heterologous elements. Compared to the rest of our breast phyllodes tumor cohort, PIK3CA-mutant cases showed significantly fewer pathogenic genomic alterations in MED12 (8% vs. 55%, p=0.0037) and TP53 (17% vs. 56%, p=0.0245), as well as a trend to fewer mutations in RB1 (0% vs. 22%, p=0.11). The other most frequently mutated genes in the PIK3CA-mutant group were TERTp (70%), CDKN2A (67%), and NF1 (50%). Conclusions: PIK3CA-mutant phyllodes tumors define a unique subset of tumors characterized by aggressive histologic features and largely lacking the MED12 mutations seen in many fibroepithelial neoplasms. These findings provide compelling rationale for comprehensive genomic profiling of advanced cases of this disease in an effort to inform therapeutic options including clinical trials of PI3K-targeted agents in this setting. Citation Format: Erik A Williams, Ethan S Sokol, Dean C Pavlick, Nikunj Shah, Julia A Elvin, Jo-Anne Vergilio, Jonathan K Killian, Nhu Ngo, Douglas Lin, Vincent A Miller, Jeffrey S Ross. PIK3CA-mutant breast phyllodes tumors show a uniformly aggressive histology and significant mutual exclusivity with MED12 mutation [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 P4-06-07.

Introduction: Invasive lobular carcinoma (ILC) is the most prevalent histologic special type of breast cancer, representing ~15% of invasive breast cancers. ILCs are mostly hormone receptor-positive and diagnosed histologically based on their distinctive discohesive growth pattern, dispersed single cells or line of cells invading the stroma. ILCs commonly show biallelic inactivation of a CDH1 (~65%), a tumor suppressor gene that is mapped to 16q22.1 and encodes for E-cadherin, a critical component of the epithelial adhesion complex. Massively parallel sequencing and pathologic studies have shown that a subset of ILCs (up to 15%) may lack CDH1 loss-of-function (LOF) mutations/deletions and retain E-cadherin expression, despite their distinctive lobular phenotype. The genetic and epigenetic underpinning of ILCs lacking CDH1 alterations has yet to be defined. Here we sought to define the mechanistic basis of the lobular phenotype in ILCs lacking CDH1 LOF genetic alterations or CDH1 gene promoter methylation and to determine the repertoire of epigenetic and genetic alterations of CDH1-wildtype ILCs. Materials and methods: Reanalysis of the CDH1 gene status in 364 primary ILCs, previously subjected to clinical massively parallel sequencing, was performed to identify all primary ILCs lacking bi-allelic CDH1 alterations. The hematoxylin and eosin stained slides of all identified ILCs lacking bi-allelic CDH1 alterations were reviewed by two pathologists and primary pure ILCs lacking bi-allelic CDH1 alterations were curated. We evaluated the presence of genetic alterations in genes playing essential roles in epithelial adhesion included in the clinical sequencing panel of up to 505 genes. The presence of CDH1 promoter methylation in 18 ILCs lacking bi-allelic CDH1 alterations with available formalin-fixed, paraffin-embedded (FFPE) material was assessed using methylation-specific PCR (MSP) and bisulfite sequencing. Results: We identified 23/364 (6.3%) primary ILCs lacking bi-allelic CDH1 alterations, of which 65.3% and 34.7% were classic and pleomorphic lobular variants, respectively. In 18 cases with available FFPE material, our analyses revealed that 67% (12/18) of ILCs lacking bi-allelic CDH1 alterations displayed biallelic CDH1 inactivation via promoter methylation and 16q loss. Furthermore, we observed that 1/23 ILC lacking bi-allelic CDH1 alterations had bi-allelic inactivation of AXIN2. We then extended our query to all invasive breast carcinomas subjected to clinical sequencing, including the ones initially categorized as invasive breast carcinoma “type unknown”, and observed that the three additional cases with pathogenic LOF AXIN2 alterations were ILCs lacking bi-allelic CDH1 alterations. Conclusions: The lobular phenotype in ILCs can be due to CDH1 promoter methylation or genetic alterations affecting other genes related to epithelial cell adhesion, such as AXIN2 LOF mutations. Whole-genome sequencing analyses of ILCs whose molecular basis has not been identified by targeted sequencing are warranted. Citation Format: Fatemeh Derakhshan, Higinio Dopeso, Arnaud Da Cruz Paula, Pier Selenica, Antonio Marra, Edaise M da Silva, Andrea Gazzo, Shirin Issa Bhaloo, Dara S Ross, Anne Grabenstetter, Sarat Chandarlapaty, Pedram Razavi, Hannah Y. Wen, Hong Zhang, Edi Brogi, Britta Weigelt, Fresia Pareja, Jorge S. Reis-Filho. Genetic and epigenetic basis of invasive lobular carcinomas lacking CDH1-alterations [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr PD14-03.

Genomic profiling is widely predicted to become a standard of care in clinical oncology, but more effective data sharing to accelerate progress in precision medicine will be required. Here, we describe cancer-associated genomic profiles from 18,004 unique adult cancers. The dataset was composed of 162 tumor subtypes including multiple rare and uncommon tumors. Comparison of alteration frequencies to The Cancer Genome Atlas identified some differences and suggested an enrichment of treatment-refractory samples in breast and lung cancer cohorts. To illustrate novelty within the dataset, we surveyed the genomic landscape of rare diseases and identified an increased frequency of NOTCH1 alterations in adenoid cystic carcinomas compared with previous studies. Analysis of tumor suppressor gene patterns revealed disease specificity for certain genes but broad inactivation of others. We identified multiple potentially druggable, novel and known kinase fusions in diseases beyond those in which they are currently recognized. Analysis of variants of unknown significance identified an enrichment of SMAD4 alterations in colon cancer and other rare alterations predicted to have functional impact. Analysis of established, clinically relevant alterations highlighted the spectrum of molecular changes for which testing is currently recommended, as well as opportunities for expansion of indications for use of approved targeted therapies. Overall, this dataset presents a new resource with which to investigate rare alterations and diseases, validate clinical relevance, and identify novel therapeutic targets. Cancer Res; 77(9); 2464–75. ©2017 AACR.

Background: Metaplastic breast cancers (MPBC) are rare, typically triple negative aggressive tumors composed of both, adenocarcinoma and metaplastic elements. Recent evidence that TNBC and MPBC can respond vigorously to immune checkpoint inhibitor therapy (Adams et al, ASCO 2017 and npj Breast Cancer 2017) has prompted the following comprehensive genomic profiling (CGP) and histopathologic assessment of tumor infiltrating lymphocytes (TIL) designed to uncover potential biomarkers of immunotherapy response for MPBC, including mutational burden, Microsatellite Instability (MSI) status and gene amplification of 9p21.4 (or CD274, which includes the PD-L1 locus). Methods: 12,214 locally aggressive, relapsed and metastatic breast malignancies (mBM) were subjected to CGP using DNA extracted from 40 µm of FFPE sections and adaptor ligation-based libraries to a mean coverage depth >650X for up to 315 cancer-related genes. The results were analyzed for all classes of genomic alterations (GA) including base substitutions, insertions and deletions, select rearrangements, and copy number changes. Tumor mutational burden (TMB) was determined on 1.1 Mbp of sequenced DNA. MSI status was determined by an algorithm based on the sequencing results. TIL were assessed on archived H&E tumor sections and enumerated per guidelines established by the TIL Working Group (Salgado, Ann Oncol 2015) in a subset of MPBC with the highest TMB and compared with low TMBC cases. Results: 165 of mBM cases were MPBC (1.4%) and are included in this study. All patients were female with a median age of 60 (range 24-86). 165 of the MPBC cases (100%) harbored a wide variety of GA involving more than 100 individual genes. The most common GA were identified in TP53 (65%), followed by PIK3CA (37%). No cases of MSI hi status (0/103) and only one case with amplification of 9p21.4 (1/165, 0.6%) were observed. Most MPBC had a low mutational burden, with a median TMB of 2.7 mutations/Mb (range 0-39.6). Only 11/165 tumors (6.7%) were found to have a TMB over 10 mutations/Mb, including 3 cases (1.8%) with TMB >20. Tumor sections were available for TIL review from 9/11 cases with highest TMB, as well as 11 control cases with lowest TMB. TIL were more frequently observed in high versus low TMB MPBC, with median TIL percentage of 40 and 20 (range 10-80 and range 10-60), respectively, although this difference was not statistically significant (Wilcoxon rank-sum test, p=0.15). Conclusions: Genomic profiling in the largest cohort of MPBC reported to date confirms that MPBC is enriched for TP53 and PIK3CA mutations and many tumors harbor targetable GA. The frequently observed tumoral PD-L1 expression in MPBC is not based on gene amplification as amplification of 9p21.4 is rare. Most tumors had a low mutation burden, and no significant association of TIL with TMB was observed, suggesting additional processes underlying MPBC immunogenicity. Citation Format: Taff J, Suh J, Singh B, Denkert C, Troxel AB, Ross JS, Adams S. Metaplastic breast cancers: Genomic profile, mutational burden and TILs [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P3-05-03.

Background: Pembrolizumab recently received pan-tumor FDA approval for the treatment of patients with unresectable or metastatic solid tumors with tumor mutational burden (TMB) >10 mutations per megabase (mut/Mb) and who have no satisfactory alternative treatment options. The KEYNOTE-119 and TAPUR trials have validated TMB as a predictive biomarker using TMB cutoffs of ≥ 10 mut/Mb and 9 mut/Mb, respectively, for benefit from single-agent pembrolizumab in subsets of patients with clinically advanced breast cancer (aBC). Limited data exist on which pathological subtypes of aBC are most likely to be TMB >10 mut/Mb (TMB-High). This study analyzed the frequency of TMB-High tumors in aBC subtypes to evaluate variations between subtypes and correlation with additional immunotherapy biomarkers. Methods: Utilizing the Foundation Medicine breast cancer database, 5,475 samples with aBC and pathological subtype information were identified. Pathological subtypes included in this study were ER+/HER2-negative invasive ductal carcinoma (IDC), ER-negative/HER2+ IDC (HER2+), triple negative IDC (TNBC), invasive lobular, inflammatory, metaplastic, mucinous and papillary. TMB was determined on 0.8-1.1 Mb of sequenced DNA, and microsatellite instability (MSI) was assessed across 114 loci in 564 cases. PD-L1 was determined in a subset of cases by immunohistochemistry (Ventana SP142). Results: TMB-High (>10 mut/Mb) was common in all breast cancer pathological subtypes except papillary carcinoma (Table 1). TMB-High was most often found in lobular, inflammatory and HER2+ carcinomas (16%, 14% and 12% of cases, respectively). Metaplastic, mucinous and papillary subtypes were least likely to be TMB-High (7%, 7% and 0% of cases, respectively). TMB-High was more common than MSI-High in all subtypes, excluding papillary. Emerging biomarkers that may play a role in mitigating immunotherapy response, such as MDM2 amplifications and STK11 alterations, were observed in most subtypes. The frequencies of these alterations, as well as alterations associated with FDA-approved therapies in breast cancer, are provided in Table 1 and will be further discussed. Conclusions: TMB-High (>10 mut/Mb) is common in most pathological subtypes of clinically advanced breast cancer and can identify patients not identified by MSI or PD-L1 testing who may benefit from immunotherapy. Based on the high percentage of advanced breast cancer patients who are TMB-High or have genomic alterations in other biomarkers associated with FDA-approved targeted therapies, comprehensive genomic profiling including TMB and MSI should be considered for all pathological subtypes. Table 1Pathological subtypeInvasive ductal carcinoma (IDC)ER+, HER2-negativeER-negative, HER2+Triple-negative (TNBC)Invasive Lobular CarcinomaInflammatory MetaplasticMucinousPapillaryCases (n)1,2371,9536411,180353892911Age55 (23-89)55 (20-89)53 (20-85)62 (24-89)54 (28-87)59 (28-89)55 (30-80)61 (38-78)Tumor mutational burden (TMB)TMB > 10 mut/Mb8%12%9%16%14%7%7%0%TMB > 20 mut/Mb2%2%3%7%6%2%3%0%Additional immunotherapy biomarkersMSI-High0.2%0.1%0.4%0%0%0%0%0%PD-L1+ (>1% Ventana SP142)N/AN/A47%N/AN/A26%0%N/AMDM2 amplification6%5%3%2%0%2%3%18%STK11 alteration1%1%2%1%3%3%3%0%Additional biomarkers associated with FDA-approved therapiesBRCA1/2 alteration3%/6%2%/3%7%/3%1%/4%6%/6%4%/3%3%/0%0%/9%ERBB2 amplification0%100%0%3%26%3%34%9%NTRK 1/2/3 rearrangement0%/0%/0%0%/0%/0%0%/0%/0%1%/0%/0%0%/0%/0%0%/0%/1%0%/0%/0%0%/0%/0%PIK3CA alteration38%38%19%56%34%38%21%36% Citation Format: Mason A Israel, Marni Tierno, Richard SP Huang, Kimberly McGregor, Ethan S Sokol, Prasanth Reddy, Jeffrey S Ross. High tumor mutational burden (>=10 mut/Mb) is enriched in specific breast cancer pathological subtypes [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 PD9-09.

Background: Non-colorectal MSI-H tumors are increasingly identified by CGP. Rare types such as MSI-H BC remain poorly defined with an evidence gap on how to optimally sequence or combine with standard of care treatment. MSI can be measured by either IHC, PCR, or CGP and can be caused by both sporadic and germline variants within different tumor types. Prior studies in BC have shown evidence of dMMR by IHC cases MSS based on PCR. This could be due to intra-tumor heterogeneity, specific microsatellite loci evaluated, or penetrance of germline, somatic, or epigenetic alterations. Published data suggests carriers of germline pathogenic MMR variants have a BC risk equivalent to the normal population and currently germline testing is recommended only for BRCA. Currently in advanced BC, standard tumor biomarker testing includes IHC, PCR, and FISH; however, with increasing use of CGP we demonstrate additional actionable biomarkers as well as potential germline variants in MSI-H BC. Methods: DNA was extracted and hybrid capture CGP was performed on 29,160 BC cases. TMB was determined on 0.8-1.2 Mb of DNA and MSI status on 95-114 loci. Genomic LOH was also evaluated. Comparative analysis was done with 101 MSI-H BC, 841 MSS BC and 4,988 non-breast MSI-H cancers. Histological subtype was obtained from the pathology along with orthogonal testing for ER/PR/HER2 status. Somatic-germline-zygosity (SGZ) status was predicted using a published research use algorithm. Select case reports with clinical outcomes will be presented. Results: We identified 101 (0.35% of total) MSI-H BC cases: 29 ER+/HER2-, 5 HER2+, 29 TNBC, and 28 unknown. Amongst BC cases with known subtype, TNBC was enriched for MSI-H vs MSS (53.4 vs 35.8%, p=0.005). The median TMB in MSI-H BC (26.1 mut/Mb, IQR 17.4;42.8) was significantly lower than that of MSI-H colon (46.1mut/MB) and higher than that of MSI-H uterine tumors (22.6mut/Mb) in our comparison group (p Citation Format: Kimberly McGregor, Natalie Danzinger, Jeffrey S. Ross, Kyle Gowen, Alexa B. Schrock, Garrett M. Frampton, Dean C. Pavlick, Jan W. Davis, Carl R. Gray, Jeffrey M. Venstrom. Therapeutic considerations in microsatellite instability high (MSI- H) breast cancers (BC) identified by comprehensive genomic profiling (CGP) [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 PS5-04.

Introduction Metastatic breast cancer is a clinically challenging disease with poor outcomes. Invasive lobular carcinoma (ILC) is a rarer subtype of breast cancer with distinct patterns of metastasis, including frequent GI and female reproductive (FR) metastases (mets). Due to their relative rarity, little is known about the genomic characteristic of ILC across met sites. Here we explore the genomic landscape of 1909 ILC, stratified by met site, with an examination of immune checkpoint inhibitor (ICPI) biomarkers, therapeutic alterations, and resistance mutations. Methods CGP was carried out at Foundation Medicine with hybrid capture for exons from up to 395 cancer-related genes plus select introns from up to 31 genes (PMID 24142049). Tumor mutational burden (TMB) was determined on 0.8-1.1 Mb (PMID: 28420421). Ventana PD-L1 IC staining (SP142; positive >= 1% staining) was available for a subset of samples. 1071 breast-biopsied and 1909 met-biopsied ILC and 6926 breast-biopsied and 1901 met-biopsied IDC were available for analysis. Results High TMB (>= 10 muts/mb) and PD-L1 IC staining are associated with response to ICPI. ILC mets overall had a greater rate of high TMB relative to IDC mets (21% v 9%, p = 7E-25) and breast-biopsied (breast) ILC (21% v 10%; p = 9E-15) with the highest frequency in GI (23%) and skin (21%). PD-L1 IC+ rates were lower in ILC mets (18%) relative to IDC mets (34%) and breast ILC (31%), but were variable across sites, with relatively high rates of positivity in GI (48%), skin (29%), and FR (18%) mets, and no positive staining in bone mets (0/37). Alterations in PIK3CA were higher in ILC mets (58%) relative to IDC mets (34%) and generally exhibited a similar frequency across ILC met sites, with modestly lower prevalence in skin (48%, p = 0.005). Pathogenic alterations in BRCA1/2 were observed in 4.8% of ILC mets overall, with a lower frequency in GI mets (1.3%, p = 0.03). A comparison of ILC breast biopsies to ILC mets revealed 19 genes with higher prevalence in at least one ILC met site, most with known roles in therapy resistance (eg ESR1, NF1, RB1, KRAS, ERBB2, BRAF), though significant heterogeneity was observed across sites. Met-enriched (ME) alterations were highest in ILC from the liver (71%) and lowest in FR (33%). ERBB2 mutations, which are may be targetable with HER2 kinase inhibitors, were predominantly found in liver mets (21%) with significantly lower prevalence in skin (11%), bone (10%), GI (3%), and FR (3%). ESR1 alterations were common in most ILC sites, with the highest prevalence in liver (26%) and low frequency in FR (4%). While FR ILC harbored few ME alterations, the rare alterations were primarily found in NF1 (5%) and NCOR1 (5%). Conclusions CGP revealed significant heterogeneity in ILC mets across tissues. ICPI biomarkers were variable across sites with the highest frequency in ILC GI mets, offering additional potential treatment avenues for these tumors. Alterations in PIK3CA were common in ILC mets with high prevalence across sites, suggesting utility for PIK3CA inhibitors. Therapy-resistant alterations were common in ILC mets but varied across sites. Notably, ERBB2 alterations were most prevalent in ILC liver mets, but less common at other sites. The high prevalence of therapeutic and resistance alterations suggests value in profiling metastatic lesions. breast_ILCmet_ILCbreast_IDCmet_IDCgi_ILCliver_ILCfemale_repr_ILCbone_ILCskin_ILCsample count (N)1071190969261901154639114268199ICPI biomarkerTMB-H (>=10muts/mb)10.2%21.0%4.4%9.2%23.4%18.9%14.0%19.0%21.1%PDL1-IC+ SP142 (subset)31.5% (39/124)17.8% (38/213)57.3% (480/837)34.2% (54/158)47.6% (10/21)11.6% (8/69)18.2% (2/11)0% (0/37)28.6% (6/21)Therapy AssociatedPIK3CA50.0%57.6%29.3%34.3%59.1%57.3%62.3%60.1%48.2%BRCA1/24.9%4.8%9.2%7.7%1.3%4.1%7.0%5.2%5.5%Met-EnrichedESR1_mut3.8%18.3%2.2%17.0%20.8%25.5%3.5%11.9%7.5%ERBB2_mut9.7%12.5%2.0%2.7%3.2%20.7%2.6%9.7%11.1%ARID1A_mut7.5%12.2%4.0%5.5%14.3%11.4%7.9%9.0%10.6%NF1_mut4.4%8.7%3.5%4.2%5.8%8.5%5.3%10.4%8.0%RB1_mut2.7%6.3%4.3%3.5%7.8%5.9%2.6%4.5%11.6%KRAS2.3%5.5%3.9%3.6%5.2%6.7%2.6%6.3%2.0%PTEN_del3.1%4.1%5.7%4.7%1.9%4.7%4.4%1.9%8.0%FGFR20.7%3.1%2.9%2.6%1.3%4.1%1.8%4.1%2.0%NCOR1_mut1.2%2.9%1.1%1.1%3.2%3.0%5.3%1.9%1.0%SMAD42.0%2.9%1.2%1.9%1.9%4.7%1.8%2.6%1.5%BRAF1.6%2.1%1.2%1.4%0.0%1.4%2.6%4.5%2.5%FOXP1_mut0.4%1.6%0.3%0.3%2.6%1.1%0.0%1.9%1.0%APC_mut0.9%1.4%1.1%1.1%0.6%0.6%1.8%3.4%2.0%SOX90.3%1.3%0.5%0.4%0.0%1.1%0.0%1.5%0.5%CASP8_mut0.0%0.8%0.5%0.5%0.0%1.1%0.9%1.1%1.0%PTPN11_mut0.1%0.7%0.2%0.3%0.0%0.6%1.8%1.5%1.0%TERT_mut0.2%0.6%0.4%0.7%2.6%0.0%0.0%1.1%0.5%ALK_mut0.1%0.4%0.2%0.4%2.6%0.0%0.9%0.4%0.5%KMT2D_RE0.1%0.3%0.4%0.3%1.9%0.2%0.0%0.0%0.5%any_ME32.5%60.1%30.1%42.4%53.9%71.4%33.3%53.4%52.3% Citation Format: Ethan Sokol, Dexter X Jin, Jeffrey Ross, Garrett M Frampton, Steffi Oesterreich. Comprehensive genomic profiling (CGP) of metastatic invasive lobular carcinomas reveals heterogeneity in immune biomarkers and resistance alterations across biopsy sites [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 PD8-01.

Oncogenic kinase activation by gene fusions can promote cancer development and tumor progression, however, kinase fusions have not been characterized extensively in breast cancer. In this study, we characterized kinase fusions in a large cohort of 4854 breast cancer patients using clinical DNA and/or RNA next generation sequencing platforms. Twenty-seven cases harboring kinase fusions were identified, including 11 FGFR (5 FGFR2, 3 FGFR3, 3 FGFR1), 5 BRAF, 4 NTRK1, 2 RET, 2 ROS1, 1 ALK, 1 ERBB2, and 1 MET. Eight patients with a history of endocrine therapy had available pre-treatment samples, of which six were negative for kinase fusion, and ESR1 hotspot mutations were not observed in any of these kinase fusion-positive samples. These findings suggest a potential role for kinase fusions in endocrine therapy resistance and prompted us to model the kinase fusions in human breast cancer cell lines. Ectopic expression of LMNA-NTRK1 fusion kinase activated growth factor signaling cascades, including PI3K-AKT and MAPK-ERK pathways, and promoted hormone-independent growth in MCF7 and T47D cells. Enforced expression of the LMNA-NTRK1 fusion conferred resistance to the ER antagonist fulvestrant and combined treatment of fulvestrant and the Trk inhibitor larotrectinib completely blocked the growth of LMNA-NTRK+ breast cancer cells. Similarly, expression of the EML4-ALK fusion also activated growth factor signaling pathways and caused resistance to estrogen depletion and induced sensitivity to the ALK inhibitor, Ceritinib. Treatment of xenografted LMNA-NTRK1 expressing tumors confirmed the efficacy of the combined treatment of antiestrogen and NTRK1 inhibition in vivo. Two patients with acquired LMNA-NTRK1 fusions and metastatic disease received larotrectinib and demonstrated clinical benefit. Overall, our findings demonstrate that kinase fusions promote endocrine resistance in ER-positive breast cancer, and suggest that fusion screening in advanced breast cancer, particularly those with ER-positive breast cancer at progression on hormone therapy can identify rare tumors harboring targetable kinase fusions. Citation Format: Bo Liu, Dara S. Ross, Alison M. Schram, Pedram Razavi, Stephen M. Lagana, Yanming Zhang, Maurizio Scaltriti, Jacqueline F. Bromberg, Marc Ladanyi, David M. Hyman, Alexander Drilon, Ahmet Zahir, Ryma Benayed, Jaclyn F. Hechtman, Sarat Chandarlapaty. Kinase fusions drive endocrine resistance in estrogen receptor-positive breast cancer [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 5280.

Background: Sex differences in the prevalence of cancer subtypes can inform patient screening, diagnosis, and treatment, but for many disease ontologies (DO), sex disparities have not been fully explored, especially with regards to ancestral and genomic subsets. Furthermore, existing data and reports may lack centralized expert pathology review or comprehensive genomic profiling (CGP) that can enable better characterization of biological and environmental etiologies. Design: The cohort consisted of 276,696 de-identified patient cancers, 45.7% male and 54.3% female, submitted to Foundation Medicine for CGP. Each patient was assigned an ancestry of either African, admixed American, East Asian, European, or South Asian, using a SNP-based machine learning methodology (J. Newberg et al., AACR 2019). Within the dataset 321 DOs originate at a known tissue site of origin common to both sexes. Binomial testing for significance of expected (50%) versus observed ratio of the sexes was performed for each DO, with correction for multiple hypothesis testing. For each DO, Fisher's exact tests were used to identify associations between ancestry and sex and associations between sex and gene-level genomic biomarkers (including the detection of a viral-derived gene) with ≥ 1% prevalence within the DO. Results: Of the 321 DOs, significant sex differences (corrected p-value ≤ 0.05) were seen in 106 DOs, of which 84 were significantly skewed towards males and 22 towards females. Among all DO and ancestry combinations, 14 DO-ancestry pairings showed a statistically significant (corrected p-value ≤ 0.05) skew towards one sex when compared to all other patients in the disease group. Notably, there was an enrichment of lung small cell undifferentiated carcinoma in East Asian men (OR = 2.75, corrected p-value = 0.027) and esophagus squamous cell carcinoma in East Asian men (OR = 2.55, corrected p-value = 0.028). 33 DOs exhibited sex biases in gene alterations (OR ≥ 2 and corrected p ≤ 0.05). Of particular note are a few genes that were highly enriched in a single sex across multiple distinct disease groups: CDKN2A/CDKN2B (bladder, anus, CNS non-glioma, and kidney diseases), TERT (hepato-biliary, kidney, and anus diseases), and ASXL1 (myelomas and plasma neoplasms) for males; ATRX in both lung small cell undifferentiated carcinoma and colon adenocarcinoma for females. Conclusions: Using FMI's large cancer cohort, significant differences in the occurrences of the sexes were observed in various DOs and have the potential to inform our understanding of different disease etiologies and possible courses of therapy. Further analyses incorporating covariates may help to explain the source of these imbalances. Citation Format: Benjamin G. Kaplan, Peter Halmos, Justin Y. Newberg, Ethan S. Sokol, Meagan Montesion, Lee A. Albacker, Vincent Miller, Jeff S. Ross, Garrett M. Frampton, Jonathon K. Killian. Pan-cancer analysis of sex differences and their associations with ancestry and genomic biomarkers in a large comprehensive genomic profiling dataset [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 2337.

Background: Androgen independence is the most challenging problem in the treatment of advanced prostate cancer. Although most patients with early-stage prostate cancer can be cured with surgical and/or radiation treatment, the remainder of patients ultimately recur with metastatic disease upon development of resistance to anti-androgen therapy. Epithelial mesenchymal transition (EMT) and lineage plasticity has been proposed as key mechanisms underlying resistance in treatment-induced prostate cancer with neuroendocrine features (NEPC). There has been a rise in treatment-induced NEPC with increased use of anti-androgen therapies. ABI1 regulates EMT by suppressing FYN-STAT3 activation downstream from noncanonical WNT pathway, but the interaction of ABI1 with androgen pathway is not known. We sought to determine the effect of neoadjuvant hormone treatment (NHT) on ABI1 expression levels in different tumor types. Methods: Here we analyzed ABI1 expression (IHC) in tissue microarray containing 75 tumors with no prior treatment, 198 patients subjected to NHT prior to surgery and 39 castration resistant tumors (total n=307). Androgen-induced ABI1 expression was determined upon R1881 treatment by Western blotting and immunofluorescence in hormone- sensitive LNCaP cells. Results: ABI1 expression was significantly downregulated in tumors from patients undergoing NHT(p Conclusion: ABI1 is an androgen sensitive gene. ABI1 expression is increased in CRPC, most likely secondary to androgen receptor reactivation. It is subsequently downregulated in NEPC, hence, we hypothesize that this may be a key mechanism driving the NEPC phenotype through promoting EMT. Further understanding of the role ABI1 may play in this transformation will provide a more mechanistic understanding of prostate cancer progression and maylead to improved targeted treatment options. Citation Format: Leszek Kotula, Baylee A. Porter, Xiang Li, Maria A. Ortiz, Rakesh Khanna, Fan Zhang, Sonia H. Kung, Ladan Fazli, Martin E. Gleave, Gennady Bratslavsky, Jeffrey S. Ross. ABI1 and androgen dependence in prostate cancer [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 340.

Background The prognosis of invasive lobular carcinoma (ILC) is controversial. Some studies report equivalent or better outcomes than invasive ductal carcinoma, likely due to ILC’s favorable prognostic phenotype. Classic ILC (CL) is typically of low histologic grade, estrogen receptor (ER) positive and human epidermal growth factor receptor-2 (HER2) negative. However, worse long term outcomes have been observed because ILC tends to be multicentric, bilateral, and to show higher rates of late metastases in uncommon sites. Additionally, subtypes of ILC, including solid (SOL) and pleomorphic (PL), have been reported to display more aggressive behavior. Yet current treatment paradigms are distinguished by hormone receptor profile and stage, not by tumor morphology. This study aimed to evaluate the clinicopathologic and immunohistochemical (IHC) characteristics of ILC, particularly those with unusual patterns. Methods From 2009-2018, consecutive in-house biopsy and resection specimens with the diagnosis of ILC were identified by a retrospective review of our institutional database. Demographic, morphologic, and IHC data were obtained from pathology reports. Cases with equivocal HER2 IHC results are sent for fluorescence in situ hybridization analysis at our Center. CL ILC was defined by small cells with low nuclear grade and discohesive growth arranged in single file or cords. PL ILC is composed of larger, more atypical cells, which still exhibit discohesion typical of lobular phenotype. SOL ILC consists of tumors growing in large sheets with minimal intervening stroma. Results Over the nine year study period, a total of 3,028 ILCs from 2,322 patients were identified. Most (74%, 2229/3028) cases were classified as CL while 22% (665/3028) were PL and 4% (134/3028) SOL (Table 1). The median patient age was 60 years (range 32-95). The median tumor size was larger for PL (2.3 cm) and SOL (2.4 cm) subtypes than CL (1.1 cm). IHC results were available for 896 specimens. The majority of all ILCs showed an expected phenotype: ER positive and HER2 negative (99% CL, 84% PL, 97% SOL). A smaller subset of ER positive ILC were also HER2 positive (1% CL, 8% PL, 3% SOL). No CL or SOL ILC were ER negative; in contrast 8% (21/270) of PL ILC were ER negative. Twelve (5%) PL ER negative cases were also HER2 negative. Overall, 15% of ILC were HER2 positive, with the PL variant showing the highest incidence (11%, 30/270). Lymph node (LN) metastases were seen in 22% (66/291) of CL cases, 55% (53/96) PL and 29% (17/58) SOL. Metastases in 10 or more LN (N3) was seen only in PL and SOL cases (13% and 5%, respectively). Conclusions In this study, up to 8% of ILC cases were found to be ER negative, a finding unique to the PL variant. No CL or SOL ILC was ER negative. HER2 positivity, not typically expected in ILC, was seen in 15% of cases, most commonly in the PL type. ILC that was both negative for ER and HER2 was observed in 5% of cases, and only within the PL subgroup. PL ILC showed higher median tumor size at presentation than CL and nodal metastases were also more common in PL than in CL or SOL ILC (55% vs 22% and 29%, respectively). Our findings are in keeping with prior observations that compared to CL ILC. PL ILC exhibits features associated with unfavorable clinicopathologic features, such as higher tumor and nodal stage at presentation. Additionally, this is the first study to systematically examine the prevalence of unconventional IHC phenotypes of ILC (ER negativity and/or HER2 positivity) in a large series from one institution. Table 1. Summary of Clinicopathologic FeaturesClassicPleomorphicSolidILC types (n=3028)2229665134Median age, (range)60 yrs (32-94)61 yrs (31-95)66 yrs (44-93)Median size, cm 1.12.32.4Immunophenotype (n=896)ER+/HER2-99% (551/559)84% (228/270)97% (65/67)ER+/HER2+1% (8/559)8% (21/270)3% (2/67)ER-/HER2+0% (0/559)3% (9/270)0% (0/67)ER-/HER2-0% (0/559)5% (12/270)0% (0/67)Lymph node status (n=445)N077% (225/291)45% (43/96)71% (41/58)N120% (59/291)31% (30/96)19% (11/58)N22% (7/291)11% (11/96)5% (3/58)N30% (0/291)13% (12/96)5% (3/58) Citation Format: Matthew G Hanna, Anne Grabenstetter, Dara S Ross, Lee K Tan. The prevalence of invasive lobular carcinomas with unconventional immunohistochemical phenotypes [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 P5-02-09.

Background: Inflammatory breast cancer (IBC) is a distinct clinicopathologic entity that carries a worse prognosis relative to non-IBC breast cancer even when matched for biomarkers (ER/PR/HER2). Genomic profiling of IBC cases may identify alterations that suggest response to targeted therapies, but is best implemented by an integrated NGS assay capable of detecting all classes of genomic alterations (GA). Methods: Hybridization capture of 3769 exons of 236 cancer-related genes and 47 introns from 19 genes that are frequently rearranged in cancer were fully sequenced to high, uniform coverage from a commercial CLIA-certified laboratory (Foundation Medicine). Results: Of 2,208 clinical breast cancer cases assayed, 55 IBC cases were identified, and of the 50 cases for which hormone receptor and HER2/neu status were know, 34% were ER-/PR-/HER2- (TNBC). IBC cases harbored 274 GA with an average of 5.0 GA/tumor (range 1-15). At least one alteration associated with an FDA approved therapy or clinical trial was identified in 53/55 (96%) of cases, yielding an average of 2.6 clinically relevant GA/case. Genes most frequently altered were TP53 (60%), MYC (31%), PIK3CA (25%), ERBB2 (20%), FGFR1 (18%) and PTEN (16%). MYC amplifications were present in 24% of the 2,208 clinical breast carcinoma cases. In the TNBC subset of IBC, 8/19 (42%) pt samples showed MYC amplification (median copy number 8X, range 7-20) as compared to 9/36 (25%) in non-TNBC pt samples (median copy number 7X, range 6-21). Within this prospective series, treatment decisions were made based on FoundationOne results. A58-year old pt with likely secondary IBC harbored two ERBB2 activating base substitutions (V777L and S310F), but without amplification of ERBB2), and had durable response to lapatinib (Ali et al. JCO 2014). In another case, a 53 year old pt presented with ERBB2-amplified IBC now refractory to HER2-targeted therapy. FoundationOne testing revealed an activating EGFR mutation (L858R) as well as the previously described ERBB2 amplification, suggesting that the EGFR alteration may underlie the acquired resistance. The pt responded to erlotinib monotherapy for 8 months. (Ali et al. Clin Br Ca, 2013). Clinical follow-up for additional patients is ongoing. Conclusions: 96% of IBC cases harbored at least one alteration that suggests responsiveness to agents that are FDA approved or being studied in clinical trials. IBC cases also frequently had MYC amplifications (31%), but this may reflect a high percentage of TNBC-IBC cases with MYC amplifications (42%) in this series. Given the limited treatment options and poor prognosis of patients with metastatic IBC, the FoundationOne assay with comprehensive NGS-based genomic profiling has the potential to identify new treatment paradigms and address an unmet clinical need for this disease. Citation Format: Norma A Palma, Siraj M Ali, Kai Wang, Juliann Chmeleiki, Gary Palmer, Deborah Morosini, Jeffrey S Ross, Vincent A Miller, Phil J Stephens, Massimo Cristofannilli. Genomic profiling by FoundationOne® analysis of inflammatory breast cancer cases reveals a high frequency of clinically relevant genomic alterations (GA) [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P6-14-02.

Background: Targeted ERBB2/HER2 inhibitors are FDA-approved for the treatment of breast and gastric cancers. Immunohistochemistry (IHC) and fluorescence in situ hybridization (FISH) to test for overexpression and amplification of ERBB2/HER2, respectively, are performed as part of routine clinical care. Recently, activating mutations in ERBB2 have been reported and may confer sensitivity to targeted agents. Testing for these mutations is not routine, and testing for amplifications is not done outside of approved indications. We explored the complete spectrum of activating ERBB2 mutations and amplifications across a collection of ∼7,300 solid tumor specimens, including a large number of breast cancer specimens, to determine a)how many breast cancer patients could benefit from HER2-targeted therapy due to activating mutations in ERBB2 and b)how widespread ERBB2 alterations are across the solid tumor spectrum. Methods: Extracted DNA from clinical tumor samples underwent comprehensive genomic profiling using FoundationOne . 3769 exons of 236 cancer-related genes and 47 introns from 19 genes that are frequently rearranged in cancer were fully sequenced to high, uniform coverage and results were analyzed for base substitutions, insertions and deletions, amplifications/deletions, and rearrangements. Results: Known oncogenic ERBB2 alterations were identified in approximately 6% of all solid tumors across 27 different histologies. Of all the ERBB2 alterations, activating mutations in ERBB2 were identified in 131 samples and amplifications were observed in 246 samples. Two samples harbored an ERBB2 rearrangement. Ten samples harbored multiple ERBB2 mutations, yet mutations and amplifications were mutually exclusive in 91% of mutated cases. ERBB2 amplification in breast and gastric cancers accounted for only 30% of these alterations. Breast cancer accounted for 37% of all the ERBB2 alterations detected, and included primarily amplifications. 25% of the alterations found in breast cancers were activating base substitutions. Standard tests for overexpression or amplification of ERBB2 would fail to detect these potentially treatable mutations. Non-amplification ERBB2 alterations were enriched in cases of relapsed lobular breast cancer. Multiple breast cancer patients with non-amplification ERBB2 alterations have responded to combinations of anti-HER2 targeted therapies. Conclusions: Comprehensive genomic profiling through FoundationOne identifies 25% more breast cancers that may be susceptible to HER2 targeted therapy due to the presence of activating mutations in ERBB2. Also, many more solid tumor cases could potentially benefit as well from ERBB2 testing. The current policy of testing only breast and gastric tumors for only HER2 amplifications is greatly limiting the potential value of the ERBB2/HER2 inhibitors. Citation Format: Gary A Palmer, Jeffrey S Ross, Kai Wang, Garrett M Frampton, Siraj M Ali, Norma Palma, Deborah Morosini, Vincent A Miller, Roman Yelensky, Doron Lipson, Philip J Stephens, Juliann Chmielecki. Activating mutations in ERBB2/HER2 as found by FoundationOneTM represent potential therapeutic targets in breast cancer [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P4-15-03.

In-frame deletions within exon 19 (Ex19Del) and the Leu858Arg mutation within exon 21 (L858R) are the two most common oncogenic EGFR mutations found in non-small cell lung cancer (NSCLC). Currently, there is little data differentiating Ex19Del from L858R, and patients whose tumors harbor Ex19Del and L858R are approached similar in clinic when choosing EGFR tyrosine kinase inhibitors (TKIs). Unfortunately, resistance mutations may emerge against all classes of EGFR TKIs, including osimertinib. Recently, G724S has been identified as a novel mutation which is selected in osimertinib resistant tumors. Interestingly, in a large genomic profile database of NSCLC patient samples we found that G724S co-occurs preferentially with Ex19Del (15/19 G724S cases) but not with L858R (0/19 G724S cases). We further combined in silico and in vitro investigations to demonstrate the mechanism of G724S mutation confers resistance to TKIs in an oncogenic variant-dependent manner. Our in silico investigations showed evidence of reduced stability of the osimertinib-bound G724S/Ex19Del complex but not the osimertinib-bound G724S/L858R complex. G724S/Ex19Del-transduced cells are resistant to osimertinib treatment in vitro, while G724S/L858R-transduced cells are sensitive to osimertinib treatment. These results suggest G724S/Ex19Del may emerge under selective pressure from EGFR TKIs, and the underlying oncogenic profile of EGFR can influence the development of drug resistance mutations. Our results have potential direct implications for the treatment of patients with EGFR-mutant NSCLC. Citation Format: Yunkai Zhang, Benjamin P. Brown, David Westover, Yingjun Yan, Huan Qiao, Vincent Huang, Zhenfang Du, Jarrod A. Smith, Jeffrey S. Ross, Vincent A. Miller, Siraj Ali, Lyudmila Bazhenova, Alexa B. Schrock, Jens Meiler, Christine M. Lovly. On-target resistance to mutant selective EGFR inhibitors develops in an allele-specific manner dependent on the original EGFR activating mutation [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 930.

Background: Immune checkpoint inhibitors (ICPIs) have led to dramatic improvement in outcome of several cancers. Program death ligand1 (PD-L1) staining and tumor mutational burden (TMB) have emerged as independent predictive biomarkers of ICPIs in lung cancer. Here we examine the landscape of TMB and PD-L1 expression in breast cancer and present a case of patient with high TMB and PD-L1 negative breast cancer with exceptional response to ICPIs. Methods: Hybrid-capture based comprehensive genomic profiling of 395 cancer related genes using the FoundationOne assay was performed on 14,867 breast carcinomas sequenced in the course of routine clinical care. Ventana (SP-263) PD-L1 status (n=1425) and hormone receptor status was available for a subset of patients. Subgroup analyses were performed based on histological type [invasive lobular carcinoma (ILC, n=740)], molecular subtypes [ER-positive (ER+; n= 1371), HER2-amplified (HER2+; n=1522), and TNBC (n=917)], patient age (≤45, 46-60, ≥61), and local vs. metastatic disease (n=5241 and 6710). Results: Consistent with previous reports, the rates of positive PD-L1 staining are highest in TNBC (14%) and lowest in HER2+, ILC, and ER+ disease (6.0%, 5.1%, 2.3%). Interestingly, the rate of PD-L1 positivity, defined as ≥1% tumor staining, was significantly lower in metastatic disease vs. local disease (6.3% vs. 11.1%; p = 0.005). The frequency of high TMB, defined as >10 mutations/mb, was greatest in ILC and HER2+ disease (13.6% and 9.9%) and lowest in TNBC and ER+ disease (7.0% and 6.9%). Rates of high TMB were associated with increased patient age (3.7%, 9.3%, and 12.8% frequency in patients ≤45, 46-60, and ≥61) and were significantly higher in metastatic vs. local disease (11.1% vs 5.3%; p20 mut/mb), there was a significant association between TMB and PD-L1 positivity (OR = 2.6, p = 0.023). Nevertheless, even with high cutoff, 79% of the TMB high samples were PD-L1 negative. We also report on a stage IIIb (T4, N2, M0) ER+ HER2- breast cancer patient with high TMB (40muts/mb) and negative PD-L1 IHC who previously progressed on aromatase inhibitor with CDK4/6 inhibitor and chemotherapy but achieved durable complete response of > 1 year with nivolumab in combination with capecitabine. Conclusions: Predictive biomarkers for ICPIs are critical to identify a subset of breast cancer patients who may respond to immunotherapy. TMB high and PD-L1 positivity do not significantly co-occur and the majority of TMB high cases were PD-L1 negative. Nevertheless, this group of patients may still benefit with ICPIs. Further studies are needed to evaluate this subset of patients. Citation Format: Ethan Sokol, Lee Albacker, Aixa Soyano, Ricardo Parrondo, Brenda Ernst, Emmanuel Gabriel, Garrett Frampton, Jeffrey Ross, Siraj Ali, Jon Chung, Saranya Chumsri. Incidence of high tumor mutation burden (TMB) and PD-L1 positivity in breast cancers and potential response to immune checkpoint inhibitors (ICPIs) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 4894.

Background. FNAC are generated during the FNA procedure - for example, commonly used in conjunction with endobronchial ultrasound (EBUS-FNA) as a minimally invasive tool for the diagnosis and management of non small-cell lung cancer (NSCLC) - as a histologic QC for the presence of malignant cells. Formalin-fixed core biopsies and aspirate cell blocks (FNACB) are then created for comprehensive genomic profiling (CGP). However, the routine pathology workup may render the FNACB unable to generate a CGP result, while the FNAC remain in reserve. In the current study we performed a retrospective multi-cancer, multi-institutional analysis of FNAC performance, including computed tumor purity analysis, using a regulated commercial grade CGP assay. Experimental procedures. 119 FNAC met minimum specimen cellularity (10K total cellularity) and DNA extraction (50ng) criteria for CGP. Extracted DNA from FNACS underwent hybrid-capture-based CGP for all classes of genomic alterations in 315 genes. Tumor mutational burden (TMB) was determined on 1.1 Mbp of sequenced DNA and microsatellite instability (MSI) was determined on 114 loci. A comparison set of 182 FNACB specimens underwent the same CGP assay. Individual specimen CGPs were annotated as Pass, Fail, or Qualified per computational biology standard procedures that factor in sequencing quality metrics such as mean target coverage and computational tumor purity modeling. Results. The 119 FNAC, collected at 14 different medical facilities, encompassed 44 distinct tumor disease ontologies (DO) from 17 unique anatomic biopsy sites. 105 of 119 FNACS (88%) received a passing report, 10 (8%) were qualified, and 4 (3%) failed. In the FNACB comparison group, 155 of 182 reports were pass (85%), 22 qualified (12%), and 5 failed (3%). The number of passing reports in FNACS versus FNACB was equivalent (P=0.5). The FNAC had a significantly higher mean target coverage than FNACB (703 vs 665; P= 0.04995). The average computational tumor purity metric was significantly higher in FNACS than in FNACB (57.9% vs. 45.3%; P=0.00001). Conclusions. This study provides new data from FNAC that further support their utility for clinical testing. FNAC derived from multiple medical facilities, biopsy sites, and tumor types were found to be equivalent or superior to FNACB for CGP. Finally, the relative cellular homogeneity and high tumor cell content of both FNAC and FNACB likely confirms the discohesiveness of malignant cells and enables CGP without required enrichment given that macrodissection would be challenging for these samples in a daily practice setting. Citation Format: Jonathan K. Killian, Dean Pavlick, Lindsay Croshier, John Truesdell, Chris Wright, Tim Brennan, Vera Banning, Lazaro Garcia, Laurie Gay, Gene Stirchak, Jo-Anne Vergilio, Christine Malboeuf, Nhu Ngo, Julia Elvin, Siraj Ali, Vince Miller, Jeffrey S. Ross. A multi-institutional, multi-tumor analysis of fine needle aspiration (FNA) cytology smear (FNAC) performance in clinical comprehensive genomic profiling [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 1694.

Background: E-cadherin (ECAD), encoded by the CDH1 gene, is expressed in the majority of invasive ductal carcinomas (IDC). In contrast, ECAD expression is lost in ˜90% of invasive lobular carcinomas (ILC) due to genomic alterations of CDH1. Immunohistochemical (IHC) staining for ECAD has thus increasingly been utilized to differentiate between ductal and lobular lesions. This study examines the correlation between CDH1 mutation profile, morphology and ECAD IHC status in invasive breast carcinomas (BC). Methods: The Memorial Sloan Kettering-Integrated Mutation Profiling of Actionable Cancer Targets (MSK-IMPACT) platform is a hybrid capture next-generation sequencing assay to detect somatic alterations in exons and selected introns of 468 cancer-related genes. Invasive BC tested from 1/2014 – 5/2018 which yielded a CDH1 alteration were identified and only cases with ECAD IHC results were included. Histologic features were extracted from pathology reports. All available H&E/ECAD slides were reviewed by two pathologists to confirm the histology and immunophenotype. Results: CDH1 alterations were identified in 455 out of 4103 BC analyzed (11%) and ECAD IHC were available for 209 (46%) tumors. Most of the tumors were reported as ILC (70%, 147/209). The majority (77%, 161/209) were ECAD-negative but 23% (48/209) showed some ECAD staining. Table 1 summarizes the histologic subtypes and IHC results. Upon review of H&E/ECAD slides in cases with ECAD staining, the histologic subtype was re-classified in 52% (25/48) of cases (Table 2). For ECAD-negative BC, at least 61% (98/161) of CDH1 alterations were truncating mutations (frameshift deletion 47/161, nonsense 51/161) and 39% (63/161) were variants of unknown significance (VUS). In cases with ECAD staining, 48% (23/48) of CDH1 alterations were truncating mutations (frameshift deletion 12/48, nonsense 11/48) however 52% (25/48) were VUS. Conclusion: In this study, 88% (141/161) of cases with negative ECAD staining and CDH1 alteration were classified as ILC, confirming the known observation that CDH1 alteration and loss of ECAD expression is relatively specific for a lobular phenotype. The correlation is, however, imperfect. A subset of BC (23%; 48/209) with CDH1 alteration still exhibits ECAD expression. Morphologically this subset consists of ILC in 60% (29/48), IMC in 25% (12/48) and IDC in 15% (7/48). Our preliminary findings suggest that aberrant ECAD expression can occur in BC with morphologic features of ILC due to certain CDH1 alterations that most likely result in non-functional ECAD complex. BC with histologic features of ILC should therefore not be re-classified as IDC/IMC based solely on the status of ECAD expression. Given that ILC is clinically distinct from IDC, correct classification based on morphological and molecular features is prudent. Table 1:Histologic subtypes and ECAD IHC ILC (n = 147)IDC (n = 13)IMC (n = 49)ECAD staining Negative141020Strong, diffuse4135Heterogenous0020Reduced/weak204IMC: invasive mammary with mixed ductal/lobular features Table 2:Morphologic review of cases with ECAD stainingECAD stainingOriginal DX (n)Morphology Review DX ILCIDCIMCStrong, diffuseILC (4)301 IDC (13)670 IMC (5)500HeterogenousIMC (20)10010Reduced/weakILC (2)200 IMC (4)301 Citation Format: Grabenstetter A, Mohanty AS, DeLair DF, Tan LK, Ross DS. Correlation of CDH1 alterations and aberrant E-cadherin expression in lobular carcinomas [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 PD7-05.

Background: Traditional FDA-approved assays employing PCR, FISH, or IHC report presence or absence of single markers for therapeutic guidance. Comprehensive Genomic Profiling (CGP) via hybrid-capture NGS-based assays, however, can simultaneously interrogate hundreds of markers to offer a broader picture for clinical actionability. To support the reliability of using hybrid-capture NGS as a clinical platform, we present a comparative analysis on the variant calls detected between two clinically validated assays - FoundationOne CDx (F1CDx), which is the first FDA-approved NGS-based platform that indicates personalized therapies; and a clinically-validated NGS tumor panel assay, performed in an experienced CLIA-certified and CAP-accredited lab in the academic setting. The assays capture 324 and 262 genes, respectively. Methods: DNA from 188 clinical cases across an array of tumor types was selected to be sequenced by both assays, performed at the respective labs. For F1CDx, CGP was performed using a hybrid-capture, adaptor ligation-based NGS assay to a mean coverage depth of over 500X. Short variants - base substitutions, small insertions and small deletions - were identified and curated. A similar procedure was followed for the academic lab. Resulting variant calls were compared at FMI. Results: Across the 157 genes that overlap between the assays, 550 clinically reportable short variants that met filtering criteria were identified. Of these, 491 were called by both, 15 were unique to the academic assay, and 44 to F1CDx. This resulted in a bi-directional agreement of ~94% (97.0% with the academic NGS assay as the reference method, and 91.0% using F1CDx as the reference method). The bi-directional agreement for base substitutions alone is (97.4% (discordances due to low allele fraction (AF)), 92.5%, respectively), while for indels in non-homopolymer regions is (96.1%, 89.7%, respectively). Negative agreement between the assays was greater than 99.8%. Conclusions: Overall concordance among short variant calls between the two NGS assays is high, comparable to reproducibility results at low AF in qPCR. A further advantage over traditional assays is that variant calling is mostly automated but with some curation, whereas with use of IHC or FISH, the call is wholly dependent on pathologic review and interpretation. This supports the ability of robustly validated hybrid-capture based NGS to subsume sequential testing using other technologies commonly employed in oncology in the clinical diagnostics space. Citation Format: Ninad P. Dewal, Joel Skoletsky, Wai-Ki Yip, Caitlin Patriquin, Yuting He, Jeffrey S. Ross, Vincent A. Miller, Philip J. Stephens, Christine Burns, Christine Vietz, Yali Li, Colin Pritchard, James X. Sun. Comparative analysis of clinically validated NGS-based assays reveals high concordance across short variants [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 417.

Introduction: Increase in targeted therapies has resulted in the need for a single assay capable of detecting diverse biomarkers indicated for these agents. Comprehensive genomic profiling (CGP) provides such a solution, but due to the complexity and number of assays available today, standardization of validation has become critically important. We present FoundationOne CDx, the first NGS-based comprehensive companion diagnostics (CDx) platform developed and performed in compliance with FDA 21 CFR part 820. The assay interrogates 324 genes, and has CDx indications in five tumor types associated with 17 targeted therapies (Table 1). The versatile assay design will facilitate streamlined development of future CDx indications. Methods: DNA extracted from FFPE tumor tissue underwent whole-genome shotgun library construction and hybridization-based capture, followed by sequencing using Illumina HiSeq 4000. Sequence data were processed using a proprietary analysis pipeline designed to detect base substitutions, indels, copy number alterations, rearrangements, microsatellite instability (MSI), and tumor mutational burden (TMB). Results: Clinical validity was established such that the concordance between CGP and approved CDx were statistically non-inferior to that of two runs of approved CDx. For analytical validity, limit of detection (LoD) was at allele frequency 4% for known substitutions and indels. LoD was 16% tumor content for copy number amplifications, 30% for homozygous deletions, 11% for rearrangements, 12% for MSI, and 20% for TMB. Concordance with an orthogonal NGS platform was 94.6% for substitutions and indels. Within-assay reproducibility had PPA 99.4%. Conclusion: Rapid expansion of targeted therapies and CDx has necessitated a new approach and urgency to defining performance standards. We developed a comprehensive CDx assay and demonstrated clinical and analytical validity to support and accelerate using CGP for routine clinical care. Table 1. Companion Diagnostic IndicationsIndicationBiomarkerTherapyNon-small cell lung cancer (NSCLC)EGFR exon 19 deletions and EGFR exon 21 L858R alterationsafatinib, gefitinib, or erlotinibEGFR exon 20 T790M alterationsosimertinibALK rearrangementsalectinib, crizotinib, or ceritinibBRAF V600Edabrafenib in combination with trametinibMelanomaBRAF V600Edabrafenib, vemurafenibBRAF V600E and V600Ktrametinib, cobimetinib, in combination with vemurafenibBreast cancerERBB2 (HER2) amplificationtrastuzumab, ado-trastuzumab-emtansine, or pertuzumabColorectal cancerKRAS wild-type (absence of mutations in codons 12 and 13)cetuximabKRAS and NRAS wild-type (absence of mutations in exons 2, 3, and 4)panitumumabOvarian cancerBRCA1/2 alterationsrucaparib Citation Format: James X. Sun, Yali Li, Coren Milbury, Joel Skoletsky, Christine Burns, Wai-ki Yip, Jun Luo, Ninad Dewal, Adrienne Johnson, Kyle Gowen, Jing Tong, Yuting He, Jie He, Pei Ma, Jared White, Steve Roels, John Truesdell, Eric Peters, Houston Gilbert, Charlie Wu, Erica Schleifman, Johannes Noe, Carl Barrett, Kenneth Thress, Suzanne Jenkins, Julia Elvin, Geoff Otto, Doron Lipson, Jeffrey Ross, Vincent Miller, Philip Stephens, Michael Doherty, Christine Vietz. A clinically validated comprehensive companion diagnostic platform for care of patients with advanced cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 4757.

Introduction: Patients with HER2-positive breast cancer may benefit from targeted therapies including trastuzumab and pertuzumab. While FDA has approved companion diagnostics (CDx) using FISH or IHC for identifying ERBB2 amplifications, molecular diagnostic testing in breast cancer is rapidly evolving towards comprehensive genomic profiling (CGP) due to an increasing number of biomarkers. However, clinical validity of ERBB2 testing in CGP has not yet been demonstrated in an FDA-approved manner. We present the first ERBB2 test using NGS as a follow-on CDx, which is part of FoundationOne CDx (F1CDx). Methods: Clinical validity was established against the FDA-approved HER2 FISH test, using retrospective breast cancer patient samples. For each sample, two tests using the FISH assay were performed, and clinical validity was established such that the F1CDx result is statistically non-inferior to the performance between two runs of FISH. Concordance was calculated on samples that agree between the two FISH tests. Results: Results are shown in Table 1. An exploratory analysis including low-level ERBB2 amplifications (CN=4; ploidy+2) detected by F1CDx demonstrated an improved PPA of 93.8%. A non-inferiority test demonstrated a margin of 8.0%. Further analysis showed that the discordance is most significant among samples with a low FISH ratio. In samples that are FISH-positive but F1CDx-negative, a second test of FISH only yields a FISH-FISH concordance of only 55%, suggesting significantly decreased FISH-FISH reproducibility. Conclusion: We presented a follow-on CDx for ERBB2. The F1CDx platform provides a single assay that identifies cancer patients who may be eligible to receive FDA-approved targeted therapeutics, conserves tissue by avoiding serial testing and can serve as clinical trial assay for studies requiring a molecular biomarker for eligibility. The data demonstrating clinical and analytical validity of ERBB2 may accelerate the use of CGP for routine clinical use. Table 1. The ERBB2 companion diagnostic in FoundationOne CDxCompanion diagnosticERBB2 amplificationIndicated useTrastuzumab, ado-trastuzumab-emtansine, or pertuzumab in breast cancerApproved CDx comparatorHER2 FISH PharmDx® Kit (Dako Denmark A/S)Positive percent agreement (PPA)89.4% (101/113)Negative percent agreement (NPA)98.4% (180/183)Median unique sequence coverage>500XPrecision100%Limit of Detection19.7% tumor content Citation Format: Wai-ki Yip, Joel Skoletsky, Pei Ma, Jun Luo, Coren Milbury, Christine Burns, John Truesdell, Julia Elvin, Geoff Otto, Doron Lipson, Jeffrey Ross, Vincent Miller, Philip Stephens, Michael Doherty, Christine Vietz, James X. Sun, Yali Li. An ERBB2 follow-on companion diagnostic for clinical care of patients with breast cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 1607.

Introduction Patients with PI3K/Akt-pathway activation may be sensitive to selective Akt-inhibitors that are currently under development. We have developed a novel next-generation sequencing (NGS)-based composite biomarker assay that identifies patients with PI3K/Akt-pathway activated tumors by identifying activating PIK3CA and AKT1 alterations, and inactivating alterations in PTEN. This assay was analytically validated, and applied to triple-negative breast cancer (TNBC) patients in the LOTUS trial (NCT02162719), a placebo-controlled phase II clinical trial to assess the safety and efficacy of adding ipatasertib to paclitaxel treatment in patients with metastatic TNBC (Kim et al., 2017). Methods DNA extracted from FFPE tumor tissue underwent whole-genome shotgun library construction and hybridization-based capture, followed by sequencing using Illumina HiSeq 4000. Sequence data were processed using a proprietary analysis pipeline designed to detect base substitutions, indels, copy number alterations, genomic rearrangements, microsatellite instability, and tumor mutational burden. The assay further evaluated the PI3K/Akt-pathway activation biomarker status that consists of six features: 1-2) AKT1 and PIK3CA activating mutations, 3) PTEN homozygous deletion, 4) PTEN heterozygous deletion (HE), 5) PTEN dominant negative mutations, and 6) bi-allelically inactivated (BI) PTEN mutations defined as mutation plus loss of heterozygosity (LOH). We evaluated the limit of detection (LoD) and the precision of the biomarker for two novel genomic features: HE and BI, with the other four features previously validated. Results Analytical validation of novel biomarker features: The LoD of detecting PTEN HE and BI was determined to be 30%, the lowest tumor content at which the features can be detected at 90% probability. In the precision study, 100% (81/81) agreement was achieved across different replicates within the same sequencer run and across different sequencer runs for biomarker positive samples, demonstrating high reproducibility in calling PTEN HE and BI. Conclusions We developed and analytically validated an NGS-based assay that identifies complex and novel genomic alterations (heterozygous deletion and bi-allelic inactivation) in PTEN that is part of a composite PI3K/Akt-pathway activation biomarker. This assay identified patients that appeared to derive greater benefit in the Phase II LOTUS study as compared to using PTEN IHC to only identify patients with PTEN protein loss (Kim et al., 2017). This assay could be generalized to identify other biomarkers with similar types of genetic alterations. It also demonstrates that NGS-based CGP can broaden the intent to treat population to be more specifically related to the mechanism of action of a drug, while also being more selective to patients with potential to respond. Citation Format: Yuting He, Matthew Wongchenko, Joel Skoletsky, Christine Burns, Yali Li, Paula Maness, Doris Kim, Doron Lipson, Philip Stephens, Vincent Miller, Jeffrey Ross, Steven Gendreau, James Sun. A novel PI3K/Akt-pathway activation biomarker using comprehensive genomic profiling (CGP) for clinical trial assay [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 2582.

Introduction: In patients with advanced-stage ovarian cancer, FDA has recently granted approval of treatment with PARP inhibitors (PARPi) in patients harboring deleterious BRCA mutations (~25% of population). However, there is clear evidence of “BRCA-like” patients who respond to PARPi without BRCA mutations. To identify such patients, it has been shown that deficiency in homologous recombination repair leads to a common phenotype of genome-wide loss of heterozygosity (LOH). Depending on the cutoff employed, genomic LOH could identify more than twice the number of ovarian cancer patients who could benefit from PARPi than measuring BRCA alone. We present here an NGS-based platform developed and performed in compliance with FDA 21 CFR part 820. The assay provides a tumor measurement of BRCA1/2 (FDA-approved as FoundationFocus CDx BRCA) as well as genomic LOH, and is on the same platform as the comprehensive FoundationOne CDx, which interrogates 324 genes. Methods: DNA extracted from FFPE tumor tissue underwent whole-genome shotgun library construction and hybridization-based capture, followed by sequencing using Illumina HiSeq 4000. Sequence data were processed using a proprietary analysis pipeline designed to detect base substitutions, indels, copy number alterations (CNA), genomic rearrangements, microsatellite instability (MSI), and tumor mutational burden (TMB). A genome-wide LOH profile based on SNPs is measured as part of the CNA pipeline, and is summarized as the percentage of the tumor genome displaying LOH (scored from 0-100%), with ≥16% being considered LOH high based on clinical data derived from ARIEL2 Part 1, a phase II study of the PARPi rucaparib for the treatment of platinum-sensitive ovarian cancer (ARIEL2; NCT01891344) Results: For analytical validity, BRCA limit of detection (LoD) was at allele frequency 5.9% for substitutions and non-repetitive indels, and 30% tumor content for LOH. Overall percent agreement with comparator NGS assay was 97.3% for BRCA. No orthogonal platform concordance was established for LOH as no validated test exists. Within-assay reproducibility was measured with overall concordance of 100% for BRCA, and 98% for LOH. Conclusion: We developed a novel diagnostic assay (in compliance with FDA 21 CFR part 820) that can measure BRCA and genomic LOH simultaneously, and established robust analytical validation data. Citation Format: James X. Sun, Kevin Lin, Yali Li, Kyle Gowen, Yuting He, Coren Milbury, Christine Burns, Jun Luo, Steve Roels, Murtaza Mehdi, John Truesdell, Pei Ma, Lakshman Ramamurthy, Christine Vietz, Jeri Beltman, Thomas Harding, Doron Lipson, Jeffrey Ross, Vincent Miller, Philip Stephens, Michael Doherty, Julia Elvin. A validated diagnostic assay for identifying ovarian cancer patients with deleterious BRCA mutations and high genomic loss of heterozygosity (LOH) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 4544.

Huntingtin interacting protein 1 (HIP1) is a 116-kDa endocytic protein, which is necessary for the maintenance of several tissues in vivo as its deficiency leads to degenerative adult phenotypes. HIP1 deficiency also inhibits prostate tumor progression in mice. To better understand how deficiency of HIP1 leads to such phenotypes, we analyzed tumorigenic potential in mice homozygous for a Hip1 mutant allele, designated Hip1Δ3-5, which is predicted to result in a frame-shifted, nonsense mutation in the NH2 terminus of HIP1. In contrast to our previous studies using the Hip1 null allele, an inhibition of tumorigenesis was not observed as a result of the homozygosity of the nonsense Δ3-5 allele. To further examine the contrasting results from the prior Hip1 mutant mice, we cultured tumor cells from homozygous Δ3-5 allele–bearing mice and discovered the presence of a 110-kDa form of HIP1 in tumor cells. Upon sequencing of Hip1 DNA and message from these tumors, we determined that this 110-kDa form of HIP1 is the product of splicing of a cryptic U12-type AT-AC intron. This event results in the insertion of an AG dinucleotide between exons 2 and 6 and restoration of the reading frame. Remarkably, this mutant protein retains its capacity to bind lipids, clathrin, AP2, and epidermal growth factor receptor providing a possible explanation for why tumorigenesis was not altered after this knockout mutation. Our data show how knowledge of the transcript that is produced by a knockout allele can lead to discovery of novel types of molecular compensation at the level of splicing. [Cancer Res 2008;68(4):1064–73]

We have analyzed mammary tumors arising in transgenic mice expressing a novel, multifunctional RNA-binding protein. The protein, which we call the c-myc mRNA coding region instability determinant binding protein (CRD-BP), binds to c-myc, insulin-like growth factor II, and β-actin mRNAs, and to H19 RNA. Depending on the RNA substrate, the CRD-BP affects RNA localization, translation, or stability. CRD-BP levels are high during fetal development but low or undetectable in normal adult tissues. The CRD-BP is linked to tumorigenesis, because its expression is reactivated in some adult human breast, colon, and lung tumors. These data suggest the CRD-BP is a proto-oncogene. To test this idea, the CRD-BP was expressed from the whey acidic protein (WAP) promoter in mammary epithelial cells of adult transgenic mice. The incidence of mammary tumors was 95% and 60% in two lines of WAP-CRD-BP mice with high and low relative CRD-BP expression, respectively. Some of the tumors metastasized. Nontransgenic mice did not develop mammary tumors. H19 RNA and insulin-like growth factor II mRNA were up-regulated significantly in non-neoplastic WAP-CRD-BP mammary tissue. WAP-CRD-BP mice are a novel model for mammary neoplasia and might provide insights into human breast cancer biology.

Background: AR mutations have been described as a mechanism of resistance to AR antagonists in prostate cancer. There are limited data regarding the presence of AR mutations in breast cancer. We aim to describe the presence of tumor AR mutations in a cohort of patients (pts) with breast cancer who are candidates for targeted cancer therapy, who underwent tumor genomic profiling as part of a clinical trial at MSK (NCT01775072). Methods: MSK-IMPACT is a targeted tumor sequencing assay capable of detecting mutations and other critical genetic aberrations in 410 cancer genes; these data are available through an institutional database. Following IRB approval and using an electronic medical record, we examined the subset of pts in this database with breast cancer for the presence of AR mutations and performed a chart review for clinicopathologic features and outcomes. Statistics are descriptive. Results: As of 03JUN2015, 628 of 4,379 samples that underwent MSK-IMPACT testing since 2012 were from invasive breast cancers. 6 of 628 (1%) harbored AR mutations (Table 1), none of which contribute to a reported or predicted functional alteration. Table 1MutationTypeAllele FrequencyFunctional Impact According to MutationAssessorR31HMissense0.21LowR856CMissense0.31MediumA430TMissense0.12NeutralG409RMissense0.18MediumQ445PMissense0.47MediumQ73LMissense0.25Neutral Patient/tumor characteristics are shown in Table 2. Five out of 6 patients recurred between 7.4 and 117.4 months (mo) from the date of initial diagnosis. One patient is without disease recurrence at 3.9mo from diagnosis at date of last contact, 7.9mo ago. Table 2MutationStage at DiagnosisHIstologyER%/PR%/HER2AR%Sites of MBCSite of IMPACTR31HT1N0Lobular90/20/1+-SkinSkinR856CT3N3Ductal20/0/1+0Chest wall, LNLNA430TT2N1Ductal95/0/--Bone, liverBreastG409RT2N0Ductal90/80/FISH 1.3-Bone, liverLiverQ445PT3N3Ductal99/50/1+-LNBreast/LNQ73LT3N0Malignant Phyllodes--N/ABreastMBC metastatic breast cancer; LN lymph node; N/A not applicable Conclusions: AR mutations were uncommon in this dataset of 628 breast cancers. Similarly, TCGA sequencing data has revealed only 12 invasive breast cancers with AR mutations (2.2% of 973 samples). The functional significance of these mutations has not been demonstrated. As tissue from therapeutic studies using AR antagonists for the treatment of breast cancer become available, mutations or amplification in AR may be more readily identified as a potential mechanism of resistance to AR-targeted therapy. Acknowledgments: We gratefully acknowledge the members of the Molecular Diagnostics Service in the Department of Pathology and the Marie-Josée and Henry R. Kravis Center for Molecular Oncology. Citation Format: Gucalp A, Proverbs-Singh TA, Razavi P, Chandarlapaty S, Patil S, Ross DS, Zehir A, Baselga J, Hudis CA, Traina TA. Androgen receptor (AR) mutations in a cohort of patients with breast cancer (BC) who have undergone tumor genomic profiling. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P6-07-08.

Precision medicine has evolved as an individualized approach for treating cancer patients and has become standard in an ever-increasing number of clinical settings. It is predicated upon matching targeted-/immuno-therapy to genomic alterations detected in patients' tumors. However, widespread feasibility/adoption has been limited by: 1) high rates of insufficient tumor DNA (reaching 25%); 2) panels limited to few genes that are unable to detect multiple classes of genomic alterations; 3) testing patients late in the disease course; and 4) low molecular matching rates, which may be in part due to limited access to trials and the unpredictable nature of genomic alterations detected in each individual. We evaluated the feasibility of investigating molecular profile-related evidence for determining individualized cancer therapy (I-PREDICT) in patients with lethal tumors (NCT02534675). This navigation trial was performed under the auspices of 2 precision medicine programs (UCSD and Avera Cancer Institute) and an IRB-approved protocol. Treatment-naïve and previously treated patients with ECOG Citation Format: Jason K. Sicklick, Brian Leyland-Jones, Shumei Kato, Casey Williams, Pradip De, Gregory Heestand, Steven Plaxe, Benjamin Solomon, Vincent Miller, Adam Benson, Jennifer Webster, Jeffrey Ross, Michael Scur, Robert Porter, Shelby Jepperson, Paul Fanta, Razelle Kurzrock. High rates of personalized molecular matching are achievable in a precision oncology navigation trial: the I-PREDICT study [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr CT053. doi:10.1158/1538-7445.AM2017-CT053

Background: Receptor tyrosine kinase alterations have played a significant role in therapeutic decisions for cancer due to their oncogenic nature and response to targeted small molecule kinase inhibitors. Increased genomic profiling of tumors using hybrid-capture based next-generation sequencing approaches now reveal the presence of previously unknown fusions and alterations involving kinases in a diverse set of cancers. Here we report the presence and therapeutic significance of recurrent and novel fusions involving RET, a known oncogenic tyrosine kinase receptor, in breast cancer. Methods: Comprehensive genomic profiling on formalin-fixed, paraffin embedded patient tumor tissues was performed using FoundationOne platform that covers the entire coding region for 315 cancer-related genes and introns of 28 genes involved in rearrangements at a depth of 500-1000X (Foundation Medicine, MA). Out of 23 rearrangements, two representative RET fusion expression vectors were synthesized and expressed in non-tumorigenic cell lines (breast MCF10A and mouse 3T3 fibroblasts) and were evaluated for RET kinase signaling, drug response, and tumorigenicity. Results: RET gene fusions, the canonical NCOA4-RET and a novel, noncanonical RASGEF1A-RET fusion, were identified in two separate breast cancers and both include exons required to retain the intact kinase domain of Ret. The novel RASGEF1A-RET fusion includes the non-coding region of RASGEF1A potentially resulting in a truncated RET protein using an alternate internal start site in exon 11 of RET. In vitro characterization of both fusions expressed in mouse 3T3 and human MCF10a cell lines revealed constitutive kinase activation and subsequent downstream signaling as evidenced by phosphorylation of Ret, Erk and Akt. This is the first reported noncanonical RET rearrangement resulting in a 5’ truncated but functional RET kinase. Non-tumorigenic cell lines with stable expression of either rearrangement showed transformed phenotypes assessed by changes in morphology, enhanced growth rate, colony forming ability, and tumor formation in mice. RET fusion-transformed cells were exquisitely sensitive to treatment with RET inhibitors when evaluated in both short-term and long-term functional assays. NCOA4-RET was found by CGP in an index case of metastatic ER+/HER2+ breast cancer that had radiographic evidence of disease progression while on trastuzumab, pertuzumab, and anastrazole. Subsequent treatment with cabozantinib plus anastrazole led to a rapid clinical and radiographic response. Conclusion: CGP techniques involving hybrid-capture based approaches can identify previously unreported but recurrent RET gene fusions in breast cancer. Here, we show that RET fusions including both canonical and non-canonical complex rearrangements are functional and may represent promising therapeutic targets in selected breast cancer patients. Citation Format: Bhavna S. Paratala, Jeffrey S. Ross, Casey B. Williams, Whitney Petrosky, Kirstin A. Williams, Jon Chung, Sonia C. Dolfi, Shridar Ganesan, Siraj Ali, Brian Leyland-Jones, Kim M. Hirshfield. RET rearrangements as promising therapeutic targets in breast cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3018. doi:10.1158/1538-7445.AM2017-3018

Background: The success of anti PD–1 therapy in Hodgkin lymphoma has led to clinical trials in other types of lymphomas. However, there is limited data on PD–L1 expression in acute leukemias. In this study, we analyzed PD–L1 protein expression and tumor mutational burden in various hematologic malignancies including leukemias, lymphomas and myelomas. Methods: In the IHC cohort, formalin fixed paraffin embedded sections from 92 hematologic malignancies including acute myeloid and lymphoblastic leukemia (AML, ALL), Hodgkin lymphoma (HL), diffuse large B-cell lymphoma (DLBCL), low-grade non-Hodgkin lymphoma (LGNHL) and myeloma were immunostained using the FDA approved PD-L1 IHC 22C3 pharmDx assay [monoclonal mouse anti-PD–L1 antibody (DAKO clone 22C3)]. Any perceptible partial or complete membrane staining in all viable tumor cells was scored as follows: 50%-high level PD-L1 expression. In the second cohort of 2064 cases, comprehensive genomic profiling (CGP) was performed using a hybrid–capture, adaptor ligation assay to a mean depth of >672X. Tumor mutation burden (TMB) was calculated from a minimum of 1.11MB of sequenced DNA as previously described and reported as mutations/MB. High PD–L1 expression was noted in 100% of HL. DLBCL showed significantly higher PD–L1 positivity compared to LGNHL (p=0.01). No correlation for PD–L1 expression was noted between AML and ALL (p = 0.2). On CGP analysis of the second cohort, TMB was significantly higher in DLBCL as compared to LGNHL (p Conclusion: The high PD-L1 expression in both HL and DLBCL and high TMB in DLBCL may be linked to enhanced responsiveness to check point inhibitor therapy in these cancers. PD-L1 expression and TMB are relatively infrequent and at low levels in acute leukemias. Additionally, novel anti PD–1/PD–L1 therapeutic strategies are worthy of consideration for both advanced stage refractory/relapsed acute myeloid and lymphoblastic leukemias. Results: PD-L1 expression and TMB are tabulated below.PD-L1 IHCTMB (mutations/MB)Tumor typenLow PositiveHigh PositiveTotal Positive (%)n>=10.0>=20.0AML21505/21 (24%)50%0%ALL3212012/32 (38%)10%0%HL1001010/10 (100%)368%8%DLBCL10448/10 (80%)25344%15%LGNHL10202/10 (20%)930%0%Myeloma9202/9 (22%)NANANA Citation Format: Lina Abdul Karim, Peng Wang, Jose de Guzman, Brandi Higgins, Joeffrey Chahine, Christine Sheehan, Bhaskar Kallakury, Jeffrey S. Ross. PDL1 protein expression and tumor mutation burden in hematologic malignancies: correlation with Hodgkin and high grade lymphoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3724. doi:10.1158/1538-7445.AM2017-3724

Background Patients across a range of disease types have demonstrated robust and durable responses using checkpoint inhibitor therapies (CPITs). Given the limitations of immuno-histochemical based testing, identifying a unified, quantitative metric to determine potential response to CPITs remains an urgent need. Tumor mutational burden (TMB) measures the number of somatic protein coding mutations per target sequence in a tumor specimen. This measure has been associated with response and survival for multiple CPITs across an array of indications. In this study we describe Foundation Medicine’s (FMI) work to develop and validate a TMB result as part of our comprehensive genomic profiling assays and summarize clinical feasibility in NSCLC, melanoma and bladder cancer. Methods We developed an analysis method to determine TMB based on data from our comprehensive genomic profiling assays. TMB is calculated by counting all synonymous and non-synonymous somatic variants across 315 or 405 genes, excluding germline alterations and known or likely driver alterations. The mutation count is normalized by the coding target territory to achieve a mutation density of mutations per megabase (mut/Mb). To determine accuracy, we compared TMB values from our comprehensive genomic profiling assay against a CLIA-validated whole-exome sequencing (WES) method on 29 patients. Precision was assessed over 10 clinical samples replicated 4-6 times. Lower limit of sample tumor purity was determined through dilutions of tumor/normal pairs from 80% to 5% tumor. Clinical feasibility was assessed by analyzing TMB versus immunotherapy-based survival in a cohort of 65 metastatic melanoma patients, 150 urothelial carcinoma patients and 463 NSCLC patients. Additionally, we examined the relationship between TMB and microsatellite instability status (MSI), an independent biomarker associated with response to CPITs. Results Foundation Medicine’s TMB measure provides accurate and precise results across a range of tumor mutational burden values on samples with as little as 20% tumor purity. Using cohort specific thresholds, TMB was significantly associated with improved survival to CPITs in NSCLC, melanoma and bladder cancer. Using data from over 40,000 patient samples, we also show significant overlap between high TMB and high MSI samples and show that MSI-High specimens represent a subset of TMB-High specimens. Conclusions We have developed and validated the tumor mutational burden (TMB) biomarker as part of our comprehensive cancer genomic profiling assays. Initial clinical feasibility results demonstrate that TMB can be used to predict the likely response to anti-PD-1/PD-L1 CPITs across a growing number of indications including NSCLC, melanoma and bladder cancer. Citation Format: Daniel S. Lieber, Mark R. Kennedy, Douglas B. Johnson, Jonathan E. Rosenberg, Marcin Kowanetz, Joel R. Greenbowe, Garrett M. Frampton, Caitlin F. Connelly, Alexa B. Schrock, Jeffrey S. Ross, Philip J. Stephens, Siraj M. Ali, Vincent A. Miller, David A. Fabrizio. Validation and clinical feasibility of a comprehensive genomic profiling assay to identify likely immunotherapy responders through tumor mutational burden (TMB) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 2987. doi:10.1158/1538-7445.AM2017-2987

Background: Genomic alterations in the ALK tyrosine kinase, such as copy number gains, point mutations, and ALK gene fusions, have been described in a broad spectrum of malignancies. Recently a novel mechanism of ALK activation in melanoma was discovered in which ALK transcription was initiated from a de novo alternative transcription initiation (ATI) site in ALK intron 19 (ALK-ATI) [1]. In this study, we used whole transcriptome sequencing (WTS) to interrogate a cohort of clinical melanoma specimens in order to identify those that express ALK-ATI. Method: RNA-seq libraries of 33 BRAF-negative clinical metastatic melanoma specimens were prepared using the KAPA Stranded RNA-Seq Kit with RiboErase with 200ng RNA as input and sequenced on Illumina HiSeq 4000 at 2x75bp. The specimens were sequenced to an average of 69.4m total read pairs with 11.3m on-target distinct read pairs. Sequencing reads were mapped to the human genome and transcriptome, gene- and exon-level expressions were quantified. An algorithm to identify the ALK-ATI was developed based on assessing 1) transcription of intron 19 putative ATI region, 2) ALK relative RNA expression and 3) ALK 5’ to 3’ differential expression. In addition, immunohistochemistry (IHC) staining for ALK (using D5F3 antibody) was done on 10 of the 33 samples with adequate tumor material. Results: High coverage (median 30x) of intron 19 ATI region and high ALK relative expression (log ratio > 2) were observed in 4 of the 33 specimens tested, in which the ATI site is consistent with what had been reported in [1]. Among the 4 potential ALK-ATI positive cases, 3 have shown significant 3’ to 5’ overexpression of ALK. IHC staining were performed on the 4 putative ALK-ATI positive cases and 6 putative negative cases, and has confirmed ALK-ATI calls detected by WTS assay. Two of the ALK IHC positive cases harbor the highest ALK expression, and the other two ALK-ATI likely cases are ALK IHC weak positive. ALK-ATI occurred in cutaneous, mucosal, and unknown primary melanoma, and co-occurred with other driver mutations (NRAS, NF1, KIT). None of the above specimens had ALK rearrangements detected from DNAseq (by FoundationOne) or WTS. Conclusions: Our findings confirmed previously reported ALK activation through ALK-ATI in an independent melanoma cohort, and provided direct evidence that such events can be detected using whole transcriptome sequencing and align with IHC results in clinical FFPE tumor specimens. This confirmation together with future functional validation and clinical evidence may provide new therapeutic opportunities for BRAF-negative metastatic melanoma patients. 1. Wiesner, T. et al. Alternative transcription initiation leads to expression of a novel ALK isoform in cancer. Nature 526, 453-457 (2015). Citation Format: Shan Zhong, Christine Lovly, Christine Malboeuf, Kristina Brennan, Douglas Johnson, Kelsie Riemenschneider, Catherine Meador, Emily White, Geoff A. Otto, Doron Lipson, Philip Stephens, Vincent Miller, Jeffrey Ross, Jie He. Identification of ALK alternative transcription initiation in BRAF-negative metastatic melanoma patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 2367. doi:10.1158/1538-7445.AM2017-2367

Introduction: In a cohort of non-small cell lung cancers (NSCLC), we identify new potential biomarkers that may predict response to checkpoint inhibitors through an analysis of genomics, PD-L1 protein expression, and tumor mutational burden (TMB). Motivation: In cancer therapy, anti-PD-1/PD-L1 checkpoint inhibitors provide long term responses in a subset of patients. The current biomarker, PD-L1 IHC, is only partially successful in identifying responders. TMB, a measure of the number of mutations in a tumor’s genome, is an emerging biomarker that can predict response rates to checkpoint inhibition in addition to PD-L1 IHC. Higher TMB is thought to confer a higher probability of expressing a neoantigen protein recognizable by the immune system as non-self and therefore responding to checkpoint inhibition. While TMB represents the likelihood that tumor cells express immune-triggering neoantigens, PD-L1 expression represents one of many ways a tumor can evade immune response. We investigated the patterns of PD-L1 IHC staining and TMB in NSCLC. Methods: We tabulated PD-L1 IHC stain statuses, TMB, and genomic alterations for 789 unique NSCLC tumor samples. Tumor cell PD-L1 expression was measured by IHC and scored for distribution (0, 1-24%, > 24% of cells) and intensity (0, 1+, 2+) as Negative, Low Positive, or High Positive. Samples were tested with a comprehensive genomic profiling assay that identifies alterations in exons and select introns in 405 genes. TMB was calculated over 1.1 Mb by counting somatic, non-driver genomic alterations and scored as Low (< 6 mut/Mb), Medium (< 20, ≥ 6 mut/Mb), or High (≥ 20 mut/Mb). Results: PD-L1 tumor staining was significantly correlated with TMB (P = 3.52e-4), although the correlation was not strong (Spearman’s rho = 0.12). To better explain discordant TMB/PD-L1, we ran Fisher’s exact tests to identify genes enriched for alterations in TMB-High/PDL1-Negative or TMB-Low/PD-L1-High Positive samples. The TMB-High/PD-L1-Negative subset preferentially harbored STK11 loss-of-function alterations with an odds ratio of 3.65 (P = 1.20e-3). The TMB-Low/PD-L1-High Positive subset exhibited enrichment for BRAF alterations with an odds ratio of 5.46 (P = 1.71e-3). Conclusions: TMB-High tumors have high immunogenicity and often rely on PD-L1 expression as a means of immune evasion. Alterations in BRAF and other genes enriched in TMB-Low/PDL1-High Positive samples may trigger immune responses moderated by PD-L1 expression. Samples with these alterations, although often TMB-Low, may nevertheless respond to checkpoint inhibitors. Alterations in STK11 and other genes enriched in TMB-High/PD-L1-Negative samples may be implicated as non-PD-L1-based mechanisms for immune evasion. Although further ongoing analysis is required, our study suggests that the combination of TMB, PD-L1 staining, and genomics may provide the best prediction of response to checkpoint inhibitors. Citation Format: Michael E. Goldberg, Laurie M. Gay, David Fabrizio, Garrett M. Frampton, Julia A. Elvin, Jeffrey S. Ross, Vincent A. Miller, Philip J. Stephens, Lee A. Albacker. The interaction of PD-L1, TMB, and genomic alterations in NSCLC [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 1775. doi:10.1158/1538-7445.AM2017-1775

Background: Effective therapies for the management of castrate resistant prostate cancer are lacking. We performed comprehensive genomic profiling (CGP) on advanced prostate carcinomas (PC) in the course of clinical care to identify genomic alterations that could suggest benefit from targeted, immune- and PARP inhibitor therapeutic strategies. Methods: DNA was extracted from 40 microns of FFPE specimen from 721 clinically advanced PC cases. CGP was performed using a hybrid-capture, adaptor ligation based next generation sequencing assay to a mean coverage depth of over 500X. All four classes of genomic alterations (GA) - base substitutions, insertions and deletions, gene fusions, and copy number alterations (amplifications and losses) - were identified. Results: Of the 721 PC patients - men with median age of 65 (range 34 - 88) - CGP was performed on 335 (46.5%) prostate specimens and 386 (53.5%) specimens from metastatic sites. The most common genes altered were TP53 (N=336, 46.6%), PTEN (N=254, 35.2%), TMPRSS2-ERG (N=214, all fusions, 29.7%), AR (N=173, 24%), and MYC (N=105, 14.6%). Median tumor mutational burden (TMB) for this series was 3.6 mut/Mb (range 0 - 305). AR and MYC amplifications were enriched in metastatic tumors compared to primary tumors (p < 0.001 for both). Alterations characteristic of homologous recombination deficiency (HRD) were found in >15% of cases, including homozygous deletions and truncating mutations in BRCA2 in 79 (11%) samples, as well as truncating mutations in ATM in 41 (5.7%) samples, with only 3 samples possessing alterations in both genes. Cases with these HRD-relevant alterations possessed overall higher genomic loss of heterozygosity (LOH) content than did those without (p = 0.02). HRD and non-HRD cases had median TMB of 4.8 versus 2.7 mut/Mb, respectively, and mean TMB of 10.4 versus 4.0 mut/Mb (p < 0.01). Only 2.6% of non-HRD cases were TMB-high (20 mut/Mb or more), versus 8.4% of HRD cases, a small but significant enrichment in the latter. Conclusions: CGP for advanced PC cases identifies largely non-overlapping TMB-high and HRD positive cases, suggesting benefit from immunotherapeutics and PARP inhibitors respectively. Further investigation will assess whether HRD pathway alterations result in genomic LOH in the context of a hypermutated tumor, and enrichment or exclusivity with other GA in prostate carcinoma. Citation Format: Ninad Dewal, Yuting He, Richard J. Lee, Alexa B. Schrock, Jon Chung, Christopher Hoimes, Zachary R. Chalmers, Garrett M. Frampton, James X. Sun, Primo N. Lara, Neeraj Agrawal, Paul Matthew, Philip J. Stephens, Vincent A. Miller, Jeffrey S. Ross, Siraj M. Ali. CGP identifies largely non-overlapping high tumor mutational burden and HRD genomic alterations in 721 clinically advanced prostate acinar adenocarcinoma cases [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3394. doi:10.1158/1538-7445.AM2017-3394

Background: The TSC1/TSC2 complex inhibits the mTOR pathway. Loss of function of this complex leads to hyperperactivation of the mTOR pathway which promotes growth and survival. Few studies have examined the role of TSC1 and TSC2 mutations in breast cancer and as a clinically relevant target for molecularly targeted therapy, such as everolimus, a rapamycin analog. Methods: Hybridization capture of 3769 exons of 236 cancer-related genes and 47 introns from 19 genes that are frequently rearranged in cancer were fully sequenced to high, uniform coverage from a commercial CLIA-certified laboratory (Foundation Medicine). Results: 30 of 2,208 breast cancer pt samples (1.3%) harbored either TSC1 (13/30, 43%) or TSC2 (17/30, 56%) genomic alterations (GAs). The samples with TSC1/2 consisted of breast carcinoma n.o.s. (17/30) and invasive ductal carcinoma (12/30); one invasive lobular carcinoma (ILC) harbored a TSC2 alteration. Sequencing was performed on metastatic (18/30, 60%) and primary (12/30, 40%) tumors. All TSC1 alterations resulted in truncation of the protein product; TSC2 GA consisted of truncations (14/17) and point mutations (3/17). TSC1/2 GAs frequently co-occurred with GAs in TP53 (19/30; 63%) and amplifications of MYC (9/30; 30%) and FGFR1 (7/30; 23%). Of the TSC1/2-mutant cases where hormone receptor and HER2/neu status was known (23/30), 65% were ER-/PR-/HER2- (TNBC), 26% were ER+/HER2- and 9% were ER-/HER2+. Patient case: We describe a postmenopausal node positive ILC (ER+/PR+/HER2-) pt who received adjuvant chemotherapy followed by 9 yrs adjuvant anastrazole then with disease recurrence. On FoundationOne testing a TSC2 A1141T missense mutation, along with GAs in CDH1, ETV6, and CSF1R, were found in her metastatic disease in the soft tissues of her left orbit associated with substantial swelling resulting in loss of vision. While TSC2 A1141T has not been functionally characterized, in vitro and in vivo data suggest sensitivity to mTOR inhibitors. Her disease stabilized but did not decrease in size with vinorelbine + capecitabine. She was treated with letrozole + everolimus and had a near clinical complete response with restoration of vision for 5 + months. Collection of clinical data on additional pts is currently ongoing. Conclusion: 1.3% of breast cancer cases harbored GAs in either TSC1 orTSC2. The majority of TSC1/2 GAs cause inactivation of these negative regulatory proteins leading to activation of the mTOR signaling pathway. TSC1/2 GAs were enriched for TNBC (65%) versus non-TNBC (35%) (p= 0.077; Fisher’s exact test). Truncation of TSC1/2 occurred at multiple codons, underscoring the importance of examining the entire coding region of tumor suppressor genes for these inactivating events. Comprehensive genomic profiling has the potential to identify the broad spectrum of these inactivating events, a subset of which may be clinically relevant. Citation Format: Norma A Palma, Juliann Chmielecki, Garrett Frampton, Siraj Ali, Maren Levin, Jeffrey S Ross, Deborah Morosini, Gary Palmer, Vincent A Miller, Phil Stephens, Joyce O'Shaughnessy. FoundationOne profiling of TSC1 and TSC2-mutated advanced breast cancers [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P2-03-06.

Purpose Tumor mutation load is an emerging prognostic and diagnostic marker for many cancers. Sensitivity to immunotherapeutic agents, which stimulate an antitumor immune response by selectively inhibiting immunosuppressive cell surface ligands, is known to correlate with high mutation load in the tumor. Numerous somatic and germline defects can cause genomic instability, including alterations affecting the mismatch repair pathway, DNA polymerases, and cell cycle regulators. We describe here the discovery of previously unreported mutations in the promoter of the PMS2 gene that are associated with significantly increased tumor mutation load. Methods Comprehensive genomic profiling by hybridization capture of exonic regions from either 236 or 315 cancer-related genes and select introns from 19 genes commonly rearranged in cancer was used to characterize more than 60,000 clinical FFPE cancer specimens. At least 50 ng of extracted DNA was analyzed per sample and the constructed libraries were sequenced to high, uniform median coverage (>500x). Samples were assessed for base substitutions, short insertions and deletions, copy number alterations and gene fusions/rearrangements. Mutation load is assessed as the number of somatic coding point mutations per megabase of targeted territory. Results Mutation load from targeted cancer gene analysis recapitulates previous results evaluating whole exome and whole genome mutation load in tumors and cell lines. A novel mutation hotspot was identified in the promoter of PMS2, which codes for the PMS2 protein, a dimerization partner of MLH1 and integral to the DNA mismatch repair complex. Promoter mutations were found in 7.5% of melanoma specimens (n = 101/1348) and 17% of skin squamous cell carcinomas (n = 30/175). In both diseases, PMS2 promoter mutations are the most significant genomic correlate of high mutation load. In melanoma, PMS2 promoter mutant specimens have 4x the median mutation load of the general melanoma population. Skin squamous cell carcinomas show a 2.5-fold increase. Functional characterization is underway to support the hypothesis that these mutations lead to modified PMS2 transcriptional activity, which is known to cause hypermutation. Conclusions The growing corpus of cancer genome data continues to enable novel discoveries in cancer biology. Non-coding and regulatory mutations have not been the target of focused study, and our findings extend the small set of regulatory mutations thought to affect tumor development. Our discovery highlights the power of large-scale genomic analysis to uncover additional disease mechanisms, and furthers our understanding of how cells maintain genome integrity. With the documented association of mutation load and immunotherapy sensitivity, investigation into the response of PMS2 promoter mutant tumors is warranted. Citation Format: Zachary R. Chalmers, Franklin W. Huang, Laurie M. Gay, Siraj M. Ali, Roman Yelensky, Juliann Chmielecki, Jeffery S. Ross, Vincent A. Miller, Philip J. Stephens, Levi A. Garraway, Garrett M. Frampton. Broad analysis of recurrent somatic mutations in cancer reveals a common novel non-coding mutation in the promoter of PMS2 associated with greatly increased tumor mutation load. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 3576.

Most large genomic profiling efforts in breast cancer have focused on primary breast tumors. Profiling of metastatic breast cancer (MBC) could more accurately define actionable genomic alterations and reveal novel alterations that arise under the selective pressure and clonal evolution of prior therapy. Utilizing a 410-gene targeted capture-based sequencing platform (MSK-IMPACT), we analyzed 920 tumors (62% metastatic, 38% primary) from 874 MBC patients for somatic mutations, DNA copy number alterations, and structural rearrangements (planned final analysis 1000 patients). Detailed clinical data including treatment outcomes was collected for all patients. The cohort was representative of the well characterized clinical subtypes of breast cancer with 71% ER+ or PR+ and HER2-, 17% HER2+, and 12% triple negative breast cancer (TNBC). Our analysis revealed recurrent alterations in multiple pathways including PI3K/AKT/mTOR (56%), cell cycle regulation (42%), RTK signaling (40%), epigenetic regulation (33%) and MAPK/ERK (19%) pathways. The most frequent actionable alterations include: PIK3CA mutation (36%), ERBB2 amplification (15%), FGFR1 amplification (12%), ESR1 mutation (11%), PTEN mutation/deletion (10%), AKT1 mutation (6%), and ERBB2 mutation (5%). The genomic landscape was significantly different across breast cancer subtypes. For example, the PI3K/AKT/mTOR pathway was altered in ER+ MBC mainly through activating PIK3CA mutations whereas PTEN deletion/mutations were most common in TNBC. We also identified significant differences in the genomic profiles of metastatic and primary tumor samples. Gene enrichment analyses revealed a subset of genes more frequently altered in metastatic tumors (STK11: 13 vs 2; ROS1: 14 vs 2; FGFR4: 15 vs 1) suggesting that mutations in these genes may play a role in breast tumor metastasis and/or therapy resistance. ESR1 mutations were predominantly present in metastatic tumors (88%) and were significantly associated with duration of prior hormonal therapy (P In summary, our genomic analyses of this large cohort of MBC patients revealed actionable alterations in over 60% of the patients. 29% of patients with these alterations were enrolled in clinical trials of matched targeted therapies to date (PIK3CA 27%, AKT1 30%, ESR1 23%, ERBB2 39%) suggesting that prospective genomic characterization can accelerate enrollment of patients with MBC onto therapeutic clinical trials. Citation Format: Pedram Razavi, Matthew T. Chang, Sumit Middha, Dara S. Ross, Ahmet Zehir, Tracy A. Proverbs-Singh, Cyriac Kandoth, Sarat Chandarlapaty, Maura N. Dickler, Jorge S. Reis-Filho, Sujata Patil, Venkatraman Seshan, Lillian Smyth, Neil M. Iyengar, Komal Jhaveri, Shanu Modi, Chau T. Dang, Mark E. Robson, Larry Norton, Clifford A. Hudis, Marc Ladanyi, Maurizio Scaltriti, Nikolaus Schultz, David Hyman, Michael F. Berger, Barry S. Taylor, David B. Solit, José Baselga. Clinical genomic profiling of 1000 metastatic breast cancer patients: actionable targets, novel alterations, and clinical correlations. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 4509.

Pediatric cancers are rare and diverse entities characterized by low mutational burden. Multiple studies investigating the genomic events driving oncogenesis have identified characteristic diagnostic lesions as well as targets suitable for therapeutic exploitation in common tumor subtypes. Here, we present genomic profiles from 1239 pediatric tumors (ages 0-18) from 51 disease subtypes representing sarcomas (26.6%), other blastomas (22.4%), brain tumors (20.4%), hematological malignancies (19.5%), carcinomas (9.8%), and gonadal tumors (1.4%). This collection represents real-world tumor specimens in which select cancer-associated genes were analyzed for all classes of alterations as part of routine clinical care. Samples were sequenced to a median unique depth of >500x to allow for optimal detection of oncogenic events in low purity specimens. Additionally, 635 samples (51%) were analyzed by targeted RNA-seq for fusions in 265 genes. Alterations with proven clinical actionability in pediatric cancers (BRAF V600E and ALK, NTRK1, and ABL1 fusions) were found in 3.9% of samples from brain, sarcoma, and heme cases. Three novel ALK fusions were identified in a neuroblastoma (BEND5-ALK), a soft tissue sarcoma (IGFBP5-ALK), and an astrocytoma (PPP1CB-ALK). Two novel BRAF fusions were also found in an astrocytoma (BCAS1-BRAF) and a ganglioglioma (TMEM106B-BRAF). In rhabdomyosarcoma, we identified a novel PAX3-GLI2 fusion and confirmed a second occurrence of the rare PAX3-NCOA1 fusion. This large data set also challenged the paradigm of “disease-specific” alterations as previously characterized fusions involving ALK, NTRK1, and PAX3 were observed in novel diseases from which they were originally reported. We investigated variants of unknown significance for potentially novel cancer mutations, and identified four recurrent and likely somatic variants in KMT2C, PRSS1, and CTNNB1 predicted to have functional impact by in silico analyses. Compared to other large published data sets in neuroblastoma, ALL, AML, rhabdomyosarcoma, and medulloblastoma, similar frequencies of clinically significant alterations were observed, thereby validating this dataset as biologically relevant. While many common and diagnostic alterations have been described in pediatric cancers, researchers are becoming increasingly aware of other rarer, yet equally important, clinically relevant genomic alterations. As demonstrated by the identification of novel fusions and mutations, this data set is ripe with discovery potential, and can be used to generate hypotheses, validate rare findings, and investigate the genomic landscape of rare tumors for which only small studies exist. Data from these 1239 tumors will be publicly available to the research community for exploration and incorporation into future experiments that will ultimately improve the treatment and prognosis for children with cancer. Citation Format: Juliann Chmielecki, Mark Bailey, Jie He, Julia Elvin, Jo-Anne Vergilio, Shakti Ramkissoon, James Suh, Garrett M. Frampton, Siraj Ali, Jeffrey Ross, Vincent A. Miller, Philip J. Stephens, Doron Lipson. Genomic profiling of 1239 diverse pediatric cancers identifies novel discoveries across tumors. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr LB-178.

Introduction: At least 70,000 patients will be diagnosed with bladder urothelial carcinoma (UC) in the United States this year. Current pathologic staging divides UC into muscle invasive (MIBC) and non-muscle invasive bladder cancer (NMIBC). Existing molecular classifications reveal a heterogeneous mutational landscape especially in the MIBC group, with studies finding that UC breaks down to anywhere from two to six classes defined by molecular features. These studies have utilized either select somatic alterations or high-dimensional mRNA expression data. Classifications of UC based on comprehensive genomic profiling (CGP) and subsequently detected somatic tumor alterations are conspicuously absent. Methods: To create this CGP-based classification, 638 cases of mostly refractory or metastatic UC patients were assayed with CGP in the course of clinical care sequencing the coding exons of 287 cancer-related and 47 introns of 19 genes frequently rearranged in cancer and analyzing base-pair substitutions, indels, copy number alterations, and select gene fusions. Hierarchical latent class analysis (HLCA) a mixed-model classification method that finds highly statistically significant combinations of gene alterations, was applied to the resultant alteration profiles. Results: HLCA detected the structure of somatic alteration combinations in UC, finding two main super classes driven by either CDKN2A/B co-deletion or TP53 and RB1 mutations, and subclasses further defined by altered genes in signal transduction (FGFR3, ERBB2, PIK3CA) and chromatin modifying (KDM6A, MLL2, ARID1A) pathways. These classes are predictive of clinical characteristics and tumor progression, and outperform existing molecular-based UC classifiers. Conclusions: Identifying robust molecular subtypes of UC allows for appropriate assignment of patients to relevant clinical trials and can aid in electing suitable therapeutic strategies. Investigating the clinical context and targeted and immunotherapy implications of these subclasses will be essential for diagnostic, prognostic and therapeutic purposes. Citation Format: Alyna Khan, Alex Fichtenholtz, Jeffrey Ross, Garrett Frampton, Eric Neumann, Siraj M. Ali, Phil J. Stephens. Subtypes of urothelial carcinoma of the urinary bladder defined by genomic alteration combinations predict clinical characteristics. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr LB-328.

Purpose: NTRK1-3 fusions are oncogenic drivers in lung and other carcinomas in adults, suggest benefit from targeted therapy. We examined a large series of pediatric/adolescent advanced cancer cases to identify those with NTRK fusions. Background: The NTRK, neurotrophic tyrosine kinase receptor, genes encode proteins essential for normal neuronal growth and development. Fusions of NTRK family members have recently been linked to oncogenesis as well as identified as potential targets of therapy. In this study, a hybrid capture based comprehensive genomic profiling (CGP) assay was used to study a large series of pediatric solid tumors to search for NTRK gene fusions. Methods: CGP using hybridization capture of at least 3,320 exons from 182 cancer-related genes and 37 introns of 14 genes commonly rearranged in cancer (previous version of the test) was applied to ≥ 50ng of DNA extracted from 1351 pediatric/adolescent/young adult tumors ( Results:, From 1351 pediatric/adolescent young adult advanced cancer patients, 7 (0.52%) harbored NTRK family member fusions. Of these, 5 were pediatric ( Conclusions: 0.5% pediatric/adolescent/young advanced cancers harbor NTRK1 and NTRK3 fusions, and all such tumors are mesenchymal in origin. Of these, an index case benefitted from crizotinib treatment, suggesting pathways to clinical benefit for such cases. Citation Format: Dean Pavlick, Siraj Mahamed Ali, Julia A. Elvin, Phil J. Stephens, Vincent A. Miller, Jeffrey S. Ross, James H. Suh, Jo-Anne Vergilio, Juliann Chmielecki, Tim Brennan, John R. Crawford, Denise M. Malicki, Hyunah Ahn, Victor N. Wong. Identification of NTRK fusions in pediatric tumors via comprehensive genomic profiling. [abstract]. In: Proceedings of the AACR Special Conference on Advances in Pediatric Cancer Research: From Mechanisms and Models to Treatment and Survivorship; 2015 Nov 9-12; Fort Lauderdale, FL. Philadelphia (PA): AACR; Cancer Res 2016;76(5 Suppl):Abstract nr A46.

Early detection and diagnosis of prostate cancer is key to designing effective treatment strategies. Microarrays have resulted in the discovery of hepsin (HPN) as a biomarker for detection of prostate cancer. In this study, we explore the development of HPN imaging probes for detection of prostate cancer. We used phage display to isolate HPN binding peptides with 190 + 2.2 nmol/L affinity in monomeric form and high specificity. The identified peptides were able to detect human prostate cancer on tissue microarrays and in cell-based assays. HPN-targeted imaging agents were synthesized by conjugating multiple peptides to fluorescent nanoparticles to further improve avidity through multivalency and to improve pharmacokinetics. When injected into mouse xenograft models, HPN-targeted nanoparticles bound specifically to HPN-expressing LNCaP xenografts compared with non–HPN-expressing PC3 xenografts. HPN imaging may provide a new method for detection of prostate cancer. [Cancer Res 2008;68(7):2286–91]

Huntingtin interacting protein 1 (HIP1) is an inositol lipid, clathrin, and actin binding protein that is overexpressed in a variety of epithelial malignancies. Here, we report for the first time that HIP1 is elevated in non–Hodgkin's and Hodgkin's lymphomas and that patients with lymphoid malignancies frequently had anti-HIP1 antibodies in their serum. Moreover, p53-deficient mice with B-cell lymphomas were 13 times more likely to have anti-HIP1 antibodies in their serum than control mice. Furthermore, transgenic overexpression of HIP1 was associated with the development of lymphoid neoplasms. The HIP1 protein was induced by activation of the nuclear factor-κB pathway, which is frequently activated in lymphoid malignancies. These data identify HIP1 as a new marker of lymphoid malignancies that contributes to the transformation of lymphoid cells in vivo. [Cancer Res 2007;67(18):8923–31]

Huntingtin interacting protein 1 (HIP1) is a multidomain oncoprotein whose expression correlates with increased epidermal growth factor receptor (EGFR) levels in certain tumors. For example, HIP1-transformed fibroblasts and HIP1-positive breast cancers have elevated EGFR protein levels. The combined association of HIP1 with huntingtin, the protein that is mutated in Huntington's disease, and the known overexpression of EGFR in glial brain tumors prompted us to explore HIP1 expression in a group of patients with different types of brain cancer. We report here that HIP1 is overexpressed with high frequency in brain cancers and that this overexpression correlates with EGFR and platelet-derived growth factor β receptor expression. Furthermore, serum samples from patients with brain cancer contained anti-HIP1 antibodies more frequently than age-matched brain cancer–free controls. Finally, we report that HIP1 physically associates with EGFR and that this association is independent of the lipid, clathrin, and actin interacting domains of HIP1. These findings suggest that HIP1 may up-regulate or maintain EGFR overexpression in primary brain tumors by directly interacting with the receptor. This novel HIP1-EGFR interaction may work with or independent of HIP1 modulation of EGFR degradation via clathrin-mediated membrane trafficking pathways. Further investigation of HIP1 function in brain cancer biology and validation of its use as a prognostic or predictive brain tumor marker are now warranted. [Cancer Res 2007;67(8):3609–15]

Background: CREB-binding protein (CBP), encoded by the CREBBP gene functions as an activator of a variety of transcription factors that play critical roles in normal embryonic growth and development, tumorigenesis and homeostasis by coupling chromatin remodeling to transcription factor recognition. Whether CBP functions as a tumor suppressor or an oncogene in cancer progression remains unclear. In this study, we assessed the prognostic significance of CBP protein expression in PACs. Design: Formalin-fixed paraffin-embedded tissue sections from 115 PACs were immunostained by a manual method (DAKO EnVision+ Dual Link System-HRP) using rabbit polyclonal CBP (Santa Cruz Biotech, Santa Cruz, CA). Cytoplasmic and/or nuclear immunoreactivity was scored based on intensity and percentage of positive cells in both the tumor (T) and adjacent benign (B) epithelium in each case. Cases were assessed as tumor>benign (T>B), tumor = benign (T = B), tumor Results: Cytoplasmic CBP immunoreactivity was observed as follows: T>B 41%, T = B 37%, T = Gleason 7) vs low (LG< = Gleason 6) tumor grade [T>B 52% HG vs 33% LG, T = B 19% HG vs 51% LG, TB 51% adv vs 31% early, T = B 23% adv vs 52% early, TB 53% recurr vs 28% non recurr, T = B 26% recurr vs 49% non recurr, TB 42%, T = B 7%, T = Gleason 7) vs low (LG< = Gleason 6) tumor grade [T>B 58% HG vs 30% LG, T = B 6% HG vs 7% LG, N 36% HG vs 63% LG; p = 0.008] and advanced (stage III, IV) vs early (stage I, II) tumor stage [T>B 54% adv vs 29% early, T = B 7% adv vs 7% early, N 39% adv vs 64% early; p = 0.019]. On multivariate analysis, high tumor grade independently predicted biochemical disease recurrence (p = 0.006). Conclusion: CBP immunoreactivity is a biomarker of adverse prognosis in PAC, correlating with high tumor grade and advanced stage disease. Cytoplasmic CBP expression is also associated with biochemical disease recurrence. Further study of CBP expression and its potential role in prostate cancer appears warranted. Citation Format: Jeffrey S. Ross, Olga Voronel, Siddhartha Dalvi, Gregory M. Sheehan, Christine E. Sheehan, Tipu Nazeer, Bhaskar VS Kallakury. Prognostic significance of CBP (CREB-binding protein) in prostatic adenocarcinomas (PACs): CBP is associated with high grade, advanced stage and biochemical disease recurrence. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4343. doi:10.1158/1538-7445.AM2015-4343

The genomic landscape of glial tumors is complex, and the identification of effective targeted treatments has proven challenging. To find potential therapeutic vulnerabilities, we analyzed the spectrum of genomic alterations in over 700 confirmed glial tumors. Extracted DNA was subjected to hybrid capture for 236 cancer-related genes plus 19 genes frequently rearranged in cancer, and sequenced to a median depth of >500x. Resultant sequences were assessed for base substitutions, insertions and deletions, copy number changes, and select rearrangements. Glioblastomas showed frequent alterations in CDKN2A/B (55%), TP53 (40%), PTEN (40%), and EGFR (40%), including EGFRviii (18.6%). Anaplastic astrocytomas harbored frequent alterations in TP53 (65%), IDH1 (55%), and ATRX (51%). Astrocytomas, including both low grade and mixed grade tumors, had a similar genomic landscape to that of anaplastic astrocytomas with frequent alterations in TP53 (53%), IDH1 (45%), and ATRX (36%). In addition to known fusions involving EGFR, FGFR3, and BRAF, we also identified rare, but therapeutically relevant, fusions involving ROS1 and PDGFRA that have been reported exclusively in lung cancer and hypereosinophilic syndrome, respectively. Unsupervised clustering of all glial tumors based on genomic alterations identified six distinct classes each driven by unique combinations of genomic events. Tumors driven by concurrent alterations in IDH1, ATRX, and TP53 were associated with younger age and anaplastic astrocytoma histology (p = 2.2×10−16). The addition of NF1 alterations to IDH1, ATRX, and TP53 mutations was modestly enriched in glioblastoma samples (p = 0.04). Other classes showed distinct drivers (MAPK or PI3K pathway alterations) coupled with unique mechanisms of cell cycle deregulation (CDKN2A/B loss, CDK4/MDM2 amplification, or RB1 alterations). The potential clinical importance of these genomic findings was demonstrated in three patients with surgically confirmed recurrent glioblastoma who received molecularly-matched targeted inhibitors as part of second-line therapy. Patients with activation of EGFR via amplification or mutation (R108K), activation of the mTOR pathway via inactivation of NF1, and constitutive activity of BRAF (V600E) were treated with erlotinib, everolimus, and vemurafenib, respectively, administered in combination with chemotherapy. In all three cases, imaging responses and overall survival were deemed far greater than expected by the treating physician. In conclusion, complex genomic data from glial tumors can be grouped into distinct sub-classes driven by unique combinations of alterations. Anecdotal clinical data suggest that subsets of patients may benefit from molecularly matched targeted therapies; larger studies investigating the efficacy of these agents in selected patients are warranted. Citation Format: Juliann Chmielecki, Michael E. Goldberg, Alex Fichtenholtz, Julia Elvin, Garrett M. Frampton, Siraj M. Ali, Jeffrey S. Ross, Deborah Morosini, Vincent A. Miller, David Piccioni, Santosh Kesari, Philip J. Stephens. The complex genomic landscape of glial tumors reveals distinct subclasses and potential therapeutic targets associated with clinical responses to targeted inhibitors. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4667. doi:10.1158/1538-7445.AM2015-4667

Background: RUNX2, a member of the RUNX family of transcription factors, regulates normal bone, cartilage and skeletal muscle morphogenesis. Aberrant RUNX2 expression has correlated with malignant cell migration, invasion and bone metastases. In the following study, we evaluated the prognostic significance of RUNX2 protein expression in a large cohort of tumors from each of three sites of origin - prostate, colon and breast. Design: Formalin-fixed paraffin-embedded tissue sections from 121 PAC, 104 CRC and 82 invasive BC [64 ductal (IDC); 18 lobular (ILC)] were immunostained by a manual method (DAKO LSAB+ System-HRP) using mouse monoclonal RUNX2 (Santa Cruz). Nuclear and/or cytoplasmic immunoreactivity was scored based on intensity and percentage of positive cells in both the tumor (T) and adjacent benign (B) epithelium in each case. Cases were assessed as tumor>benign (T>B), tumor = benign (T = B), tumor Results: RUNX2 immunoreactivity was predominately nuclear in PAC, CRC and BC and noted as follows: in PACs: T>B 58%, T = B 31%, T = Gleason grade 7] vs low [< = Gleason grade 6] tumor grade (p = 0.001), advanced [stage III, IV] vs early [stage I, II] tumor stage (p = 0.039) and biochemical disease recurrence (p = 0.008). In CRCs, immunoreactivity was noted as: T>B 61%, T = B 22%, TB 65% and T = B 35%; and correlated overall with early [stage I, II] vs advanced [stage III, IV] tumor stage (p = 0.05); ER positive status (p = 0.028); and on Cox univariate analysis, correlated with both lengthened recurrence free (p = 0.033) and overall (p = 0.042) survival; and showed a trend toward association with PR positive status (p = 0.075). Within the ER negative BC subgroup, RUNX2 immunoreactivity correlated with early tumor stage (p = 0.008) and PR positive status (p = 0.033). Conclusion: While RUNX2 protein expression correlates with adverse prognostic variables in prostate and colon cancers, RUNX2 immunoreactivity portends favorable prognosis in breast cancer. Further study of RUNX2 expression and its potential role as a therapeutic target particularly in PAC and CRC appear warranted. Citation Format: Bhaskar V.S. Kallakury, Albert Huho, Ann B. Boguniewicz, David M. Jones, Tipu Nazeer, Hwa Jeong Lee, Christine E. Sheehan, Jeffrey S. Ross. Prognostic significance of RUNX2 protein expression in carcinomas of the prostate (PAC), colon (CRC), and breast (BC). [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4355. doi:10.1158/1538-7445.AM2015-4355

Oncogenic kinase fusions are validated targets for cancer therapy. With the advent of next generation sequencing (NGS) based clinical diagnostic tools, the detection of these kinase fusions is rapidly increasing across multiple adult and pediatric tumor types. However, there remains a need to better understand the functional impact of these fusions in order to help direct clinical therapies. To study fusion biology, we have made use of a large collection of inflammatory myofibroblastic tumor (IMT) samples. IMT is a rare mesenchymal malignancy, which typically occurs in children. We have recently demonstrated that IMTs harbor multiple therapeutically actionable kinase fusions, including ALK, ROS1, and PDGFRB fusions, using a targeted capture-based NGS assay in a CLIA laboratory (Foundation Medicine). Our initial results demonstrated both previously described ALK fusions, including TPM3-, TPM4-, TFG-, and RANBP2-ALK as well as novel ALK fusions, including PRKAR1A-ALK and LMNA-ALK. Although the presence of ALK fusions within a tumor has been correlated with response to ALK inhibitor therapy, the role that the 5′ prime partner gene may play in the functional biology of the fusion has not been systematically investigated. To address this, we stably transfected cDNAs encoding LMNA-ALK, RANBP2-ALK, FN1-ALK, TFG-ALK, and PRKAR1A-ALK into BA/F3 cells. All X-ALK (X = the fusion partner) variants were tyrosine phosphorylated and their subcellular distribution was in agreement with that observed in the primary tumors harboring the identical fusion. Subcellular localization was altered as a function of the fusion partner. For example, LMNA-ALK was predominantly cytoplasmic while RANBP2-ALK was predominantly perinuclear. Additionally, we compared proliferation rates, downstream signaling, and sensitivity to various structurally different ALK inhibitors amongst all of the X-ALK fusions. Overall, our results suggest that the specific fusion partner may affect the properties of the ALK fusion protein, including sensitivity to ALK inhibitors currently in clinical use. To date, most ALK fusions are detected by immunohistochemistry for ALK overexpression or by “break-apart” fluorescence in situ hybridization (FISH), techniques which may be falsely negative in some settings and in others cannot discern specific fusion present. As the role of NGS increases in clinical diagnostics, our findings may provide further biological and clinical insights into these kinase fusions. Citation Format: Merrida A. Childress, Abha Gupta, Doron Lipson, Geoff Otto, Tina Brennan, Catherine T. Chung, Scott C. Borinstein, Jeffrey S. Ross, Phillip J. Stephens, Vincent A. Miller, Cheryl M. Coffin, Jason L. Hornick, Christine M. Lovly. Understanding oncogenic fusions: Lessons learned from inflammatory myofibroblastic tumor. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 497. doi:10.1158/1538-7445.AM2015-497

Background: Estrogen receptor (ER) inhibition is an important treatment option for advanced breast cancer (BC) pts. Recent studies describe recurring somatic mutations in codons 537 and 538 within the ligand-binding domain (LBD) of ∼ 20% of ER+ metastatic disease, but not treatment-naïve ER+ cancer or ER-negative disease, that render the ER constitutively active and confer partial resistance to endocrine agents. Few studies have described the clinicopathologic and genomic covariates that accompany the ESR1 genomic alterations (GAs). Correlating genomic events as captured by FoundationOne in pts with ESR1-mutated BCs with clinical history may provide clinical insight into these alterations. Methods: Hybridization capture of 3769 exons of 236 cancer-related genes and 47 introns from 19 genes that are frequently rearranged in cancer were fully sequenced to high, uniform coverage using FoundationOne (Foundation Medicine, Cambridge, MA). Results: 176 of the 2,208 (7.9%) BC pt cases harbored ESR1 alterations. 1127 short variants (SV) were detected in 176 patient samples for an average of 6.4 GA/sample. 16.5% SV (186/1127) were ESR1 GAs with an average of 1.0 ESR1 GAs/sample. ESR1 GAs consisted of base substitutions (77%) and amplifications (20%, median copy number 9X, range 6-28). Base substitutions occurred at codons 538 (71/145) and 537 (70/145) and at two other novel sites, 341 (2/145) and 563 (1/145). A patient-derived xenograft study suggested that tumors refractory to ER inhibitors and harboring ESR1 amplification could be responsive to higher doses of estradiol. (Li et al, Cell Reports, 2013). The most frequently co-occurring GAs were PIK3CA (37.5%), GATA3 (22.7%), TP53 (24.4%), MAP3K1 (10.2%) and CDH1 (7.9). Collection of data on tissue sites sequenced, treatments and outcomes on these pts with ESR1 GAs are ongoing. Of note, one pt had recurrent disease following adjuvant tamoxifen and letrozole. FoundationOne testing of her recurrent bony disease revealed an ESR1 Y537 alteration along with mutations in PIK3CA, APC, and RAD51. She has been on fulvestrant monotherapy with stable metastatic disease 13 + months. Conclusions: We found through FoundationOne testing that ESR1 GAs are reasonably common in advanced BCs, with base substitutions accounting for 77% of ESR1 GAs and amplifications for 20%. Identification and characterization of ESR1-mutated advanced BC pts by comprehensive genomic profiling capable of detecting both base substitutions and copy number changes may identify clinically relevant GAs. Clinical correlation is pending. Citation Format: Norma A Palma, Siraj Ali, Garrett Frampton, Kai Wang, Hannah Gilmore, Julio Peguero, Lyndsay N Harris, Massimo Cristofanilli, Juliann Chmielecki, Jeffrey S Ross, Deborah Morosini, Vincent A Miller, Phil J Stephens, Gary Palmer, Joyce O'Shaughnessy. Profiling of ESR1-mutated metastatic breast cancers by FoundationOne® allows a broad genomic understanding for potential clinical implications [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr PD6-5.