Ali, Siraj M.

PURPOSE Combined hepatocellular cholangiocarcinoma (cHCC-CCA) is a rare, aggressive primary liver carcinoma, with morphologic features of both hepatocellular carcinomas (HCC) and liver cholangiocarcinomas (CCA). METHODS The genomic profiles of 4,975 CCA, 1,470 HCC, and 73 cHCC-CCA cases arising from comprehensive genomic profiling in the course of clinical care were reviewed for genomic alterations (GA), tumor mutational burden, microsatellite instability status, genomic loss of heterozygosity, chromosomal aneuploidy, genomic ancestry, and hepatitis B virus status. RESULTS In cHCC-CCA, GA were most common in TP53 (65.8%), TERT (49.3%), and PTEN (9.6%), and 24.6% cHCC-CCA harbored potentially targetable GA. Other GA were predominantly associated with either HCC or CCA, including, but not limited to, TERT, FGFR2, IDH1, and presence of hepatitis B virus. On the basis of these features, a machine learning (ML) model was trained to classify a cHCC-CCA case as CCA-like or HCC-like. Of cHCC-CCA cases, 16% (12/73) were ML-classified as CCA-like and 58% (42/73) cHCC-CCA were ML-classified as HCC-like. The ML model classified more than 70% of cHCC-CCA as CCA-like or HCC-like on the basis of genomic profiles, without additional clinico-pathologic input. CONCLUSION These findings demonstrate the use of ML for classification as based on a targeted exome panel used during routine clinical care. Classification of cHCC-CCA by genomic features alone creates insights into the biology of the disease and warrants further investigation for relevance to clinical care.

Abstract 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%; p<2E-23). PD-L1 positive and TMB high populations were not significantly co-occurrent (OR = 1.02, p = 0.87). Similar percentages of PD-L1 positivity were observed in both TMB low (9.3%) and TMB high (9.5%). However, among patients with very high TMB (>20 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.

4087 Background: The management of CCA has evolved as targeted and immune checkpoint inhibitor (ICPI) therapies have emerged. We used comprehensive genomic profiling (CGP) to characterize the genomic alterations (GA) that have potential to personalize therapy for CCA. Methods: 3634 CCA underwent hybrid capture based CGP on 0.8-1.1 Mb of the coding genome to identify GAs in exons and select introns in up to 404 genes, TMB, microsatellite status (MSI) and % monoallelic genome (gLOH). PD-L1 expression was determined by IHC (Dako 22C3). Results: 52% of CCA were female with a median age of 62 years (range 16 - > 89). The most common biopsy sites were liver (74%), lymph node (4%), bile duct (3.3%), and lung (2%). MSI-high was rare (1%), 118 and 47 cases had TMB > 10 and > 20 mut/mb respectively. Of the latter, 51% (24/47) were MSI-H. PD-L1 amplification (AMP) was present in 0.27%. Of 490 CCA tested, 43 (9%) were positive for PD-L1 expression. 11% of cases had gLOH > 16%, only 2 cases had both TMB > 20 and gLOH > 16%. GA were most common in TP53 (31%), CDKN2A (29%), KRAS (20%) and ARID1A (17%). Potentially targetable GAs included FGFR2 (11%, 85% fusions), BRAF (5%, 50% V600E), ERBB2 (5%, 72% AMP), MET (2%, 90% AMP), EGFR (0.52%) and rarer ( < 0.5%) FGFR3, RET, FGFR1, ALK, and ROS1 fusions. The FGFR2 fusions had 128 unique 3’ partner genes including BICC1 (26%), CCDC6 (3.2%), AHCYL1 (2.6%) and KIAA1217 (2.6%). FGFR2 fusions occurred in a mutually exclusive fashion from high gLOH (p < 0.002), but not high TMB. GA in IDH1 (15%) were mutually exclusive of FGFR2 fusions (p < 1e-13), but co-occurred with PBRM1 GA (23%, p < 1e-21), ARID1A (26% p < 1e-10). IDH1 GA had gLOH similar to the overall CCA population but were enriched for low TMB (p < 1e-3). Conclusions: Nearly 20% of CCA cases harbor targetable kinase GA, half of which were FGFR2 fusions. Independently, an additional 10% (gLOH) and 1% (high TMB, MSI and/or PD-L1 AMP) may benefit from PARP inhibitors and ICPI respectively. Independently, co-mutation of IDH1 and PBRM1/ARID1A defines a class of CCA that warrants further investigation for sensitivity to PARP inhibitors and may serve as a paradigm for other tumors (ie. gliomas) with a similar co-occurrence landscape.

263 Background: DNA repair genomic alterations (GAs) have been identified in 12% of Intrahepatic cholangiocarcinoma (IHCCA), 26% of extrahepatic CCA (EHCCA) and 8% of gallbladder cancer (GBC) patients (pts) ( Cancer 2016;122:3838–3847). Recently, the Cancer Genome Atlas (TCGA) has described more than 20 mutated DNA repair genes, many of which were not previously represented in prior reports. DNA repair GAs including MSI are associated with higher tumor mutational burden (TMB). Our study aim is to identify variations in the frequency of DNA repair GAs in IHCCA, EHCCA, and GBC. Methods: Hybrid capture-based next-generation sequencing of 422 fixed formalin paraffin embedded (FFPE) tissue blocks of BTC including; 270 IHCCA, 60 EHCCA, and 92 GBC was performed. We included 20 DNA repair genes and classified them as "direct" DNA repair genes ( ATM, ATR, BRCA1, BRCA2, FANCA, FANCD2, MLH1, MSH2, MSH6, PALB2, POLD1, POLE, PRKDC, RAD50, SLX4) and "caretaker" genes that induce genomic instability ( BAP1, CDK12, MLL3, TP53, BLM). We calculated TMB on 1.1 Mb of sequenced DNA in 205 tissue specimens [138 IHCCA, 23 EHCCA, and 44 GBC] and classified into three groups; high (TMB-H; ≥ 20 mut/Mb), intermediate (TMB-I; 6-19mut/Mb) and low (TMB-L; < 6mut/Mb). Results: A distinct pattern of DNA repair GAs in each tumor type was detected. (Table 1). DNA repair GAs were more commonly noted in both EHCCA and GB (63%) as compared with IHCAA (45.2%) (p= .002). Direct DNA repair GAs were highest within EHCCA (25%), while indirect DNA GAs were predominant in GBC (59.8%) (p= 0.04 and 0.001, respectively). The frequency of TMB-H and TMB-I differed significantly between BTC subtypes. Pts with EHCCA and GBC had significantly higher TMB-H and TMB-I versus IHCCA. Conclusions: The frequency of DNA repair GA’s is higher in EHCCA and GBC as compared with IHCCA. These results may have implications for clinical trials with DNA repair inhibitors and immune checkpoint blockers. [Table: see text]

4086 Background: Mutations in DNA repair pathway were identified in 13% of Biliary Tract Cancers (BTC) [ Cancer2016;122:3838–3847]. High TMB tumors including melanoma, lung cancer and those with microsatellite instability (MSI-H) are associated with susceptibility to immune blockade using checkpoint inhibitors. TMB data in BTC is limited and its association with actionable somatic mutation (mut) profiles in BTC is unknown. Methods: Comprehensive genomic profiling (CGP) of 309 FFPE tissue blocks of BTC pts with a hybrid capture of all coding exons of 236 cancer-related genes and 47 introns of 19 genes rearranged in cancer was done using FoundationOne. Base substitutions, indels, gene fusion/rearrangements, TMB, and MSI status were assessed. TMB was calculated by counting mutations across a 1.25Mb region and classified into high (TMBH; ≥20 mut/Mb), intermediate (TMBI; 6 - 19mut/Mb) and low (TMBL; < 6mut/Mb). MSI high (MSIH) and Stable (MSS) status was assigned by a computational algorithm examining 114 intronic homopolymer loci. Patients with TMB ≥6 mut/Mb (N = 60) were included in the clinical correlative portion of this study. Results: Sixty patients with TMB ≥6 mut were identified out of 309 pts of which 9 (15%) were TMBH and 51 (85%) were TMBI. These included 3 (5%) MSIH and 18 (30 %) MSS. The median age was 59 years (range: 29-86), 35 (58%) were females, majority were intrahepatic cholangiocarcinoma (n = 31; 52%) and 28 (47%) presented with advanced disease at diagnosis. Twenty three (38%) pts had received radiation therapy, 28 (47%) surgery and 3 (5%) received immunotherapy. Most frequent co-existing mut seen was TP53 (N = 35; 58%). APC mut was seen in 7 (12%) pts. DNA repair pathway muts ( MSH6, BRCA1, BRCA2, ATM, MLH1, or MSH2 genes) were identified in 78% of TMBH versus 16% in TMBI cases (p < 0.0001). Frequency of PIK3CA mut differed significantly between TMBH and TMBI (44% vs 10%, p < 0.0001). Pts with TMBI had a significantly better median OS (110 weeks) as compared to TMBH (43 weeks) (p = 0.003). Conclusions: DNA repair pathway and PIK3CA mut maybe associated with TMBH in BTC. A better understanding of TMB and associated actionable mutations in BTC may be of value for the management of BTC patients with targeted agents and immunotherapy.

4076 Background: Gallbladder carcinoma (GBC) is commonly diagnosed at an advanced, unresectable disease stage and has a poor prognosis. Comprehensive genomic profiling (CGP) has a developing role in guiding systemic, precision anti-cancer therapy. Methods: We performed hybrid capture-based CGP on FFPE samples for 491 consecutive advanced GBCs. Mean coverage depth was >550X for up to 315 cancer-related genes plus 37 introns from 28 genes frequently rearranged in cancer. All 4 classes of genomic alterations (GA) were detected. Tumor mutational burden (TMB) was determined on up to 1.1 Mb of sequenced DNA. Results: Median age was 64 (range 25-88) and 69% (337/491) of patients were female. 96% (470), 2.9% (14), and 1.4% (7) of GBC patients had the diagnosis of adenocarcinoma, adenosquamous, or carcinoma not specified (NOS), respectively. Commonly altered genes were TP53 (62%), CDNK2A (31%), ARID1A (18%), and SMAD4 (15%), and most of the genetic aberrations (GA) were short variants. Potentially targetable SCNAs (including ERBB genes, MET, FGFRs, and the CCND1-FGF3/4/19 11q13 amplicon) were identified in 21% of cases (Table). Oncogenic BRAF, ALK, or FGFR2/3 rearrangements were found in 7 cases (1.4%). Moreover, 7.8% of cases had BRCA2 or ATM GA, and 0.8% had INI1loss suggesting benefit from PARP or EZH2 inhibitors, respectively. TMB was low; the 25th, 50th, and 75th percentiles were 2.5, 3.8 and 6.3 mutations/Mb, respectively. Less than 1% of cases had microsatellite instability. Radiological response to TKIs and immunotherapy was noted. Conclusions: In addition to the significant opportunity for anti-HER targeted therapies, other subsets of GBC cases harbored kinase GA (particularly SCNA), the 11q13 amplicon, or BRCA2/ ATM/ INI1mutations that are linked to therapeutic benefit. How the frequency of both driver SCNA and DNA repair alterations in GBC can be linked with inflammation awaits additional investigation. [Table: see text]

Abstract Background Esthesioneuroblastoma (ENB), also known as olfactory neuroblastoma, is a rare malignant neoplasm of the olfactory mucosa. Despite surgical resection combined with radiotherapy and adjuvant chemotherapy, ENB often relapses with rapid progression. Current multimodality, nontargeted therapy for relapsed ENB is of limited clinical benefit. Materials and Methods We queried whether comprehensive genomic profiling (CGP) of relapsed or refractory ENB can uncover genomic alterations (GA) that could identify potential targeted therapies for these patients. CGP was performed on formalin-fixed, paraffin-embedded sections from 41 consecutive clinical cases of ENBs using a hybrid-capture, adaptor ligation based next-generation sequencing assay to a mean coverage depth of 593X. The results were analyzed for base substitutions, insertions and deletions, select rearrangements, and copy number changes (amplifications and homozygous deletions). Results Clinically relevant GA (CRGA) were defined as GA linked to drugs on the market or under evaluation in clinical trials. A total of 28 ENBs harbored GA, with a mean of 1.5 GA per sample. Approximately half of the ENBs (21, 51%) featured at least one CRGA, with an average of 1 CRGA per sample. The most commonly altered gene was TP53 (17%), with GA in PIK3CA, NF1, CDKN2A, and CDKN2C occurring in 7% of samples. Conclusion We report comprehensive genomic profiles for 41 ENB tumors. CGP revealed potential new therapeutic targets, including targetable GA in the mTOR, CDK and growth factor signaling pathways, highlighting the clinical value of genomic profiling in ENB.