Brendan Reardon

Comprehensive Characterization of Cancer Driver Genes and Mutations

Identifying molecular cancer drivers is critical for precision oncology. Multiple advanced algorithms to identify drivers now exist, but systematic attempts to combine and optimize them on large datasets are few. We report a PanCancer and PanSoftware analysis spanning 9,423 tumor exomes (comprising all 33 of The Cancer Genome Atlas projects) and using 26 computational tools to catalog driver genes and mutations. We identify 299 driver genes with implications regarding their anatomical sites and cancer/cell types. Sequence- and structure-based analyses identified >3,400 putative missense driver mutations supported by multiple lines of evidence. Experimental validation confirmed 60%-85% of predicted mutations as likely drivers. We found that >300 MSI tumors are associated with high PD-1/PD-L1, and 57% of tumors analyzed harbor putative clinically actionable events. Our study represents the most comprehensive discovery of cancer genes and mutations to date and will serve as a blueprint for future biological and clinical endeavors.

The impact of tumor profiling approaches and genomic data strategies for cancer precision medicine

BackgroundThe diversity of clinical tumor profiling approaches (small panels to whole exomes with matched or unmatched germline analysis) may engender uncertainty about their benefits and liabilities, particularly in light of reported germline false positives in tumor-only profiling and use of global mutational and/or neoantigen data. The goal of this study was to determine the impact of genomic analysis strategies on error rates and data interpretation across contexts and ancestries.MethodsWe modeled common tumor profiling modalities—large (n = 300 genes), medium (n = 48 genes), and small (n = 15 genes) panels—using clinical whole exomes (WES) from 157 patients with lung or colon adenocarcinoma. We created a tumor-only analysis algorithm to assess germline false positive rates, the impact of patient ancestry on tumor-only results, and neoantigen detection.ResultsAfter optimizing a germline filtering strategy, the germline false positive rate with tumor-only large panel sequencing was 14 % (144/1012 variants). For patients whose tumor-only results underwent molecular pathologist review (n = 91), 50/54 (93 %) false positives were correctly interpreted as uncertain variants. Increased germline false positives were observed in tumor-only sequencing of non-European compared with European ancestry patients (p < 0.001; Fisher’s exact) when basic germline filtering approaches were used; however, the ExAC database (60,706 germline exomes) mitigated this disparity (p = 0.53). Matched and unmatched large panel mutational load correlated with WES mutational load (r2 = 0.99 and 0.93, respectively; p < 0.001). Neoantigen load also correlated (r2 = 0.80; p < 0.001), though WES identified a broader spectrum of neoantigens. Small panels did not predict mutational or neoantigen load.ConclusionsLarge tumor-only targeted panels are sufficient for most somatic variant identification and mutational load prediction if paired with expanded germline analysis strategies and molecular pathologist review. Paired germline sequencing reduced overall false positive mutation calls and WES provided the most neoantigens. Without patient-matched germline data, large germline databases are needed to minimize false positive mutation calling and mitigate ethnic disparities.

Mutational patterns in chemotherapy resistant muscle-invasive bladder cancer

Despite continued widespread use, the genomic effects of cisplatin-based chemotherapy and implications for subsequent treatment are incompletely characterized. Here, we analyze whole exome sequencing of matched pre- and post-neoadjuvant cisplatin-based chemotherapy primary bladder tumor samples from 30 muscle-invasive bladder cancer patients. We observe no overall increase in tumor mutational burden post-chemotherapy, though a significant proportion of subclonal mutations are unique to the matched pre- or post-treatment tumor, suggesting chemotherapy-induced and/or spatial heterogeneity. We subsequently identify and validate a novel mutational signature in post-treatment tumors consistent with known characteristics of cisplatin damage and repair. We find that post-treatment tumor heterogeneity predicts worse overall survival, and further observe alterations in cell-cycle and immune checkpoint regulation genes in post-treatment tumors. These results provide insight into the clinical and genomic dynamics of tumor evolution with cisplatin-based chemotherapy, suggest mechanisms of clinical resistance, and inform development of clinically relevant biomarkers and trials of combination therapies.The impact of cisplatin-based chemotherapy on tumor genomes is complex. Here, the authors study matched pre- and post-chemotherapy primary samples in muscle-invasive bladder cancer, finding a cisplatin-based mutational signature, and highlighting the impact of intratumor heterogeneity on survival.

Genomic Evolution after Chemoradiotherapy in Anal Squamous Cell Carcinoma

Purpose: Squamous cell carcinoma of the anal canal (ASCC) accounts for 2% to 4% of gastrointestinal malignancies in the United States and is increasing in incidence; however, genomic features of ASCC are incompletely characterized. Primary treatment of ASCC involves concurrent chemotherapy and radiation (CRT), but the mutational landscape of resistance to CRT is unknown. Here, we aim to compare mutational features of ASCC in the pre- and post-CRT setting. Experimental Design: We perform whole-exome sequencing of primary (n = 31) and recurrent (n = 30) ASCCs and correlate findings with clinical data. We compare genomic features of matched pre- and post-CRT tumors to identify genomic features of CRT response. Finally, we investigate the mutational underpinnings of an extraordinary ASCC response to immunotherapy. Results: We find that both primary and recurrent ASCC tumors harbor mutations in genes, such as PIK3CA and FBXW7, that are also mutated in other HPV-associated cancers. Overall mutational burden was not significantly different in pre- versus post-CRT tumors, and several examples of shared clonal driver mutations were identified. In two cases, clonally related pre- and post-CRT tumors harbored distinct oncogenic driver mutations in the same cancer gene (KRAS or FBXW7). A patient with recurrent disease achieved an exceptional response to anti-programmed death (PD-1) therapy, and genomic dissection revealed high mutational burden and predicted neoantigen load. Conclusions: We perform comprehensive mutational analysis of ASCC and characterize mutational features associated with CRT. Although many primary and recurrent tumors share driver events, we identify several unique examples of clonal evolution in response to treatment. Clin Cancer Res; 23(12); 3214–22. ©2016 AACR.

Integrated genomic characterization of oral carcinomas in post-hematopoietic stem cell transplantation survivors

OBJECTIVES Secondary oral squamous cell carcinoma (OSCC) is a late complication in allogeneic hematopoietic stem cell transplantation (HSCT) patients, but little is known about long-term outcomes and prognostication. Additionally, molecular alterations and immunologic insights unique to this disease remain largely unexplored. METHODS We present a cohort of 31 patients with post-HSCT OSCC and reported on clinicopathologic predictors of survival. Whole-exome sequencing was performed on 6 (19%) matched pairs of peripheral blood (post-conditioning, pre-HSCT) and tumor samples. The entire cohort had archival tumor available for immunoprofiling with PD-1/L1 immunohistochemistry. RESULTS Five-year overall survival (OS) was 57% (95% CI: 46.1-69.8) with a median disease-free survival (DFS) of 13.3 months. Advanced initial staging, a buccal or oral tongue subsite, chronic oral graft-versus-host disease (GVHD) and smoking all negatively impacted survival. High tumor mutational burden (TMB) (median 11.3 vs. 5.0) and unique mutational signatures were noted between unrelated and related donor groups - with a strong correlation between infiltrating PD-1+ lymphocytes and TMB (R = 0.98, p < 0.01). Some differences were observed when comparing commonly mutated genes among our cohort and TCGA, with a predominance of TP53 events. CONCLUSION Survival outcomes appear similar in HSCT survivors with OSCC compared with non-HSCT OSCC populations. We identified somatic alterations in genes with therapeutic potential unique to this subpopulation of oral cancers.

Inherited DNA-Repair Defects in Colorectal Cancer

Colorectal cancer (CRC) heritability has been estimated to be around 30%. However, mutations in the known CRC-susceptibility genes explain CRC risk in fewer than 10% of affected individuals. Germline mutations in DNA-repair genes (DRGs) have recently been reported in CRC, but their contribution to CRC risk is largely unknown. We evaluated the gene-level germline mutation enrichment of 40 DRGs in 680 unselected CRC individuals and 27,728 ancestry-matched cancer-free adults. Significant findings were then examined in independent cohorts of 1,661 unselected CRC individuals and 1,456 individuals with early-onset CRC. Of the 680 individuals in the discovery set, 31 (4.56%) individuals harbored germline pathogenic mutations in known CRC-susceptibility genes, and another 33 (4.85%) individuals had DRG mutations that have not been previously associated with CRC risk. Germline pathogenic mutations in ATM and PALB2 were enriched in both the discovery (OR = 2.81 and p = 0.035 for ATM and OR = 4.91 and p = 0.024 for PALB2) and validation (OR = 2.97 and adjusted p = 0.0013 for ATM and OR = 3.42 and adjusted p = 0.034 for PALB2) sets. Biallelic loss of ATM was evident in all individuals with matched tumor profiling. CRC individuals also had higher rates of actionable mutations in the HR pathway, which can substantially increase the risk of developing cancers other than CRC. Our analysis provides evidence for ATM and PALB2 as CRC-risk genes, underscoring the importance of the homologous recombination pathway in CRC. In addition, we identified frequent complete homologous recombination deficiency in CRC tumors, representing a unique opportunity to explore targeted therapeutic interventions such as poly-ADP ribose polymerase inhibitor (PARPi).

Abstract 558: Computational analysis of clinically actionable genomic features: precision heuristics for interpreting the alteration landscape (PHIAL)

Background: PHIAL (Precision Heuristics for Interpreting the Alteration Landscape) was developed as a heuristic clinical interpretation algorithm for cancer genomic data to inform treatment decisions at the point of care and provide researchers with rapid assessment of tumor actionability. This approach used somatic whole exome sequencing data and a database of tumor alterations relevant for genomics driven therapy (TARGET). However, PHIAL was limited to first order genomic relationships, could not distinguish relative actionability given multiple actionable variants, did not maximize the richness of somatic-germline interactions, and could not leverage both exome and transcriptome data to move towards feature-based actionability. Towards that end, we developed a new interpretation methodology to address these areas and improve clinical actionability algorithms. Methods: We revised PHIAL to predict actionable alterations based on the presence of SNVs (in the context of allele specific expression from RNA-seq), indels, SCNAs, fusions, and global features (e.g., context-specific mutational burden) that imply actionability. Additionally, we refined and expanded the TARGET database to enable PHIAL to produce scores on multiple dimensions and reflect newly discovered relationships between genomics and clinical actions. Predictive implication values were assigned to reflect the validities of TARGET’s drug sensitivity, drug resistance, and prognostic claims. Results: We applied both the original (PHIAL1) and an updated version of PHIAL (PHIAL2) to a 255 patient cohort with whole exome/transcriptome sequencing data (146 castration-resistant prostate cancer and 109 metastatic melanoma samples). PHIAL1 identified 1,342 clinically actionable/biologically relevant events across the cohort with a median of 3 events per patient and 95% of patients having at least one event. PHIAL2 identified 2,508 events, with a median of 6 events per patient and 98.5% of patients harboring at least one event. Of these events, 8.12% were associated with an FDA-approved therapy and 2.09% with a clinical trial. PHIAL2 identified events in 9 patient samples that PHIAL1 associated with no events. Conclusion: PHIAL2 was able to identify and rank more putatively actionable alterations than PHIAL1, and effectively transitioned from a variant-based to a feature-based approach. This strategy may inform the utility of point-of-care whole-exome/transcriptome sequencing in larger contexts as these data emerge in clinical settings, and may bridge towards machine learning based approaches as patient outcomes are linked to genomic and transcriptomic features. Finally, PHIAL2 may ultimately provide a deeper understanding of, and suggest clinical actions for, cases in which there is no clear single genomic alteration associated with oncogenesis. Citation Format: Brendan Reardon, Nathanael Moore, Eliezer VanAllen. Computational analysis of clinically actionable genomic features: precision heuristics for interpreting the alteration landscape (PHIAL) [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 558. doi:10.1158/1538-7445.AM2017-558

Abstract 5647: Intron retention as a novel source of tumor neoantigens associated with response to checkpoint inhibitor therapy

Background: Development of immune checkpoint inhibitors has substantially improved outcomes in patients diagnosed with metastatic melanoma. However, only a minority of patients treated experience long-term clinical benefit, and clinicians have limited ability to predict which patients will respond. Recent studies have demonstrated that the burden of tumor neoantigens generated by expressed somatic mutations is predictive of response to immunotherapy. Intron retention, which is widespread in cancer transcriptomes, represents a putative source of tumor neoantigens by generating peptides that are available for presentation through the MHC I pathway. Methods: We developed a neoantigen prediction pipeline to identify patient-specific neoantigens from transcriptome sequencing data, which enables identification of retained intron neoantigens from clinical cohorts receiving checkpoint inhibitor therapy. This pipeline incorporates published methods for detecting intron retention events from transcriptome data, detects open reading frames that extend from normal transcripts into intronic sequences, and identifies neoepitopes predicted as strong binders based on the patient’s HLA molecules. We applied this pipeline to a cohort of 41 melanoma patients receiving checkpoint inhibitor therapy and classified patient outcomes as receiving clinical benefit (CB) (n=14), no clinical benefit (NCB) (n=22), or long-term survival without clinical benefit (LS) (n=5). Results: Our initial analysis identified a mean retained intron neoantigen burden of 7709 per sample, without significant difference between response groups. In one patient who derived clinical benefit from checkpoint inhibition, neoantigen load from nonsynonymous mutations was low (407, 0.34 standard deviations (SD) below a mean of 1,015 among CB patients), while retained intron neoantigen load was high (14579, 1.7 SDs above a mean of 7517 among CB patients), suggesting that retained intron neoantigen load may explain response in some patients with low mutational burden. Preliminary analysis of specific neoantigens suggests that a retained intron in ZNF880 identified in patients expressing HLA-A01:01 is present in 6 of 6 patients experiencing clinical benefit, but only 2 of 7 patients not experiencing clinical benefit. The same analysis was performed on two additional cohorts of melanoma tumor samples to assess whether a larger sample size could aid in the identification of recurrent neoepitopes generated by retained introns. Conclusions: Application of this approach to data from patients receiving checkpoint blockade with selective response identifies response-associated neoantigens that may warrant further investigation. Identification of a novel source of neoantigens associated with immunotherapy response will provide valuable prognostic information to patients and inform the development of next generation immunotherapeutics. Citation Format: Alicia C. Smart, Claire Margolis, Diana Miao, David Liu, Jihye Park, Meng Xiao He, Brendan Reardon, Stephanie Mullane, Bastian Schilling, Levi A. Garraway, Dirk Schadendorf, Eliezer M. Van Allen. Intron retention as a novel source of tumor neoantigens associated with response to checkpoint inhibitor therapy [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 5647. doi:10.1158/1538-7445.AM2017-5647