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

Garofalo et al. Genome Medicine (2016) 8:79 DOI 10.1186/s13073-016-0333-9 RESEARCH Open Access The impact of tumor profiling approaches and genomic data strategies for cancer precision medicine Andrea Garofalo1,2, Lynette Sholl3, Brendan Reardon1,2, Amaro Taylor-Weiner2, Ali Amin-Mansour2, Diana Miao1,2, David Liu1,2, Nelly Oliver1, Laura MacConaill1,3, Matthew Ducar3, Vanesa Rojas-Rudilla3, Marios Giannakis1,2, Arezou Ghazani1, Stacy Gray1, Pasi Janne1, Judy Garber1, Steve Joffe4, Neal Lindeman3, Nikhil Wagle1,2,5, Levi A. Garraway1,2,5*† and Eliezer M. Van Allen1,2,5*† Abstract Background: The 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. Methods: We 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. Results: After 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. Conclusions: Large 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. Keywords: Genomics, Precision medicine, Disparities, Immuno-oncology, Neoantigens, Panel testing * Correspondence: ; † Equal contributors 1 Department of Medical Oncology, Dana-Farber Cancer Institute, Harvard Medical School, 450 Brookline Avenue, Boston, MA 02115, USA Full list of author information is available at the end of the article © 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Garofalo et al. Genome Medicine (2016) 8:79 Background The mapping of the human genome, together with the advent of massively parallel sequencing, has accelerated discovery of driver genetic alterations in cancer and the development of drugs to target or otherwise exploit these events [1]. Multiple tumor profiling approaches that leverage these advances have entered the clinic. Such assays often consist of targeted sequencing panels that query a subset of typically 200–500 genes implicated in cancer biology or clinical management [2–8]. Alternatively, panels that emphasize rapid turnaround time by profiling smaller gene sets (n = 15–48 genes) have also emerged [9, 10]. On the other end of the spectrum, clinical whole-exome sequencing (WES; n ~ 20,000 genes) of matched tumor and germline samples has been studied through prospective sequencing efforts [11–13]. However, the benefits and limitations of these different sequencing strategies remain incompletely understood. Understanding the differences in genomic results between different tumor profiling approaches will become increasingly important as the cancer genome is leveraged to stratify patients for new therapeutic strategies. For example, unlike targeted therapies linked to specific genetic lesions (e.g., epidermal growth factor receptor mutations and inhibitors), immune targeting strategies, such as checkpoint blockade or personalized cancer vaccines, may require large-scale ascertainment of mutational and neoantigen loads and individual mutationassociated neoantigens for personalized cancer vaccine development [14–18]. One effort demonstrated the ability of two large gene panels (315 or 573 genes) to predict mutational load for immunotherapy response in pilot patient cohorts [19], and another effort demonstrated the ability of one large gene panel (341 genes) to predict DNA mismatch repair protein deficient tumors through mutational load [20], although a systematic characterization of different tumor profiling strategies for both mutation load and personal neoantigen identification should inform their relative utilities for stratifying patients in emerging cancer precision medicine frameworks. Moreover, although sequencing of paired normal blood or tissue samples is standard practice for research-oriented WES applications, many targeted panel approaches do not include matched normal samples [2, 3, 9, 21, 22]. Together with the limited ancestral diversity in many existing germline databases, this absence of paired normals has raised concerns for the potential of increased false positive somatic mutation calls that are actually germline [23, 24]. To investigate these issues, we analyzed clinical sequencing data from 157 patients with advanced lung and colon adenocarcinoma to ascertain the relative merits of distinct tumor profiling approaches. Page 2 of 10 Methods Patients and tumor specimens All patients consented to an institutional review board-approved protocol that allows comprehensive genetic analysis of tumor and germline samples (Dana-Farber Cancer Institute #12-078). Ancestry status was self-reported. Samples were selected from pathology archi (...truncated)