Abstract LB-105: Characterization of total mutational burden in the GENIE cohort: Small and large panels can provide TMB information but to varying degrees

Abstract

Abstract Introduction: Total mutation burden (TMB) correlates with response to immune checkpoint inhibition. AACR Project Genomics Evidence Neoplasia Information Exchange (GENIE) recently released clinical sequencing data for ~19000 cases from 8 institutions. We determined the ability of diverse sequencing panels to assess TMB. Methods: De-identified sequencing data from eight international academic cancer centers were used to investigate the per sample non-silent total somatic mutational burden (TMB) per 1Mb sequenced for all submitted samples. Summary: We first limited TMB characterization to three sites with sequencing data from large gene panels (sampling greater than approximately 1Mb) comprising 14472 samples. The distribution of the TMB normalized per Mb sequenced demonstrated a wide range of mutational rates both within and between tumor types (Table 1). The other five sites utilized smaller targeted sequencing panels. We then restricted the large panel data to the footprint of the smaller targeted panels and compared the number of mutations in the footprint of the small panels to the TMB of each sample from the larger panel data. We found that samples with >5 mutations in the footprint of the smaller targeted panels almost always had a high TMB (94%). In contrast, samples with no mutations in the footprint of the small panels almost never had a high TMB (2%). However, in samples with 1-5 mutations (which represented 2/3 of the samples) the footprint of the smaller targeted panels was not predictive of TMB. Conclusions: While whole exome analyses remain the gold standard for determination of total mutational burden, large sequencing panels (covering greater than approximately 1Mb) can be used to estimate TMB in diverse tumor types. Smaller targeted panels can also predict TMB for a subset of samples, with either 0 or >5 mutations, but this approach appropriately characterizes only just under a third of cases. The majority of cases need the increased genomic sampling of the larger gene panels (or whole exome analyses) to accurately quantify TMB. In summary, while the detection of high mutational burden cases by smaller targeted panel data is quite specific it lacks sensitivity. [A. S. B. and T. S. contributed equally to this work.] Table 1Mutations per MbsMedianSpectrum[0,1)[1,10)>10nMelanoma7.804%54%42%408Skin Cancer, Non-Melanoma7.6521%32%47%182Bladder Cancer7.585%60%35%587Colorectal Cancer6.632%81%18%1170Small Cell Lung Cancer6.635%69%26%104Endometrial Cancer6.222%72%26%446Non-Small Cell Lung Cancer6.187%70%23%2069Cancer of Unknown Primary5.307%74%19%255Non-Hodgkin Lymphoma4.7811%71%18%188Cervical Cancer4.789%79%13%80Esophagogastric Cancer4.749%82%9%440Head and Neck Cancer4.309%77%14%300Breast Cancer3.868%86%6%1739Ovarian Cancer3.868%88%4%603Leukemia3.8615%84%1%349Glioma3.798%86%6%858 Table 2.Mutations per Mbs (based on full panel)[0,1)[1,10)>10n# Mutations (based on footprint of amplicon hotspot panels)033%65%2%4521[1-5]3%81%15%9806>50%6%94%145 Citation Format: Alexander S. Baras, Thomas Stricker, on behalf of the AACR Project GENIE Consortium. Characterization of total mutational burden in the GENIE cohort: Small and large panels can provide TMB information but to varying degrees [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 LB-105. doi:10.1158/1538-7445.AM2017-LB-105