Abstract 2250: Tumor mutational burden: Guidelines for derivation and robustness of measurement

• Published in: Cancer research (Chicago, Ill.) Vol. 78; no. 13_Supplement; p. 2250
• Main Authors: Mola, Natalie, Schu, Matthew, Stiegelmeyer, Suzy, Jones, Wendell, Weigman, Victor J.
• Format: Journal Article
• Language: English
• Published: 01.07.2018
• ISSN: 0008-5472; 1538-7445
• DOI: 10.1158/1538-7445.AM2018-2250

Abstract Tumor mutational burden (TMB) is defined as the number of somatic mutations per Mb of the genome and is used to represent accumulation of somatic mutations over the life of the tumor. Recent evidence has shown that high TMB scores are associated with improved response to immune therapies across multiple indications. However, the derivation of this score is not well characterized, nor is this score universally performed across laboratories. Here we describe our methodology for calculating TMB, which considers an estimation of true germline mutations, the efficacy of different somatic callers, and the identification of regions within a capture panel that most reliably contribute to robust TMB scoring. To accomplish this, we only include positions that fall within high confidence regions of the genome as defined by NIST GIAB datasets HG001-HG005 v3.3.2. The high confidence regions encompass 88% of the exome. Somatic variant calling is performed utilizing the tumor and normal samples using only positions where there is sufficient depth of coverage in both samples. Any remaining germline variants are excluded, leaving only somatic variants in the targeted, high-confidence regions for the final TMB calculation. TMB calculations utilizing high-confidence regions reduce the contribution of false positive variants in the mutational burden estimate. We have compared TMB scores using this approach to those provided from evaluating somatic calls from TCGA, where matched tumor/normal pairs exist. In addition, we introduce a method to determine the effect on TMB scores when in silico determination of germline variants is used to derive somatic calls when matching normal samples are not available. Defining the robustness of this score across and criteria for its derivation will help better position defining analytical range and future utility across indications and NGS panels Citation Format: Natalie Mola, Matthew Schu, Suzy Stiegelmeyer, Wendell Jones, Victor J. Weigman. Tumor mutational burden: Guidelines for derivation and robustness of measurement [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 2250.