Abstract 3167: Improving and standardizing TMB assay performance

Abstract

Abstract Determining and using the most effective and safest treatment is of great importance in cancer disease management. Checkpoint inhibitors that target immune regulatory molecules such as PD-1, PD-L1, and CTLA-4, have successfully improved PFS and OS - but only in some patients and for some cancers. In the case of PD-1 and PD-L1 inhibitors, it is believed that PD-L1 expression by a solid tumor allows it to escape attack from the immune system and that by inhibiting the PD-L1/PD-1 interaction immune evasion is no longer possible. This may then cause some of the mutations that give rise to expressed neoantigens to act as targets for T cells and allow for the gradual elimination of the tumor. Since the risk for developing autoimmune adverse events is not insignificant with the use of checkpoint inhibitors, determining which patients are likely to respond favorably to their use is imperative. Similarly, expanding approved uses to new indications also benefits from the identification of populations that are most likely to show improvement in PFS and OS - generally, through clinically-validated biomarkers. Current indications for the use of PD-1 and PD-L1 inhibitors rely on cancer type and several biomarkers. One of these is the expression (or level of expression) of PD-L1 on the tumor. Another - usually in colorectal cancer - is the presence of microsatellite instability (MSI), which indicates that DNA is not being copied with high fidelity and resulting mutations may lead to the emergence of potential neoantigens. While MSI can be assessed by the PCR amplification of several loci, the presence or absence of neoantigens cannot, and they can also be created in the absence of MSI. This has given rise to a potential biomarker - tumor mutational burden (TMB) - that is an assessment of the number of relevant mutations in a tumor. TMB measurement is challenging since different targeted next generation sequencing (NGS) panels look at different regions and percentages of the genome and use different criteria as to what constitutes a relevant mutation. Presently, there is poor correlation between different TMB assays at mutation levels that may be relevant for a companion diagnostic where patients may be denied treatment. Here, we describe the characterization of a panel of reference materials for the assessment, harmonization, and improvement of TMB measurements by NGS assays. The panel is comprised of cancer cell lines and their SNP-matched normals. DNA from unfixed cells was used to establish a baseline truth set of somatic mutations and germline SNPs, and DNA from FFPE cells was used to determine how TMB assessment is affected by fixation artifacts - especially, at lower variant allele frequencies. We present data from high coverage whole exome sequencing - the current gold standard in TMB assessment - and from targeted NGS panels that are more typical of what may be used clinically. We show that accurate measurements at what may be clinically relevant TMB cutoffs remain a challenge. Citation Format: Matthew G. Butler, Yves Konigshofer, Lequan Nguyen, Shikha Kalotra, Omo Clement, Russell K. Garlick, Bharathi Anekella. Improving and standardizing TMB assay performance [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 3167.