Background: Tumor mutational burden (TMB) has emerged as a new biomarker for response to checkpoint inhibitor immunotherapy. Recently, we described a novel assay to calculate TMB from the circulating tumor DNA (ctDNA) in the blood (bTMB) and demonstrated validation to support its application in a prospective phase III trial in 1L NSCLC comparing the anti-PD-L1 agent, atezolizumab, against standard of care chemotherapy. Herein, we demonstrate clinical utility of the bTMB assay across a retrospective study of 794 NSCLC patients from two randomized clinical trials evaluating progression free survival (PFS) and overall survival (OS) between atezolizumab and chemotherapy. Furthermore, we compare TMB between tissue and blood, and evaluate variant-level concordance between the bTMB assay and the CLIA-validated ctDNA assay, FoundationACT (FACT).

Methods: Cell free DNA (cfDNA) from plasma was isolated from retrospective clinical samples. The bTMB assay delivered a count of somatic base substitutions in the ctDNA down to 0.5% allele frequency across 394 genes from as little as 1% tumor content, yielding a bTMB score. Tissue TMB was evaluated from formalin-fixed, paraffin embedded specimens using the FoundationOne assay.

Results: Patients with bTMB ≥14 mut/Mb were considered high, and demonstrated significant enrichment for both PFS and OS in the phase II POPLAR study evaluating atezolizumab vs. chemotherapy (PFS HR = 0.57, OS HR = 0.56, n=211). Survival results were validated in the larger phase III study, OAK (PFS HR = 0.65, OS HR = 0.64, n=583). Significant enrichment in PFS was observed between bTMB-high and non-high patients receiving atezolizumab, however this effect was not observed for patients receiving chemotherapy. Amongst 259 patients with blood and tissue TMB, overall and positive percent agreement (PPA) in the TMB categorical result were 81.5% and 63.6%, respectively. When bTMB was restricted to the 62 gene FACT assay, the PPA dropped to 17.0%, suggesting a sufficiently sized panel is required to sensitively identify patients with high TMB. When samples were TMB-high in blood, the median percent overlap of TMB variants between blood and tissue was 70%. Blood samples tested in both the bTMB and FACT assays revealed that 93% of variants were detected in both assays across overlapping regions, with ≥99% of variants detected in both platforms with variant allele frequency ≥1.0%.

Conclusions: The bTMB assay is analytically validated for high accuracy and precision, and demonstrates a high degree of variant level concordance with the FACT assay. High bTMB is significantly associated with improved survival to atezolizumab vs. chemotherapy in 2L NSCLC, and the survival benefit is specific to atezolizumab and not prognostic. Finally, TMB between blood and tissue is correlated and largely explained by overlapping variants present in both sample types.

Citation Format: David Fabrizio, Daniel Lieber, Christine Malboeuf, Jacob Silterra, Emily White, Michael Coyne, Tina Brennan, Jie Ma, Mark Kennedy, Erica Schleifman, Sarah Paul, Yan Li, David Shames, Craig Cummings, Eric Peters, Marcin Kowanetz, Doron Lipson, Geoff Otto. A blood-based next-generation sequencing assay to determine tumor mutational burden (bTMB) is associated with benefit to an anti-PD-L1 inhibitor, atezolizumab [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 5706.