Tumor genomic analysis for 128 renal cell carcinoma (RCC) patients receiving first-line everolimus: Correlation between outcome and mutations status in MTOR, TSC1, and TSC2.

Abstract

484 Background: MTOR inhibitors are standard agents in the management of advanced RCC. Case series have suggested somatic mutations in MTOR, TSC1, and TSC2 may sensitize tumors to everolimus (EVE) [Voss et al., CCR 2014; Kwiatkowski et al., CCR 2016]. We sought to test this hypothesis through next generation sequencing (NGS) of tumors from a large cohort of patients (pts) treated with everolimus (EVE) on the randomized RECORD3 trial of first-line EVE vs. sunitinib. Methods: Somatic mutations were investigated using a custom exon-targeted NGS platform with deep sequence coverage. This analysis focused on mutations in TSC1, TSC2, PTEN, PIK3CA, and MTOR. Association between genotypes and progression free survival (PFS) was assessed by Cox PH models and log-rank tests. Mutations in other pathways and non-genetic covariates were investigated. Functional investigation of individual mutations in MTOR, TSC1, and TSC2 were undertaken. Results: 238 pts received 1st line EVE with median PFS of 7.85 mo (95% CI 5.6-8.2). Tumor sampleswere analyzed for 128 pts. Median depth of coverage was ~540X. There was no correlation between PFS and mutation status of mTOR, TSC1, or TSC2 (15 pts [12%]; HR 0.99 [95% CI 0.47-2.07]; p=0.4697); the objective response rate (ORR) in this group was 0%. Individual PFS for all 15 pts clustered outside the 95% CI for the population of 128 everolimus-treated pts (8 pts PFS<2.8mo; 7pts PFS>8.3mo). 3 of 8 pts with PFS<2.8mo had >1 mutation in MTOR, TSC1, or TSC2, while none among the 7 pts with PFS >8.3mo. When including pts with mutations in PTEN, PIK3Ca, MTOR, or TSC1 (30pts [23%]), we found no association between mutation status and PFS (HR 1.02 [95% CI 0.61-1.71]; p=0.3025). Functional investigations suggest varying biologic effects of individual mutations. Conclusions: "One-dimensional" mutation status for core components of the MTOR signaling pathway did not correlate with PFS in this dataset. Grouping based on more detailed characterization incorporating copy number status and functional annotation may provide better insights and should be considered for future biomarker development. Clinical trial information: NCT00903175.