Correlation of survival in colorectal cancer with tumor mutational burden and mutations identified by next-generation sequencing.

Abstract

e18861 Background: Next-generation sequencing (NGS) can identify specific mutations and molecular abnormalities such as high tumor mutational burden (TMB), defined as 10 or more mutations per megabase. Immune checkpoint inhibitors (ICI) have been approved for all solid tumors with high TMB, including colorectal cancer (CRC), though studies have been inconclusive regarding prognosis of high TMB as well as the efficacy of ICI’s in treating CRC specifically. We aimed to use data collected from next-generation sequencing (NGS) of CRC at our institution to correlate survival to TMB, correlate survival to presence of mutations of interest, and to investigate efficacy of ICI in this patient population. Methods: This was a retrospective cohort analysis of patients at a single institution, analyzing NGS of patients who were identified as TMB-high and microsatellite stable (MSS) (n = 161) from January 20218 to December 2020.Demographics, clinical-pathological factors, and NGS results were extracted along with clinical outcomes. We conducted a survival analysis looking at TMB status and selected mutations of interest. A descriptive subset analysis of the patients who underwent treatment with immunotherapy was performed. Results: Patients with high TMB MSS CRC trended to have worse survival with a HR of 1.45, though this was not significant (p = 0.22). Median survival for high TMB patients was 28.1 months and was 77.5 months for non-high TMB. Analysis of NGS revealed 112 unique mutations, with TP53 (82.0%), APC (80.1%), and KRAS (43.5%) occurring the most frequently. Patients were found to have statistically significant worse overall survival with KRAS (HR = 1.81, p = 0.023), CDKN2A (HR = 4.81, p = 0.0002), RB1 (HR = 2.83, p = 0.047), and SMAD4 (HR = 1.858, p = 0.039) mutations. Twelve patients underwent treatment with ICI with nine patients having between 0.7 to 4.1 months of progression-free survival (PFS) and three patients with long PFS of 13.2, 24.7, and 26.3 months. The two patients with the longest PFS were not high TMB. Conclusions: Presence of high TMB in MSS CRC did not show a statistically significant difference in outcome. Presence of KRAS, CDKN2A, RB1, and SMAD4 correlated with significantly worse prognosis. Treatment with ICI’s was effective in some CRC patients though survival did not correlate with high TMB. There is a need to identify a marker beyond TMB which better indicates which patients with MSS CRC will have good response to ICI therapy.