The prevalence of common KRAS variants and associated outcomes in patients with metastatic colorectal cancer (mCRC).

Abstract

173 Background: KRAS is the most common driver oncogene in mCRC. Comprehensive analysis of KRAS variants prevalence and the relationship between variants and outcomes are lacking. Herein, we aimed to examine the impact of KRAS variants on outcomes in patients (pts) with mCRC. Methods: A retrospective review of pts with mCRC with known KRAS mutation status was performed. Fisher’s exact test was used to analyze the association between KRAS variants. Cox Proportional Hazard modeling was used to study the relationship between KRAS variants and overall survival (OS). Kaplan-Meier method was used to assess OS. Results: A total of 423 pts diagnosed with mCRC from 2014-2020 with available extended KRAS status were evaluated. Median age at diagnosis was 59.8 yrs (22.3-92 range), 57% were male, 22% were Black, and 75% presented with de novo metastatic disease. A majority (56%) of tumors harbored KRAS mutations. The most frequent KRAS variants were G12D 47% (111), G12V 12% (28), G12C 13% (31), G13D 9% (21), and 20% (47) were other variants. In univariate analyses, the presence of a KRAS mutation, Black race, de novo metastatic disease, age, receipt of chemotherapy, and receipt of surgery were associated with OS. Tumor location was not associated with OS. In multivariable analyses, mutation type was a significant predictor of death and the presence of G12D was associated with an increased risk of death compared to G12V and KRAS wildtype. There was no increased risk of death in pairwise comparisons of G12D and G13D or other KRAS variants. Black race, de novo metastatic disease, and no receipt of surgery were additional independent predictors of death (Table). With a median follow-up of 25.7 months (mo.), median OS was 35.5 mo. (95% CI 10.5-50.9) with G12C, Not Reached (NR) (95% CI 21-NR) with G12V, 36.2 mo. (95%CI 14.9-58.5) with G13D, 26.2 mo. (95% CI 21.8-37) with G12D, 39.6 mo. (95% CI 22.4-47.9) for other KRAS mutations, and 59.6 mo. (95%CI 48.2-NA) for KRAS wildtype, respectively (p=0.0003). Conclusions: Our findings highlight differences in unadjusted overall survival when comparing G12D to some other KRAS variants. Codon and amino acid-specific mutations of KRAS should be considered when evaluating prognosis and further study on treatment effects and sequencing is warranted. [Table: see text]