Background
The validity of KRAS mutation as a predictor of recurrence free survival (RFS) or overall survival (OS) is unclear. This study investigated whether KRAS mutation decreases RFS or OS in patients with colorectal cancer who underwent liver resection.
Methods
Patients with resected colorectal liver metastases who were treated with adjuvant hepatic arterial infusion (HAI) plus systemic therapy and in whom KRAS data was available were evaluated. Correlation between KRAS and clinical factors was evaluated using Fisher's exact test. Kaplan-Meier methods were used to estimate median overall RFS and OS.
Results
169 patients were evaluated: 118 KRAS wild type (WT) and 51 mutated (MUT). Three year RFS was 46% for KRAS WT [95%CI: 35-56%], and 30% [95%CI: 16-44%] for MUT patients (p=0.005). Three year OS was 95% [87%-98%] and 81% [62%-95%] for KRAS WT and MUT patients, respectively (p=0.07). By multivariate analysis, KRAS remained a significant predictor of RFS (HR: 1.9). Cumulative recurrence by 3 years in sites showed: 2% versus 13.4% for bone metastases (p=<0.01), 2% versus 14.5% for brain (p=0.05), 33.2% versus 58% for lung (p=<0.01), and 30 versus 47% for liver (p=0.10) in KRAS WT versus MUT patients, respectively.
Conclusions
In patients with resected colorectal liver metastases treated with adjuvant HAI plus systemic therapy, patients with KRAS mutation had a significantly worse 3 year RFS of 46% versus 30% (p=0.005) for KRAS WT vs MUT, respectively. Cumulative incidence of bone, brain, and lung metastases was significantly higher for KRAS MUT patients as compared to WT patients.