[651] KRAS c.35G>T (p.G12V) and c.34G>C (p.G12R) Mutations Predict Shorter Survival in 1925 Colorectal Cancers

Yu Imamura, Charles Fuchs, Shuji Ogino. Dana-Farber Cancer Institute, Boston, MA; Brigham and Women's Hospital, Boston, MA

Background: KRAS mutation test has become routine for colorectal cancer. Experimental evidence indicates stronger oncogenic effects of KRAS c.35G>T (p.G12V) and c.34G>C (p.G12R) mutations, compared to other codon 12 and 13 mutations. We hypothesized that colorectal cancers with KRAS c.35G>T and c.34G>C mutations might behave more aggressively than tumors with other KRAS mutations and wild-type KRAS.
Design: We sequenced KRAS codons 12 and 13 in 1925 colorectal cancers, including 1267 stage I-IV colorectal cancers (cohort A) and a validation set of 658 stage III colon cancers in a clinical trial (cohort B). Patient survival was assessed by Kaplan-Meier method, and by Cox proportional hazards model, which was adjusted for other variables, including stage and status of MSI and BRAF mutation.
Results: KRAS mutations were detected in 684 (36%) of 1925 tumors. In cohort A, Kaplan-Meier and multivariate Cox regression analyses showed that c.35G>T mutants (N=93) and c.34G>C mutants (N=8) were associated with significantly high colorectal cancer-specific mortality.

Individual KRAS mutations and colorectal cancer-specific mortality (cohort A)
KRASNUnivariate HR (95% CI)P valueMultivariate HR (95% CI)P value
Wild-type6351 (reference) 1 (reference) 
c.34G>A (p.G12S)122.61 (1.15-5.91)0.0221.03 (0.44-2.44)NS
c.34G>C (p.G12R)84.22 (1.72-10.4)0.0023.39 (1.28-9.00)0.014
c.34G>T (p.G12C)441.70 (1.01-2.85)0.0451.56 (0.92-2.65)NS
c.35G>A (p.G12D)1551.47 (1.07-2.04)0.0191.08 (0.76-1.51)NS
c.35G>C (p.G12A)191.36 (0.60-3.08)NS0.56 (0.24-1.30)NS
c.35G>T (p.G12V)931.84 (1.28-2.64)0.0012.00 (1.38-2.90)0.0003
c.38G>A (p.G13D)1031.33 (0.90-1.96)NS0.88 (0.59-1.30)NS
CI, confidence interval; HR, hazard ratio; NS, not significant. BRAF mutants were excluded to controll for confounding by BRAF. Tumors with two or more different KRAS mutations were exculded.

Findings were essentially replicated in our validation cohort B (stage III colon cancers in the trial)(Figure, the bottom panel).
Conclusions: Colorectal cancers with KRAS c.35G>T and c.34G>C mutations are aggressive tumor subtypes, and KRAS status serves as a prognostic biomarker. Our data highlight the importance of accurate molecular characterization of colorectal cancer.
Category: Gastrointestinal

Monday, March 4, 2013 1:00 PM

Poster Session II # 132, Monday Afternoon


Close Window