KRAS discordance between primary and recurrent tumors after radical resection of colorectal cancers.

Background: Although KRAS shows high concordance between primary and metastatic colorectal cancers, recent studies have reported discordance and intra-tumoral heterogeneity. To evaluate KRAS concordance between primary colorectal cancers and recurrent tumors after radical resection, we performed this study.

Methods: Between January 2007 and August 2013, among patients underwent radical resection for primary colorectal cancers and tissue sampling of recurred tumors including resection or biopsy, 74 patients whose both primary and recurred tumor tissues were available for KRAS analysis were enrolled. The clinical and pathologic data were retrospectively revised and KRAS analyses were performed.

Results: The patients with initial M1 stage showed significantly higher KRAS discordance rate (54.5%). The KRAS concordance rate was 79.7% (n = 59). Forty-two patients (56.8%) showed the wild-to-wild type and 17 (22.9%) showed the mutant-to-mutant type. The discordance rate was 20.3% (n = 15). Eight patients (10.8%) showed the wild-to-mutant type, and 7 (9.5%) showed the mutant-to-wild type. Among 15 discordance cases, intra-tumoral heterogeneity was found in 26.7% (n = 4).

Conclusions: There is 20.3% KRAS discordance between primary and recurrent tumors, which is higher rate than is generally known. For selection of the effective target agent, KRAS analysis of recurred tumors will be necessary, if it is available.