Is there a specific phenotype associated with the different subtypes of KRAS mutations in patients with advanced...

KRAS mutations occur in 20 to 25% of non-small-cell lung cancers (NSCLC) and seem to predict a poor prognosis. There is heterogeneousness in the frequency and spectrum of KRAS mutations, which can be categorized in transitions and transversions. We wondered if subtypes of KRAS mutation were associated with specific clinical phenotypes and specific survival.

Between July 2007 and May 2012, patients with advanced NSCLC and KRAS mutation diagnosed in two university hospitals were included. Clinical and histological characteristics, therapeutics and survival data were collected.

Among 635 patients screened for KRAS mutations, 90 were found to be mutated and were included. Median age was 59 years (range: 54-69). Most were males (60%), current or former smokers (63% and 33%, respectively) and had an adenocarcinoma (ADC) (80%). Eighty patients were stage IV and 10 were stage IIIB. Eighty percent of the KRAS mutations were transversions and 20% were transitions. In uni- and multivariate analyses, there was a trend for fewer smokers among patients with transitions than among those with transversions (Odds Ratio [OR]=0.28, 95% CI [0.079-0.999], p=0.05). No significant difference was noted between transitions and transversions for other clinical characteristics. Patients with transitions had more frequently squamous-cell carcinoma (SCC) compared to those with transversions, who had more frequently adenocarcinomas (OR=16.7, 95% CI [2.76-100.8], p=0.002). Seventy-nine patients (86%) had received first-line chemotherapy. No significant difference was seen for disease-control rate, median progression-free survival or overall survival between transitions and transversions.

A higher proportion of non-smokers and SCC subtypes were observed in the transitions compared to transversions. This confirms the heterogeneity of KRAS mutations and could suggest to expand KRAS testing in SCC to assess impact of RAS in SCC, which remains poorly investigated.