Natalia Rodon1, Ruth Roman1, Montse Verdu1,3 ,Miquel Calvo4, Beatriz Garcia-Pelaez1, Marta Gonzalez1, Carme Pubill3 and Xavier Puig1,2,3.
1BIOPAT. Biopatologia Molecular SL, Grup Assistència. 2Hospital de Barcelona-SCIAS, Grup Assistencia. 3Histopat Laboratoris. 4Departament d’Estadística, Facultat de Biologia. UB. Barcelona.
KRAS oncogenic mutations, present in approximately 30-50% of colorectal carcinomas (CRC), result in a lack of response to EGFR inhibitors. The aim of this study was to confirm with a multivariate analysis the correlations observed in our initial univariate approach between KRAS mutational status and clinicopathological parameters.
A cohort of 308 patients with primary CRC was used for this retrospective study. Specimens were routinely fixed and embedded in paraffin. The histopathological features evaluated included as cathegorical variables: gender, tumor location, tumor configuration, extent of invasion (pT), intramural and extramural thin-walled vessel invasion (TWVI), venous vessel invasion (VVI), perineural invasion (PNI), growth pattern, peritumoral Crohn-like reactivity, presence of tumor infiltrating lymphocytes (TIL), presence of residual adenomas, cytological grade, V600E BRAF mutation, P53 mutations and microsatellite instability (MSI) using a panel of 11 microsatellites. The numerical variables evaluated were: age, tumor size, number of affected lymph nodes, percentage of solid, mucinous, cribriform, micropapillary and microglandular patterns as well as expression of Mismatch Repair (MMR) genes, Ki-67 and p53 by immunohistochemistry.
The statistical selection procedure used to eliminate unrelated variables of the multivariate model is based on the glmnet algorithm and a resampling cross-validation approach. Only P53 mutations, overexpression of P53 and V600E BRAF were found finally associated with KRAS status. We also confirmed, as published, that KRAS and V600E BRAF mutations are mutually exclusive. Confirming our previous results this study supports the hypothesis of P53 inactivation and KRAS mutations arising through independent pathways. The association of KRAS status and solid or mucinous percentage seen in our initial univariate study was not confirmed. The lack of association between KRAS and high grade parameters such as pT, tumour size and presence of lymph node metastasis, sustains the fact that KRAS mutations are an early event in colorectal carcinogenesis.