Clinicopathological analysis of KRAS mutation in an Irish population.
467 Background: Clinical trials have shown that tumor KRAS status predicts response to anti-EGFR monoclonal antibodies in metastatic colorectal cancer (mCRC). We sought to compare distribution of wild-type and mutated KRAS in an Irish population and identify clinicopathologic correlates. Methods: From our prospectively maintained database we retrospectively identified patients with mCRC and documented KRAS status between Jan 2007 to June 2010. Medical notes were examined for patient demographics and disease characteristics. Variables were extracted and compared using unpaired t test and chi2 test. Results: 52 patients were identified, 29 (55.8%) with mutated (mt) and 23 (44.2%) with wild type (wt) KRAS from tumor tissue. Males accounted for 61.5% (n=31). Median age at diagnosis of metastatic disease for the KRASmt group was 66.4 years (range 56.7-82.1) and for the KRAS wt group was 64.2 years (40.1-76.8), p= 0.08. 21 (72.4%) of the KRASmt group and 16 (69.6%) of the KRASwt group had metastatic disease at presentation (p=0.81). For patients who presented initially with localised disease, time to development of metastases was 22.5mo (range 14-37.6) for the KRASmt group and 16.7mo (3.2-123.7) for the KRASwt (p=0.30). 21 (72.4%) of KRASmt and 17 (74.1%) of KRASwt tumors had left-sided primary (p = 0.84) with equal numbers of primary tumor resection in both groups. There was no statistical difference in TN-stage or the presence of liver metastases. Numbers of patients with KRASmt tumors with one, two and three or more different sites of metastases was 22 (75.9%), 5 (17.2%), and 2 (6.9%), and in KRASwt tumors there was no significant difference: 18 (78.3%), 4 (17.4%), and 1 (4.3%), respectively (p = 0.93). Median CEA at diagnosis for both groups were 19.7 μg/l (range 1.2-5958) and 6.1 (1.3-696.9; p = 0.11). Conclusions: We found that the ratio of KRASmt to KRASwt tumours in an Irish population is comparable to that in large international trials. Our analysis revealed no association between KRAS status and clinicopathologic variables. The inclusion of BRAF status, not readily available in our institution, may help define poor prognosis tumors. At present there is no validated molecular biomarker that is superior to standard clinicopathologic variables. No significant financial relationships to disclose.