This study performed on 59 MCRC patients confirms that the presence of KRAS
mutation in tumour is highly predictive of a non-response to treatment based on cetuximab plus chemotherapy, as shown previously in a series of 30 patients (Lievre et al, 2006
). It is important to highlight that, in our series, the proportions of CR/PR, SD and PD patients were 20.3, 32.2 and 47.5%, respectively, and this distribution is similar to that reported in the randomised cetuximab trial (Cunningham et al, 2004
). The relationship between KRAS
status and sensitivity to anti-EGFR monoclonal antibodies had not been found previously by Moroni et al (2005)
in a series of 31 patients. This discrepancy might probably be explained, at least in part, by the limited number of patients in these latter series. Direct sequencing allowed us to detect a KRAS
mutation in 16 out of 59 patients (27%) with MCRC, and among the 43 patients without detectable KRAS
mutations, 15 presented a PD. We hypothesized that we missed some KRAS
mutations by direct sequencing of tumour DNA, since malignant tumours are genetically heterogeneous. Furthermore, it is important to highlight that our study was based on paraffin-embedded tumours from which it is more difficult to obtain high-quality DNA. Using two independent sensitive methods, respectively, based on SNaPshot and PCR-LCR assays specifically designed to detect KRAS
mutation, we detected additional mutations in two patients with SD, four with PD but none in 11 with CR/PR. This demonstrates the need to use highly sensitive molecular techniques to ensure detection in tumours of mutations conferring resistance to treatments. Considering the heterogeneity of tumour cells, sampling tissue is particularly important. We determined using SnaPshot assay the KRAS
status in two to three different areas of three tumours including two with KRAS
mutation. These analyses showed that the results did not differ according to the site of the analysis. The absence of KRAS
mutation in 11 out of 28 patients with PD in our series is probably explained by the fact that KRAS
mutation is not the only genetic alteration conferring resistance to anti-EGFR antibodies. Indeed somatic alterations hitting other downstream effectors of the EGFR transduction cascade, such as RAF
, may have a similar effect.
In conclusion, these results should prompt further studies on larger MCRC series to definitely establish the clinical relevance of KRAS mutation detection in anti-EGFR antibodies based on chemotherapy. They also highlight the need to use sensitive molecular methods to detect mutations conferring resistance and the two assays presented in this study should facilitate the detection of KRAS mutations in CRC, on a routine basis. A major criticism that should be made to all studies on predictive markers of clinical response to anti-EGFR agents in MCRC, including ours, is that the clinical response is evaluated on the metastatic disease whereas the presence of the molecular marker is assessed from the primary tumour. In the present study we had the opportunity to compare the KRAS mutational status between primary tumour and metastases in five patients whom samples were available. In these five patients, SnaPshot assay had indicated that a KRAS mutation was present in one case and absent in the remaining five cases. For these five patients, analysis of the corresponding metastatic site showed that the KRAS mutation status was identical between the primary tumour and metastases. Considering the genetic evolution of metastases compared to primary tumour, we think nevertheless that it will be important in the future to screen directly metastases for the presence of alterations conferring either sensitivity or resistance to these targeted therapies.