Sir,

We appreciate the comment from Dr Vauthey and colleagues (Vauthey et al, 2012) on our recently published article by Kawamoto et al (2012) titled ‘KRAS mutations in primary tumours and post-FOLFOX metastatic lesions in cases of colorectal cancer’. The KRAS mutation has been regarded as one of the major ‘driver mutations’ of colorectal cancer, and its frequency is high (approximately 40%) regardless of the Dukes stage (Andreyev et al, 1998). However, the prognostic and predictive usefulness of KRAS mutations is still controversial (Karapetis et al, 2008; Van Cutsem et al, 2009; Yoshino et al, 2012). Although the suggestion raised by Dr Vauthey that adjuvant FOLFOX may provide a selection pressure favouring a chemotherapy-resistant subset enriched for KRAS mutation is interesting, we should consider it carefully. We closely examined the seven cases of stage III colorectal cancer (metastases appeared ‘after’ FOLFOX therapy) in our data set and found that only two cases (28.6%) harboured KRAS mutations. Meanwhile, among the 14 cases of stage IV cancer (metastases appeared ‘before’ FOLFOX therapy), 10 cases (71.4%) harboured KRAS codon 12 mutations and 2 additional cases had KRAS A146V or NRAS Q61H mutations in both primary and metastatic lesions. Therefore, we could not conclude that KRAS mutations were enriched in the cases that relapsed ‘after’ adjuvant FOLFOX therapy in our data set.

However, we would like to emphasise that the frequency of KRAS mutations among the cases with ‘resectable’ metastases after FOLFOX therapy reported in our article was quite high. As we reported previously, the KRAS mutation frequency of oxaliplatin-refractory ‘unresectable’ metastatic colorectal cancer in our institute was 70 out of 159 (44.0%), which is equivalent to the rates described in the previous reports (Bando et al, 2011). We cannot deny that the mutant KRAS affected the biological features of metastasised tumours, and we plan further investigations.