We read with great interest the article of Stintzing et al. published in Annals of Oncology on July 2012  and would make a few remarks.
The results of this unplanned retrospective analysis, carried out on 96 KRAS-mutated colorectal cancer patients, did not show any substantial difference in terms of response rate, progression-free and overall survival between anti-epidermal growth factor receptor- and anti-vascular endothelial growth factor-based treatments. These results may be just related to operative bias: for example, the similar response rate observed in the cetuximab and bevacizumab arms is explainable by the lack of independent radiological review and the limitation of standard radiological assessment for anti-angiogenic treatments, which may delay disease progression mainly through disease stabilization: in fact, biological response is often characterized by a reduction of lesions' density, without a substantial RECIST response .
Overall survival seemed to favour cetuximab over bevacizumab (22.7 versus 18.7 months), although this difference was not statistically significant (HR = 0.86, 95% CI, 0.55–1.35; P = 0.55). Since cetuximab arm was affected by a higher and earlier dropout rate, it is possible that lower median treatment duration in the cetuximab arm may have led to earlier initiation of effective bevacizumab-based second-line treatment. In fact, half of the patients treated with FOLFIRI plus cetuximab received subsequent bevacizumab-based regimens. Not surprisingly, a combination of bevacizumab and oxaliplatin-based regimens was shown to prolong overall survival in the second-line setting .
Even though the addition of cetuximab to FOLFIRI regimen seems to be at least not detrimental for progression-free survival in KRAS-mutated patients,  the apparent lack of benefit from the addition of bevacizumab could also derive from confounding factors and imbalance between prognostic factors rather than KRAS mutation itself. For example, the rate of patients undergoing surgery was twice as high in the cetuximab arm: although this is clearly not attributable to treatment (given the identical response rate in both the arms), a multimodality strategy may have improved progression-free and overall survival in patients treated with the cetuximab-based combination.
The authors stated that KRAS G13D mutations, found in 20% of cases, may be associated with poorer prognosis independently of the treatment arm. However, in the recently published pooled analysis of CRYSTAL and OPUS trials, KRAS G13D-mutated patients receiving first-line chemotherapy alone failed to show a statistically significant difference in terms of progression-free survival and overall survival, when compared with other KRAS-mutated subtypes .
We believe that the study of Stintzing et al. is particularly valuable because it stimulates the investigation on the predictive role of KRAS mutation in bevacizumab-based treatment. However, given the small sample and evidence of great benefit of bevacizumab independently of KRAS status,  the preliminary nature of these results add a few more to what was previously known and any conclusions about the present analysis should be drawn very carefully.