BRAF mutation increased mortality risk in melanoma, colorectal cancer patients

Jose Lutzky, MD

Nearly half of melanomas and about 10% of colorectal carcinomas have been found to harbor an activating mutation in the BRAF gene. The role that these mutations play in driving the proliferative stimulus in a given malignancy will be influenced by the configuration of the remainder of the tumoral molecular milieu.

Therefore, a BRAF V600 mutation in melanoma appears to be a driving event, and inhibition of the mutated protein has been shown to result in clinically important survival benefit. On the other hand, the same mutation in colorectal carcinomas has not been shown to affect survival.

The well-executed meta-analysis by Arekani et al confirms the contention of smaller studies that a BRAF mutation increases the risk of mortality in colorectal cancer by a factor of 2.2 and in melanoma by a factor of 1.7. The study does not take into account the effect of the advent of BRAF inhibitors for melanoma in the clinical arena, which is likely to attenuate this effect. This study provides provocative data that suggests that BRAF mutation status should be incorporated as a prognostic factor in future studies of these malignancies. Its interaction with other known prognostic factors should be prospectively evaluated in such studies.

Jeffrey S. Weber, MD, PhD

The review by Ardekani et al on the survival outcome of patients with either colorectal cancer or melanoma that are BRAF mutated encompasses a total of seven published studies in melanoma. Three of the studies are clinical intervention trials that contained small numbers of patients, although reference 23 by Flaherty et al from 2010 was a phase 1/2 study of vemurafenib, the FDA approved mutated BRAF inhibitor that has been shown to prolong survival in BRAF mutated patients with metastatic melanoma, so its interpretation with respect to overall survival would be compromised. In the trial of reference 18,432 primary melanomas from China were assessed for a BRAF mutation. The median follow-up period was 24.0 (range: 3.0–229.0) months. The median survival time for patients with BRAF mutations (33.0 months) was significantly shorter than that for patients with wild-type tumors (53.0 months, P=.005), although there was an association of BRAF status with ulceration. There was no indication that a multivariate analysis was performed, yet these are the strongest data supporting the idea that BRAF mutation is a negative prognostic factor in melanoma.

The data from reference 72 from 2003 employed only 38 metastatic tumors, for which BRAF mutation status was not associated with survival, with a P=.13. Fifty-one primary nodular tumors were subsequently assessed in 2005 by the same group from reference 72, demonstrating no prognostic importance for BRAF mutation status in Akslen LA. J Invest Dermatol. 2005;125:312-317. Ninety-seven tissue biopsies from patients with stages III and IV melanoma that were tested for BRAF mutations in 2007 from the same research group did not suggest a prognostic role for the BRAF mutation in a multivariate analysis, in Ugurel S. PloS One. 2007;doi:10.1371/journal.pone.0000236.

In reference 73, the median overall survival time from diagnosis of first distant metastases in the complete cohort of patients, according to BRAF mutation status and treatment with a BRAF inhibitor, was 46.1 months for BRAF wild-type patients and 11.1 months for BRAF-mutant patients who had no treatment with a BRAF inhibitor; P=.006. This would suggest that, in fact, BRAF mutation status was a negative prognostic factor, although it is colored by the fact that noninhibitor treated patients in that analysis of 197 tumors probably had a poor performance status that prohibited their entry to a BRAF inhibitor trial, and thus had a worse outcome that might be unrelated to BRAF status. That same group from reference 73 just published a follow-up study in which 308 metastatic tumors were analyzed for BRAF, in which the outcome for the mutated patients who received a BRAF inhibitor was significantly better than for those who were BRAF wild-type or for those who were mutated but never received a BRAF inhibitor, in Menzies AM. Clin Cancer Res. 2012;18:3242-3249. Nonetheless, there was no difference between the survival of BRAF wild type or BRAF mutated patients who were not treated with a BRAF inhibitor, with the same limitations as in reference 73.

In reference 71, 62 patients were treated in a phase 2 study with a combination of bevacizumab and temozolomide, and retrospective analysis of BRAF mutation status was undertaken. OS (12.0 versus 9.2 months; P=.014) was higher in BRAF V600E wild-type patients in that study, but given that patients were all treated with a regimen that might or might not have some clinical activity, or an interaction with BRAF mutation status, the actual interpretation is difficult, and cannot directly support the notion that BRAF mutation status is a prognostic factor in melanoma.

So, where does that leave us in assessing the available data suggesting that a BRAF mutation in either a primary or a metastatic lesion is a negative prognostic factor in melanoma independent of other prognostic factors, and independent of treatment? The data are suggestive in melanoma, but hardly definitive. Given the paucity of available cohort studies with long-term follow-up and reliable survival data, as well as the contradictory information noted, I would conclude that the jury is still out on the negative prognostic value of having a BRAF V600 mutation in melanoma.

It certainly is true that having a BRAF V600E/K mutation is a predictive marker for clinical response to BRAF inhibition, but I am unable to agree with the authors’ statement that “In addition, we revealed that BRAF-V600E mutation also significantly increases the risk of mortality in melanoma patients.”