A validated prognostic classifier for BRAF-mutated metastatic colorectal cancer: the 'BRAF BeCool' study.

BACKGROUND

Despite the well-known negative prognostic value of the BRAF mutation in patients with metastatic colorectal cancer (mCRC), its outcome is quite heterogeneous, and the basis for this prognostic heterogeneity should be better defined.

METHODS

Two large retrospective series of BRAF-mutated mCRC from 22 institutions served as an exploratory and validation set to develop a prognostic score. The model was internally and externally validated.

RESULTS

A total of 395 BRAF-mutated mCRCs were included in the exploratory set. Performance status, CA19.9, lactate dehydrogenase, neutrophil/lymphocyte ratio, grading and liver, lung and nodal involvement emerged as independent prognostic factors for overall survival (OS). Two different scoring systems were built: a 'complete' score (0-16) including all significant covariates and a 'simplified' score (0-9), based only on clinicopathological covariates, and excluding laboratory values. Adopting the complete score, proportions of patients with a low (0-4), intermediate (5-8) and high (9-16) score were 44.7%, 42.6% and 12.6%, respectively. The median OS was 29.6, 15.5 (hazard ratio [HR] for intermediate vs low risk: 2.16, 95% confidence interval [CI]: 1.44-3.22, p < .001) and 6.6 months (HR for high vs low risk: 4.72, 95% CI: 2.72-8.20, p < .001). Similar results were observed also after adjusting for the type of first-line treatment and adopting the simplified score. The simplified prognostic score derived from the exploratory set was then applied to the validation set for external confirmation.

CONCLUSIONS

These scoring systems are based on easy-to-collect data and defined specific subgroups with relevant differences in their life expectancy. These tools could be useful in clinical practice, would allow better stratification of patients in clinical trials and may be adopted for proper adjustments in exploratory translational analyses.