Clinical prognostic score of BRAF V600E mutated (BM) metastatic colorectal cancer (mCRC): Results from the “BRAF, BeCool” platform.

Abstract

639 Background: BM mCRCs account around 10% of mCRCs, with a dismal prognosis. Intra-group heterogeneity has been recently studied, however, an extensive prognostic clinical analysis has not been provided yet. Methods: BM mCRCs tested on primary tumor (T) or on metastases (mets) were collected from 13 Italian Oncology Units. A prognostic score was derived by an internal cross-validation procedure: the whole population was splitted in a training (67%) and in a testing (33%) sample; this process was repeated 10 times. Primary endpoint was OS. Multivariate analysis (MA) was performed on each training sample, and covariates with independent prognostic value were included in the scoring system, assigning rounded scores to the covariates. Results: A total of 395 BM mCRC patients were included. At MA, independent prognosticators for OS were ECOG PS (1 vs 0; 2-3 vs 0), Ca19.9 (high vs normal); LDH (≥300 vs low); neutrophil/lymphocyte ratio ( > 3 vs low); T grading (3-4 vs 1-2); liver mets (yes vs no); lung mets (yes vs no); lymphnode mets (yes vs no). Two different scoring systems were built: a «complete» score (0-18) with all significant covariates; a «simplified» score (0-11), selecting only significant clinico-pathological covariates, excluding laboratory values. With «complete» score, proportion of patients with low (0-4), intermediate (5-8) and high (9-18) score was 39%, 46% and 15%, respectively. Median OS was 27.6, 18.7 (HR interm. vs low 1.89, 95%CI 1.25 – 2.86, p = 0.003) and 6.6 months (HR high vs low 4.95, 95%CI 2.89 – 8.47, p < 0.0001), respectively. Similar results were observed adjusting for type of first-line treatment. Median PFS was 11.1, 8.6 (HR interm. vs low 1.36, 95%CI 0.94 – 1.97, p = 0.11) and 4.1 months (HR high vs low 3.50, 95%CI 1.98 – 6.20, p < 0.0001). Similar results were obtained with the «simplified» score. Conclusions: A simple and ready-to-use prognostic score for BM mCRC was developed. The robustness of the internal cross-validation justifies the effort of a validation study. Identification of specific prognostic factors in distinct molecular subgroups is crucial for adjusting exploratory translational analyses and future clinical trial design.