Consensus molecular subtypes in metastatic colorectal cancer treated with sequential versus combined fluoropyrimidine, bevacizumab and irinotecan (XELAVIRI trial).

Abstract Background The XELAVIRI trial compared sequential (fluoropyrimidine and bevacizumab; irinotecan (Iri) at progression) versus initial combination therapy (fluoropyrimidine, bevacizumab, Iri) of treatment-naive metastatic colorectal cancer (mCRC). In the confirmatory analysis, the primary end-point (non-inferiority of sequential therapy regarding time to failure of strategy, TFS) was not met. Nevertheless, significant differences regarding treatment efficacy were observed according to RAS status. Here, we evaluate the consensus molecular subtypes (CMS) as additional biomarkers for sequential versus combination therapy. Material and methods Gene expression was measured using NanoString after mRNA extraction from formalin-fixed paraffin-embedded tumour specimens. CMS were predicted using multinomial regression and correlated with updated data for TFS, overall (OS) and progression-free survival. Results CMS were predicted in 337 of 421 (80.0%) patients (CMS1: 18.4%; CMS2: 51.6%; CMS3: 2.7%; CMS4: 27.3%). CMS2 together with RAS/BRAF wild-type status was identified as potential predictive marker of benefit from initial combination therapy for OS (HR 0.56, 95% CI 0.33–0.96, p = 0.036) and progression-free survival (HR 0.28, 95% CI 0.29–0.79, p = 0.004) and also trending in TFS (HR 0.63, 90% CI 0.41–0.95, p = 0.066). In patients with RAS-mutated mCRC, CMS1 was associated with longer OS after initial combination therapy (HR 0.43, 95% CI 0.20–0.95, p = 0.038). Interaction testing (two-sided) of CMS and RAS/BRAF status in favour of the combination treatment strategy was significant for OS (p = 0.012) Conclusions In patients with RAS/BRAF wild-type mCRC, CMS2 may serve as an additional biomarker of benefit from the initial combination therapy, including Iri. Trial registration Trial registration ID ( clinicaltrials.gov ) NCT01249638 .