Upfront vs deferred monoclonal antibodies in metastatic colorectal cancer: A target trial emulation using the GEMCAD 14-01 prospective cohort.

Abstract

3597 Background: There are no randomized trials comparing the addition monoclonal antibodies (MAB: bevacizumab, cetuximab, panitumumab) to first line chemotherapy (upfront use) versus deferring their addition to the second-line chemotherapy (deferred use) in pts with metastatic colorectal cancer (mCRC). We emulated a target trial comparing upfront vs deferred use of MAB using the GEMCAD 14-01 observational registry. Methods: We first specified the (hypothetical) target trial to fully articulate the research question and then emulated it using real-world data. The eligibility criteria of the target trial were a diagnosis of mCRC, being treatment naïve, and a ECOG PS <2. The target trial would randomize pts to the following strategies: (1) initiation of MAB within 2 months of starting first line chemotherapy (“upfront MAB”) and (2) initiation of MAB within 2 months of starting second line chemotherapy (“deferred MAB”). The primary outcome of the target trial would be overall survival and the causal contrast (or estimand) would be the effect under complete adherence. We emulated this target trial using data from the GEMCAD 1401 registry (ClinicalTrials.gov identifier: NCT02254941), which collected data prospectively from 47 Spanish centers from June 2014 to June 2018. The emulation used the same definitions of eligibility criteria and treatment strategies, and classified individuals according to their baseline data using clones. The effect under complete adherence was estimated by censoring pts when they deviated from the assigned treatment strategy and by using time-varying weights to adjust for baseline and post-baseline confounding. Results: A total of 627 pts were eligible in the ‘’upfront MAB’’ and 397 pts in the ‘’deferred MAB’’. Median age was in the ‘’upfront vs deferred’’ 64.6 (interquartile range: 56-71) vs 67.8 (61-75) years, 96% vs 87% had an ECOG 0-1, 80% vs 79% had a Charlson score < 3, 46% vs 60% had a RAS mutation, 6% vs 4% had a BRAF mutation, 72% vs 69% had left side primary location, 74% in both strategies had liver metastasis and 42% vs 50% had LDH levels above the normal threshold. Pts in the “upfront MAB” group contributed a total 16,057 months of follow-up and 502 if them died. Pts in the “deferred MAB” contributed a total of 7,774 months of follow-up, and 222 of them died. The 48-month overall survival was 26.7% (95% CI 21.4-38.2%) in the “upfront MAB” group and 21.6% (14.2-41.7%) in the “deferred MAB” group, corresponding to a 48-month survival difference (“upfront MAB” is the reference) of -5.0% (95% CI -17.9 to 17.95) and a hazard ratio or 1.15 (0.88-1.40). Conclusions: Our study suggests little or no survival detrimental effect of deferring the use of MAB to the second line of treatment compared with the use of MAB as part of the first line of treatment among pts with mCRC. Clinical trial information: NCT02254941 .