Effect modification in network meta-analyses of treatments for relapsing refractory multiple myeloma (RRMM): systematic review, meta-analysis, and simulation

AimsNetwork meta-analysis (NMA) has been used in several systematic reviews on relapsing refractory multiple myeloma (RRMM). NMAs have been questioned on the basis that effect modification may invalidate the underpinning assumptions. We aimed to systematically review and meta-analyze the evidence for effect modification of hazard ratios (HRs) for overall survival (OS) and progression-free survival (PFS) with respect to refractory status and number of treatment lines. MethodsWe extracted stratified HR estimates from 42 phase 2 and 3 randomized controlled trials (RCTs). We tested for within-study effect modification and used meta-analyses to estimate ratios of hazard ratios (RHRs) across trial under assumptions that strongly favor the modification hypothesis. RHR estimates were used in simulations to estimate how many NMA results would be expected to differ in the presence versus absence of effect modification. ResultsMost (95%) publications could have reported stratified estimates but only 14% (OS) and 43% (PFS) did. Within-study evidence for effect modification is very weak (p > 0.05 for 47 of 49 sets of stratified estimates). The largest RHR estimated was 1.31 (95% CI 1.16-1.47), for the modifying effect of refractory status on HR for PFS. Simulations suggest that, in the worst case, effect modification would result in 4.48% (95% CI 4.42%-4.53%) of NMA estimates differing statistically significantly in the presence versus absence of effect modification. ConclusionsEffect modification is essentially undetectable in phase 2 and 3 trials. In the worst case, it is unlikely to affect more than about 5% of random-effects NMA estimates.