Systematic review reveals urgent need to homogenize endpoints choices and definitions in marginal zone lymphomas trials.

Marginal zone lymphoma (MZL) is a heterogeneous disease and has various end-point measures. Our aim was to describe the endpoints used in trials involving patients with MZL. We searched over the last 35 years via PubMed, The Cochrane Library, clinicaltrials.govandclinicaltrialsregister.eu for published and registered clinical trials using the keyword "marginalzone lymphoma." We excluded studies focusing on pediatric populations, cutaneous MZL and on use of allogenic stem cell transplant. Endpoints were reviewed as well as their influencing factors and their definitions. Among 1192 references Q7 dentified by initial screening, 309 references were included (111 published, 198 registered), with 213 (69%) phase 2, 65 (21%) phase 1/2 and 31 (10%) phase 3 trials. The majority were open-label (n¼295, 95%) non-randomized (n¼256, 83%) trials, concerned all subtypes of MZLs at once (n¼239, 77%), and were often merged with non-MZL patients (n¼232, 75%). Among phase 1/2 and 2 trials, Overall/complete response rate (ORR/CRR) (n¼196, 70.5%) and progression-free survival (PFS,n¼28, 10.1%) were the most used primary endpoints; in phase 3 trials PFS was the most used primary endpoint (n¼18, 58.1%; ORR/CRR n¼6, 19.4%, p<0.001). Overall, the most frequent secondary endpoints were overall survival (OS, n¼153, 50%), PFS (n¼142, 46%) and ORR/CRR (n¼116, 38%). Distribution was similar when considering trials with only patients with MZL. Endpoints definitions were inconsistent across published trials (up to 9 definitions per endpoint). Trials involving patients with MZL showed marked heterogeneity both in the choice and definitions of primary and secondary endpoints, thus hampering comparability between trials.