Use of Endpoints in Phase III Randomized Controlled Trials for Hematopoietic Stem Cell Transplantation over the Last 15 Years: A Systematic Review

Background: The use of objective endpoints is critical for the generalization and clinical implications of a study. Overall survival (OS) has traditionally been used as the gold standard for demonstrating the true clinical benefit of a therapy or intervention. Use of clinically meaningful pre-specified endpoints is essential for a successful clinical trial. In this systemic review, we aimed to investigate different primary and secondary endpoints used in phase III randomized controlled trials (RCTs) for hematopoietic stem cell transplantation (HCT), and their trends over a span of 15 years. Methods: Following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines, a comprehensive literature search was conducted on three databases (PubMed, Cochrane, and Clinical trials.gov) using MeSH terms and keywords for “hematopoietic stem cell transplantation” AND “Randomized Controlled Trials as Topic” from January 2006 to May 2021. We screened 2223 articles. After excluding duplicates, reviews, and irrelevant articles, 170 articles reporting only phase III RCTs with primary and secondary endpoints on HCT were included for our systematic review. Primary and secondary endpoints data were extracted from the included studies and the frequency of various endpoints as well as their yearly frequencies were calculated. Disease-free survival (DFS) was used to represent similar outcomes including event-free survival (EFS), progression-free survival (PFS), leukemia-free survival (LFS), and relapse-free survival (RFS). Results: Our study included 154 phase III RCTs on HCT. DFS was the primary endpoint in 40 (26%) studies, while OS and graft versus host disease (GvHD) were reported as primary endpoints in 40 (26%) and 13 (8%) studies, respectively. Infections, response rate, relapse rate (RR), toxicity/adverse events, quality of life (QoL), non-relapse mortality (NRM)/transplant-related mortality (TRM), GvHD-free survival, and hematological improvement (HI) were used as primary endpoints in 13 (8%), 13 (8%), 12 (8%), 8 (5%), 7 (4.5%), 7 (4.5%), 6 (4%), and 5 (3%) studies respectively. Other infrequently reported primary endpoints are listed in Table 1. Secondary endpoints followed a similar pattern as detailed in Table 2. OS (n=26, 40%), toxicity/adverse events (n=33, 21%), and TRM/NRM (n=33, 21%) were commonly reported secondary endpoints. DFS, GvHD, RR, QoL, infections, response rate, disease progression, and hematopoietic engraftment were us