Prevention of graft-versus-host disease (GvHD) without malignant relapse is the overall goal of allogeneic haemopoietic cell transplantation (HCT). We aimed to evaluate regimens using either maraviroc, bortezomib, or post-transplantation cyclophosphamide for GvHD prophylaxis compared with controls receiving the combination of tacrolimus and methotrexate using a novel composite primary endpoint to identify the most promising intervention to be further tested in a phase 3 trial.

In this prospective multicentre phase 2 trial, adult patients aged 18–75 years who received reduced-intensity conditioning HCT were randomly assigned (1:1:1) by random block sizes to tacrolimus, mycophenolate mofetil, and post-transplantation cyclophosphamide (cyclophosphamide 50 mg/kg on days 3 and 4, followed by tacrolimus starting on day 5 and mycophenolate mofetil starting on day 5 at 15 mg/kg three times daily not to exceed 1 g from day 5 to day 35); tacrolimus, methotrexate, and bortezomib (bortezomib 1·3 mg/m² intravenously on days 1, 4, and 7 after HCT); or tacrolimus, methotrexate, and maraviroc (maraviroc 300 mg orally twice daily from day −3 to day 30 after HCT). Methotrexate was administered as a 15 mg/m² intravenous bolus on day 1 and 10 mg/m² intravenous bolus on days 3, 6, and 11 after HCT; tacrolimus was given intravenously at a dose of 0·05 mg/kg twice daily (or oral equivalent) starting on day −3 (except the post-transplantation cyclophosphamide, as indicated), with a target level of 5–15 ng/mL. Tacrolimus was continued at least until day 90 and was tapered off by day 180. Each study group was compared separately to a contemporary non-randomised prospective cohort of patients (control group) who fulfilled the same eligibility criteria as the trial, but who were treated with tacrolimus and methotrexate at centres not participating in the trial. The primary endpoint (GvHD-free, relapse-free survival [GRFS]) was defined as the time from HCT to onset of grade 3–4 acute GvHD, chronic GvHD requiring systemic immunosuppression, disease relapse, or death. The study was analysed by modified intention to treat. The study is closed to accrual and this is the planned analysis. This trial is registered with ClinicalTrials.gov, number NCT02208037.

Between Nov 17, 2014, and May 18, 2016, 273 patients from 31 US centres were randomly assigned to the three study arms: 89 to tacrolimus, methotrexate, and bortezomib; 92 to tacrolimus, methotrexate, and maraviroc; 92 to tacrolimus, mycophenolate mofetil, and post-transplantation cyclophosphamide; and six were excluded. Between Aug 1, 2014, and Sept 14, 2016, 224 controls received tacrolimus and methotrexate. Controls were generally well matched except for more frequent comorbidities than the intervention groups and a different distribution of types of conditioning regimens used. Compared with controls, the hazard ratio for GRFS was 0·72 (90% CI 0·54–0·94; p=0·044) for tacrolimus, mycophenolate mofetil, and post-transplantation cyclophosphamide, 0·98 (0·76–1·27; p=0·92) for tacrolimus, methotrexate, and bortezomib, and 1·10 (0·86–1·41; p=0·49) for tacrolimus, methotrexate, and maraviroc. 238 patients experienced grade 3 or 4 toxicities: 12 (13%) had grade 3 and 67 (73%) grade 4 events with tacrolimus, mycophenolate mofetil, and post-transplantation cyclophosphamide; ten (11%) had grade 3 and 68 (76%) had grade 4 events with tacrolimus, methotrexate, and bortezomib; and 18 (20%) had grade 3 and 63 (68%) had grade 4 events with tacrolimus, methotrexate, and maraviroc. The most common toxicities were haematological (77 [84%] for tacrolimus …