Background
Phase III, open-label, 3-arm RCT (BMT CTN 1703). 431 patients with hematologic malignancies undergoing HLA-matched related (MRD) or unrelated (MUD) donor PBSC allo-HCT after RIC. Standard GVHD prophylaxis (tacrolimus + methotrexate or CD34+ graft selection) compared to post-transplant cyclophosphamide-based (PTCy) prophylaxis, extending the haploidentical Baltimore protocol to matched donors.
Interventions and follow up
Arm A (PTCy/Tac/MMF): cyclophosphamide 50mg/kg IV days +3 and +4 + tacrolimus (target 5–15 ng/mL days +5–180) + mycophenolate mofetil 15mg/kg TID days +5 to +35
Arm B (Tac/MTX): tacrolimus + methotrexate 15mg/m² day +1, 10mg/m² days +3, +6, +11
Arm C (CD34+ selection): ex vivo T-cell depletion via CD34+ positive selection (≥5×10⁶ CD34+/kg target)
Primary endpoint: GVHD-free, relapse-free survival (GRFS) at 1 year
Median follow up: 22 months
Arm B (Tac/MTX): tacrolimus + methotrexate 15mg/m² day +1, 10mg/m² days +3, +6, +11
Arm C (CD34+ selection): ex vivo T-cell depletion via CD34+ positive selection (≥5×10⁶ CD34+/kg target)
Primary endpoint: GVHD-free, relapse-free survival (GRFS) at 1 year
Median follow up: 22 months
Results
1-year GRFS (PTCy vs Tac/MTX): 52.7% vs 34.9% — HR 0.64, 95% CI 0.48–0.86, P=.003
1-year GRFS (PTCy vs CD34+): 52.7% vs 35.5% — HR 0.69, 95% CI 0.51–0.93, P=.02
Grade II–IV acute GVHD: 18.3% (PTCy) vs 35.3% (Tac/MTX) vs 22.6% (CD34+)
1-year chronic GVHD: 16.3% (PTCy) vs 42.4% (Tac/MTX) vs 12.5% (CD34+)
1-year OS: 85.1% vs 80.8% vs 83.7% — NS
1-year relapse: 12.8% vs 14.9% vs 22.7% (CD34+ numerically higher)
1-year GRFS (PTCy vs CD34+): 52.7% vs 35.5% — HR 0.69, 95% CI 0.51–0.93, P=.02
Grade II–IV acute GVHD: 18.3% (PTCy) vs 35.3% (Tac/MTX) vs 22.6% (CD34+)
1-year chronic GVHD: 16.3% (PTCy) vs 42.4% (Tac/MTX) vs 12.5% (CD34+)
1-year OS: 85.1% vs 80.8% vs 83.7% — NS
1-year relapse: 12.8% vs 14.9% vs 22.7% (CD34+ numerically higher)
Adverse events
Hematologic/organ: PTCy arm had higher grade ≥3 cytopenias around days +3 to +5 (expected, cyclophosphamide-related); engraftment delayed median 1–2 days
Infections: grade ≥3 infections similar across arms (~35–40%)
Non-relapse mortality: no excess NRM with PTCy
Infections: grade ≥3 infections similar across arms (~35–40%)
Non-relapse mortality: no excess NRM with PTCy
Conclusions
PTCy-based prophylaxis significantly improved GRFS vs both Tac/MTX and CD34+ selection in HLA-matched donor PBSC allo-HCT, driven by substantially lower acute and chronic GVHD without compromising OS or relapse. This established PTCy/Tac/MMF as the standard of care for GVHD prophylaxis in matched-donor allo-HCT.
Key Limitations
Median follow-up 22 months — OS and relapse require longer-term confirmation. GRFS is a composite endpoint; GVHD improvements may not translate to OS benefit. PTCy requires 2-day cyclophosphamide infusion with hydration, adding logistical complexity; CD34+ selection needs specialized infrastructure. Conducted in matched-donor PBSC only — extrapolation to BM grafts or mismatched donors requires separate evidence.
Clinical Context
BMT CTN 1703 represents the most significant advance in GVHD prophylaxis in decades. PTCy/Tac/MMF is now the preferred prophylaxis regimen for matched-donor PBSC allo-HCT per ASTCT practice, widely adopted since 2023, extending the PTCy approach from the haploidentical setting to the matched-donor standard. Ongoing trials evaluate PTCy combined with newer agents (abatacept, ruxolitinib) for further GVHD reduction.