Background
Phase III, open-label RCT (BMT CTN 0201). 551 patients with hematologic malignancies (AML, CML, MDS, ALL, NHL, CLL) undergoing allogeneic HCT from 8/8 HLA-matched unrelated donors. Question: does peripheral blood stem cell (PBSC) graft offer superior engraftment and OS over bone marrow (BM) with acceptable chronic GVHD risk? PBSC was widely adopted despite limited randomized data.
Interventions and follow up
Arm A (PBSC): G-CSF mobilized allogeneic PBSC from unrelated donors (target CD34+ ≥4×10⁶/kg)
Arm B (BM): unmanipulated bone marrow from unrelated donors (target nucleated cells ≥3×10⁸/kg)
Conditioning/GVHD prophylaxis: per institutional standard (tacrolimus + methotrexate predominant), identical between arms
Primary endpoint: 2-year overall survival
Median follow up: 36 months
Arm B (BM): unmanipulated bone marrow from unrelated donors (target nucleated cells ≥3×10⁸/kg)
Conditioning/GVHD prophylaxis: per institutional standard (tacrolimus + methotrexate predominant), identical between arms
Primary endpoint: 2-year overall survival
Median follow up: 36 months
Results
2-year OS: 46% (PBSC) vs 51% (BM) — HR 1.05, 95% CI 0.81–1.37, P=.67, NS
2-year graft failure: 3% (PBSC) vs 9% (BM) — P=.002
Neutrophil engraftment: 15 days (PBSC) vs 21 days (BM) — P<.001
2-year chronic GVHD: 53% (PBSC) vs 41% (BM) — P=.01
2-year graft failure: 3% (PBSC) vs 9% (BM) — P=.002
Neutrophil engraftment: 15 days (PBSC) vs 21 days (BM) — P<.001
2-year chronic GVHD: 53% (PBSC) vs 41% (BM) — P=.01
Adverse events
Acute GVHD: grade II–IV 67% (PBSC) vs 66% (BM), NS; grade III–IV 21% vs 20%, NS
Chronic GVHD: higher with PBSC 53% vs 41% (P=.01); extensive chronic GVHD 34% vs 25% (P=.06)
Non-relapse mortality: 2-year NRM 21% (PBSC) vs 23% (BM), NS
Chronic GVHD: higher with PBSC 53% vs 41% (P=.01); extensive chronic GVHD 34% vs 25% (P=.06)
Non-relapse mortality: 2-year NRM 21% (PBSC) vs 23% (BM), NS
Conclusions
PBSC and BM yielded equivalent 2-year OS in unrelated donor allo-HCT. PBSC offered faster engraftment and lower graft failure but significantly higher chronic GVHD. Graft source selection should weigh relapse risk (favoring PBSC for graft-versus-tumor) against chronic GVHD burden (favoring BM).
Key Limitations
OS equivalence masks competing risks: PBSC patients had more chronic GVHD but fewer graft failures, yielding neutral net survival. No MRD-stratified or disease-specific analyses were pre-specified. The PTCy era (BMT CTN 1703) has substantially changed GVHD prophylaxis, potentially narrowing the chronic GVHD gap between graft sources. Findings derive from a single GVHD-prophylaxis paradigm (predominantly Tac/MTX).
Clinical Context
BMT CTN 0201 established PBSC as a valid alternative to BM for unrelated donor allo-HCT, with the chronic GVHD trade-off as the key consideration. ASTCT guidance presents both as acceptable; most adult centers favor PBSC for faster engraftment and graft-versus-leukemia effect in high-risk disease, while BM is preferred for aplastic anemia and pediatric recipients. PTCy-based prophylaxis (BMT CTN 1703) partially mitigates PBSC chronic GVHD risk.