Comparison of bone marrow and PBPC FOU allogeneic sct in myelodysplastic syndromes. A retrospective study on behalf of the EBMT

Recent studies, mainly in the setting of leukemia, have shown that peripheral blood progenitor cells (PBPC) could be used instead of bone marrow (BM) for allogeneic transplantation (SCT). This multicentric retrospective study compared results of 95 PBPC-SCT with those of 134 BM-SCT performed myelodysplastic syndromes using HLA identical siblings from 1995 to 2000. There were 229 patients with RA (n=64 , RAEB (n=70) or RAEB-t (n=95) at SCT. 67 patients were in complete remission, 100 had received no chemotherapy, and 62 were in more advanced phases. 174 kaiyotypes weie available, 52 with a poor prognosis. T cell depletion was used in 50 cases. The 2 grou[ s were comparable excepted for age at SCT (age < 30 years, 21% for BM vs 9% for PBPC, p=.02) and the use of TBI (55% for BM vs 31% for PBPC, p<.01). Median follow-up was 639 days (PBPC: 522 days vs BM: 726 days, p=.03). Multivariate analyses focused on 2-year outcomes. Engraftment occurred earlier with PBPC (median time to ANC.5 GA.: 14 vs 18 days, p<.01 - engraftment by day21: 68% with BM vs 90% with PBPC). The incidence of grade H-1V acute graft-versus host disease (GVHD) was similar whatever the type of graft (BM: 30% vs 33% for PBPC, p=.75), even after adjustment in a Cox model. There was also no difference when considering grade III-IV acute GVHD incidence (BV : 14% vs 15% for PBPC, p=.96). More patients developed chronic GVHD in the PBPC group (BM: 63% vs 39% for PBPC, p<.02). 2-year transplant-related mortality was not different (BM: 35% vs 41% with PBPC, p=.28), even in multivariate analysis. 2-yen relapse rate was lower with PBPC in univariate analysis (BM: 40% vs 14% with PBPC, p<.001) and in Cox models (RR: .23, 95%CI: .11-.50, p<.001 without cytogenetic - RF.: .21, 95%CI: .07-.61, p=.004 with cytogenetic). 2-year EPS was 50% with PBPC versvs 39% with BM (p=.23, non proportional hazard), and was significantly better with PBPC in Cox models, mainly after third month post-SCT (RR: .43, 95Á: .2S-.72, p=.002l. This result was also observed when cytogenetic was included in the model (RR: .3!i, 95%CI: .18-.70, p=.003). 2-year overall survival was 58% with PBPC versus 49%. PBPC were associated with an improved survival after 6 months post-SCT (RR: .39, 95%C : .18-.87, p=.02)in multivariate analysis. These results favored the use of PBPC for SCT in myelodysplastic syndromes as it is associated with a more rapid neutrophil recovery, a lower relapse rate, and improved event-free and overall survivals. A prospective randomized study is needed to confirmed these results