Abstract
Use of reduced intensity conditioning (RIC) and nonmyeloablative (NMA) regimens in allogeneic transplantation (allo HCT) for AML and MDS has rapidly increased over the past decade. Efficacy of these approaches compared to myeloablative (MA) conditioning has not been clearly established. We compared disease status, donor, graft and recipient characteristics and outcomes of 3731 MA HCT with 1500 RIC/NMA procedures performed at 217 centers between 1997 and 2004 (median f/u 58 vs. 40m, MA vs. RIC/NMA). Patients in the MA group were younger (median age 42 (18–68) vs. 55 (18–69) years), more frequently in CR1, and more often were transplanted from matched sibling grafts, compared to recipients of RIC/NMA. NMA recipients were less often transplanted with >10% blasts and had a lower performance scores. RIC recipients received fully matched unrelated donor (URD) grafts and PBSC as hematopoietic source more often than MApatients. Proportion with adverse cytogenetics, IPSS categories for MDS patients, median time to first relapse in AML CR2 patients, and marrow blasts at HCT were similar in the conditioning groups. 5-years univariate probabilities and multivariate relative risk (RR) outcomes of relapse, transplant-related mortality (TRM), leukemia-free (LFS) and overall survival (OS) are shown below:
| Outcome: . | MA . | RIC PBSC . | RIC BM . | NMA . |
|---|---|---|---|---|
| * overall p<0.001 | ||||
| Univariate: | ||||
| Relapse, % | 32 (31–34) | 39 (35–43) | 42 (36–48) | 43 (38–48) |
| TRM,% | 34 (32–36) | 34 (30–38) | 37 (31–43) | 36 (31–41) |
| Adjusted LFS, % | 33 (31–34) | 30 (26–35) | 28 (23–34) | 24 (20–29) |
| Adjusted OS, % | 34 (33–36) | 33 (29–37) | 32 (27–38) | 26 (21–30) |
| Multivariate: | ||||
| Relapse* | 1.00 | 1.05 (0.92–1.21) | 1.51 (1.23–1.85) | 1.64 (1.38–1.94) |
| TRM* | 1.00 | 1.26 (1.03–1.54) | 1.36 (1.01–1.83) | 1.55 (1.20–1.99) |
| Treatment | ||||
| 1.00 | 1.03 (0.90–1.17) | 1.17 (1.01–1.36) | 1.28 (1.12–1.45) | |
| failure* | ||||
| Mortality* | 1.00 | 1.03 (0.91–1.16) | 1.12 (0.96–1.30) | 1.20 (1.05–1.36) |
| Outcome: . | MA . | RIC PBSC . | RIC BM . | NMA . |
|---|---|---|---|---|
| * overall p<0.001 | ||||
| Univariate: | ||||
| Relapse, % | 32 (31–34) | 39 (35–43) | 42 (36–48) | 43 (38–48) |
| TRM,% | 34 (32–36) | 34 (30–38) | 37 (31–43) | 36 (31–41) |
| Adjusted LFS, % | 33 (31–34) | 30 (26–35) | 28 (23–34) | 24 (20–29) |
| Adjusted OS, % | 34 (33–36) | 33 (29–37) | 32 (27–38) | 26 (21–30) |
| Multivariate: | ||||
| Relapse* | 1.00 | 1.05 (0.92–1.21) | 1.51 (1.23–1.85) | 1.64 (1.38–1.94) |
| TRM* | 1.00 | 1.26 (1.03–1.54) | 1.36 (1.01–1.83) | 1.55 (1.20–1.99) |
| Treatment | ||||
| 1.00 | 1.03 (0.90–1.17) | 1.17 (1.01–1.36) | 1.28 (1.12–1.45) | |
| failure* | ||||
| Mortality* | 1.00 | 1.03 (0.91–1.16) | 1.12 (0.96–1.30) | 1.20 (1.05–1.36) |
MA regimens were associated with significantly less relapse (RR with RIC/NST 1.32 (0.19–1.46), p <0.001). While early TRM was substantially less with RIC/NMA approaches, 5 year TRM is equivalent (p=0.48). This leads to marginally better 5 year LFS (RR 1.08, p=0.05) but similar OS (RR 1.05, p=0.25) with MA vs. RIC/NST. These data indicate that late TRM negates any early advantage offered by RIC and NMA regimens, resulting in similar but slightly better 5 year LFS in recipients of MA conditioning. Similar outcomes with these RIC/NMA vs. MA regimens suggest the need for prospective trials comparing these transplant intensities in patients eligible for either approach.
Disclosures: No relevant conflicts of interest to declare.
Author notes
Corresponding author
