Donor Lymphocyte Infusions (DLI) After Peripheral Blood Stem Cell Transplantation. A Retrospective Analysis of 357 Patients by the Chronic Leukemias Working Party EBMT.

Abstract 3309

Poster Board III-197

Donor Lymphocyte Infusions (DLI) constitute a potent therapeutic option for treating relapse of chronic myelogenous leukemia (CML) after hematopoietic stem cell transplantation (SCT) inducing durable remissions in the majority of patients. A number of factors is known to influence the efficiency of DLI. A preliminary analysis of EBMT data had suggested that DLI efficiency might be inferior after peripheral blood stem cell transplantation (PBSCT) as compared to DLI following bone marrow transplantation (BMT) (Schmid et al. ASH, 2005). To control for a number of other factors that were not known at the time of the previous analysis, we repeated this analysis based on the results of 357 patients treated with DLI following PBSCT (N=108) and BMT (N=249). We limited the analysis to patients who relapsed after standard intensity conditioning SCT from HLA-identical family donors in first chronic phase of disease. The median age of patients was 39 years (range 18-60) with predominance of males (59%). 53% of patients with known data were CMV positive and in 44% of the patients there was a sex-mismatch with the stem cell donor. SCTs have been performed between 1994 and 2007 (median year: 1998) and the conditioning treatment included total body irradiation (TBI) in 68% and T cell depletion in 44% of patients. 92% of patients with known data achieved complete remission after SCT while grade II-IV acute GvHD occurred in 18% of patients and extensive chronic GvHD in 17% of patients. Median time to relapse was 17 months (range 0.6-129) and median time from SCT to first DLI infusion was 23 months (range 0.6-142). Looking at the patients with known data at the time of first DLI infusion, the relapse could be classified as molecular/cytogenetic in 63%, hematologic in 27% and transformed in 10% of patients. The median year of first DLI was 2000 ranging from 1995 to 2007. As the initial DLI infusion, 9% of patients received <1×10e6, 62% 1.1-10×10e6 and 29% received >10×10e6 CD3+ cells/kg. However, the comparative analysis of groups based on the stem cell source revealed that the group of patients transplanted with PBSCs included significantly more males (68 vs. 56%), were older (median age 42 vs. 39) and underwent more frequently T cell depletion at SCT (72 vs. 34%,). PBSCTs have been performed more recently (median year 1999 vs. 1997) and both duration of remission and time from SCT to first DLI were shorter after PBSCT (median duration: 12 vs. 21 months and 14 vs. 26 months respectively). The initial cell dose in patients from PBSCT arm was significantly lower than in BMT group (≤10×10e6 CD3+/kg in 89% vs. 65% of patients). Similarly to the previous study, we also observed a trend towards superior overall survival after DLI in BMT group compared to PBSCT group, especially in the early post-transplant period. The actuarial probability of survival at five years from DLI was 77% in PBSCT group and 79% in BMT group. However the differences were not statistically significant (p=0.77). The source of stem cells did not influence the occurrence of molecular/cytogenetic remissions after DLI (80% vs. 77%) grade II-IV acute GVHD (16% vs. 16%), chronic GVHD (23% vs. 30%) and myelosuppression (10% vs. 16%). In order to search for factors having impact on survival of analyzed patients, we performed both univariate and multivariate survival analyses. The univariate analysis revealed that interval from SCT to DLI longer than 2 years (p=0.001), date of DLI after 2000 (p=0.026) and molecular/cytogenetic stage of relapse at DLI (p<0.001) were associated with favorable survival. Similarly, the multivariate Cox analysis identified interval between SCT and DLI (HR= 0.50, CI: 0.3-0.8; p=0.01 for after 2 years), date of DLI (HR=0.63, CI:0.4-1.0; p=0.07 for after 1999), and stage of relapse (HR=2.8, CI:1.2-6.5; p=0.02 for HemCR and HR=3.6, CI:1.8-7.0; p<0.001 for missing data group), but not for stem cell source (HR=0.95, CI 0.56-1.6 ;p=0.86) as independent factors affecting survival. Based on our retrospective data from EBMT registry covering a period of 14 years of SCT and DLI, it seems that the PBSCT does not affect the efficiency of DLI compared to BMT. Therefore, keeping all limitations of a retrospective analysis in mind, it seems that differences in efficacy of DLI do not influence the decision whether PBSC or BM should be used as stem cell source for allogeneic SCT in CML in first chronic phase.