Bone Marrow Mixed Chimerism Is a Risk Factor for Relapse After T-Cell Depleted Allogeneic Stem Cell Transplant

Abstract

AbstractAbstract 4145 Background:Full donor chimerism following non myeloablative unmodified stem cell transplants (SCTs) is important for disease control and SCT outcome. The impact of mixed chimerism (MC) after myeloablative unmodified SCT remains controversial, and there are limited reports about its prognostic significance after T-cell depleted (TCD) SCT. Treatment of MC, including reduction of immunosuppression or infusion of donor leucocytes (DLI), can lead to the development of graft versus host disease (GVHD). In order to optimize the graft vs leukemia (GVL) effect of a TCD allograft while minimizing GVHD risk, we sought to determine the significance of MC on TCD SCT outcomes and to identify a subgroup that would benefit from intervention. Methods:From 2000–2010, 447 adult patients with acute myeloid and lymphocytic leukemia (AML, ALL) and myelodysplastic syndrome (MDS) underwent bone marrow (BM) or peripheral blood stem cell (PBSC) TCD SCT. Conditioning regimens were total body irradiation (TBI)/Thiotepa with Cyclophosphamide or Fludarabine, Clofarabine/Melphalan/Thiotepa or Busulfan/Melphalan/Fludarabine. Like-gender donor-recipient chimerism was analyzed using semi-quantitative PCR for polymorphisms; PCR and/or cytogenetics were used for unlike gender pairings. Relapse was cytogenetic or hematologic evidence of disease. BM chimerism and disease status were evaluated at approimately 6, 12, 18 and 24 months. BM chimerism and outcome data were reviewed to determine whether 6 month BM MC (defined as < 95% donor chimerism) was associated with 2 year post-SCT relapse, transplant-related mortality (TRM) and disease-free survival (DFS). The 6 month landmark was chosen as a reasonable time point for intervention, such as DLI, if needed. Patients who received DLI prior to landmark were excluded. Patient and SCT characteristics were compared using Wilcoxon’s rank-sum test or Fisher’s exact test. DFS was estimated using Kaplan-Meier methods, and time-to-relapse and TRM were estimated using cumulative incidence functions. Results:300 of the 447 patients survived, were disease free, and had BM chimerism data at 6 months. Median age was 50.6 years. Most patients received PBSC grafts (89%). 131 (44%) and 9 (3%) related, and 86 (29%) and 74 (24%) unrelated donors, were matched and mismatched, respectively. Only patients in remission or with low level disease (BM blasts < 10%) were transplanted with TCD grafts. The majority of patients were conditioned with Busulfan/Melfalan/Fludarabine and TBI/Thiotepa/Fludarabine, 45% and 32%, respectively. 251 pts (84%) received antithymocyte globulin. 25% demonstrated MC in the BM at 6 mos. The majority (60%) with MC had a donor component of 75–95%. Diagnosis was not associated with MC at 6 months. Stem cell source and degree of HLA-match had no impact on BM chimerism. There was a significant difference in chimerism based on donor-recipient gender pairing (P = 0.04): male grafts into female patients having the greatest risk of MC and female grafts into female patients having the least. There was also a significant difference in the degree of chimerism based on conditioning regimen: Busulfan/Melphalan/Fludarabine was most likely to result in MC (P = <0.001). TRM was not significantly different between the mixed and full donor chimerism groups. However, relapse was significantly higher in the MC group (P = 0.03): 10% (95% CI: 6–14%) with full chimerism and 18% (95% CI: 7–28%) with MC at 2 years (Figure 1). Diagnosis had no impact on relapse among patients with MC. Conclusions:25% of TCD SCT recipients had MC in the BM. Although the incidence of relapse was low for patients who survived to 6 months in the full and MC groups (10% and 18%, respectively), the difference was statistically significant. There was also a significant difference in incidence of MC based on conditioning regimen and donor-recipient gender pairings. The low percentage of relapse in the MC group precluded subset analyses to identify a group who would benefit from the GVL effect of DLI. Lineage-specific PB chimerism analysis is ongoing and may provide information about subgroups that would benefit from intervention for MC. Analyses of outcomes, i.e., GVHD, in recipients of DLI or TCD stem cell boost for MC may help to determine the risk-benefit ratio of such interventions. [Display omitted] Disclosures:Goldberg:SOBI Biovitrum: Research Funding. Perales:SOBI Biovitrum: Research Funding.