Acute GVHD May Play a Role in Preventing Relapse Following Reduced Intensity Allografts for Acute Myeloid Leukemia.

Abstract

The role of reduced intensity allografting in the treatment of patients (pts) with acute myeloid leukemia (AML) and the role that graft versus host disease (GVHD) plays in preventing relapse in this setting is unclear. While more pts may survive the initial period following allografting in comparison to those receiving full intensity transplants, the risk of relapse may be greater, particularly in those not in remission at the time of transplantation. Here we report the results of allogeneic transplants following a reduced intensity conditioning (RIC) regimen given to 158 consecutive pts with AML. All were all conditioned with an identical regimen containing Cyclophosphamide at 120mg/kg iv over 2 days (days −3 and −2) followed by a single dose of total body irradiation at 550cGy on day −1. HLA-identical (HLA-ID) siblings donated cytokine mobilized peripheral blood (PB) to 68 recipients and unrelated donor (UD) bone marrow (BM) was used for 90 patients. GVHD prophylaxis consisted of single agent cyclosporine (CSP) for recipients of HLA-ID allografts and CSP, methotrexate, and methylprednisolone in recipients of UD BM. The median age of recipients was 44 years (range 6–70), with 74 males and 84 females. 43 pts were in CR1, 47 in CR2 or 3, and 68 had relapsed or refractory disease. The majority of UD BM recipients were serologically identical with their donors at 6/6 antigens. There were no graft failures in either group.The median follow-up in survivors was long at 1903 days (range 938–2757). Overall survival (OS) was evaluated in relation to the development of any grade of acute GVHD. There were no significant differences in the incidence of grade 2–4 (30+/− 6%) or grade 3–4 (7+/−3%) acute GVHD in the HLA-ID PB or UD BM groups. Since OS was not significantly different between CR1 and CR2 pts, the results were pooled. The Kaplan-Meier (KM) estimate of 3 year OS in UD BM pts in CR was 45+/−9% in those with no acute GVHD compared to only 25+/−10% with any grade of acute GVHD (p=0.24). In recipients of HLA-ID PB who were in CR, 3 yr OS was 39+/−12% without acute GVHD and interestingly, 54+/−10% with acute GVHD. UD BM recipients not in remission at the time of transplant had dismal outcomes with or without acute GVHD, with less than 5% OS. No recipients of HLA-ID PB with active disease at the time of transplant remained alive if they did not develop acute GVHD versus 19+/−10% if they did (p=0.17).These data obtained in a large group of homogeneously treated pts suggest that acute GVHD may play an important role in preventing relapse after RIC and HLA-ID sibling PB transplantation. Following UD BM, acute GVHD appears to adversely affect OS. RIC should not be proposed in the unrelated donor setting if pts with AML are not in remission. Future efforts aimed at harnessing graft versus leukemia effects in the HLA-ID setting following RIC may lead to improvements in OS and are clearly warranted.