Chimerism Analysis Is Predictive of Relapse-Free Survival After Allogeneic Transplantation for MDS

Abstract 1332

Chimerism analysis permits evaluation of the extent of donor engraftment following allogeneic hematopoietic cell transplantation (HCT) via differentiation between donor- and recipient-derived cells. Though relapse remains a major cause of treatment failure post-HCT for hematologic malignancies, the clinical utility of chimerism analysis for the early detection of morphologic relapse varies between different diseases and remains controversial. The predictive value of chimerism analysis for relapse rates and mortality in myelodysplastic syndromes (MDS) is not well-described. We reviewed serial chimerism results for 72 consecutive patients (pts) who underwent allogeneic HCT for MDS between 1999 and 2009; 9 pts were excluded from analysis due to lack of appropriately timed chimerism studies. Donor engraftment was initially assessed 28 days post-HCT and then at 2-week intervals through day 100. Chimerism studies were performed with peripheral blood using a short tandem repeat assay by PCR-based analysis. Acute and chronic GVHD rate, relapse-free survival (RFS), and overall survival (OS) were assessed for patients with donor leukocyte chimerism and T-cell chimerism ≥95% and <95%. The median age was 51 years (range 20–70) and 52% were male. The median time from MDS diagnosis to HCT was 5.4 months; 9 pts (14%) were in complete remission at the time of HCT, 11 (18%) were in partial remission, 17 (27%) had relapsed or refractory disease, and 26 (41%) were untreated. HCT-comorbidity index was low in 25 pts (40%), intermediate in 18 (29%), and high in 20 (32%). 31 pts (49%) received sibling donor HCT and the remainder received an unrelated donor graft. 40 (63%) received bone marrow and 23 (37%) received peripheral stem cells. Myeloablative (MA) busulfan- or cyclophosphamide-based preparative regimens were used in 47 pts (75%) and a non-myeloablative (NMA) regimen with fludarabine and TBI was given to the remaining pts. 61 pts achieved ≥95% donor leukocyte chimerism at a median of 29 days and 39 evaluable pts achieved ≥95% donor T-cell chimerism at a median of 42 days. Two pts did not achieve donor leukocyte chimerism ≥95% and 9 did not achieve donor T-cell chimerism ≥95% at any timepoint. Univariable analysis of prognostic factors for relapse showed that donor leukocyte chimerism ≥95% was significantly associated with lower probability of relapse (hazard ratio [HR] 0.11, 95% confidence interval [CI] 0.02–0.51, p=.005), whereas prior exposure to radiation therapy (excluding exposure during HCT preparative regimen) was associated with increased probability of relapse (HR 3.48, 95% CI 1.14–10.60, p=.028). Multivariable analysis implicated donor leukocyte chimerism <95% as the only independent risk factor for relapse. Transplant type (MA vs NMA) and cell source did not significantly impact the likelihood of relapse. Donor leukocyte chimerism ≥95% was not associated with acute or chronic GVHD. Univariable analysis of risk factors for survival showed that donor leukocyte chimerism ≥95% was associated with improved RFS (HR 0.29, 95% CI 0.09–0.97, p=.045) but not OS, both findings of which were confirmed on multivariable analysis. Donor T-cell chimerism ≥95% was not significant in univariable analysis; however, in multivariable analysis, controlling for comorbidity scores, donor T-cell chimerism <95% was associated with lower risk of chronic GVHD (HR 0.18, 95% CI 0.04–0.88, p=.034) but did not significantly impact relapse, RFS, or OS. In conclusion, achievement of a high donor leukocyte chimerism post-HCT for MDS is associated with improved RFS. Donor T-cell chimerism, however, is not predictive of outcome and its routine use for MDS pts should be reevaluated. Pts with donor leukocyte chimerism <95% might be considered for immunologic interventions such as withdrawal of immunosuppression or donor lymphocyte infusion.