Multicenter propensity score matched study on allogeneic stem cell transplantation in adult sickle cell disease: Comparison of donor vs. no donor outcomes

Abstract Background Despite the lack of high-level scientific evidence, allogeneic hematopoietic stem cell transplantation (Allo HPSCT) is routinely performed successfully for adult sickle cell disease (SCD). However, studies on transplanted vs. non-transplant patients for the development of SCD indications are lacking. A controlled evaluation of outcomes among adults undergoing allo-HSCT compared to matched controls without a donor is critical to inform clinical decision-making. This study was to assess the outcome of transplanted vs. non-transplant SCD groups by propensity matched pair analysis. Methods This multicenter, retrospective cohort study on behalf of the Turkish Society of Hematology Hemoglobinopathy Study Group (TSH-HSG) included consecutive SCD patients with donor undergoing Allo HSCT or SCD patients with no donor in 14 centers from Turkiye. Propensity scores were calculated based on nine pretransplant variables (age, vasooclusive crisis, pulmonary hypertension, proteinuria, acute chest syndrome and cerebral event, which are known to be associated mortality). Transplant and non-transplant groups were matched in a 1:1 ratio. Outcomes including complications and mortality were compared using paired test. Results: The group sample sizes of 102 in each group achieved 100% power at a significance level of alpha=0.05 based on group survival means. Among 805 patients with SCD patients, 102 pairs were matched with transplanted group receiving non-myeloablative Fu Bu ATLG and posttransplant cyclophosphamide-containing regimen. The mean follow up time for transplanted group was shorter [47 (IQR 24-75) vs. 202 (IQR 84-241); P< 0.001]. Ninety-seven (95%) of transplanted patients achieved cure with no secondary graft failures. Two (2%) patients developed chronic graft-versus-host disease. Hemoglobin, hemoglobin S value, vasooclusive crisis, acute chest, proteinuria, transfusion need and hospitalization rate were significantly better in transplant group vs. non-transplant group (p< 0.001, for all). Overall survival curve analysis of the groups were estimated as similar via log-rank test (P=0.298). When patients who received transplants from alternative donors are excluded, the difference of survival curves did not change. Conclusion This multicenter propensity score matched study confirms the superiority of transplant over non-transplant group based on hematological parameters, vasooclusive crisis, acute chest, proteinuria, transfusion requirement, and hospitalization rate thus validating current guideline advice for adult SCD.