The use of immunosuppressive therapy in MDS: clinical outcomes and their predictors in a large international patient cohort

Most studies of immunosuppressive therapy (IST) in myelodysplastic syndromes (MDS) are limited by small numbers and their single-center nature, and report conflicting data regarding predictors for response to IST. We examined outcomes associated with IST and predictors of benefit in a large international cohort of patients with MDS. Data were collected from 15 centers in the United States and Europe. Responses, including red blood cell (RBC) transfusion independence (TI), were assessed based on the 2006 MDS International Working Group criteria, and overall survival (OS) was estimated by Kaplan-Meier methods. Logistic regression models estimated odds for response and TI, and Cox Proportional Hazard models estimated hazards ratios for OS. We identified 207 patients with MDS receiving IST, excluding steroid monotherapy. The most common IST regimen was anti-thymocyte globulin (ATG) plus prednisone (43%). Overall response rate (ORR) was 48.8%, including 11.2% (95% confidence interval [CI], 6.5%-18.4%) who achieved a complete remission and 30% (95% CI, 22.3%-39.5%) who achieved RBC TI. Median OS was 47.4 months (95% CI, 37-72.3 months) and was longer for patients who achieved a response or TI. Achievement of RBC TI was associated with a hypocellular bone marrow (cellularity < 20%); horse ATG plus cyclosporine was more effective than rabbit ATG or ATG without cyclosporine. Age, transfusion dependence, presence of paroxysmal nocturnal hemoglobinuria or large granular lymphocyte clones, and HLA DR15 positivity did not predict response to IST. IST leads to objective responses in nearly half the selected patients with the highest rate of RBC TI achieved in patients with hypocellular bone marrows.