Antithymocyte Globulin (ATG) Plus Cyclosporine (CSA) for Aplastic Anemia (AA) and Low-Risk Myelodysplastic Syndrome (MDS).

We investigated the effectiveness of immunosuppressive therapy using ATG plus CSA in adult patients with AA and low-risk MDS. Between APR 2000 and JAN 2005, 93 patients, 82 AA and 11 MDS, from 4 institutes in Korea were included in this prospective study. Patients received one of 3 types of ATG according to the availability in Korea: horse ATG (Atgam®, Upjohn, Kalamazoo, MI; 40 mg/kg/d × 4 days), rabbit ATG (Thymoglobuline®, IMTIX-SANGSTAT, Lyon, France; 2.5 mg/kg/d × 5 days), or horse ALG (Lymphoglobuline®, IMTIX-SANGSTAT; 10 mg/kg/d × 5 days). Methylprednisolone 2 mg/kg/d was given before each dose of ATG. Oral CSA 3 mg/kg twice a day was administered for 3–6 months. In 38 patients, the HLA phenotyping results were available. The median disease duration before ATG/CSA therapy was 31 days (range, 3–3126). Eight patients were not evaluable for response because of early deaths in 4, early follow-up loss in 3, and clonal evolution to AML in 1 within 60 days after treatment. Of 85 patients who were evaluable for response, 46 (54.1%) showed response to ATG/CSA therapy: partial response in 23 and complete response in 23. Responses were observed in 44 (56.7%) of 75 AA patients and in 3 (30.0%) of 10 low-risk MDS patients (P=0.103). Patients with HLA B40 phenotype showed higher response (66.7% vs. 32.1%, P=0.043), but other HLA phenotypes did not show any significant influence on response. Cumulative incidence of response was 55.5% at 2-year. Multivariate analysis indicated that disease duration before ATG/CSA therapy (≤ 90 days vs. > 90 days; OR 0.154; P=0.001) and type of ATG (horse ATG vs. rabbit ATG; OR 0.235; P=0.006) were independent predictive factors for response. Nine patients, 7 AA and 2 MDS, relapsed and cumulative incidence of relapse was 25.9% at 5-year. Clonal or disease evolution occurred in 3 patients with AA (PNH, MDS, and AML in 1 of each) and in 3 patients with MDS (RAEB-1 in 2 and RAEB-2 in 1). Eighteen patients died and actuarial survival was 69.8% at 5-year: 76.9% in AA and 36.0% in MDS. No patient who attained at least PR died. Our results demonstrated the effectiveness of ATG/CSA therapy in AA and low-risk MDS. The effects of rabbit ATG was significantly inferior in our study probably due to dose problem rather than ATG type itself. The influence of HLA phenotype on response to ATG/CSA should be further investigated.