Neutrophil Counts Could Predict the Response to Immunosuppressive Treatment with Antithymocyte Globulin and Cyclosporine in the Patients with Severe Aplastic Anemia

Background: Standard therapy in patients with severe aplastic anemia (SAA) without available HLA-matched sibling donor is immunosuppressive treatment (IST) including antithymocyte globulin (ATG) plus cyclosporine (CSA). However, the predictive marker of the response following IST is still to be fully elucidated.

Methods and patients: The current study was aimed to determine the predictive factor of response to IST including ATG plus CSA for the patients with SAA, and to identify the prognostic factor of the patients following IST. Overall, 62 patients were included in the present study, who were diagnosed as SAA and received IST with either rabbit ATG (n=33) or horse ATG (n=29) plus CSA between Oct 1994 and Dec 2007. Patients’ characteristics were as follows: median age 49 years old, male:female 29:33. The dosage of ATG was as follows: rabbit ATG (Thymoglobuline^®) 3.5mg/kg/day for 5 consecutive days or horse ATG (ATGAM^®) 40mg/kg/day for 4 consecutive days. CSA was started at the dose of 5mg/kg/day initially, then adjusted to maintain the trough level between 150 and 400 ng/ml till day 180. CSA was tapered slowly after then. Pneumocystis jirovechi and Herpes Simplex virus prophylaxis was done. Antifungal prophylaxis was done locally and G-CSF support was not performed routinely.

Results: With median follow up duration of 23 months (range, 0.5–122months), the rate of overall response (OR) was 50% (complete response (CR) 32%, 95% CI [19–44%] and partial response (PR) 18%, 95% CI [8–27%]). There was no difference of response rate between the groups receiving rabbit ATG versus horse ATG (p=0.571). The 2 year overall survival (OS) was 77±6%. Allogeneic hematopoietic stem cell transplantation (HSCT) was performed in 7 patients failing IST of whom 5 patients are alive and disease free. Another 4 patients failing IST received second line IST, but all of them failed. The IST-related mortality within 3 months was noted in 4 patients (6%). The causes of mortality were infection in all 4 cases. The only predictive factor for the response to IST was neutrophil count over 0.5×10⁹/L prior to IST (p=0.01). Favorable prognostic factors for OS following IST was absolute reticulocyte count (ARC) over 16,000 (p=0.007, hazard ratio [HR] 0.162, 95% CI [0.036–0.727]) and an achievement of CR (p=0.02, HR 0.121 [0.015–0.959]) or OR (p=0.0001, HR 0.05 [0.006–0.396]). In a multivariate analysis for OS using Cox’s proportional hazard model, an achievement of OR was confirmed as an independent favorable prognostic factor for OS (p=0.006, HR 0.053 [0.007–0.428]).

Conclusion: IST with ATG plus CSA could achieve successful outcomes in a half of SAA patients with acceptable treatment related mortality. The response to IST could be predicted by the neutrophil count of 0.5×10⁹/L. The achievement of ORR is an independent prognostic factor for OS. However, further study is strongly warranted to reach a clear conclusion for the predictive factor for OR or OS based on a larger number of patients.