[Preliminary study on predictors for selection of immunosuppressive therapy or androgens in treating aplastic anemia patients]

Abstract

Objective: To evaluate the potential usefulness of a multivariable model, established mainly upon peripheral T-cell subsets, Th1/Th2, T-bet and GATA-3 gene expressions as well as TCM and Western medical diagnostic criteria, in predicting the selection of immunosuppressive therapy (IST) or androgens in treating patients with aplastic anemia (AA).

Methods: Peripheral blood T cell subsets in 85 patients with AA were serially analyzed by flow cytometry before and after treatment, and their T-bet and GATA-3 gene expressions were assessed meantime by Real-time PCR. Then analysis of Logistic regression equation and ROC curves were performed based on the cases responding to IST or androgens.

Results: (1) According to the logistic equation and ROC curve of SPSS, setting the false positive rate as 0.10, the P value was 0.832. When P> or =0.832, patients were judged in the immunosuppressive dominant state, IST should be applied; when P<0.832, it means patients in the bone marrow failure dominant state, androgens should be added. (2) A novel theory is raised by the above-mentioned analysis, which indicated that the genesis and development process of AA could be divided in 2 stages, the abnormal immune dominant stage and the bone marrow failure dominant stage. For treatment of patients in the two stages, IST and androgens is the preference respectively.

Conclusion: The multivariable model could be used for indicating which stage the AA patient is in, the abnormal immune stage or the bone marrow failure stage, and thus to guide the proper selection of IST or androgens in clinical practice.