Abstract
The biphasic shape of the survival curve of 99 patients with aplastic anemia suggested that there may be at least two subgroups of patients with this disease, one with a very short survival and another with a longer survival. Patients who survived for 4 mo or less after the first clinic visit (group A) were different from the patients who survived longer (group B) with respect to their modes of onset, sex, intervals from the onset of symptoms to first clinic visit, and initial hematologic values. These differences suggested that short survival could be predicted from data available at the first contact with the physician. From these measurements, a prognostic index could be calculated which was useful in identifying the patients in group A. Although this method of prognostication needs further testing, if validated, it may prove useful in selecting patients for therapeutic trials and could explain the divergent results in previous studies of androgen treatment of aplastic anemia. When our androgen-treated subjects were compared with subjects with a similar prognostic index who had not received androgens, a beneficial effect of androgen therapy on survival could not be demonstrated.
