Validation of a model to predict survival in elderly patients with acute myeloid leukaemia.

Abstract

Intensive chemotherapy is the treatment of choice for selected elderly patients with acute myeloid leukaemia (AML). We recently developed a model to predict survival, thereby providing objective information upon which to select appropriate therapy for such patients. Such models, however, must be validated on a cohort of patients not used during the development of the model. We have tested the model using a series of 61 elderly patients consecutively treated with intensive chemotherapy. Using several statistical techniques, we have shown that the model is of value in predicting prognosis, though two patients did markedly better than the model prediction. This model may be useful for predicting survival in elderly patients with AML and warrants more extensive validation.