Recent reports suggested that central nervous system (CNS) involvement (CNS+) in patients with acute myeloid leukemia (AML) before allogeneic hematopoietic stem cell transplantation (allo-HSCT) is not an independent predictor of survival after allo-HSCT. However, these studies did not analyze minimal residual disease in the CNS at the time of allo-HSCT. We evaluated the effect of residual CNS+ on the transplant outcomes of 214 AML patients in a single institution. Twenty-one (10%) patients were diagnosed with CNS+ prior to allo-HSCT. Of these, 13 patients had CNS disease at the time of allo-HSCT. The patients in CNS+ AML remission at the time of allo-HSCT had better overall survival (OS) than the patients who were not in remission (2-year OS: 55% vs. 7.7%, p = 0.0001). In multivariate analyses, CNS+ at the time of allo-HSCT (hazard ratio (HR), 1.9; 95% confidence interval (CI), 1.05-3.59; p = 0.04), age over 50 years at the time of allo-HSCT, and non-complete remission disease status in bone marrow at the time of allo-HSCT were independent adverse factors for OS. However, a prior history of CNS+ before allo-HSCT did not independently affect OS (HR, 1.27; 95% CI 0.53-2.07; p = 0.6). Early diagnosis and eradication of CNS+ at the time of allo-HSCT may be necessary to improve the outcome for patients with CNS+ AML.
Keywords: Acute myeloid leukemia; Allogeneic hematopoietic stem cell transplantation; Central nervous system involvement.