Clinical significance of central nervous system involvement at diagnosis of childhood t-cell acute lymphoblastic leukemia

Abstract

Patients with T-cell acute lymphoblastic leukemia (T-ALL) frequently present with unfavorable features at diagnosis. Therefore, they are considered to have a higher risk to relapse. We sought to correlate initial central nervous system (CNS) disease at diagnosis with shortened survival in childhood T-ALL. A retrospective analysis of 48 children with T-ALL was performed. The group consisted of 32 boys and 16 girls whose median age was 8 years. Their CNS status was classified as CNS-1 (no blast cells in cerebrospinal fluid (CSF); n = 44), CNS-2 (<5 WBC/microl of CSF with blast cells; n = 0), or CNS-3 (> or =5 WBC/microl of CSF with blast cells or signs of CNS involvement; n = 4). For univariate prognostic analyses, we used the log-rank test to determine the influence of patient characteristics (age, sex, lymphomatous presentations, initial leukocyte count, CNS disease, and newer therapeutic strategies) on each point. Complete remission was induced in 87.5% of patients. Median survival was 37 months, and 5-year overall survival and disease-free survival rates were 49.5% +/- 8.1% and 47.1% +/- 8.2%, respectively. Patients without initial CNS involvement seemed to have a trend toward longer overall survival (P = 0.036). Disease-free survival was not influenced by age, leukocyte count, or other factors analyzed. Patients who present with initial CNS involvement have a prognosis worse than that of patients without CNS disease. The introduction of early and effective CNS-directed therapy might no longer portend a poor prognosis for CNS leukemia.