Adding induction intrathecal cytarabine in newly diagnosed CNS2 B-acute lymphoblastic leukemia does not improve outcomes.

Patients with central nervous system (CNS) disease at diagnosis on Children's Oncology Group (COG) B-acute lymphoblastic leukemia (B-ALL) trials AALL0331 and AALL0232 had inferior outcomes, largely secondary to CNS relapse. In response, for patients with newly diagnosed B-ALL enrolled on COG studies AALL0932 and AALL1131, therapy was adjusted to incorporate additional intrathecal cytarabine during induction for patients with low-level CNS involvement (CNS2). We evaluated the impact of this intervention.Event-free survival, overall survival, and cumulative incidence of relapse were compared among patients with CNS2 B-ALL before vs after amendment, stratified by receipt of a 3-drug (standard-risk) or 4-drug (high-risk) induction. Multivariable models were constructed to adjust for demographic and disease variables. When stratified by trial, preamendment and postamendment patients with CNS2 status did not differ significantly by demographic or disease factors. Additional intrathecal cytarabine doses during induction did not improve outcomes for patients with B-ALL. Multivariate analyses adjusting for disease prognosticators, race/ethnicity, and leukemia cytogenetics did not identify any subpopulation that significantly benefited from additional intrathecal cytarabine. Additional intrathecal cytarabine during induction for patients with CNS2 B-ALL did not improve outcomes or mitigate the adverse prognostic impact of CNS2 status in B-ALL. Future COG B-ALL studies will not include additional intrathecal cytarabine in induction for patients with CNS2 disease. Alternative treatment strategies are needed for patients with CNS2 disease.