Significance of Initial Lumbar Puncture on Day8 of Remission Induction Therapy in Childhood Acute Lymphoblastic Leukemia: the Results from the Tokyo Children'S Cancer Study Group L99-15 Protocol

Traumatic lumbar puncture with the presence of leukemic blasts (TLP+) in the cerebrospinal fluid (CSF) has been shown to adversely affect the outcome of children with acute lymphoblastic leukemia (ALL). In the Tokyo Children's Cancer Study Group (TCCSG), we adopted a strategy to defer the initial lumbar puncture (LP) and intrathecal (IT) chemotherapy until 1 week after prednisolone (PSL) monotherapy so that circulating blasts were substantially reduced when the initial LP/IT was performed. The result of L89-12 study was previously reported, in which the incidence of TLP+ was significantly diminished but the impact on the rate of central nervous system (CNS) relapse was not determined because over 80% of patients received cranial irradiation (CRT). We analyzed the result of L99-15 study in which CRT was employed in only 16% of patients. Seven hundred and fifty-five children with newly diagnosed ALL were enrolled onto the TCCSG L99-15 study between April 1999 and June 2003. Patients received the initial IT administration of methotrexate, hydrocortisone, and cytarabine on day 8 following PSL monotherapy. Patients with CNS2 (<5 leukocytes/microL of CSF with leukemic blasts) received additional IT therapy. Patients with CNS3 (>5 leukocytes/microL of CSF with leukemic blasts) or CNS1s (negative CSF finding, with cranial nerve palsy) received additional IT therapy and CRT. Patients with TLP+ did not require any reinforcement of CNS-directed therapy. The frequency of CNS1s, CNS2, CNS3, or TLP+ was 0.5%, 1.1%, 0.5% and 0.8%, respectively. These figures were almost equal to the results from L89-12 study and much lower than those reported by other study groups such as BFM. Of 22 patients (2.9% of all) classified as CNS positive group (CNS-1s, -2, -3 and TLP+), only one patient had a CNS relapse concurrently with bone marrow (BM) relapse, whereas 22 of 723 patients classified as CNS negative had CNS relapse (isolated 12, combined with BM 10). Overall, 4-year event-free survival rate was 78.2 +/− 1.6%, and 4-year cumulative incidence of overall CNS relapse was 3.3 +/− 0.7%. Our strategy may underestimate the frequency of CNS leukemia at diagnosis, however, unnecessary intensification of CNS treatment may be avoided while overall CNS relapse rate is quite low in spite of restricted indication for cranial irradiation.