Effectiveness of intrathecal rituximab in patients with acute lymphoblastic leukaemia relapsed to the CNS and resistant to conventional therapy

Abstract

Results pertaining to non-Hodgkin lymphoma (NHL) (Schulz et al, 2004) and B-cell chronic lymphocytic leukaemia (Watanabe et al, 2005) with leptomeningeal infiltration suggest that intrathecal rituximab may have an important therapeutic role on B-lymphocyte diseases affecting the central nervous system (CNS). We tested the efficacy of rituximab on seven patients with B-cell acute lymphoblastic leukaemia (ALL) CD20+ relapsed in the CNS and refractory to triple intrathecal therapy (methotrexate, 12·5 mg; cytarabine, 70 mg; and hydrocortisone, 37·5 mg). With the exception of Patient 1, a 4-year-old female, the patients had previously received cranial irradiation (24 cGy) and spinal irradiation (18 cGy). As initial therapy for ALL, patients received a modification of the Paediatric Oncology Group (POG) 9201protocol (Chauvenet et al, 2007). After CNS relapse, reinduction therapy included 4 doses of weekly vincristine (1·5 mg/m2), prednisone (60 mg/m2 PO, days 1–30), 6 doses of asparaginase (6000 IU/m2), 3 doses of adriamycin (40 mg/m2) and 4 of weekly triple intrathecal therapy. Consolidation was accomplished with systemic high dose of methotrexate (2 g/m2) and cytarabine (2 g/m2, 4 doses). Afterwards patients resumed maintenance therapy with 6-mercaptopurine (50 mg/m2, days 8–30), and weekly methotrexate (30 mg/m2/PO), with monthly intensifications consisting of vincristine (1·5 mg/m2 i.v., day 1), prednisone (60 mg/m2, days 1–7), and monthly triple intrathecal chemotherapy; in this phase refractory patients were included in the rituximab protocol. During intrathecal therapy with rituximab, only 6-mercaptopurine and weekly methotrexate were used as systemic therapy; afterwards, triple intrathecal chemotherapy, vincristine and prednisone, was resumed on a monthly basis. The study was approved by the Institutional Review Board, and patients or parents provided written informed consent for the administration of rituximab. Rituximab, 10 mg diluted in 6 ml of saline, was delivered intrathecally every week for four consecutive weeks (days 1, 7, 14, and 21). Prior to each administration, a cerebrospinal fluid (CSF) sample was drawn and examined after high-speed centrifugation for the presence of lymphoblasts. Seven days after the last dose of rituximab, patients restarted triple intrathecal therapy as previously described. The study group was heterogeneous for age (median 10 years), gender, time of evolution, and treatment stage. Median time of evolution for ALL was 36 months (range: 4–60 months) with CNS relapse occurring at a median of 10 months (range: 4–48 months). Patients were asymptomatic at the time of diagnosis of CNS relapse. Five patients were diagnosed after routine CSF examination during maintenance therapy, and two patients were diagnosed while under evaluation for testicular and bone marrow relapse. Median number of lymphoblasts/mm3 in the CSF was 15 (range: 5–115). Salient characteristics and current status of patients in the study group are shown in Table I. Before rituximab administration, patients received a median of 23 intrathecal therapeutic lumbar punctures (range 14–26) within a median time of 9 months after the diagnosis of CNS relapse (range 6–30 months). Cytospin analysis of CSF after the fourth dose of rituximab was negative for lymphoblasts in the seven patients. Low numbers of non-malignant cells were detected in only Patients 1 and 2 (2 and 12 cells respectively). Adverse neurotoxic effects of rituximab were not observed. After 24 months of CSF serial analysis and follow-up, 5/7 patients remained in complete remission, free of CNS leukaemic infiltration. Systemic relapse, however, developed in two cases. One of these patients died 6 months after the last dose of rituximab due to bone marrow relapse but was free of CNS disease. The other patient suffered testicular and bone marrow relapses 9 months after completion of treatment. He received an allogeneic hematopoietic stem cell transplant but died 15 months later from chronic graft-versus-host disease but with a cell-free CSF. Optimum therapy for ALL CNS relapse resistant to conventional treatment has not been established. CNS irradiation has been delivered as salvage therapy except for those patients at an especially high risk for CNS relapse-T-cell immunophenotype, CNS3 status at diagnosis, high-risk cytogenetic features, or poor response to remission of induction therapy- (Pui & Howard, 2008). Recently, intrathecal liposomal cytarabine has successfully been used for ALL patients with refractory relapse to the CNS, although with considerably more neurotoxicity than in our rituximab-treated group. In one report, 5/31 ALL patients receiving this agent as part of their CNS prophylaxis developed serious unexpected neurotoxicity including seizures, papilloedema, cauda equine syndrome, and encephalitis, after a median of four intrathecal doses. One patient died as a result of progressive encephalitis (Jabbour et al, 2007). Our results are similar to those reported by a German group in patients with NHL and leptomeningeal infiltration. They observed CNS clearance 4 weeks after intrathecal infusion of rituximab (10–40 mg/dose in four to five doses) (Schulz et al, 2004). In addition to an isolated single case report of successful intrathecal rituximab administration for an optic nerve relapse of CD20+ ALL (Kraal et al, 2005), additional information regarding intrathecal administration of rituximab in ALL patients suffering CNS relapse is lacking. In our small study group, the addition of rituximab to the therapeutic scheme appeared to be effective in eradicating CNS infiltration. CNS eradication, however, did not prevent systemic relapse in two of our patients. Given the prolonged CNS remission status after the administration of rituximab, it is reasonable to assume that sustained CNS relapse-free status is due, at least in part, to its therapeutic effect. Remarkably, rituximab administration was well tolerated in our heavily-treated study group, with no clinical evidence of neurotoxicity after a 24-month follow-up period.