Posterior reversible encephalopathy syndrome and other severe central nervous system adverse events in the NOPHO ALL2008 protocol : clinical and radiological findings, genetic risk factors and prognosis

Abstract

Background: The advances in the therapeutic protocols for pediatric acute lymphoblastic leukemia (ALL) have led to a current survival rate of more than 90% in developed countries. Treatment periods are, however, long and marked by complications and toxicity that may challenge treatment outcomes and quality of life for patients. Central nervous system (CNS) toxicity is common during pediatric ALL treatment and may implicate treatment postponement as well as long-term adverse effects. The aim of this thesis was to map CNS toxicities in pediatric ALL in patients treated according to the Nordic Society of Pediatric Hematology and Oncology (NOPHO) ALL2008 protocol.

Methods: Patients aged 1 to 17.9 years at diagnosis of B-cell-precursor and T-cell ALL who were treated according to the NOPHO ALL2008 protocol between 2008 and 2015 were included. Detailed data on CNS toxicity were collected from the NOPHO ALL2008 registry with seven participating countries and a complementary questionnaire addressing phenotypical and work-up details. Genome-wide association studies (GWAS) and candidate single nucleotide polymorphism (SNP) analyses were performed. A validation study of significant findings from GWAS and candidate SNP analyses was made in an independent Australian cohort of pediatric ALL patients (n=797) including patients with diverse CNS toxicities (n=103) and methotrexate-related CNS-toxicity (n=48). The role of minimal CNS leukemia in CNS toxicity risk was further examined by detecting leukemic blasts in cerebrospinal fluid (CSF) by flow cytometric immunophenotyping (FCI) in addition to cytomorphological analysis (CM), which is the CSF examination method specified in the NOPHO ALL2008 protocol.

Results: 1464 patients were included in the study of whom 52 (3.8%) had posterior reversible encephalopathy syndrome (PRES), and 135 (9.2%) had at least one form of CNS toxicity. Overall, 82/135 patients had at least one seizure episode (60.7%). PRES was the most common form of CNS toxicity in this cohort (38.5%). Older age, defined as each extra year of age and/or as patient group >10 years of age was a significant risk factor for PRES, seizures, and all CNS toxicities. T-cell immunophenotype was significant risk factor for PRES in univariate analysis and after adjustment for age. Leukemic blasts in CSF by CM were significantly related to PRES during induction and high-risk block treatment was related to PRES after induction. Minimal CNS leukemia, detected by FCI, was a significant risk factor for PRES, seizures, and all CNS toxicities in patients without CNS leukemia by CM in univariate analyses and for PRES and seizures after adjusting for induction therapy. Genome-wide association studies did not demonstrate any significant associations with CNS toxicities, but candidate SNP analyses showed that the ATXN1rs68082256 SNP, related to epilepsy, was associated with seizures in patients <10 years. ATXN1rs68082256 was replicated in the Australian cohort in the patient group with diverse CNS toxicities. At the last follow-up, 11.7% of survivors (12/103) who had displayed CNS toxicity were reported to have had an epilepsy diagnosis. Clinical suspicion of neurocognitive impairment was reported for 10.9% of survivors (12/110) with CNS toxicity at their last follow-up, but neuropsychiatric testing was performed in only two cases.

Conclusion: Central nervous system toxicity was common during pediatric ALL treatment and PRES was the most common form of CNS toxicity in this cohort. Older patients had a greater risk of CNS toxicity as well as patients with minimal CNS leukemia. The role of ATXN1rs68082256 SNP in CNS toxicity warrants further studies. Epilepsy is rather common in ALL survivors, while the neurocognitive outcome warrants more systematic follow-up.