Twenty-Year Survival Analysis of Adeno-Associated Virus Vector Serotype 2-Mediated Gene Therapy to the Central Nervous System for CLN2 Disease

Abstract

CLN2 disease (late infantile neuronal ceroid lipofuscinosis) is an autosomal recessive, neurodegenerative lysosomal storage disease that results from loss of function mutations in the CLN2 gene, which encodes tripeptidyl peptidase 1. It affects the central nervous system (CNS) with progressive neurodegeneration and early death, typically at ages from 8 to 12 years. Twenty years ago, our phase I clinical trial treated subjects with CLN2 disease by a catheter-based CNS administration of an adeno-associated virus vector serotype 2 (AAV2) expressing the CLN2 gene. Here we present an analysis of the survival of the 10 treated children and find 2 distinct survival groups with a wide disparity in survival. Group 1 (n = 7) had the typical mean survival of 8.8 ± 0.5 years of age, 3.8 ± 0.6 years post-therapy. Group 2 (n = 3) had a markedly longer mean survival of 23.4 ± 2.4 years of age and 14.9 ± 2.8 years post-therapy (p < 0.00002, survival of group 1 vs. group 2). Long survivors (group 2) at the time of treatment were older (group 1: 5.0 ± 0.6 years; group 2: 8.5 ± 0.9 years; p < 0.02); had similar disease severity (Hamburg clinical score group 1: 4.7 ± 0.5, group 2: 3 ± 0.0, p > 0.05); and had larger CNS ventricular volume (81.1 ± 22.2 cm³ vs. 27.3 ± 7.2 cm³ for group 1; p < 0.02). While the genotype of 3, group 2 subjects, had one allele (509-1G>C) identical to that of three in group 1, the second allele was different. This was unlikely to explain the survival difference, as alleles for both groups were equally predicted deleterious by the Combined Annotation-Dependent Depletion score: 34.9 ± 0.7 and 32.8 ± 0.3 for groups 1 and 2, respectively (a score of >20 is considered deleterious). This represents one of the longest survival studies (up to 20 years) of AAV-treated individuals with hereditary disorders and demonstrates variability of therapeutic efficacy where the genotype on its own has no apparent survival advantage.

Protocol registration numbers for the original study: NCT00151216 and NCT00151268; www.clinicaltrials.gov.