O34: Precision medicine models for rare and undiagnosed Mendelian disorders

likelihood of neurodegeneration and/or type A disease.The best-characterized pathogenic variant in ASMD is the deletion of arginine at amino acid 610, considered neuroprotective in homozygous and heterozygous patients.In homozygous patients, p.Arg610del is also often associated with less severe somatic disease, a more attenuated disease course, slightly higher residual enzyme activity, and normal growth and bone development in comparison to p.Arg610del heterozygotes and to patients with other genotypes.To better understand the impact of this "protective" genotype, we compared baseline characteristics and treatment response in patients with and without this genotype in the pivotal olipudase-alfa ASCEND trial (NCT02004691, funded by Sanofi) of 36 adults with ASMD treated with olipudase alfa.Methods: Five ASCEND patients were homozygous for the p.Arg610del pathogenic variant.We evaluated burden of disease and treatment response over time in these patients and the 31 patients with other genotypes using linear mixed effects models with random intercept and unstructured variance-covariance structure.Variables included percent predicted diffusing capacity of the lungs for carbon monoxide (DL CO ), spleen volume, liver volume, high-resolution computed tomography (HRCT) lung ground glass appearance (GGA) score, liver transaminases, platelet count, HDL, LDL, and triglycerides.Models included genotype status, time on olipudase alfa, age at first infusion, and two-way interactions of genotype category and age at first infusion with time on olipudase alfa.Results: Mean enzyme activity, age at diagnosis, and age at first infusion were higher in p.Arg610del homozygous patients compared to patients without this genotype; however, only the difference in age at diagnosis was statistically significant (35 vs 15 years).Mean baseline disease burden was mostly comparable, with marginally less severe splenomegaly (12.6 vs 13.7 multiples of normal [MN]) and hepatomegaly (1.5 vs 1.7 MN) and marginally more severe mean percent predicted DL CO (52.3 vs 56.0) in the pArg610del cohort.However, the mean baseline lung HRCT GGA score was significantly worse in the pArg610del cohort (0.98 vs 0.56).This cohort included the patient with the worst baseline GGA score (35-year-old, who also had the second worst baseline DL CO ) and the patient with the worst DL CO (59-year-old).The time course and magnitude of response to treatment was similar in both cohorts, with improvements across all clinical domains, including GGA scores, which decreased to normal or near normal levels in all patients.With respect to individual pulmonary response, the 35-year-old pArg610del patient's GGA score decreased from three to zero within 1 year, and their DL CO improved from 29% to 53% over 1.5 years, transitioning from severe to moderate lung impairment.In contrast, DL CO in the 59-year-old pArg610del patient improved only slightly (from 25% to 29% at 1.5 years), although treatment response in all other domains was robust.Conclusion: This analysis suggests that although disease manifestations in p.Arg610del-homozygous ASMD patients may be less apparent/have later onset (resulting in later diagnosis), disease burden can be comparable to patients lacking this neuroprotective variant.P.Arg610del homozygosity did not protect against pulmonary disease, which can progress "silently" in ASMD patients.Earlier treatment may improve their pulmonary outcomes.