The incidence of hydrocephalus among patients with and without spinal muscular atrophy (SMA): Results from a US electronic health records study

Background The incidence of hydrocephalus in the spinal muscular atrophy (SMA) population relative to the general population is currently unknown. Since the approval of nusinersen, an intrathecally administered drug for SMA, a small number of hydrocephalus cases among nusinersen users have been reported. Currently, the incidence of hydrocephalus in untreated SMA patients is not available, thereby making it difficult to determine if hydrocephalus is a side effect of nusinersen or part of SMA’s natural history. This retrospective, matched cohort study used electronic health records (EHRs) to estimate and compare the incidence of hydrocephalus in both SMA patients and matched non-SMA controls in the time period prior to the approval of nusinersen. Methods The U.S. Optum® de-identified EHR database contains records for approximately 100 million persons. The current study period spanned January 1, 2007–December 22, 2016. Patients with SMA were identified by one or more International Classification of Diseases (ICD)-9 and/or ICD-10 codes for SMA appearing as primary, admission, or discharge diagnoses, without a pregnancy diagnostic code in the 1-year time before and after the first occurrence of SMA. The first occurrence of SMA defined the index date and non-SMA controls were matched to cases. Incident cases of hydrocephalus were identified with one or more ICD-9 and/or ICD-10 code for any type of hydrocephalus following the index date. Hydrocephalus incidence rates per person-months and the incidence rate ratio comparing SMA cases with non-SMA controls were calculated. Results There were 5354 SMA cases and an equal number of matched non-SMA controls. Incident hydrocephalus events were identified in 42 SMA cases and 9 non-SMA controls. Hydrocephalus incidence rates per 100,000 person-months were 15.5 (95% CI: 11.2–20.9) among SMA cases and 3.3 (95% CI: 1.5–6.3) among non-SMA controls. The incidence rate ratio was 4.7 (95% CI: 2.4–10.2). Conclusions Based on this retrospective analysis utilizing US EHR data, SMA patients had an approximately fourfold increased risk of hydrocephalus compared with non-SMA controls in the era preceding nusinersen treatment. This study may assist in properly evaluating adverse events in nusinersen-treated SMA patients.