Background: Risdiplam is an oral small molecule approved for the treatment of patients with spinal muscular atrophy, with approval for use in patients with type 2 and type 3 spinal muscular atrophy granted on the basis of unpublished data. The drug modifies pre-mRNA splicing of the SMN2 gene to increase production of functional SMN. We aimed to investigate the safety and efficacy of risdiplam in patients with type 2 or non-ambulant type 3 spinal muscular atrophy.

Methods: In this phase 3, randomised, double-blind, placebo-controlled study, patients aged 2-25 years with confirmed 5q autosomal recessive type 2 or type 3 spinal muscular atrophy were recruited from 42 hospitals in 14 countries across Europe, North America, South America, and Asia. Participants were eligible if they were non-ambulant, could sit independently, and had a score of at least 2 in entry item A of the Revised Upper Limb Module. Patients were stratified by age and randomly assigned (2:1) to receive either daily oral risdiplam, at a dose of 5·00 mg (for individuals weighing ≥20 kg) or 0·25 mg/kg (for individuals weighing <20 kg), or daily oral placebo (matched to risdiplam in colour and taste). Randomisation was conducted by permutated block randomisation with a computerised system run by an external party. Patients, investigators, and all individuals in direct contact with patients were masked to treatment assignment. The primary endpoint was the change from baseline in the 32-item Motor Function Measure total score at month 12. All individuals who were randomly assigned to risdiplam or placebo, and who did not meet the prespecified missing item criteria for exclusion, were included in the primary efficacy analysis. Individuals who received at least one dose of risdiplam or placebo were included in the safety analysis. SUNFISH is registered with ClinicalTrials.gov, NCT02908685. Recruitment is closed; the study is ongoing.

Findings: Between Oct 9, 2017, and Sept 4, 2018, 180 patients were randomly assigned to receive risdiplam (n=120) or placebo (n=60). For analysis of the primary endpoint, 115 patients from the risdiplam group and 59 patients from the placebo group were included. At month 12, the least squares mean change from baseline in 32-item Motor Function Measure was 1·36 (95% CI 0·61 to 2·11) in the risdiplam group and -0·19 (-1·22 to 0·84) in the placebo group, with a treatment difference of 1·55 (0·30 to 2·81, p=0·016) in favour of risdiplam. 120 patients who received risdiplam and 60 who received placebo were included in safety analyses. Adverse events that were reported in at least 5% more patients who received risdiplam than those who received placebo were pyrexia (25 [21%] of 120 patients who received risdiplam vs ten [17%] of 60 patients who received placebo), diarrhoea (20 [17%] vs five [8%]), rash (20 [17%] vs one [2%]), mouth and aphthous ulcers (eight [7%] vs 0), urinary tract infection (eight [7%] vs 0), and arthralgias (six [5%] vs 0). The incidence of serious adverse events was similar between treatment groups (24 [20%] of 120 patients in the risdiplam group; 11 [18%] of 60 patients in the placebo group), with the exception of pneumonia (nine [8%] in the risdiplam group; one [2%] in the placebo group).

Interpretation: Risdiplam resulted in a significant improvement in motor function compared with placebo in patients aged 2-25 years with type 2 or non-ambulant type 3 spinal muscular atrophy. Our exploratory subgroup analyses showed that motor function was generally improved in younger individuals and stabilised in older individuals, which requires confirmation in further studies. SUNFISH part 2 is ongoing and will provide additional evidence regarding the long-term safety and efficacy of risdiplam.

Funding: F Hoffmann-La Roche.