Efficacy and safety of risdiplam in patients with type 1 spinal muscular atrophy: a 3-year open-label extension of the two-part, phase 2 FIREFISH trial.

BACKGROUND: FIREFISH was a multicentre, open-label, two-part, phase 2 trial that assessed the safety, tolerability, pharmacokinetics, pharmacodynamics, and efficacy of risdiplam in children with type 1 spinal muscular atrophy and two SMN2 copies, aged 1-7 months at enrolment. Part 1 assessed safety and determined the dose for part 2. In part 2, treatment with risdiplam for 24 months resulted in continual improvements in motor function and achievement of developmental motor milestones. The aim of the 3-year open-label extension reported here was to assess long-term safety and efficacy of risdiplam. METHODS: The 3-year extension of the FIREFISH trial was a multicentre, open-label study in 17 centres across 12 countries (in Asia, Europe, Brazil, and the USA). Children with a confirmed genetic diagnosis of spinal muscular atrophy and two copies of the SMN2 gene who had completed 2 years of risdiplam treatment were eligible to continue in the open-label extension study, during which risdiplam was administered once daily by oral syringe or feeding tube at the pivotal dose of 0·20 mg/kg for infants younger than 2 years, at 0·25 mg/kg for children aged 2 years and older with bodyweight below 20 kg, and at 5 mg for children aged 2 years and older with bodyweight of 20 kg or more. Adverse events were assessed as part of physical assessments every 13 weeks. More comprehensive assessments every 26 weeks included level of respiratory support and efficacy assessments of motor function in addition to safety and physical assessments. Outcomes assessed during the open-label extension study included adverse events, survival and event-free survival (defined as being alive with no permanent ventilation), growth measures, swallowing, feeding, and motor function. Motor function was assessed with the Children's Hospital of Philadelphia Infant Test of Neuromuscular Disorders (CHOP-INTEND), the gross motor subscale of the Bayley Scales of Infant and Toddler Development, third edition (BSID-III), and the Hammersmith Infant Neurological Examination, Module 2 (HINE-2). For all children, safety data including adverse events, serious and non-serious adverse events of special interest; ophthalmology assessments; laboratory assessments; vital signs; ECGs; and other protocol-specified tests deemed critical to the safety evaluation of the study, were collected up to 30 days after the last dose. Survival and event-free survival are reported at year 5 for all enrolled children. Adverse events are reported at year 5 for all children who received at least one dose of risdiplam (safety population). All other efficacy endpoints are reported at year 5 for the population who received the pivotal dose. FIREFISH is registered with ClinicalTrials.gov, NCT02913482. FINDINGS: Between Dec 23, 2016, and Nov 19, 2018, 62 children aged 1-7 months were enrolled in FIREFISH parts 1 and 2. After 2 years in the study (parts 1 and 2), 55 children continued risdiplam treatment into the open-label extension for a further 3 years. The study was completed on Dec 22, 2023, at which time 52 children (84%) had completed 5 years of risdiplam treatment. After 5 years of risdiplam treatment, 56 children (90%) were alive, and 50 (80%) were alive without the need for permanent ventilation. In the pivotal-dose population (n=58), 17 children (29%) did not require invasive or non-invasive respiratory support at year 5 and 13 (22%) did not require hospitalisations during the study. Upward trajectories for motor function responses were observed between years 1 and 5. The number of children who could sit without support for at least 5 s and 30 s was maintained during the 3-year extension, with 36 children (62%) and 34 children (59%), respectively, doing so by year 5. Similarly, for HINE-2, eight children (14%) reached the stable sit milestone and 26 (45%) met pivots at year 5. By year 5, four children (7%) could stand without support as assessed by both the BSID-III and HINE-2. No children could walk independently, but six (10%) met the HINE-2 cruising milestone. The mean HINE-2 total score increased from baseline (0·93 [90% CI 0·72-1·14]) to year 4 (14·22 [12·66-15·79]), remaining stable thereafter. Mean CHOP-INTEND scores increased from 22·47 at baseline (90% CI 20·97-23·96) and stabilised at around 50 at year 5, with scores of ≥40 for 38 children (66%), ≥50 for 29 children (50%), and ≥60 for 11 children (19%) at year 5. Most children maintained oral feeding (n=42; 72%) and swallowing (n=46; 79%) abilities, and growth was consistent with typical patterns seen in type 1 spinal muscular atrophy. Safety findings over 5 years were consistent with the results from the primary analysis. Most adverse events were mild or moderate. The most frequently reported adverse events were pyrexia (65% [n=40]), upper respiratory tract infection (63% [n=39]), and pneumonia (50% [n=31]). Pneumonia was the most frequently reported serious adverse event in 28 children (45%). No new safety findings were reported and no treatment-related events led to withdrawal from study treatment. INTERPRETATION: Unlike the natural history of children with type 1 spinal muscular atrophy, after 5 years of risdiplam treatment, most children in FIREFISH were alive without needing permanent ventilation, were able to swallow and feed orally, and had reached motor milestones not typically observed in untreated patients. These results show long-term continuous efficacy and safety of risdiplam in children with type 1 spinal muscular atrophy. FUNDING: F Hoffmann-La Roche.