Background

A phase 1 study reported reduced cerebrospinal fluid SOD1 protein and initial safety findings after repeated doses of the small interfering RNA (siRNA) candidate RAG-17 in six people with amyotrophic lateral sclerosis (ALS, also known as Lou Gehrig’s disease) associated with SOD1 variants. A marker of nerve injury also declined, but the study did not establish delayed loss of motor function or a survival benefit. Preclinical and clinical findings were published in Nature Medicine.

SOD1 is the gene that encodes superoxide dismutase 1. Some variants are thought to contribute to motor neuron injury through a toxic gain of function of misfolded protein. This has led to the development of strategies to lower SOD1 expression for this patient group. Tofersen, an antisense oligonucleotide (ASO) drug, targets the same gene.

siRNA suppresses target gene expression through RNA interference, but delivering it effectively to the central nervous system is difficult. RAG-17 combines SOD1-targeting siRNA with a non-targeting, chemically modified auxiliary oligonucleotide (ACO). It uses the SCAD delivery platform, designed to support cellular uptake and tissue distribution.

Key findings

Preclinical results varied by administration route and tissue. In symptomatic female SOD1 G93A mice given two intracerebroventricular doses on postnatal days 126 and 151, median survival in the 400 μg high-dose group increased by 75.8% compared with controls. This was a survival extension in an animal model and should be distinguished from treatment effects in people.

Cynomolgus monkeys received two intrathecal doses, on experimental days 1 and 15. At 50 mg per dose, SOD1 messenger RNA (mRNA) in the lumbar spinal cord fell by up to 91%. At the same dose, the maximum reduction in the cerebral cortex exceeded 70%; suppression was not uniform across the brain and spinal cord.

The human study was a single-center, open-label, nonrandomized, dose-escalation phase 1 trial whose primary objectives were safety and tolerability. Six adults with confirmed SOD1 variants were divided into groups of three, starting at either 60 mg or 90 mg. Doses were increased in 30 mg increments. Five participants received a maintenance dose of 150 mg, and one received 180 mg. The cohorts received seven and six doses in total, respectively, with maintenance doses given at two-month intervals.

No serious adverse events were reported. Two participants had transient, mild muscle twitching; one of them, who had pre-existing fatty liver and abnormal liver function, also had elevated alanine aminotransferase (ALT). Investigators assessed the twitching as treatment-related and the ALT elevation as potentially related to both the drug and the participant’s underlying condition.

Cerebrospinal fluid (CSF) SOD1 protein fell from baseline by a mean of 69% at day 240 in the first cohort and 56% at day 210 in the second. Mean end-of-study reductions in plasma neurofilament light chain (NfL), a marker of axonal injury, were reported as 62% and 52%, respectively. One participant’s NfL initially declined but rebounded to around baseline by day 240; investigators excluded this participant from the longitudinal NfL analysis. These averages should not be interpreted as a consistent response in all six participants.

Meaning and outlook

The findings demonstrate pharmacologic activity of an siRNA directed at a target in the human central nervous system, as reflected by lower CSF SOD1 and changes in biomarkers. Changes in functional assessments varied among participants, and respiratory scores declined in some. No deaths or invasive mechanical ventilation occurred during follow-up, but without a control group this cannot be taken as evidence of a survival benefit from the drug.

The study was small, SOD1 variants and disease courses were heterogeneous, and pretreatment progression rates were estimated retrospectively. Because serial CSF cell counts were not available, an asymptomatic inflammatory response could not be ruled out. Long-term safety, delayed functional decline, and the possibility of longer dosing intervals remain to be tested in subsequent randomized controlled trials. There is also no direct comparative evidence that RAG-17 is superior to tofersen.

The study design and findings can be reviewed in the full text of the original paper and its abstract and bibliographic record.