Indirect treatment comparisons of nusinersen versus onasemnogene abeparvovec-brve in treatment-naïve and -experienced spinal muscular atrophy patients.

Abstract

Nusinersen 12 mg was the first regulator-approved disease-modifying therapy (DMT) for spinal muscular atrophy (SMA); intrathecal onasemnogene abeparvovec-brve (OAV101 IT) and nusinersen 50/28 mg have gained recent approval. In absence of head-to-head evidence, this research assessed the comparative efficacy of these treatments in DMT-naïve and DMT-experienced SMA patients using indirect treatment comparisons (ITCs). An ITC feasibility assessment was conducted. Individual patient data (IPD) from CHERISH (NCT02292537) and DEVOTE Part B (NCT04089566) clinical trials assessing nusinersen in DMT-naïve, later-onset participants were trimmed to align with STEER (NCT05089656). DMT-experienced IPD from DEVOTE Part C (NCT04089566) were trimmed to align with STRENGTH (NCT05386680). Matching adjusted indirect comparisons (MAIC) were conducted to assess motor function improvement using the Hammersmith Functional Motor Scale-Expanded (HFMSE) and Revised Upper Limb Module (RULM). The anchored MAIC in DMT-naïve participants numerically favored nusinersen 12 mg compared with OAV101 IT in change from baseline (CFB) in HFMSE (mean difference [MD]: 1.86; 95% confidence intervals [CIs]: -1.08, 4.80; starting sample [N]/effective sample size [ESS]: 97/58), and RULM (MD: 1.52; CIs: -0.57, 3.61; N/ESS: 97/58) in participants aged 2 to <5 years. The unanchored MAIC in DMT-experienced participants numerically favored nusinersen 50/28 mg compared with OAV101 IT in CFB in HFMSE (MD: 1.02; CIs: -1.31, 3.34; N/ESS: 10/3), and RULM (MD: 1.41; CIs: -3.64, 6.46; N/ESS: 10/5) in participants aged <18 years. Trends were consistent across weighting scenarios. After applying participant weights, the ITC of DEVOTE Part B and STEER was deemed infeasible. As is common in rare diseases, small sample sizes from available trial data may have limited the precision of the estimates despite consistent numerical trends. Nusinersen resulted in numerical functional gains compared with OAV101 IT in both DMT-naïve and DMT-experienced patients. Further studies should be undertaken to validate any between-treatment difference. People with Spinal Muscular Atrophy (SMA) have a genetic mutation that means they do not make enough of the survival motor neuron (SMN) protein. This results in muscle weakness and decreased function that gets worse over time. Nusinersen and onasemnogene abeparvovec-brve (OAV101 IT) are currently used to treat SMA by increasing SMN protein; yet it is unclear which therapy is more efficacious because they have not been directly compared. In the absence of data comparing these treatments, statistical methods known as indirect treatment comparisons (ITCs) can be conducted to estimate the relative efficacy of treatments on SMA symptoms if they were directly compared. We conducted ITCs comparing nusinersen and OAV101 IT in trial participants who had never received SMA treatment (i.e. treatment-naïve) and participants who had previously received SMA treatments (i.e. treatment-experienced) by comparing measures of participants’ motor function. These measures were the Hammersmith Functional Motor Scale–Expanded (HFMSE), assessing whole-body motor function, and the Revised Upper Limb Module (RULM), assessing upper-body motor function. We found nusinersen may improve HFMSE by up to 2.28 points and RULM up to 1.90 points compared with OAV101 IT across both treatment-naïve and -experienced patients, where a 2-point change may reflect acquiring one new motor activity or improving two existing activities. These findings contribute to the limited comparative evidence available for SMA treatments. Though the samples of trial participants were small, the direction of results was consistent across multiple analyses. Future work in larger patient samples would be beneficial to confirm the validity of results.