Enhanced clinical trial stratification in spinal cord injury: the value of electrophysiology

There are no approved interventional therapies, aside from neurorehabilitation, that enhance neurological recovery after acute traumatic spinal cord injury. A key challenge is the lack of biomarkers surpassing clinical standards for optimal stratification. We evaluated electrophysiological markers of preserved neuronal function to improve enrichment strategies over clinical measures. We hypothesized that participants with preserved electrophysiological markers would achieve greater neurological and functional recovery in response to a plasticity-inducing intervention. We conducted a retrospective stratification analysis of data from the recently completed randomized, placebo-controlled, phase 2b Nogo Inhibition in spinal cord injury (NISCI) trial (NCT03935321) investigating the efficacy of NG101, a recombinant human antibody that neutralizes the neurite outgrowth-inhibiting protein Nogo-A. Participants aged 18-70 years with acute (4-28 days) cervical spinal cord injury were eligible. At screening, all participants underwent clinical neurological examination and electrophysiological recordings (i.e. somatosensory evoked potentials). Treatment effect sizes for the recovery of upper extremity motor scores and spinal cord independence measure of self-care (6-month change) between NG101 and placebo groups were compared for stratification based on clinical versus electrophysiological criteria. Power analyses were conducted to estimate the required sample sizes needed for each method. The cohort included 116 participants (45.5 ± 16.8 years old, 74 NG101 and 41 placebo). Clinical stratification showed greater functional recovery in motor-incomplete participants treated with NG101 versus placebo [estimate 0.02 (95% confidence interval: 0.006-0.038), P = 0.007]. Electrophysiological stratification revealed greater functional recovery in participants with preserved somatosensory evoked potentials treated with NG101 versus placebo [0.04 (0.015-0.054), P < 0.001]. Effect sizes were large for electrophysiological stratification (Cohen's d = 0.94) but small for clinical stratification (Cohen's d = 0.46). Power analyses demonstrated smaller required sample sizes for electrophysiological stratification (required n = 32) versus clinical stratification (required n = 120). This study shows the value of electrophysiology in comparison to clinical measures for biomarker-driven enrichment and improved power in acute spinal cord injury trials. We emphasize the importance of functionally spared neuronal pathways in promoting recovery in response to plasticity-inducing interventions, such as anti-Nogo-A antibodies.