Outcome measures for trials of remyelinating agents in multiple sclerosis: retrospective longitudinal analysis of visual evoked potential latency

Objective Visual evoked potentials (VEP) may be suitable surrogate outcome measures in multiple sclerosis (MS) remyelination trials. The extent of spontaneous changes of subclinically delayed VEP is unknown, whereas VEP improve after acute optic neuritis (ON). Methods In all, 124 patients with three VEP recordings at least 3 months apart: 71 patients with MS who had never suffered clinical ON; 53 patients with ON (isolated ON or ON as an attack of MS at first recording). Latencies of P100 were analyzed by multivariate analysis of variance. Results Eyes of patients with MS had a mean P100 latency of 110.2 ms, worsening mildly over time (n = 104 eyes, P = 0.022). MS patients' eyes with subclinical demyelination (delayed P100 latency at first recording > 116 ms) showed no significant evidence of remyelination (n = 24 eyes, P = 0.27). By contrast, in ON patients' affected eyes, mean P100 latency decreased (P = 0.001), whereas unaffected eyes remained stable (P = 0.26). Clinically non-affected eyes from both diagnostic groups with subclinically prolonged latencies remained stable (n = 32: mean P100 at 124.8 +/- 10.7, 123.5 +/- 13.6, and 122.8 +/- 13.1 ms; P = 0.57), whereas non-affected eyes with normal latency at baseline deteriorated slightly (P = 0.001). A subgroup with more homogeneously defined follow-up periods confirmed this observation. Non-affected eyes selected for stability (difference <5 ms) between first and second recording deteriorated (normal baseline, n = 66 eyes, P = 0.013) or remained stable (prolonged baseline, n = 18 eyes, 95% confidence interval of change -5.42 to +6.89 ms, P = 0.805). Conclusion Prolonged P100 latencies in eyes never affected by clinical ON remain stable and thus can be used as surrogate outcome measure for remyelination trials. Multiple Sclerosis 2009; 15: 68-74. http://msj.sagepub.com