Purpose: To evaluate longitudinal changes of visual function in relapsing-remitting multiple sclerosis (RRMS).
Methods: Multifocal visual evoked potential (mfVEP), contrast sensitivity (CS), and Humphrey visual fields (HVFs) were obtained at two visits (mean follow-up, 1.5 [±0.9] years) in both eyes of 57 RRMS patients (53 eyes with optic neuritis [ON]: 14 ON within 6 months of first visit [ON < 6 months] and 39 ON ≥ 6 months; 57 non-ON). Longitudinal changes were assessed using mfVEP amplitude (log signal-to-noise ratio [logSNR]), latency, CS, and HVF mean deviation based on established 95% tolerance limits of test-retest variability.
Results: A significant percentage of eyes in the ON < 6 months group exceeded 95% tolerance limits for mfVEP logSNR (21%, p < 0.05), latency (35%, p < 0.01), and CS (31% p < 0.001); more improved than worsened over time (14% vs. 7% for logSNR, 21% vs. 14% for latency, and 31% vs. 0% for CS). Multifocal visual evoked potential latency decreased in 11% of non-ON eyes and in 10% of eyes in the ON ≥ 6 months group, and increased in 21% and 10%, respectively (p < 0.01 for all). Latency changes correlated negatively with baseline latency (r = -0.43 and -0.45 for non-ON and ON ≥ 6 months; p = 0.0008). Although a nonsignificant percentage of non-ON and ON ≥ 6 months eyes exceeded tolerance limits for logSNR, CS, or HVF, logSNR and latency changes correlated, and both measures correlated with changes in CS (r = 0.47 to 0.79, p < 0.01).
Conclusions: Multifocal visual evoked potential, particularly latency, is potentially useful for assessing neuroprotective and remyelinating strategies in RRMS.