Multifocal visual evoked potentials and contrast sensitivity correlate with ganglion cell-inner plexiform layer thickness in multiple sclerosis

Abstract

Objective: To examine the relationship between optical coherence tomography (OCT) macular ganglion cell-inner plexiform layer thickness (GCIPLT), peripapillary retinal nerve fiber layer thickness (RNFLT) and visual function in relapsing-remitting multiple sclerosis (RRMS).

Methods: Cirrus OCT, VERIS 60-sector multifocal visual evoked potential (mfVEP) and Pelli-Robson contrast sensitivity (CS) were obtained for 53 eyes with last optic neuritis (ON) > 6 months and 105 non-ON eyes in 90 patients. One eye (43 ON, 73 non-ON) was used for correlations when both had the same history. Global (G, 60 sectors) and central 5.6° (C, 24 sectors) mfVEP amplitude and latency were calculated as mean logSNR and median latency.

Results: Eyes showing abnormal mfVEP (amplitude or latency) vs OCT (GCIPLT or RNFLT) was 77% vs 69% (p = 0.33) in ON, 45% vs 22% (p < 0.0005) in non-ON. In ON and non-ON, mfVEP measures and CS correlated with GCIPLT and RNFLT (r = -0.24 to 0.78, p = 0.03-0.0001). In ON, mfVEP amplitude (C,G) correlated better with GCIPLT (r = 0.78, 0.76) than RNFLT (r = 0.43, 0.58; p < 0.001, 0.01).

Conclusions: MfVEP measures and CS correlated well with GCIPLT and RNFLT in ON and non-ON. MfVEP amplitudes were more highly correlated with GCIPLT than RNFLT in ON. MfVEP detected significantly more defects than OCT in non-ON.

Significance: GCIPLT, mfVEP and CS provide useful measures of optic nerve integrity in RRMS.

Keywords: Contrast sensitivity; Multifocal visual evoked potential; Multiple sclerosis; Optic neuritis; Remyelination; Visual function.

Figure 1.

(a) The mfVEP stimulus contains 60 sectors of black and white checks (one sector shown in red). (b) The mfVEP responses from the right eye (blue) and the left eye (red) of a normal subject are essentially identical. (c) The mfVEP responses from an MS subject show prolonged latencies in the right eye (blue) compared to those form the left eye (red). (d) The interocular latency probability plot shows a cluster of prolonged latencies in the right eye (dark blue for p < 0.01, light blue for p < 0.05) compared to the left eye of an MS patient who had an ON attack in the right eye two years prior and no history of ON in the left eye. Black symbols denote no significant differences compared to the ‘Portland’ normative database, small gray symbols for responses too small for comparisons (Hood et al., 2003, Fortune et al., 2004, Hood et al., 2004b).

Figure 2

Scatter plots illustrating correlations between ganglion cell-inner plexiform layer thickness (GCIPLT) and functional measurements in ON (filled triangles) and non-ON eyes (open circles). (a) GCIPLT vs central mfVEP amplitude (logSNR), (b) GCILPT vs central mfVEP latency, (c) GCIPLT vs CS (log units), (d) GCILPT vs 10–2 RVS. The dashed and solid lines are fitted linear regression lines for ON and non-ON eyes, respectively. The tick marks on the X and Y axis indicate previously established normative values (see Table 2). In ON eyes, mfVEP amplitude showed stronger correlation with GCIPLT (r = 0.78, 0.76 for central and global amplitude) than RNFLT (r = 0.43, 0.58 in Figure 3; p < 0.001, 0.01 after Bonferroni correction for central and global amplitude).

Figure 3

Scatter plots illustrating correlations between retinal nerve fiber layer thickness (RNFLT) and functional measurements in ON (filled triangles) and non-ON eyes (open circles). (a) RNFLT vs mfVEP global amplitude (logSNR), (b) RNFLT vs mfVEP global latency, (c) RNFLT vs CS (log units), (d) RNFLT vs 24–2/30–2 RVS. The dashed and solid lines are fitted linear regression lines for ON and non-ON eyes, respectively. The tick marks on the X and Y axis indicate previously established normative values (see Table 2).