Introduction

Since the first reports by Halliday’s group in the early 1970s [1, 2], the changes in pattern reversal visual evoked potential (PVEP) in optic neuritis (ON) and multiple sclerosis (MS) have been well characterised and have been summarised in many review articles e.g. [3, 4]. Although it is well known that MS can be associated with retinal abnormalities [5], there has been a recent increase in interest in the vascular changes that can occur in MS and neuromyelitis optica (NMO) and, thus, a renewed interest in the electroretinogram (ERG) changes in these conditions [6].

An early report in 1966, of retinal function in advanced MS, described a reduced ERG in the majority of cases [7]. Auerbach subsequently reported changes in the ERG in patients with optic nerve disease (18/69 with optic neuritis); “enhanced ERGs” were present in 29 of 69 cases (42%) usually in both eyes, but a lower ERG amplitude in 38 cases (55%) [8]. The enhanced ERG was characteristically associated with a reduced or absent visual evoked potential (VEP). Speculation regarding the physiological substrate of this phenomenon initially focused upon the existence of inhibitory centrifugal optic nerve fibres synapsing with the amacrine cells. It was later reported that 22% of a series of “supra-normal” ERGs were patients with optic nerve pathology [9].

We have been unable to identify any reports addressing this issue since the introduction of the ISCEV standards. The present study describes the conventional full-field ERG, pattern ERG and PVEP findings in a retrospectively ascertained group of patients with unilateral optic neuritis, the data having been acquired since the introduction of the ISCEV standards.

Methods

Patients with unilateral optic neuritis were retrospectively identified from the Moorfields Eye Hospital Electrophysiology Department database (between 1998 and 2010). Patients were excluded if they did not have ERG, PERG and PVEP performed to ISCEV standards; had known co-existing retinal disease; had evidence of bilateral optic nerve disease clinically or electrophysiologically, or had been seen acutely (<3 weeks). The diagnosis was made by a consultant neuro-ophthalmologist and was based on typical clinical findings such as a sudden onset of unilateral visual acuity loss, pain on eye movement, a relative afferent pupillary defect and colour vision deficiency [10]. Patients were only included if the clinical course and electrophysiological findings were consistent with a diagnosis of optic neuritis and it was their first episode. None of the patients had either the primary or secondary progressive type of MS. An Ethics Committee opinion was not required for this retrospective data review.

ERG, PERG and VEP incorporated the ISCEV standard recommendations [1113]. After 25 min of dark adaptation, ERGs were recorded to white flash strengths of between 0.01 and 11.0 cd s m−2 (DA 0.01–DA 11.0, approx 0.6 log unit steps). After 10 min of light adaptation (30 cd m−2), generalised photopic cone system function was assessed by recording 30 Hz flicker (LA 30 Hz) and single flash cone ERGs (LA 3.0). Pattern ERGs evoked by high-contrast black and white checkerboard reversal were recorded using a field size of 15° × 10°; check size 45′; white luminance 100 cd s m−2, contrast ~95%; reversal rate 4.4 s−1. Pattern-reversal VEPs were elicited by 50 min checks in a field size of 13° × 10° using the same luminance and contrast parameters. Reversal rate was 2.2 s−1.

Regular calibration and service ensured that the stimulus parameters remained stable over the period of time covered by the study.

Results

Forty-six patients were identified, 63% of who were women, with an average age of 45 (range 23–68 years). Just over half (59%) of all patients carried a diagnosis of relapsing remitting MS as the cause of their optic neuritis. The Student’s t-test was used to compare the ERG, PERG and PVEP data from the clinically affected eyes with those from the fellow eye. Statistical significance was set at P < 0.05. The electrophysiological data are summarised in Table 1. See Fig. 1 for a representative example.

Table 1 Means, standard deviations and ranges of the principal parameters
Fig. 1
figure 1

34-year-old man presented with sub-acute vision loss in the right eye to 6/24, 6/6 in the left. Electrophysiology testing was performed 4 months after recovery. All left eye (LE) traces are normal. Right eye (RE) shows marked delay in the PVEP P100 component (arrow) and reduction in the PERG N95 component (arrow). RE ERGs are normal and do not significantly differ from those of the LE

ERG

Sixteen of the 46 patients had higher ERG DA 0.01 b-wave amplitudes in the clinically affected eye; only 5/16 showed a difference of greater than 20 μV and in no case did the magnitude of the inter-ocular asymmetry reach 30%, the criterion for an abnormal inter-ocular asymmetry within our laboratory. Of the remaining patients, 24 patients had lower amplitude ERGs in the affected eye; 6 showed no asymmetry. For the brighter flash ERGs, only 6 patients had an ERG DA 11.0 b-wave amplitude >550 μV, with 3 recordings being >600 μV. In the cone system, there were no abnormalities of LA 3.0 30 Hz flicker, nor the LA 3.0 ERG.

Overall, there was no statistically significant inter-ocular asymmetry in any ERG parameter. Further, there was no statistically significant difference between the ERGs of the patients with and without a diagnosis of MS.

PERG

The PERG N95 amplitude was significantly lower in the clinically affected eyes (P < 0.001). See Fig. 1. Nine patients had a mild abnormality of P50 (1.5–1.9 μV) in the clinically affected eye. No P50 abnormalities were found in the fellow eye. There was no significant difference in P50 between eyes (P = 0.09) and between patients with and without MS (P = 0.3).

PVEP

The mean PVEP P100 component peak time was 134.9 ms (SD 13.2) in the affected eye compared with a mean value of 108.5 ms in the unaffected fellow eye (P < 0.001). The mean P100 amplitude in the affected eye was 6.4 μV (SD 4.0) compared with 10.4 μV (SD 5.1) in the fellow eye (P < 0.001). The mean PVEP P100 peak time in the clinically affected eye was 140 ms in those patients with MS, compared with 125 ms in those without MS (P < 0.001). There was no significant difference in peak time between the clinically unaffected eye in those patients with and without MS (P = 0.12).

Discussion

It has been suggested by previous authors that ERG amplitudes are higher in eyes with optic nerve dysfunction. The present study of patients with unilateral optic nerve dysfunction, with or without a diagnosis of MS, is unable to confirm those observations.

No statistically significant difference was present in any ISCEV Standard ERG parameter between the ON eyes and the fellow eyes, and no patient showed an inter-ocular asymmetry of >30%. The possibility that subclinical optic nerve dysfunction in the fellow eye may dampen any asymmetry in the ERG was minimised by confining the study to patients with unilateral symptoms and a unilateral PVEP abnormality. Although MS is a systemic disease that may have separate subclinical effects in the retina of the ON and fellow eyes [14], there was no difference in the mean ERG and PERG values between the eyes in those patients with and without MS. There are some reports of reduced b-wave amplitudes in optic neuritis patients with and without MS, but no difference in a-wave amplitudes or peak-time [15, 16]. However, one of those studies [15] did not record ISCEV Standard ERGs, using surface electrodes, and the other [16] only performed photopic ERGs and found “reduction” in both affected and unaffected eyes compared with a normal control group. No such changes were observed in the present study population.

Auerbach and colleagues [8, 9, 17] used a b-wave cut-off value of 600 μV to define a “supernormal ERG” when they reported supernormal ERGs in optic nerve disease. Notwithstanding the need for each laboratory to establish its own normal ranges, that series was ascertained prior to the introduction of ISCEV Standard ERGs making direct comparison with more modern data even more difficult. The present study utilised the “suggested” DA 11.0 response and a value of >600 μV is not supernormal in a young person. By laboratory normals, no individual patient in the present series had a supernormal ERG.

It has long been established that a PERG abnormality confined to the N95 component can occur in optic nerve disease [18, 19] There are two main components of the PERG, N95, which arises in the retinal ganglion cells, and P50, approximately 70% of which arises in the retinal ganglion cells and which is “driven” by the macular photoreceptors [20, 21]. A large study reported that approximately one-third of 382 eyes with optic nerve demyelination had an abnormal N95:P50 ratio [22]. As unilateral disease was an inclusion criterion for the present study, the finding of a statistically significant difference in the PERG N95 component between the ON and fellow eyes was predictable, reflecting retrograde degeneration to the central retinal ganglion cells consequent upon the optic nerve insult. In addition, the present study did not allow the inclusion of patients with disease of less than 3 weeks duration in an attempt to ensure that the P50 data were not inappropriately affected by the transient reduction in that component that occurs in an acute ON [22], and which suggests macular involvement. Of clinical relevance, the degree of initial P50 abnormality in that earlier study related to final visual outcome and thus had prognostic value.

More recently, the photopic negative response (PhNR) has been associated with retinal ganglion cell function, and changes in the PhNR have been described in patients with optic atrophy [23]. A recent publication addressing the focal PhNR in optic neuritis [24] found attenuation of the focal macular ERG a- and b-wave amplitudes at onset of ON which had recovered by 6 months. The authors suggested that “inflammation at the onset of optic neuritis results in dysfunction extending to at least the inner nuclear layers, with recovery in all but the retinal ganglion cell layers”. The PERG data in acute optic neuritis, referred to above, [22] question that interpretation. That study also reported reversible macular involvement as part of acute optic neuritis, with recovery of PERG P50 paralleling that of the more recent focal macular ERG a- and b-waves, but also suggested, however, that there is direct ganglion cell involvement due to retrograde degeneration following the initial optic nerve insult rather than an acute trans-retinal inflammatory process with failure of ganglion cell recovery. That had previously been suggested by recordings from a patient examined, serendipitously, during the developing stage of an optic neuritis [25]. There has been recent confirmation of the possibility of macular involvement in optic neuritis, demonstrated both by optical coherence tomography (OCT) and by multi-focal ERG [26].

The present study, being retrospective, did not incorporate the most recent developments in imaging, particularly high resolution OCT. It will clearly be highly relevant and of great interest for future studies to examine high-resolution OCT data to determine any possible structural changes that may occur in conjunction with the functional changes demonstrated by electrophysiology, particularly given those recent observations of possible macular OCT changes in optic neuritis [26]. Reviews of OCT data in optic nerve disease have appeared [2730], as have studies on the relationship between structure and function in optic neuritis [3133].

The patients in this cohort were older, and contained a higher proportion of men (37% vs. 23% in the Optic Neuritis Treatment Trial [10]), than might be anticipated in a population of optic neuritis patients. This is likely to relate to referral patterns; in the “classical” young female patient with symptoms of sudden monocular visual loss accompanied by retrobulbar pain made worse on lateral eye movement, it is unlikely that electrophysiological confirmation would be requested. Nonetheless, the data should still be applicable to a more general series. Given a mean follow-up of 5 years, the incidence of 59% of patients with ON having a diagnosis of MS established prior to the final discharge letter is within the expected range [34].

To conclude, ISCEV Standard ERG data from eyes with optic neuritis did not significantly differ either from the uninvolved fellow eye or from normal, in contrast to previous studies performed prior to the introduction of the ISCEV Standard for ERG. Also in contrast to those early reports, none of the patients in the present series has a “supernormal” ERG.