Evolution of Cortical Visual Evoked Potential Abnormalities after Optic Neuritis

A typical episode of optic neuritis entails acute monocular visual impairment, ranging in severity from mild cloudiness to complete blindness. This is usually associated with the appearance of a local region of high T2 signal in magnetic resonance imaging (MRI) of the optic nerve that shows enhancement with gadolinium (Youl et al., 1991) and is believed to reflect blood-brain barrier leakage and inflammation. If cortical visual-evoked potentials (VEPs) are recorded at this time, in response to the sudden reversal of a high-contrast grating or checkerboard pattern, they are usually found to be attenuated, roughly in proportion to the visual acuity deficit (reviewed by Halliday, 1993). That the period of VEP attenuation coincides approximately with that during which gadolinium enhancement occurs (Youl et al., 1991), and that VEPs recorded immediately after that period are typically delayed, suggest a likely mechanism of temporary axonal conduction block resulting from factors associated with inflammation and/or demyelination. Furthermore, the observation that both visual acuity and VEP amplitude usually recover to near-normal values within a period of a few weeks indicates that neither axonal transection nor primary neurodegeneration is the prime cause of acute visual failure in optic neuritis.

In a minority of cases, visual function remains persistently impaired and VEPs remain attenuated, even after the resolution of inflammation. This can probably be ascribed to axonal damage, although whether it occurs as an intraneuronal process or as a consequence of demyelination and inflammation is unclear. Attempts have been made to discover electrophysiological or anatomical (MRI) factors that may predict a good or poor recovery (Kapoor et al., 1998), so that patients with a poor prognosis might be targeted for treatment, but so far no such indices have been found. In any case, the bulk of evidence suggests that the benefits of anti-inflammatory treatment are transitory, causing an acceler ated recovery but with few if any positive consequences in the longer term (Beck and Cleary, 1993; Kapoor et al., 1998).

Once the visual acuity has improved to some degree and the VEP amplitude has recovered sufficiently for a response to be measured, the latency of the major positive peak (P100) is usually found to be prolonged, typically by between 10 and 40 ms (10% to 40%; Halliday et al., 1972). It seems reasonable to assume that the magnitude of the delay should reflect the longitudinal extent of demyelina-tion, although to date no significant correlation has been demonstrated between VEP latency and the length of the lesion seen on MRI. For a typical lesion length on the order of 30 mm, and assuming that the extent of increased T2 signal reflects the length of the demyelinated axons, a VEP delay of 30 ms would imply a conduction velocity through the lesion of approximately 1 m/sec. This velocity is comparable to the conduction velocity of normally unmyelinated fibers, although their mechanism of action potential propagation may not be entirely the same. It also seems reasonable to assume that the velocity of conduction through the lesion should be related to the thickness of the individual myelin sheaths. The conduction velocity of normal nerve fibers is roughly proportional to both axonal diameter and myelin thickness, but these two parameters are closely associated and there is no directly relevant evidence as to how conduction velocity may be affected when they become dissociated by disease.

Both acutely and in the longer term, the degree of VEP amplitude reduction is loosely related to the severity of visual acuity impairment (e.g., Jones, 1993a); however, no such correlation has been shown between the severity of the visual deficit and the magnitude of the VEP latency delay (e.g., Sanders et al., 1987; Jones, 1993a). The latency may remain prolonged for months or years, even after the visual acuity has recovered virtually to normal. It appears that, as long as axons of the optic nerve remain in functional continuity, the time required to conduct impulses from the retina to the brain is virtually irrelevant as far as the spatial resolving power of vision (at least for static objects) is concerned. The tendency for VEP delay to persist after the partial or complete recovery of visual acuity is one important reason why VEPs remain a useful diagnostic test for a past episode of optic neuritis. It is also a clear sign that the myelin sheath may be affected independently of the optic nerve axon.

Despite the fact that the VEP usually remains delayed long after the visual acuity has recovered, however, there is a tendency for its latency to decrease over the ensuing months and years (Fig. 1). In a minority of the MS cases studied serially by Matthews and Small (1979, 1983), initially prolonged VEP latencies returned to normal, sometimes after several years. Normalization has since been shown to be the prevalent tendency, although it is not seen in every individual (Hely et al., 1986).

Weeks Left eye Right eye after

Weeks Left eye Right eye after

Figure 1 Serial VEPs to central field (4-degree radius) stimulation after an episode of right optic neuritis showing progressive shortening of the P100 latency, particularly between 50 and 109 weeks after onset. The right eye responses obtained after 109 and 160 weeks were within normal latency limits in absolute terms, although the interocular latency difference was still abnormal. For guidance the vertical bar is placed at 100 ms; the upper limit of "normal" for this system is approximately 108 ms.

Figure 1 Serial VEPs to central field (4-degree radius) stimulation after an episode of right optic neuritis showing progressive shortening of the P100 latency, particularly between 50 and 109 weeks after onset. The right eye responses obtained after 109 and 160 weeks were within normal latency limits in absolute terms, although the interocular latency difference was still abnormal. For guidance the vertical bar is placed at 100 ms; the upper limit of "normal" for this system is approximately 108 ms.

In the cross-sectional data of Jones (1993a), more than 70% of patients had delayed VEPs when tested more than 2 years after the onset of symptoms, compared with more than 90% whose responses were delayed during the first 6 months. Latency reduction is not usually apparent during the first 3 months after the onset of symptoms (during which time any such effect may be masked by changes resulting from the resolution of inflammation), but is most clearly manifested between 3 and 12 months (Jones, 1993a) (Fig. 2). Latency shortening has also been found to be significant during the second year (Brusa et al., 2001), associated with only a mild improvement in visual acuity. In a small patient group monitored for 3 years, latencies were on average still shorter in the final examination (Brusa et al., 1999).

Although other possible mechanisms need to be considered, it seems reasonable to suppose that, if the VEP latency delay reflects the degree of demyelination, a gradual reduction of that delay may reflect a long-term process of remyeli-nation. There seems to be no way in which such a change can be ascribed to axonal regeneration. Alteration in the distribution of sodium channels may be an important adaptive process enabling conduction to be restored in demyelinated axons (Craner et al., 2003) (see Chapter 7), but would not be expected to result in a complete normalization of conduction velocity. Latency normalization is also extremely difficult to explain in terms of other adaptive mechanisms such as cortical reorganization (Werring et al., 2000).

0 0

Post a comment