Inherited Retinal Diseases

Characteristic changes in the ERG, EOG, and/or VEP occur in some inherited diseases and may be useful in diagnosing or evaluating these conditions.

RETINITIS PIGMENTOSA

Since the late 1940s, it has been knoWU that the ERG shows marked changes in patients suffering from retinitis pigmentosa. Even at an early stage of the development of this type of retinal degeneration, the ERG is more or less completely abolished; the light rise of the EOG is similarly affected.

Most patients who show the typical fundus changes of retinitis pigmentosa have all extinguished ERG (Fig. 11.10). Patients in the early stages of this disorder clearly have some response; in fact, a close relationship probably exists between ERG sensitivity and the functioning area of the retina. Patients with a relatively well preserved ERG often have a history of late onset of a mild form or the disease. In some series in which early cases have been examined, the ERG has been almost normal. The early changes aro usually seen in the scotopic ERG, particularly if this is recorded with a red-light stimulus.

Figure 11.10. The ERG and visual field in retinitis pigmentosa. Typically, the ERG response is absent and the field is constricted, as shown.

The ERG has been used to distinguish types of retinitis pigmentosa presenting with predominant damage to the rods from those showing early damage to the cones (rod/ cone or cone/rod degenerations). Now that a number of mutations and deletions in the rhodopsin gene have been discovered in patients with these conditions, attempts are being made to link phenotype and genotype more closely. Carriers may also show a reduced amplitude of the ERG and the EOG, and this may be important confirmatory evidence for a mother-to-be. Often, of course, a firm diagnosis of retinitis pigmentosa can be achieved with the ophthalmoscope alone. In spite of this, one cannot claim that such a case has been properly documented until a record has been made of the visual field and ERG. If the ERG is totally abolished, there is little point in repeating the test subsequently, unless some new form of treatment is being tried and it is necessary to monitor the disease objectively.

Patients with retinitis pigmentosa have a normal VER unless the disease is advanced or unless there is early macular involvement.

SYSTEMIC DISEASE ASSOCIATED WITH RETINITIS PIGMENTOSA

A wide variety of diseases have been as-sociated with retinitis pigmentosa and some of the other progressive degenerations related to it. Although electrodiagnostic investigations cannot now distinguish between classic retinitis pigmentosa and that associated with systemic disorders, they have a particular value when applied to the latter. Sometimes, the fundus changes are minimal in these patients, and they may be referred from the neurologist or the physician with the question "Are the eyes normal?" The ERG and the EOG can provide an immediate answer. Apart from this, the vision may be affected by other pathologic processes.

Figure 11.11 shows the ERG report for a patient with ill-defined peripheral constriction of the visual fields and enlargement of the pituitary fossa. She also had polydactyly, having had the extra finger removed in infancy. The fund! showed one or two irregular flecks of pigment in the periphery. The problem was deciding whether the field defect was caused by chiasmal compression or early retinitis pigmentosa in association with polydactyly, the Laurence-Moon-Biedl syndrome being suspected.
The systemic disorders associated with progressive retinal degeneration include the following conditions:

1. Metabolic disorders
   a. Lipid abnormalities
       i.   a, b-lipoproteinemia
       ii. Refsum's disease
       iii. Familial amaurotic idiocy
2. Neurologic disorders
   a. Laurence-Moon-Beidl syndrome
   b. Hereditary ataxias
   c. Ocular myopathy
   d. Syndromes involving mental retardation
3. Occasional associations
   a. Dermatologic disorders
   b. Megacolon
   c. Marfan syndrome
   d. Familial nephropathies
   

Figure 11.11. The ERG report for a patient with ill-defined peripheral constriction of the visual fields, enlargement of the pituitary fossa, and polydactyly. The absence of an ERG response, consistent with retinitis pigmentosa, and the other signs point to a diagnosis of Laurence-Moon-Biedl syndrome.

Familial amaurotic idiocy, although characterized by progressive degeneration of the retina, is probably a completely separate entity. The ERG in the infantile form has been reported as normal by several workers. This finding is consistent with the pathologic changes, which are restricted at first to the ganglion cell layer and spare the outer parts of the retina whence the ERG arises. In late infantile and juvenile amaurotic idiocy, however, the ERG may be reduced or absent.
All the other conditions listed above are associated with retinitis pigmentosa, and the ERG and the EOG are affected accordingly, Several of these conditions show specific biochemical abnormalities, which may throw some light on the cause of the pigmentary degeneration. For example, the association of a, b-lipoproteinemia, retinitis pigmentosa, ataxia, and acanthocytosis of the red cells also involves a lowering of the blood level of the fat-soluble vitamins, including vitamin A. It has been claimed that vitamin A supplements in sufficient dosage to raise the vitamin A level to normal also restore the dark-adaptation curve and the ERG to normal. In the mucopolysaccharidoses, vision may be impaired by infiltration of the cornea, and the fundus may not be visible, Electrodiagnostic tests may thus be the only way of detecting associated retinitis pigmentosa. It should be apparent that because retinitis pigmentosa may sometimes present with minimal changes in the fundi, an ERG is essential if any of the above conditions are suspected.

Figure 11.12 shows the serial ERGs in a patient with Refsum's disease. Here the ERG was recorded annually because the patient was maintained on a special diet. No improvement in the response is visible.

Figure 11.12. Periodic ERGs recorded from each eye in a patient with proven Refsum's disease. In this patient, the ERG was used to monitor the effectiveness of treatment; the ERG has started to decline between 1984 and 1992.

The ERG and the EOG have been investigated in a wide variety of conditions related to retinitis pigmentosa. For more details, the reader is referred to the more specialized textbooks listed in Suggested Readings.

LEBER'S AMAUROSIS

Leber's amaurosis was first described by Leber in 1869. He classed the condition with
pigmentary degenerations of the retina, even though the pigmentation was minimal or appeared at a late stage. The condition is usually congenital, with the affected child being blind from birth, but some become blind during the first year of life. The fundi may appear normal, but a variety of minor changes have been described from fine pigment stippling to choroidal sclerosis. On occasion, the classical appearance of retinitis pigmentosa may be observed.

The advent of electroretinography has made it possible to distinguish children with Leber's amaurosis from those with optic nerve atrophy and Tay-Sachs disease. Since it is usually necessary to examine the fundi under a general anesthetic, it is important to take the opportunity to perform electroretinography at the same time. If a general anesthetic is contraindicated, it is often surprisingly easy to insert a contact lens without an anesthetic in a child up to 3 or 4 months old. The ERG is abolished or very small in Leber's amdurosis but normal in optic nerve atrophy and Tay-Sachs disease.

Sometimes, these children present as having congenital nystagmus. In such cases, the ERG under general anesthesia may be crucial in making the diagnosis. The diagnosis must be made as soon as possible, so that the parents can be properly advised about the future education of the affected child.

HEREDOMACULAR DYSTROPHIES

The heredomacular dystrophies are a group of diseases characterized by bilateral macular degeneration, a hereditary tendency, and the absence of associated disease in the central nervous system. These diseases have been classified and named according to the age of the onset, but they may all be one and the same condition. They may be listed as follows:

Infantile heredomacular dystrophy (Best's disease or "vitelline dystrophy")
Juvenile heredomacular dystrophy (Stargardt's disease)
Adult heredomacular dystrophy (Behr's disease)
Presenile and senile heredomacular dystrophies

As a general rule, these patients present with a gradual deterioration of their central vision. Children may have difficulty in reading or seeing the blackboard at school, which is not corrected by wearing glasses. The fundus appearance varies considerably from case to case. In Best's disease, a round or oval lesion is seen at the macula, which has a yellowish color and has been likened to the yolk of an egg—hence, the term "vitelline dystrophy." The vitelline lesion evolves into a pigmented scar. Rather surprisingly, the vision of these patients may remain normal in spite of the fundus appearance. In Stargardt's disease, which usually appears between the ages of 8 and 11 years, the vision may be impaired when the fundus is normal. The earliest change is disappearance of the normal foveal reflex; gray, yellow, or brown spots may appear at the macula. Eventually, an oval circle of pigment stippling is seen; occasionally, this may spread to involve the entire posterior pole, Senile macular degeneration is associated with degenerative changes in the underlying choroid and Brach's membrane. Although It may bear some resemblance to the types that occur at an earlier age, it may be complicated by the presence of hemorrhages and subretinal exudates. Sometimes, senile macula degenerations are divided into dry and wet types, referred to as degeneration of Haab and disciform degeneration, respectively.

It has been shown that if the macular area in the monkey is photocoagulated, the ERG obtained from this damaged eye is normal. Furthermore, a normal ERG has been described in cases of solar retinopathy. It is not surprising, therefore, that early reports revealed normal ERGs in patients with heredomacular dystrophy. However, when examination is carried out using a red-light stimulus, a high percentage of patients show a reduced amplitude of the 'b' wave. The photopic ERG is normal, but its spectral sensitivity curve may be displaced toward shorter wavelengths. The foveal ERG is subnormal in all these patients, including those whose visual acuity is still fairly good. The VER has also been shown to be abnormal. This might be expected considering the relatively large macular representation on the occipital cortex.

Some slightly unexpected changes have been described in the EOG in cases of macula degeneration. Several different sources report that the EOG may be markedly abnormal in patients with vitelline dystrophy; the ERG in these patients is usually normal. In Stargardt's disease, the EOG is usually normal unless the retinal periphery is involved.

Therefore, in contrast to early reports, the ERG and the EOG may be abnormal in patients with macular degeneration, but the difficulty still remains that a physically minute lesion can cause a serious disturbance of vision in macular disease. Thus one sometimes finds relatively minor changes in the electrical responses even with severe impairment of visual acuity. The foveal ERG is a promising technique, but its value is limited when opacities in the media scatter the stimulus light.
The VEP may be of some help in monitoring the progressive nature of these conditions. Figure 11.13 shows the VEP obtained from a child with a lesion at one macula, with a doubtful history of contusion injury. In this particular patient, the history of injury was probably irrelevant because there was a family history of macular disease, the child's brother being affected. The VEP shows specific changes that were repeatable from month to month.

Figure 11.13. The VEP from a patient with a right-sided macular hole in response to large-check (bottom) and small-check (top) stimuli. Note the impaired response to small checks in the affected eye.

ALBINISM

Hereditary albinism became of great interest to the electrophysiologist following the work of Guillery, who showed that the normal crossover of nerve fibers at the chiasm is greatly altered in many albino mammals. VEP studies have subsequently shown that the same abnormality exists in human albinos. The normally uncrossed fibers from the temporal retina undergo decussation to a greater or lesser degree in albinos; this abnormality is seen as asymmetry in the amplitude of the VEP as recorded from each side of the scalp. Although these findings have not so far been backed up by equivalent psychophysical research, the VEP studies have been extensive.

Acquired Retinal Diseases

DIABETIC RETINOPATHY

It has been known for many years that the classic ERG waves are not affected until a late stage in diabetic retinopathy, and even then, the reduction in amplitude does not show any features that might be specific for diabetes. However, a renewed interest in the subject was created by Yonemura and Kawasaki, and others, who described the selective disappearance of the oscillatory potential. Similar changes were also described in some other circulatory disturbances of the retina. A recent report involving a large series of eyes has shown beyond doubt that this component of the ERG may be absent; furthermore, there is no doubt that ERG changes also occur early in diabetic retinopathy. The exact point in the development of the retinopathy at which the oscillatory potential is affected differs markedly in different series, but this is probably due to differences in the type of stimulus used.

Therefore, selective loss of the oscillatory potential appears to be a consistent sign of diabetic retinopathy (Fig. 11.14). This change is often accompanied by a reduction in the size of the 'b' wave and the 'a' wave; in advanced diabetic retinopathy, a small 'b' wave alone may persist. Now that vitreous surgery has become better developed, it is often important to assess the retinal function in advanced cases. The ERG must be interpreted with great care under such circumstances because sometimes a small island of healthy retina may remain at the posterior pole and yet the ERG may be poor. A very small 'b' wave is therefore not a contraindication to vitreous surgery, but a well-developed response would suggest a good prognosis.

Changes in the VEP in diabetic cerebrovascular disease so far have not been accurately assessed.

Figure 11.14. The ERG from a patient with advanced diabetic retinopathy. Note the absence of the oscillatory potential. The ERG reflects diffuse retinal function, not macular function; thus, even when small areas of healthy retina remain, the ERG may be weak or absent.

OCCLUSIVE VASCULAR DISEASE

From the early days of clinical electroretinography an interest has been shown in the effect of retinal vascular disease on the electrical response of the eye. As described above, striking changes can occur in patients with diabetic retinopathy. In this section, the electroretinographic changes that may be seen in other types of vascular disease of the retina are described.

OCCLUSION OF THE CENTRAL RETINAL ARTERY. The retina may be regarded as having a double blood supply: the inner half being supplied by the central retinal artery and the outer half being nourished from the choroidal circulation. Obstruction of the circulation of the central retinal artery might, therefore, be expected to affect that part of the ERG response derived from the inner half of the retina and spare those components that originate from the receptors and the pigment epithelium.

Two characteristic features of the ERG usually present in central artery occlusion are (a) loss of the oscillatory potential and (b) a "negative" type of ERG with enlargement of the 'a' wave and no change or slight diminution of the 'b' wave. However, the more severe types of occlusive episode, those with a poor prognosis, generally show a marked reduction in the size of the 'b' wave, whereas the milder types of occlusion may show a "negative" response.

In branch artery occlusions, the ERG may be normal or minimally affected. The oscillatory potential may be reduced in amplitude or even absent in branch artery occlusions. It has been claimed that the wavelets give a good indication of the prognosis in any given case.

OCCLUSION OF THE CENTRAL RETINAL VEIN. The findings in venous occlusion are similar to those in central retinal artery occlusion, the most common change being a subnormal negative response and diminution of the oscillatory potential. The changes, however, are milder than in arterial occlusion.

Sometimes, patients are referred to the electrodiagnostic clinic with unexplained field defects and systemic hypertension. The ERG can help to decide whether these defects are caused by retinal ischemia or proximal changes in the optic nerve.

Diseases of the Optic Nerve Head

ERG IN OPTIC NERVE ATROPHY

For many years, it has been generally accepted and confirmed that damage to the optic nerve does not affect the electrical response from the retina as measured by routine methods. Histologic evidence also indicates that the part of the retina that is thought to give rise to the ERG response is not damaged with optic atrophy. This fact creates a serious pitfall in the interpretation of ERG traces; an eye may be completely blind from glaucoma and yet show a normal ERG. In fact, minor changes in the photopic components of the ERG have been described in chronic glaucoma, but these are slight.

In traumatic optic atrophy, the VER shows changes in proportion to the amount of visual loss, whereas the ERG remains normal. The ERG has also been reported as normal in patients with congenital unilateral hypoplasia of the optic nerve. The observation that some patients with optic nerve atrophy seem to have a supernormal response has excited some interest, and there are several authentic reports of this phenomenon. It has been suggested that the increase in amplitude of the ERG may be caused by the division of centrifugal fibers in the optic nerve; these fibers may normally have an inhibitory influence on the size of the response.

In general, any changes that are seen in the ERG as evoked by an unstructured flash of light appear to be minimal in optic nerve atrophy. As explained above, the use of a patterned stimulus gives a very different result in patients with optic atrophy. It appears that this type of stimulus produces a re-sponse that may arise partly in the ganglion cells.

VEP IN RETROBULBAR NEURITIS

A high incidence of abnormal-pattern VEPs in patients suffering from retrobulbar neuritis was, described by Halliday, McDonald, and Mushin. This finding in itself would not have been of clinical interest except that more subtle changes in the VEP persisted long after the clinical signs of optic neuritis had subsided. A test of previously healed optic neuritis is of more value to the clinician than a test of active disease, which is already detectable by routine clinical methods. Therefore, it was of special interest when Halliday and several others showed that the latency of the major peaks in the transient VEP with a pattern-reversal stimulus is increased and may remain increased for several years following an acute attack. This delay in the response is best shown by comparing the response from the two eyes.

During an acute attack of retrobulbar neuritis, when the visual acuity is severely impaired, the VEP is severely affected. In some cases, the response from the affected eye is abolished altogether. As the vision recovers, the amplitude of the VEP returns toward its normal value, but a characteristic slight delay in the response remains.

The VEP recorded from a patient with acute retrobulbar neuritis is shown in Figure 11.15. It can be seen that the response from the right side is greatly impaired. The VEP recorded from the same patient 4 months later is also shown; although the response from the right eye has apparently recovered, careful inspection shows a delay in the latency of all the major components. It is this slight change that enables the electrodiagnostician to say whether a patient has suffered from retrobulbar neuritis in the past, and the change may persist after other clinical evidence of the attack has disappeared.

Figure 11.15. The VEP in acute and healed retrobulbar neuritis. a, The acute phase, showing a greatly diminished response in the affected right eye. b, Healed phase, 4 months after previous trace. Although the amplitude of the response in the right eye has recovered, the latency of all the main components is slightly greater than in the unaffected left eye.

A surprisingly large number of patients with multiple sclerosis who have supposedly normal eyes give abnormal results when the VEP is checked. This abnormal response is not, of course, specifically seen in demyelinating disease; other possible causes of altered response such as amblyopia or macular disease must be excluded. Other evoked responses have also been investigated in multiple sclerosis, and measurement of a variety of evoked responses can provide diagnostic evidence in suspected cases.

VEP IN TOBACCO AMBLYOPIA

The amplitude of the VEP is sensitive to changes in visual acuity during an acute attack of optic neuritis, in which the response may be abolished altogether. Similar findings have been recorded in patients with tobacco amblyopia, in whom it is possible to monitor recovery by measuring the VEP at intervals when the patient has abstained from tobacco. In some of these patients, the normal positive peak at about 100 msec may be inverted.

VEP IN CHRONIC GLAUCOMA

Attempts to relate VEP changes to visual field defects have been described above. A purely objective test for field change in glaucoma could prove useful because election of glaucoma surgery may depend on an increase in field loss. Clear-cut alterations in phase and amplitude hae been demonstrated in the VEPs of patients with glaucomatous field defects. This was achieved by examining steady-state VERs to pattern-reversal stimulation of retinal areas corresponding to discrete field quadrants.

One of the problems in using the VEP to monitor chronic simple glaucoma or to detect early disease is that the VEP only measures a small central area of the visual field. For this reason, patients with advanced chronic glaucoma may have a normal VEP.

In fact, the VEP is likely to be more useful in monitoring cases over a period of months rather than in detecting early cases. This is because the changes in the VEP are rather unpredictable: Sometimes a severe case of glaucoma exhibits few changes, and sometimes a fairly mild case may show more marked changes that do not always seem to be consistent with the field defect. The VEP tracings, however, should remain the same over a period of a few months, as long as the glaucoma has not advanced (Fig. 11.16).

Figure 11.16. The VEPs from a patient with chronic simple glaucoma recorded on two separate occasions. Note the consistency in these tracings, which were obtained about 8 months apart; the visual field also remained unchanged. These findings indicate the disease had not progressed noticeably.

Following the discovery that the pattern ERG arises in part from the ganglion cell layer, there has been considerable research interest in the early detection of glaucoma, Clinicians specializing in glaucoma wish to know which patients with ocular hypertension are likely to go on to develop optic nerve damage. Some progress has been made in this respect. It appears that the pattern ERG can be useful if combined with other psychophysical tests.

VEP IN OTHER DISEASES OF THE OPTIC NERVE

In Tay-Sachs disease, the VEP usually is absent, whereas the ERG is normal. One might expect these findings from a lesion at the level of the retinal ganglion cells, which Involves central vision.
The function of the optic nerve can be monitored by means of the VEP, and this technique has been praCticed in orbital surgery. Permanent loss of vision following Surgery is a recognized risk, particularly when hypotensive anesthesia is being used. Wright, Arden, and Jones have described monitoring the function of the optic nerve by continuously stimulating the retina with an unstructured stimulus and recording the VEP throughout the operation.

Knowledge of the electrodiagnostic changes in diseases of the optic nerve is essential when interpreting these responses in general. A surgeon who is encouraged to remove a cataract on the grounds of a normal ERG alone may be unpleasantly surprised at the poor visual result if he or she has been led astray by an inadequate report.

Drug-Induced Retinopathy

A considerable amount of information is now available about the effect of various drugs on the electrical responses from the eye. These data have come from animal experiments and from cases of overdosage or accidental ingestion in humans. Electrodiagnostic tests have been applied in such circumstances to find out more about the nature and origin of the response and, in some cases, to help make a diagnosis. These techniques are also being used in an attempt to localize the site of action of drugs in the eye. Undoubtedly, the nature of these electrical changes is such that they can give useful objective clinical information, but our knowledge is still limited. Much of the work so far has concerned the ERG rather than the VEP.

For many years, much interest has been centered on drugs used to produce selective damage to different layers of the retina. Sodium Iodate, for example, causes selective damage to the pigment epithelium and abolishes the 'c' wave in rabbits. Another sub-stance that has been used in localization studies is sodium glutamate, which causes loss of vision in mice and selective loss of the inner layers of the retina, with destruction of the hipolars and ganglion cells. Several drugs have been used to assess the site of origin of the ERG, Table 11.1 shows the effect of some drugs on the ERG.

QUININE

Although the toxic effects of quinine have been well recognized for many years, the exact mode of adios of this poison is still controversial; electrodlagnostic tests help to throw some light on the problem. An over-close of quinine is usually taken in an attempt to procure an abortion, but cases have been reported In which an overdose was ingested as a prophylactic for malaria.

Symptoms may follow after taking as little as 1 g of quinine in sensitive individuals, but the usual dose to cause blindness is 1.5 to 4 g. The affected patient experiences deafness, tinnitus, and visual failure; larger doses produce coma. Although the fundus may appear normal at first, in some cases, there is retinal edema with a cherry red spot at the macula. The visual fields become grossly constricted. The symptoms and signs often improve over a period of weeks; however, the fundus then begins to show optic atrophy and narrowing of the retinal arteries.

For many years, arguments have been put forward to decide whether the toxic effect of quinine is primarily on the retinal vessels or whether it acts directly on the retina. The role of vascular spasm is not easy to assess because almost any condition that causes optic atrophy also causes constriction of the retinal vessels.

Investigation of the effect of acute quinine poisoning in animals has shown that the ERG is initially depressed but recovers in a matter of hours; after this, a further slow deterioration in the response occurs. The relatively slight degree of these changes compared with the severe visual loss indicates primary damage to the ganglion cells and the nerve fiber layer. In humans, the ERG may show a gradual decline between 10 days and 10 months after quinine intoxication; paradoxically, the vision may gradually improve during this period.

The EOG has shown an absent light rise during the first few days after intoxication; this has then recovered in parallel with the subjective improvements. The absence of a light rise on the EOG is not easy to explain in terms of a ganglion cell poison, and more cases will have to be investigated in detail to resolve this conflicting evidence.

The clinical picture of quinine intoxication may mimic that of the bilateral optic neuritis that sometimes is the presenting sign of multiple sclerosis, especially in young females. That is, in both conditions, the fundus may be completely normal, and the gradual recovery of vision is followed by optic nerve atrophy and narrowing of the retinal arterial vessels.

CHLOROQUINE (RESOCHIN)

Chloroquine was used extensively but in small doses in World War II as a prophylactic treatment for malaria. Corneal changes resulting from this drug were described at the end of the war, but the more serious retinotoxic effects were not observed until large doses were employed in the treatment of disseminated lupus erythematosus and rheumatoid arthritis. Its value in the treatment of disseminated lupus erythematosus was described in 1954, and the first case of chloroquine retinopathy was described in 1957.

Since then, reports from many different centers have confirmed that when the total annual dose of chloroquine exceeds 200 g per year, there is a risk of permanent visual disturbance with associated changes in the fundus. The earliest sign of a perimacular pigmentary disturbance is described as a "bull's eye appearance." In more advanced cases, the arteries become attenuated, and peripheral pigmentation may appear. By the time fundus changes appear, irreversible field defects can be detected, and in some cases, a progressive deterioration of vision occurs in spite of cessation of treatment. Histopathologic investigation of human eyes and animal experiments have shown that chloroquine accumulates in and damages the receptor layer and the pigment epithelium, thus confirming the cumulative nature of the toxicity.

In some electroretinographic studies of patients treated with chloroquine for more than a year, the amplitude of the 'b' wave was depressed; in other studies, however, the ERG was normal in the presence of fundus changes. One long-term follow-up study of 15 patients with chloroquine poisoning suggests that electroretinography may be of some prognostic value in such cases.
The effects of chloroquine on the EOG were first described by Arden and Kelsey in 1962. In a detailed study, Kolb showed that depression of the light rise of the EOG is often an early sign of toxicity and may some-times precede the fundus changes; however, in a series of 47 cases, there was considerable overlap between treated and control groups. Furthermore, in a group of patients who had not received treatment but who Buffered from collagen disease, the mean value of the Arden index was below normal, although not as low as in the group that had been treated with chloroquine. A study of patients in whom chloroquine therapy had been stopped revealed a return to normal in the Arden index in many cases (Fig. 11.17).

Figure 11.17. A. The electro-oculogram from a patient who had taken chloroquine for 2 years and who complained of blurred vision but had no ophthalmoscopic or slit-lamp evidence of toxicity. B. The EOG from the same patient taken several months after chloroquine was stopped. The Arden index is now within the normal range.

Both chloroquine and hydroxychloroquine (Plaquenil) produce similar toxic effects, but the toxic and the normal dosage of hydroxychloroquine is much larger. The evidence now indicates clearly that these drugs should be used with caution and that patients should be carefully monitored. Ocular toxicity is very unlikely with doses of less than 4 mg per kg lean body weight per day of chloroquine phosphate. It is important to realize that the dosage of chloroquine varies depending upon which salt is prescribed. If the EOG is depressed, the drug should be stopped, even if there are no subjective signs of toxicity. Unfortunately there is no doubt that some patients may develop severe toxicity although retaining a normal EOG.

In spite of its obvious hazards, chloroquine is still being prescribed; ocular damage can be avoided if the toxic dose is not exceeded. Recently, it has been found useful in short courses for the treatment of pulmonary sarcoidosis. Ophthalmologists must be continually on their guard for the appearance of side effects.

Opacities in the Media

The electrical response of the eye to a patterned stimulus is, of course, severely affected by the presence of even slight opacities in the media, depending on how greatly they affect the visual acuity. However, in response to a flash, both the VEP and the ERG can be obtained, even when dense; opacities are present. Indeed, providing the stimulus is bright enough and the retina is normal, there is virtually no opacity that can prevent a response from occurring. Thus the ERG
and the flash VEP can be used to assess retinal and cortical function when the fundus of the eye is not visible with the ophthalmoscope. For example, a normal ERG may be obtained using a bright flash stimulus through a dense vitreous hemorrhage. This can be of help when assessing retinal function in patients with diabetic retinopathy.

Similarly, the ERG and flash VEP can be helpful in assessing general retinal function in patients with dense cataracts who are being evaluated for surgery. However, because the ERG is a measure of diffuse retinal function, small areas of the retina can be damaged (e.g., in macular degeneration) with no effects on the ERG. For this reason, the ERG alone cannot predict the outcome of cataract surgery. The flash VEP has been shown to be a more accurate predictor of visual outcome in cataract surgery. As a general rule, both the VEP and the ERG should be obtained when assessing opacities in the media: the ERG to indicate whether widespread retinal disease or severe retinal ischemia is present, and the VEP to indicate something about the function of the macular area.

There are special instances in which electrodiagnostic tests can be of value in the presence of opaque media: in particular, vitreous hemorrhage following subarachnoid hemorrhage and head injury and also vitreous hemorrhage in which a retinal detachment is suspected. It is also extremely useful to determine whether the wavelets are present in a diabetic patient with dense cataracts.

Poor Vision with a Normal Optic Fundus

Several conditions in the eye or visual pathway may present to the ophthalmologist as poor vision with a normal fundus. Some of these have been mentioned above. In retinitis pigmentosa in its early stages and the related condition Leber's amaurosis, electro-retinography may be diagnostic. Likewise, in patients with suspected healed retrobulbar neuritis, measurement of the VEP is an effec-tive means of confirming or denying the underlying diagnosis.

AMBLYOPIA OF DISUSE

A patient with reduced visual acuity associated with squint or anisometropia, but with no visible abnormality in the retina, would seem to be an ideal subject for VEP studies, especially if we assume that the VEP arises In the primary visual cortex. VEP studies might be expected to tell us something about the site of the defect in amblyopia. In general, the pattern VEP recorded by stimulating the affected eye shows a reduction in amplitude; this is in contrast to the ERG as routinely recorded and the flash VEP, which are usually normal.

At present it is not possible to distinguish between the different types of amblyopia of disuse based on the VEP, but some interesting facts are emerging. For example, a "binocular negative" effect has been described in which the binocular VEP is smaller than the VEP recorded from each eye individually. In normal subjects, the binocular VEP is usually larger than the response from the individual eyes and can be seen to represent the sum of the waveform of the two. This binocular negative effect, which appears to represent some form of suppression, is found most frequently in strabismic amblyopia.

Attempts to correlate the VEP with some of the psychophysical findings in amblyopia seem to indicate defects both peripheral to and "above" the primary visual cortex. In the normal eye, the VEP shows a maximum amplitude with 15-minute checks; the response Is smaller when the stimulus is composed of either larger or smaller checks. In one careful study of an adult amblyope, the affected eye showed a maximum amplitude for the amblyopic eye with was 60-minute check stimulus, and there was a significantly larger signal for the 60-minute check from the amblyopic eye than from the normal eye. Furthermore, when a small-field stimulus was used, there was no difference in amplitude between the normal and amblyopic eyes. It has been suggested that in the amblyope, the normal central area is unable to exert sufficient lateral inhibitory effect on the surrounding retina and that the well-recognized increase in visual acuity with separate-letter testing that is found with some amblyopes may be clue to the fact that they are looking at each letter in a manner equivalent to a small-field pattern stimulus.

A different light has been thrown on this type of amblyopia by examination of subjects with high degrees of astigmatism. It is known that the resolving power of the human visual system is better in the vertical and horizontal orientation than in the two oblique orientations. The same effect can be seen In the VEP. When astigmatic subjects are examined, a high proportion show a reduction in the amplitude of the VEP when a grating stimulus is oriented in the meridian with the lower refractive error. This impairment of the VEP occurs with full spectacle correction.

When a child with amblyopia is treated by occluding the sound eye, interesting changes can be observed in the VEP as the vision of the weaker eye improves. These changes differ according to the check size of the patterned stimulus that is being applied. Figure 11.18 shows the improvement that occurs in the response. The VEP provides the only objective way of measuring the results of treatment of amblyopia; for this reason, it is likely to have considerable importance in the future. It is well recognized that visual acuity measurements by themselves can occasionally be misleading.

Figure 11.18. A. Serial recordings of the VEP produced by stimulating the right amblyopic eye during a period of left-eye occlusion. Note the increase in the amplitude of the VEP with the passage of time. B. The corresponding VEPs from the left (occluded) eye In the same amblyopic child. Little change is seen other than a possible fall in the amplitude of the response after 4 weeks.

HEMIANOPIA

Sometimes, the ophthalmic surgeon may be puzzled by an apparent reduction in visual acuity and subsequently finds it is caused by a homonymous hemianopia. In the very young and the very old, these defects may not be easy to pick up, and so it is of special interest to consider some of the relevant VEP findings.

The use of an unstructured flash stimulus together with laterally placed electrodes can give asymmetric responses from the two sides in hemianopic patients, but by and large such results have not proved reliable. The use of a patterned hemifield stimulus seems to be more promising. It has been shown that a hemifield pattern-reversal stim-ulus produces a larger response when the electrodes are placed over the ipsilateral hemisphere than when over the contralateral hemisphere. A full-field pattern stimulus tends to produce the summed effect of the two half-field responses.
Full-field and hemifield stimulations of patients with homonymous and bitemporal hemianopias give predictable results, using this technique. The exact position of the electrodes and the specifications of the stimulus appear to be important.

The rather surprising finding that a larger response is found over the inert hemisphere has been confirmed in a patient who had undergone an occipital lobectomy, which suggests that the electrical changes must be projected from the active side directly to the scalp rather than through the corpus callosum (Fig. 11.19). This paradoxic response occurs only if the stimulus is flashed at a fairly high speed. If low-frequency, pattern-onset stimulation is used and the so-called transient response is examined, the response behaves predictably; that is, it is larger over the intact hemisphere and smaller over the diseased side. In other words, with a low-frequency pattern-onset stimulus, there is a reduced response over the right occiput in a patient with a left homonymous hemianopia.

Figure 11.19. The VEPs to a pattern-reversal stimulus from a patient with a right subtotal occipital lobectomy. Note the larger responses from the damaged side. (After Blumhardt LD, Barret G, Halliday AM. The asymmetrical visual evoked potential to pattern reversal in one half field and its significance for the analysis of visual field defects. Br J Ophthalmol 1977;61:454.)

CORTICAL BLINDNESS

Since unilateral damage to the occipital Cortex can cause unilateral impairment of the VEP, one might logically conclude that total loss of the occipital cortex and occipital blindness would always be associated with the absence of any VEP. Unfortunately, the results from the limited number of published cases have been conflicting. VEPs have been recorded from some patients with apparently complete cortical blindness, but In other patients, the VEP has been abolished. Cortical blindness must be carefully defined in this kind of study. The patient should have no perception of light and no blink reflex to a flashing light. The eyes themselves must be normal as well as the ERG, and the pupils also should react normally both directly and consensually.

It is difficult to imagine how an electrical response can arise in a defunct part of the brain. However, before revising our idea about the origin of the VEP, we might be wise to review a larger store of clinical data when it becomes available. Some patients with extensive cortical damage can be left with a very small area of central field. Such patients are behaviorally blind, being unable to walk about without knocking into furniture and being unable to read the Snellen test chart
unless some time is spent locating it. Once the test chart has been found, they may have a surprisingly high level of visual acuity. The VEP in such patients should be measured, and no doubt the result of such tests will have clinical relevance in the future. We know that the stimulation of relatively small areas of the central field can produce a VEP, especially when a patterned stimulus is used.

HYSTERICAL BLINDNESS

Although in theory the VEP and the ERG seem to be ideal objective tests for the diagnosis of hysteria or malingering, in practice, such patients often refuse to cooperate during the tests, perhaps because they suspect that the results of the tests will lose them the sympathy they so much require. Of course, one would expect the ERG and the VEP to be entirely normal. One must remember, however, that some patients have very small responses and yet are still within the normal range; this particularly applies to the VEP. In these cases, the VEP should be measured with different sizes of pattern stimulus so that a clear-cut change can be noted, especially the increase in size of the second positive peak with smaller checks. Figure 11.20 shows the effect of altering the check size that one might expect to see in a normal subject.

Figure 11.20. The effect of changing the stimulus check size on the VEP in a normal subject. Note especially the increase in the amplitude of the later positive peak with smaller checks.