Ординатура / Офтальмология / Английские материалы / Electrophysiology of Vision_Lam_2005

Other predictors of neovascularization in CRVO include the relative afferent pupillary defect and the ERG (14). Hayreh et al. (15) reported that a relative afferent pupillary defect of 0.7 log units has a sensitivity of 88% and a specificity of 90% in differentiating ischemic from non-ischemic CRVO, and scotopic or photopic bright-flash b-wave amplitudes had a sensitivity of 80–90% with a specificity of 70–80%. Larsson et al. (16) in a study of 32 CRVO patients found that iris neovascularization was predicted by fluorescein angiography in 82% of the patients and with ERG scotopic 30-Hz flicker implicit time in 94% of the patients. In addition, Matsui et al. (17) found a significant correlation between b-wave=a-wave amplitude ratio and capillary dropout on fluorescein angiography.

Full-field ERG parameters proposed for predicting neovascularization include b-wave=a-wave amplitude ratio, b-wave amplitude, b-wave implicit time, and 30-Hz photopic flicker b-wave implicit time. Additional predictive ERG parameters for neovascularization include Rmax and log K, which are derived from the intensity–response function (Naka– Rushton function fit) calculated from the b-wave amplitudes of scotopic responses with increasing stimulus intensity (see Chapter 1). The Rmax is the maximal scotopic b-wave amplitude, and log K is an index of retinal sensitivity defined as the relative log value of the flash intensity that elicits a response of half of Rmax. Recommendations of ERG parameters for predicting neovascularization vary among studies, and this is due, in part, to the high ERG variability in CRVO as well as differences in ERG methodology among studies. Breton et al. (18) have demonstrated that the predictive power of b-wave implicit time and b-wave=a-wave amplitude ratio for iris neovascularization were influenced by stimulus intensity.

Regarding scotopic full-field ERG studies in predicting neovascularization, Sabates et al. (19) found that the mean b-wave=a-wave amplitude ratios from the scotopic combined rod–cone bright flash response were 1.67 in 27 non-ischemic CRVO patients compared to 0.70 in six ischemic CRVO patients. None of the patients with a b=a-wave ratio of greater

than 1 developed neovascularization, and the b=a-wave ratio was useful to categorize ‘‘undetermined’’ CRVO patients into non-ischemic or ischemic CRVO. Likewise, Matsui et al. (20) found a correlation between final visual acuity and initial ERG parameters such as b-wave amplitude and b-wave= a-wave amplitude ratio in a study of 47 CRVO patients. Matsui et al. (21) also reported b-wave=a-wave amplitude ratio improvement in some CRVO patients suggesting that retinal ischemia in CRVO may be reversible in some cases. However, Kaye and Harding (22) found that the prolonged b-wave implicit time was the parameter most significantly associated with neovascularization followed by b-wave=a- wave amplitude ratio and then b-wave amplitude on 26 CRVO patients, seven of whom developed iris neovascularization. In contrast to the studies mentioned, Johnson et al. (10) in a study of 15 CRVO patients noted that all nine eyes with neovascularization showed prolonged scotopic a-wave, b-wave, and 30-Hz flicker b-wave responses but only one eye had a reduced b-wave=a-wave amplitude ratio that was close to or less than 1. In a later report from the same investigators, ERG retinal sensitivity (log K) was found to have a sensitivity of 90% and a specificity of 91% in predicting neovascularization (23). Similarly, Moschos et al. (24) found a-wave and b- wave implicit times to be better than b-wave=a-wave amplitude ratio in predicting neovascularization. The use of scotopic 30-Hz flicker implicit time was further supported by studies by Larsson et al. (25,26). These investigators suggest that the optimal time to perform the scotopic 30-Hz flicker response to predict neovascularization may be three weeks after the onset of CRVO as there may be considerable change in ERG during the first three weeks.

Taken together, the results of these scotopic full-field ERG studies suggest that prolonged scotopic bright-flash b-wave implicit time and 30-Hz flicker b-wave implicit time may be better predictors of neovascularization than the scotopic bright-flash b-wave=a-wave amplitude ratio. Both scotopic and photopic 30-Hz bright-flash flicker elicit cone-generated responses, and the full-field ERG international standard includes a 30-Hz flicker stimulus under photopic condition.

Larsson and Andre´asson (27) compared scotopic and photopic 30-Hz flicker responses as predictor for iris neovascularization in 44 CRVO patients and both 30-Hz flicker responses had high predictive value.

With respect to other full-field ERG studies with scotopic and photopic stimuli, Breton et al. (28) studied 21 CRVO patients during initial clinical visit and found that Rmax and the scotopic bright-flash b-wave=a-wave amplitude ratio were better in predicting iris neovascularization than the retinal sensitivity (log K) and 30-Hz flicker implicit time obtained after only 2–3 min of light adaptation. A later report from Breton et al. (29) on 39 CRVO patients reached similar conclusions. Roy et al. (30) noted that ERG responses to photopic red flash are also a predictor of neovascularization. Barber et al. (31) performed ERG with undilated pupils and found the photopic cone flash responses to have prolonged a-wave implicit time and reduced b-wave amplitude in 15 CRVO patients. Taken together, the results of these studies, which included photopic stimuli, suggest that photopic ERG parameters such as the photopic 30-Hz flicker are also predictors of neovascularization.

As evident by the studies mentioned, ERG impairments are common in CRVO. However, some studies have noted that a significant percentage of CRVO patients have supernormal response in the affected as well as the fellow eye (32). The reason for this finding is unclear. Sakaue et al. (33) found that 36% of the unaffected fellow eyes of 50 patients with unilateral CRVO had full-field scotopic bright flash ERG responses above the normal range with 30% having increased b-wave amplitudes and 6% with increased a-wave amplitudes. In the same study, of the 50 affected eyes with CRVO, 42% were normal, 34% had increased b-wave amplitudes and 10% had increased a-wave amplitudes. In contrast to other studies, only 14% of affected eyes had reduced ERG amplitudes. In addition, Gouras and MacKay (34) showed that eyes with CRVO had slower but larger, supernormal full-field ERG responses to long-wavelength stimulus when compared to the unaffected eye. The authors concluded that the longwavelength cones are less able to reduce their responsiveness

to light with increasing levels of light adaptation in a retina affected by CRVO.

In a preliminary study, the P1 amplitudes and latencies of the first-order multifocal ERG response correlated significantly with the full-field 30-Hz cone flicker amplitudes and latencies (35). The authors concluded that wide-field multifocal ERG is a sensitive indicator of disease in CRVO and may potentially have a role in the clinical setting.

Electro-oculogram generally parallels ERG response in CRVO although patients with reduced EOG light- peak=dark-trough amplitude ratio and normal full-field ERG have been reported (36,37). Ohn et al. (37) performed EOG and full-field ERG on 24 CRVO patients, 13 of whom had ischemic CRVO. The mean EOG ratio for patients with ischemic and non-ischemic CRVO were 1.38 0.38 (standard deviation) and 1.92 0.43, respectively (normal 1.85), and this difference was statistically significant. However, the sensitivity and specificity for differentiating ischemic from nonischemic CRVO based on EOG was only 92% and 55%, respectively. Papakostopoulos et al. (38) in a study of 28 CRVO patients noted that EOG light-peak amplitude in the affected eyes was 48% or less than that of the unaffected eye in the eight patients who developed iris neovascularization. Pattern ERG is a measure of ganglion cell function and is reduced in CRVO (14). The VEP in CRVO parallels ERG function.

Branch Retinal Vein Occlusion

Branch retinal vein occlusion usually occurs at an arteriovenous intersection. Signs of the vein occlusion such as scattered retinal hemorrhages, edema, and venous tortuosity are visible in the region distal to the site of the occlusion. Risk factors for BRVO are similar to CRVO and include hypertension and diabetes mellitus. In contrast to CRVO, neovascularization of the retina or optic nerve head may occur but iris neovascularization is rare. Visual prognosis is dependent on the extent of retinal ischemia and macular involvement. The diagnosis of BRVO is by retinal appearance with support from fluorescein angiography if needed.

Multifocal ERG is more likely than full-field ERG to detect regional retinal dysfunction from BRVO and may be potentially useful in patients with a remote history of BRVO when retinal signs are minimal. In contrast, the effect of BRVO on full-field ERG is related to the size of the involved retina, and the full-field ERG in BRVO may be normal or mildly impaired with a selective decrease in b-wave amplitude similar to CRVO. Of interest, Hara and Miura (39) noted full-field ERG oscillatory potentials to be reduced in the affected eyes of 34 unilateral BRVO patients presumably due to inner retinal ischemia. In the same study, the EOG light-peak to dark-trough amplitude ratios and the light-peak amplitudes were significantly reduced in the affected eye compared to the unaffected eye. The investigators hypothesized that the EOG impairment was due to inner retina dysfunction that hampered the EOG amplitude rise to light. Lastly, Gu¨ ndu¨ z et al. (40) found pattern ERG and VEP to be impaired in BRVO patients without systemic disease and attributed these findings to retinal dysfunction.

OTHER PROLIFERATIVE

NEOVASCULAR DISORDERS

Retinopathy of Prematurity

Retinopathy of prematurity (ROP) is a retinal vascular proliferative disease of premature and low-birth-weight infants. Fetal retinal vascularization occurs from the optic nerve head toward the peripheral retina and is completed at approximately 36 and 40 weeks of gestation in the nasal and temporal retinal regions, respectively. The severity of ROP is categorized into five stages: stage 1—demarcation line between vascularized and non-vascularized retina, stage 2— raised demarcation line forming a ridge, stage 3—extraretinal fibrovascular proliferation of the ridge, stage 4—subtotal retinal detachment, and stage 5—total funnel-shaped retinal detachment. The diagnosis of ROP is by retinal appearance. Treatments include laser or cryotherapy to the non-vascular- ized retina. Flash VEP may be helpful as objective evidence of

visual function in infants with stage 5 ROP. Clarkson et al. (41) advocated bright-flash VEP and ocular ultrasound to evaluate stage 5 ROP infants before and after retinal reattachment surgery. Eyes with regressed ROP are more likely to develop delayed retinal detachment, cataract, glaucoma, strabismus, anisometropia, amblyopia, and myopia.

Using a-wave amplitudes of scotopic full-field ERG responses from increasing stimulus intensity to estimate the saturated maximal rod response amplitude (Rmp3) (see Chapter 1), Fulton and Hansen (42) reported mild photoreceptor dysfunction in five infants and children with complete regression of previous mild ROP. The same investigators also noted reduced oscillatory potentials in nine infants and children with previous stage 1, 2, or 3 ROP.

Diabetes Retinopathy

Diabetic retinopathy is one of the leading causes of visual impairment and blindness worldwide. Two forms of diabetes are recognized. Type 1 diabetes, also called juvenile-onset or insulin-dependent diabetes, is characterized by pancreatic beta-cell destruction leading to absolute insulin deficiency. Type 2 diabetes, also called adult-onset or non-insulin- dependent diabetes, is characterized by insulin resistance with insulin secretory defect and relative insulin deficiency. Type 2 diabetes is more common and makes up approximately 90% of diabetic patients. However, patients with type 1 diabetes are more likely to develop severe diabetic retinopathy. The risk of developing diabetic retinopathy is related to duration of the disease, and nearly all patients with diabetes for more than 20 years will have some degree of retinopathy. Diabetic retinopathy is categorized into non-proliferative or proliferative stages. In the mild to moderate non-proliferative stage, previously called background retinopathy, retinal findings include small intraretinal hemorrhages and microaneurysms. In the severe non-proliferative stage, previously called preproliferative retinopathy, signs of increasing ischemia such as severe intraretinal hemorrhages, microaneurysms, venous abnormalities, and intraretinal

microvascular abnormalities occur. In proliferative diabetic retinopathy, retinal neovascularization in response to ischemia takes place, which if left untreated may progress to vitreous hemorrhage, fibrovascular formation, and retinal detachment. Diabetic macular edema due to microvascular disease is also a major cause of visual impairment and may occur with non-proliferative or proliferative retinopathy.

Clinical electrophysiologic testing of patients with diabetic retinopathy is not commonly performed because of the reliance by most clinicians on retinal appearance. However, ERG abnormalities occur early in the disease even in the absence of any visible retinopathy, indicating that the ERG is a sensitive indicator of mild disturbance of retinal circulation. Reduced and delayed oscillatory potentials without any impairment of the a-wave and b-wave amplitudes may occur in diabetic patients with early or no retinopathy, with the most consistent finding being an increase in the implicit time of the first oscillatory potential wavelet (OP1) presumably due to early, mild ischemia of the inner retinal layers (44–47). Aside from impaired oscillatory potentials, other less consistent full-field ERG abnormalities found in diabetic patients without visible retinopathy include prolonged b-wave implicit times and reduced scotopic b-wave amplitude, with the latter observed in type 1 diabetic children (46,48). Likewise, multifocal ERG of diabetic patients without visible retinopathy showed prolonged oscillatory potentials and delayed firstorder responses (49,50). Amplitude reduction of the secondorder multifocal ERG components was also noted (51).

Clinical ERG testing in diabetic patients to assess the risk of progression to proliferative retinopathy is seldom performed because of the reliance on retinal appearance and fluorescein angiographic examinations. However, Bresnick et al. (52–54) examined the predictive value of full-field ERG findings for progression to proliferative diabetic retinopathy in 85 patients in the Early Treatment Diabetic Retinopathy Study. The investigators showed that the summed amplitudes of the oscillatory potentials, the overall severity of retinopathy, and the severity of fluorescein angiographic leakage were independent predictors of progression to severe

proliferative retinopathy. Similarly, in another longitudinal study, Simonsen (55) found that low oscillatory potential amplitudes were a predictive factor in the development of proliferative diabetic retinopathy.

With progression of diabetic retinopathy, all ERG parameters including scotopic and photopic a-wave and b-wave responses become impaired due to retinal ischemia, vitreous hemorrhage, and retinal detachment (51,56). In patients with mild to moderate non-proliferative retinopathy, scotopic and photopic b-waves are impaired in addition to impaired oscillatory potentials (57). In proliferative diabetic retinopathy, the ERG reductions are variable and dependent on the degree of microvascular damage and the presence of complications such as vitreous hemorrhage and retinal detachment.

Several studies have examined the effect of panretinal photocoagulation on ERG. Lawwill and O’Connor (58) performed ERG and EOG on diabetic patients before and after photocoagulation and noted a 10% decrease in a- and b-wave amplitudes when approximately 20% of the retina was photocoagulated. However, Ogden et al. (59) found that a decrease of ERG amplitude varied from 10% to 95% after photocoagulation. These conflicting results were clarified by Wepman et al. (60) who found a correlation between b-wave amplitude reduction and area of photocoagulated retina, but this relationship was evident only after the magnitude of the ERG response prior to treatment was taken into account. Patients with larger pretreatment amplitudes showed a greater amplitude reduction than patients with small pretreatment signals even though they had equivalent amounts of retina destroyed, because laser destruction of functionally active retina would result in a greater net reduction of ERG after treatment.

Diabetic macular edema may produce focal ERG impairment but by itself is unlikely to reduce full-field ERG responses significantly. Greenstein et al. (61,62) evaluated 11 patients with multifocal ERG before and after laser treatment for clinically significant diabetic macular edema. Their findings suggest that local ERG timing delays were not good predictors of visual field deficits and that focal laser treatment produced increases in implicit time and decreases

in amplitude of local ERG responses which were not restricted to the treated macular area.

In terms of other electrophysiologic testing, pattern ERG studies in diabetic patients without apparent retinopathy have mixed results due in part to differences in technique among studies (46). In patients with non-proliferative diabetic retinopathy, most pattern ERG studies have reported decrease in amplitude and delay in implicit time (46). Lastly, VEP may be helpful to predict final visual acuity in diabetic eyes with vitreous hemorrhage when retinal appearance is obscured (63).

Sickle Cell Retinopathy

Hereditary hemoglobulinopathies occur when mutant hemoglobins (Hgb) S and C are inherited rather than normal hemoglobin A, resulting in sickle trait (Hgb AS), sickle cell disease (Hgb SS), and hemoglobin SC disease. Intravascular sickling of red blood cells, hemolysis, and thrombosis can produce preretinal, intraretinal and subretinal hemorrhages and peripheral retinal neovascularization leading to retinal detachment. Sickle cell disease is most common in blacks. The diagnosis of sickle cell retinopathy is by retinal appearance and hemoglobulin electrophoresis if not yet performed.

Peachey et al. (64) noted normal full-field ERG responses in sickle cell disease patients without retinal neovascularization. Patients with neovascularization were found to have generalized reduced ERG components including a-wave, b-wave, and oscillatory amplitudes, presumably due to ischemia. In a later study of 44 patients with sickle cell retinopathy from the same research group, a correlation between ERG amplitude measures and capillary non-perfusion determined by fluorescein angiography was documented (65). In the same study, the maximal scotopic b-wave amplitude, Rmax, derived from the intensity–response function (Naka–Rushton function fit) (see Chapter 1), was also reduced but the retinal sensitivity, log K, defined as the flash intensity that elicits a response of half of Rmax, was unaffected in sickle cell disease patients. The findings of EOG and VEP in sickle cell disease

are rarely reported. The EOG is likely to parallel ERG responses, and VEP is likely to parallel macular function.

OTHER OCULAR VASCULAR DISORDERS

Hypertensive Retinopathy

Chronic systemic arterial hypertension produces atherosclerosis and increased vascular permeability and is a major cause of coronary arterial disease, cerebral arterial occlusion, and renal failure. A systolic blood pressure of greater than 130 mm Hg and a diastolic blood pressure of greater than 85 mm Hg on two or more clinical visits are considered elevated. Hypertension is classified as stage 1 (140–159=90–99, systolic=diastolic), stage 2 (160–179=100–109), stage 3 (180– 209=110–119), and stage 4 ( 210= 120). In over 90% of patients, the cause is unknown, and the patients are diagnosed with primary or essential hypertension. In the remaining patients, secondary hypertension results from disorders including renal vascular disease, aortic coarctation, Cushing’s disease, and pheochromocytoma.

Hypertension may cause retinopathy, choroidopathy, and optic neuropathy. Retinal signs of hypertension include arteriolar narrowing, arteriovenous crossing changes (arteriovenous nicking), arteriolar sclerosis, arteriolar tortuosity, hemorrhages, exudates, inner retinal infarcts (cotton–wool spots), and retinal edema. Hypertensive retinopathy may be categorized by Scheie’s classification. Grade I retinopathy has visible arteriolar narrowing; grade II retinopathy has arteriolar narrowing plus focal arteriolar abnormalities and arteriolar sclerosis; grade III has grade II findings plus retinal hemorrhages and exudates; and Grade IV has grade III findings plus optic disc edema. Findings of hypertensive choroidopathy include retinal pigment epithelium and choroidal infarcts (Elschnig spots), subretinal exudates, serous retinal detachments, and choroidal sclerosis. Optic disc edema may be associated with hypertensive retinopathy or occur due to anterior ischemic optic neuropathy. The diagnosis of hypertensive retinopathy and choroidopathy is based on clinical