Ординатура / Офтальмология / Английские материалы / Electrodiagnosis of Retinal Disease_Miyake_2005

The full-field ERGs elicited by increasing stimulus intensities recorded from a normal subject after 1h of dark adaptation are shown in Fig. 1.2. The ERGs elicited by relatively weak stimulus intensities are shown at the left in Fig. 1.2, and the ERGs elicited by stronger stimulus intensities are shown at the right. The calibrations for the amplitude and time are different for the weak and strong ERGs. The maximum stimulus luminance (0log unit) was 44.2cd/m2 ·s-1.

At the left, the scotopic threshold response (STR) [11], a cornea-negative wave, is first recorded at -8.2log units, approximately 0.6 unit higher than the psychophysical threshold. The maximum amplitude of the STR is 24 mV before it is masked by the developing b-wave. The implicit time of the STR near threshold is approximately 162ms, and the implicit time decreases as the stimulus intensity increases. The STR originates from retinal neurons that are postsynaptic to the photoreceptors [11]. With some types of congenital stationary night blindness, the STR has unique properties [12] (see Section 2.10.5.3)

The b-wave is first seen at an intensity of -5.8log units; the amplitude increases and the implicit time shortens as the stimulus intensity increases. The amplitude of the b-wave essentially saturates at -3.4log units; and at intensities higher than -0.8log unit, the oscillatory potentials (OPs) become clearly visible on the ascending limb of the b-wave. The a-wave is first seen at -1.7log units and increases progressively as the stimulus intensity increases.

Many studies have shown that the a-wave of the full-field ERGs recorded in the dark is the leading edge of the photoreceptor potential [13]. The b-wave originates indirectly from bipolar and Mueller cells in the middle layers of the retina [14]. The OPs are seen as a series of three or four rhythmic wavelets having almost equal amplitude with an interpeak interval of about 6.5ms in humans [15]. The best experimental evidence indicates that the OPs reflect the activity of feedback synaptic circuits within the retina and represent an inhibitory or modulating effect of amacrine cells on the b-wave [16, 17].

Fig. 1.2. ERG intensity response series recorded from a normal subject exposed to a flash of relatively low intensity (left) and relatively high intensity (right). Note that the calibration differs for the ERGs in the two columns. Arrows indicate the stimulus onset. STR, scotopic threshold response; bs, scotopic b-wave. (From Miyake et al. [12], with permission)

Although the rods outnumber the cones 13 to 1 in the normal human retina, the cone system accounts for 20%–25% of the ERG response amplitude. For purposes of diagnosis, it often becomes necessary for the examiner to evaluate rod and cone activity separately. The full-field ERGs recorded in our clinic from a normal subject are shown in Fig. 1.3. After 30min of dark adaptation, a rod (scotopic) ERG is recorded with a dim flash of light at approximately -3.9log units (Fig. 1.2). A mixed cone–rod ERG (bright) is elicited by a single flash of white light at maximum intensity (log 0 units). A cone ERG and 30-Hz flicker ERG are recorded with a stimulus intensity of -0.8log unit, under a background illumination of

40cd/m2, which is sufficient to suppress all rod activity. The photopic recordings were made after 10min of light adaptation to 40cd/m2. As shown later, the maximum photopic ERG can be obtained when recorded after light adaptation because the amplitude of the photopic ERG increases significantly after light adaptation [18] (see Section 1.1.4.1).

Our recording conditions are in accordance with the standards proposed by ISCEV [10] except that we use a higher intensity for the mixed rod–cone ERGs. We have done this because the higher stimulus intensities show the OPs much clearer, and the negative configuration of the ERG can be detected more convincingly [19].

Light-emitting diodes (LEDs) are valuable as light sources to elicit full-field ERGs [20, 21]. The LEDs are small and inexpensive, and they require low currents to drive them. They can be controlled by simple electronic circuits to give either a continuous light output or extremely brief flashes over a large range of intensities. In

addition, LEDs are available that emit different wavelengths and have different optical properties. A new contact lens electrode with built-in high-intensity white LEDs was recently constructed [22]. We use this system on special patients to test specific properties of the ERGs [22].

The relative spectral emission of the white LEDs used in the contact lens electrode is shown in Fig. 1.4. These LEDs have a relatively broad, asymmetrical spectral bandwidth with two peaks, at approximately 430 and 540nm, respectively (Fig. 1.4A). The output appears visibly white (Fig. 1.4B), and three of these white LEDs are incorporated into a standard contact lens electrode (Fig. 4.1C). The LEDs serve as the source of the stimulus and for

background illumination. The stimulus light and background illumination pass through a diffuser lens and become a broad, homogeneous light that stimulates the entire retina, similar to a Ganzfeld stimulus. The intensity, frequency, and duration of the stimulus LEDs and the background LED are controlled by electrical currents obtained from a waveform generator that drive the LEDs.

Electroretinograms can be recorded with LED contact lens electrodes that comply with ISCEV standards. ERGs recorded from a normal adult are shown in Fig. 1.5, left. All of these ERGs are similar to those elicited by conventional xenon discharge lamps in a Ganzfeld dome. This technique is useful for recording standardized ERGs

from pediatric patients under general anesthesia, as shown in Fig. 1.6. The equipment needed to obtain recordings that correspond to the ISCEV standard ERGs is compact and easily portable. The ERGs recorded from a 3-month- old baby under general anesthesia using this system are shown in Fig. 1.5, right.

Fig. 1.5. Full-field ERGs recorded with white LED contact lens electrode from a normal adult subject (left) and a normal 3-month-old baby (right)

Fig. 1.6. Standard full-field ERG recording from a baby using this system. Top: LED contact lenses are placed in both eyes under general anesthesia. Bottom: Background illumination from the contact lens is used during the recording of the photopic ERGs under the dark

As vitreoretinal surgical techniques continue to advance, close monitoring of retinal function during these procedures has become important. Although ERGs directly reflect retinal function, monitoring during surgery has proven difficult. Each recording must be made quickly under aseptic conditions, and the instruments and electrodes must be such that they do not interfere with the retinal surgeon. Furthermore, ERGs need to be cone-mediated because the eye undergoing surgery is intensively light-adapted.

The LED-contact lens electrode has been found to be highly suitable for this purpose [23]. The LED-contact lens is easily sterilized and is used as both a stimulus source and a recording electrode for 30-Hz flicker ERGs during vitreoretinal surgery (Fig. 1.7). Each recording requires approximately 7s. The changes in the ERGs during surgery in a patient with a shallow retinal detachment in the macular region, associated with a macular pucker, are shown in Fig. 1.8. The operation was performed with the

patient under local anesthesia. The ERGs recorded after local anesthesia (Fig. 1.8, start), and after the introduction of the infusion needle into the vitreous cavity (Fig. 1.8, infusion) were not significantly different in regard to amplitude and peak time. However, after vitrectomy, which required 10min, the peak time was delayed and the amplitude decreased.Additional studies have demonstrated that lowering the intravitreal temperature by applying an infusion solution kept at room temperature can alter the ERG during vitrectomy [24]. Filling the whole vitreous cavity with air after a preretinal membrane was peeled off resulted in a markedly reduced amplitude and delayed peak time. Seven days after surgery, when the air was resolved from the vitreous cavity, the ERG recovered to the preoperative amplitude and peak time. The extreme reduction of ERG amplitude following fluid–air exchange or fluid–silicone oil exchange in the vitreous cavity results from reduced electrical conductivity in the vitreous cavity [23, 25].

It is well known that there are significant potential changes when a light stimulus is turned off [26]. These potential changes are called the off response, or d-wave, of the ERG. For conventional recordings, a stroboscopic flash is used to elicit the ERGs, and the off response is embedded in the on response. Because there are retinal diseases in which the on and off responses are affected differentially, it became extremely important for us to record the on and off responses separately using long-duration stimuli [27, 28].

Figure 1.9 shows a simplified schema of the rod and cone visual pathways in the mammalian retina (Fig. 1.9, left) and the long-flash photopic ERGs (Fig. 1.9, right), which were obtained by Sieving [28]. The photoreceptors transmit visual information to the bipolar cells

with the rods containing only depolarizing bipolar cells (DBCs) through sign-inverting (-) synapses (on synapse). The cones contact both DBCs and hyperpolarizing bipolar cells (HBCs) through sign-inverting (-) synapses (on synapse) and sign-preserving (+) synapses (off synapse), respectively. The cone on and off bipolar cells contact, respectively, the on and off ganglion cells directly. The rod bipolar cells do not make synapses to the ganglion cells but contact on and off bipolar cells via A11 amacrine cells.

The two types of synapse from the photoreceptors to the bipolar cells are selectively sensitive to different glutamate analogs [29, 30]. The sign-inverting synapse (on synapse) can be blocked by 2-amino-4-phosphonobutyric acid (APB). The sign-preserving synapse (off

synapse) can be blocked by either cis-2,3- piperidine dicarboxylic acid (PDA) or kynurenic acid (KYN). Sieving demonstrated the contribution of these glutamate analogs to the DBCs and HBCs to the monkey long-flash photopic ERG [28]. As shown in Fig. 1.9 (on the right), the control ERG exhibits on responses (a-waves and b-waves) and off responses (d- wave), with a negative plateau between the b- waves and d-waves. By blocking DBC activity, the photopic b-wave was suppressed and the a- wave and d-wave were enhanced. After blocking the HBCs with KYN, the a-wave and d-wave became smaller and the plateau was elevated above the baseline (Fig. 1.9, arrowheads). Based

on these results, it has been proposed that the “push-pull” activity of the HBCs and DBCs is summated in the photopic ERGs recorded at the cornea. In summary,we have provided evidence that the a-wave of the photopic ERGs evoked by longand short-duration flashes arises not only from the neural activity of the photoreceptors but also from hyperpolarizing bipolar cells. In addition, the b-wave and d-waves of the photopic ERGs elicited by long-duration flashes are produced by an interaction of the hyperpolarizing and depolarizing bipolar cells; and the cornea-positive peak of the short-flash ERG results from a summation of the b-wave at light onset and the d-wave at light offset [28, 31].

As already mentioned, the long-flash photopic ERG can provide important information in terms of the on and off visual pathways in retinal diseases; clinically, however, the recording procedure is not a simple one to apply. Using the LED, one can record the long-flash photopic ERG easily. The stimulus duration can be regulated by a small LED control device. The photopic ERGs elicited by stimulus durations of 3–250ms with a stimulus intensity of 250cd/m2 and steady background illumination of 40 cd/m2 are shown in Fig. 1.10 [22]. When the stimulus duration is relatively long (100ms or longer), the d-wave (off wave) is clearly seen after the stimulus is turned off. One can see that with short-stimulus durations the on and off response components interact to produce a single positive deflection, called the b-wave. Interestingly, the d-wave plays a major role in shaping the main positive peak of the shortflash ERG. This is an important finding when we analyze the photopic short-flash ERG in patients with diseases where the b-wave is absent and the d-wave preserved, such as with complete type congenital stationary night blindness [27] (see Section 2.10.5.4).

Recording short-wavelength cone (S-cone) ERGs is valuable clinically because it allows us to evaluate the short wavelength S-cone system. S-cone ERGs have been recorded by stimulation with strong blue stimuli on a bright yellow background, which suppresses the middleand long-wavelength (LM) cone systems [32–34]. The procedure is not as simple and easy as it may seem; but with the LED contact lens electrode system, an inexpensive, commercially available electrode and an ordinary slide projector are required, as shown in Fig. 1.11 [35]. LEDs emitting blue light (lmax, 450nm) were used in the LED built-in contact lens electrode (NLPB 500; Nichia, Tokushima, Japan). The yellow background light was provided by a yellow filter (Kodak Wratten No. 12) placed in a

slide projector and projected onto the contact lens electrode. The diffuser in the contact lens produces a full-field homogeneous yellow background illumination. S-cone and LM-cone ERGs elicited by long-duration stimuli of LED are compared in Fig. 1.12. The amplitude is much smaller in the S-cone ERG, and the implicit time of the b-wave is longer than that of the LM-cone ERG. The components that reflect the “off” visual system (a-waves and d- waves) are essentially absent because, unlike the LM-cone system, the S-cone is connected mainly to the on visual system [36]. The intensity response series of S-cone ERGs recorded from a normal subject is shown in Fig. 1.13. The maximum response was recorded as above 3.1 log photopic trolands.

Fig. 1.12. Comparison of LM cone (top) and S-cone (bottom) ERGs with long-duration stimuli in a normal subject. The a-wave and d-wave are essentially absent in the S-cone ERGs

Fig. 1.13. Intensity response series for S-cone ERGs recorded from a normal subject. Maximum response was recorded above 3.1 log photopic trolands (phot td). (From Horiguchi et al. [35], with permission)

There are some unique properties of the retina that are seen only in cone-mediated (photopic) ERGs. The mechanisms that account for such properties are not fully understood. However,

such phenomena and our previous studies can be used to analyze the mechanisms involved in their generation.

After sufficient dark adaptation, the amplitude of the ERGs recorded during the course of light adaptation gradually increases to the point that the amplitude of the fully light-adapted photopic ERGs are sometimes as much as 200% of that recorded at the beginning of the lightadaptation process. This phenomenon has been reported in humans and other animals [37–41], but the mechanism for this increase is not fully understood. Several explanations have been proposed, such as a change of standing potential of the eye [38], an interaction between

cones and rods [40–42], and redepolarization of the cone photoreceptors [43].

This phenomenon [44] is shown in Fig. 1.14. The relative amplitudes of the 30-Hz flicker ERG in 30 normal subjects as a function of time during light adaptation shows that the amplitude gradually increased during the first 5–15 min of light adaptation, and the mean of the maximum amplitude is 1.68 ± 0.60 times larger than the amplitude of the response at 1min. We found that the eye must be completely darkadapted for this phenomenon to be observed

[41]. The 30-Hz flicker ERGs during light adaptation after 30min of dark adaptation [DA(+)] and without dark adaptation [DA(-)] in a normal subject are shown in Fig. 1.15. Increased amplitude is observed only when the dark adaptation prior to the recordings is sufficiently long.

The contribution of rods to this increase was clearly demonstrated in an isolated carp retina [42] (Fig. 1.16). After the carp retina was isolated from the retinal pigment epithelium (RPE) and placed in the dark for 60min, photopic ERGs were recorded during the course of light adaptation (Fig. 1.16, left). The isolated retina was then placed in the dark again for 60 min, after which photopic ERGs were recorded again for 5min under light-adapted conditions. In the first recording, the amplitude increased dramatically during the 5min of light adapta-

tion; in the second recording, the amplitude was large initially and did not increase (Fig. 1.16, right). These results demonstrated that RPE is not needed for eliciting this increase. More importantly, rhodopsin is necessary for this phenomenon because once rhodopsin is bleached in an isolated retina it cannot regenerate during dark adaptation. On the other hand, cone photopigments can regenerate even in isolated retinas. Thus, in the first recording the isolated retina had both rhodopsin and cone photopigments, resulting in the increased amplitude. In the second recordings the retina had only cone pigments, and rods cannot be activated, so the amplitude did not increase.

Because we found that rods contribute to the increased amplitude of photopic ERGs during light adaptation, the rods may inhibit the cones while recording the photopic ERGs. If the