Ординатура / Офтальмология / Английские материалы / Ocular Neuroprotection_Levin, Polo _2003

V.PROTOCOLS FOR DIFFERENTIATION OF THE CONE AND ROD SYSTEMS (THE ISCEV STANDARDS)

One of the main objectives when using the ERG is the differentiation of the cone and rod systems. According to the standards of the International Society for Clinical Electrophysiology of Vision (ISCEV) [15], in order to reach this goal at least five basic responses should be acquired, two in the photopic state (lightadapted retina) and three in the scotopic state (dark-adapted retina):

1.Photopic single flash response (standard flash)

2.Photopic 30 Hz flicker response (standard flash)

3.Scotopic dim flash rod response (intensity 2.5 log unit below the standard flash)

4.Scotopic mixed rod-cone response (standard flash)

5.Scotopic oscillatory potentials (standard flash, bandpass 75–500 Hz)

Examples of typical ERGs obtained according to the ISCEV standards are presented in Figure 3. A complete description of the ISCEV standards can be found on the Society’s website at www.iscev.org

A.Photopic Single Flash Response

It should be noted that ISCEV does not require the use of a specific level of background illumination nor a specific intensity of flash in order to generate the standard flash. Rather the standard suggests a range of photopic background (17– 34 cd m 2) and flash intensity (1.5–3.0 cd m 2 s), an approach that was recently questioned [35]. Responses illustrated in Figure 3 were evoked to a flash of 3 cd m 2 s delivered against a background set at 25 cd m 2. It is recommended to expose the retina to the rod desensitizing background at least 5 min before initiating the cone ERG evaluation. This period of adaptation should be increased to 10 min, at least, if the recording of the photopic responses immediately follows the dark adaptation period in order to avoid the light adaptation effect that was previously shown to affect the amplitude and the timing of most of the photopic ERG components [21,22].

As shown in Figure 4, with an increase in the strength of the stimulus (from bottom to top) there is a gradual increase in the amplitude of the ERG a- and b- waves. However, while the amplitude of the a-wave appears to grow steadily, that of the b-wave reaches a maximum at approximately 0.64 log cd m 2 s, following which further increases in flash intensity will yield b-waves of gradually smaller voltages. This unique feature of the photopic ERG luminance-response function is known as the Photopic Hill [35,36]. Also of interest is to note that

while the peak time of the a wave shortens with brighter flashes, that of the b- wave increases. Finally, as mentioned above (Sec. III.C), in response to progressively brighter flashes, the oscillatory potentials increase in number as well as in amplitude.

B. Photopic Flicker Response

Although in most instances the single-flash photopic ERG response will be sufficient to identify most of the cone-related retinal disorders, the ISCEV standard also recommends the use of a stimulus flickering at a rate of 30 flashes per second as the best means to isolate a cone ERG free from any possible rod-mediated contamination (Fig. 3, tracing C). Analysis of the resulting waveforms will include amplitude (from trough to peak) as well as peak time measurements (see Table 1).

C. Scotopic ERG Responses

As shown in Figure 5, with time spent in dark-adaptation, there is a gradual increase in the amplitude of the resulting ERG response as a result of the slow regeneration of rhodopsin, which takes approximately 30 min (Fig. 5, tracing 1). However, the ERG response becomes much more stable once the regeneration of the rod’s photopigment is completed (Fig. 5, tracing 2). Consequently, in order to assess the rod function specifically, the intensity of the flash stimulus must be significantly attenuated in order to account for the gain in sensitivity brought by the dark-adaptation process. The intensity recommended is one that is 2.5 logunit dimmer than that used in photopic condition. Typically, the ERG evoked to this intensity of stimulation will be devoid of a recordable a wave (see Fig. 3, tracing D, and Fig. 5, tracing 1), while use of an intensity of stimulation within the photopic range will yield an ERG identified as the scotopic mixed rod-cone response (Fig. 3, tracing E, and Fig. 5, tracing 2). It should be noted that if averaging is considered, the interstimulus interval should be of at least 2 s for the dimmer stimuli and at least 5 s when the brighter flashes are used. Also, responses to the first flash should be discarded as they often show a conditioning flash effect [37,38]. The ISCEV standard also recommends that the OPs should be obtained in scotopic condition (Fig. 3, tracing F) because they are usually of larger amplitudes than those obtained in photopic condition. However, the latter recommendation should not prevent one from recording the OPs against a photopic environment as well, as there is a growing body of evidence supporting the view that photopic and scotopic OPs might originate from different retinal structures or pathways [26,39].

Figure 5 Tracing 1 illustrates the progressive growth in amplitude of the b-wave as a result of gradual dark-adaptation from onset of darkness (t 0) to 30 min (t30). The waveform was obtained by the superposition of ERGs in response to flashes of light (intensity: 3.5 log cd m 2 s) delivered at the rate of 1 flash each 5 s. At tracing 2, the same procedure was adopted for the following period of 30 min [e.g., from 30 min (t 30) to 60 min (t 60) of dark adaptation]. The intensity of the flash was also raised by 1 log-unit, thus explaining the different morphology.

As mentioned above, scotopic ERG waveforms evoked to dim flashes only include a b-wave (Fig. 6, tracing 3.8 and dimmer). With increasing intensities, the amplitude of the b-wave increases rapidly until a plateau is reached. This plateau usually correlates with the appearance of an a wave suggestive of a cone contribution to the response (tracing 3.0). When plotted on a graph (see Fig. 7), the relationship between the amplitude of the b-wave and the intensity of the

Figure 6 Representative scotopic luminance-response function obtained from a normal human subject. As the intensity of the stimulus grows (from bottom to top), the amplitude of the b-wave increases gradually but, unlike in photopic condition (see Fig. 4), its peak time shortens. At threshold, the ERG includes only a b-wave; the a-wave appearing only in responses evoked to brighter flashes (brighter than3.0 log-unit). Again, as in photopic conditions, the a-wave peak time shortens with gradually brighter flashes. Also, several oscillatory potentials are seen riding on the ascending limb of the b-wave. Vertical arrow points to the onset of the flash. Calibration: 50 ms (horizontal) and 100 V (vertical).

stimulus (luminance-response function curve) adopts the shape of a sigmoidal function, which can be best described with the Naka-Rushton equation [40]

V/Vmax In /(In Kn)

where V represents the amplitude of the b-wave evoked from a flash of intensity I, n represents the slope of the function (which is usually close to 1 in scotopic condition) and K (or retinal sensitivity) the intensity of stimulation necessary to produce a b-wave half of maximal amplitude (Vmax). K and Vmax parameters have been used on several occasions to further characterize the scotopic function. These measurements were shown to be highly reproducible in test-retest conditions [41].

Figure 7 Scotopic luminance-response function curve (Naka-Rushton) obtained by plotting the b-wave amplitude data (ordinate in microvolts) against the intensity of the stimulus (abscissa in log intensity) with the use of sigmoidal curve fit software. Data obtained from both eyes (OD: right eye, OS: left eye) were plotted to illustrate high reproducibility. Vmax identifies maximal rod b-wave amplitude (as per equation) while K identifies the retinal sensitivity which is the intensity of the stimulus needed to evoke a b wave half the amplitude of Vmax.

VI. FACTORS AFFECTING THE ERG

The ISCEV standard makes use of a very brief flashing stimuli (usually 5 ms). Longer stimuli will add an OFF effect, which will contaminate the response and thus complicate the analysis; that is of course unless the experimenter wishes to separate the ON-ERG from the OFF-ERG [42,43]. Although the ISCEV standards are based on white light only, there are also specialized ERG protocols specifically aimed at isolating the activity of the short, medium, and long wavelength sensitive cones [44]. Use of these protocols might be helpful in both clinical and research settings. Normal aging also influences the ERG parameters. In humans, it is estimated the amplitude of the b-wave decreases by about 2.5 V per year from age 10 to 70 [45]. A diurnal variation in the ERG b wave has also been reported [46]. It is therefore suggested to record the ERGs at about the same time of day, especially if a follow-up study is considered.

Retinal pathologies, whether acquired or inherited, which have an impact on the normal processing of the visual stimulus can also alter the genesis of the ERG response. In order to appreciate the possible consequences, it is suggested to measure the amplitude and peak times of the major ERG components (a- and b-waves and OPs) and compare these parameters with normative data. Previous studies have shown that, depending on the retinal anomaly, the waves of the ERG could be affected in amplitude and/or timing. Also, examining each ERG component separately allows one to distinguish wave-specific anomalies. Previous studies have also shown that individual OPs can be specifically affected by a given disease process (or experimentally) while the remaining OPs are unaltered [25,26].

VII. INTERPRETATION OF THE

ELECTRORETINOGRAM

The electroretinogram is used to assess the functional integrity of the retina. With that in mind, one must remember that the retina contains two types of photorecep- tors—namely, the cones and the rods—and that the activity of each can be specifically isolated with the use of specific stimulating conditions as advocated by the ISCEV standards. This is of the utmost importance, given that most retinal disorders (acquired or inherited) will often initially affect the normal functioning of one of the two photoreceptor populations. In order to determine the normalcy of the response, two parameters are considered: namely, the peak time and the amplitude. The peak time is defined as the time separating the onset of the stimulus from the maximal amplitude (or peak) of the wave under consideration. For example the peak time of the a wave refers to the time elapsed between the onset of the flash and the first major negative peak of the ERG response (or a-wave), while the peak time of the b-wave identifies the time separating the flash onset and the highest positive peak of the ERG wave (or b-wave), and so on. The same logic also applies to the individual oscillatory potentials (OPs). Similarly, the amplitude of a given wave is always measured from the preceding trough to peak except for the a-wave, which is measured from the prestimulus baseline. Thus, the amplitude of the b-wave is measured from the trough of the a-wave to the peak of the b-wave, while the amplitude of the OP3 for example is measured from the trough between OP2 and OP3 and the peak of OP3. Once the amplitudes and peak times of the different ERG waves and OPs are obtained, they are compared to normative data to ascertain normalcy.

If one considers only amplitude and peak time measurements, the following diagnostic categories are possible: (1) normal amplitude and timing, (2) normal amplitude and delayed timing, (3) low amplitude and normal timing, and (4) low amplitude and delayed timing.

It should be noted that faster than normal responses are usually considered as a variation of normal. The above four categories can be applied to any of the ERG waves and OPs, although (as a rule) the analysis has been traditionally limited to the b-wave (photopic and scotopic). It is only recently that the a wave as well as the OPs (photopic and scotopic) have received increased attention and that their use was shown to increase the diagnostic potential of the ERG. Use of the above method of analysis of the ERG will not only allow the investigator to identify coneand/or rod-specific retinopathies but also localize the retinal site most affected by comparing a- and b-wave measurements. Similarly, since individual OPs were previously shown to be differently affected either by the stimulus parameters or pathologies, analysis of each OP individually is also strongly suggested in order to increase the diagnostic potential of the ERG.

VIII. OTHER ELECTRODIAGNOSTIC TECHNIQUES WITH MORE SPECIFIC OR LIMITED USE

A.The Pattern-ERG (PERG)

The pattern ERG is the retinal response evoked by viewing (mono or binocularly) an alternating reversible checkerboard pattern. It is claimed to be generated at the level of the retinal ganglion cell and/or optic nerve [47]. The PERG is made of two principal components: P50 and N95 (see Fig. 8), which letters and numbers correspond to the polarity (positive or negative) and usual timing of the peak or trough of the response observed in normal individuals. P50 is sensitive to the luminance and appears to be generated in part by the same generators of the fullfield ERG, whereas N95, which is sensitive to the contrast and spatial frequency of the stimulus, is more specialized to the ganglion cells. The overall amplitude of the PERG is relatively small and range from 0.5 to 8 V in normal individuals. This low amplitude accounts for the difficulty in achieving good recordings and explains why the technique is used routinely only by a handful of laboratories. PERG has received research attention because it can detect selectively macular and inner retina dysfunctions that go undetected with the full-field ERG. The pattern ERG is considered a good test for macular function but requires very good technical skills and experience with the technique. It should be noted that the PERG needs an averaging of about 100–400 responses in order to achieve a good signal-to-noise ratio. Fixation and refraction is also very important, which makes the technique difficult to use in infants and patients with low visual acuity (see the ISCEV standards for proper PERG recording).

B.The Multifocal ERG (mfERG)

The mfERG is a relatively new technique that allows for the assessment of localized retinal dysfunction within a 40 to 50° field that covers approximately 23%

Figure 8 Representative example of a PERG obtained binocularly in a female (26 years old) using a checkerboard pattern presented on a monitor with a field size of 16° 16° and reversals of 4.5/s and average of 800 responses (UTAS 3000 LKC Technologies Inc., Gaithersburg, MD). The largest amplitude, observed at N95, is close to 8 V.

of the total cone population [48]. The stimulus is a pattern (typically presented on a computer screen that is composed of an array of hexagons that alternate independently between black and white state in a pseudo-random sequence described mathematically as an m-sequence [49]. At any point in time, half of the hexagons are white and half are black, which allows for a constant luminance. The most common matrix is composed of either 61, 103, and 241 hexagons that increase in size with distance from the center to compensate for the decrease of cone density with eccentricity [50]. This allows the production of focal ERG responses of similar amplitudes independently if they are recorded in the macula or para-macula. At a fast rate of stimulation of 75 Hz, each hexagon is renewed every 13.3 ms. For the viewer, the stimuli appear as little lights flickering in a random manner. The electricity generated by the retina in response to the highspeed stimulation is recorded as a continuum (strand of ERGs) using the same electrodes that are used for the full-field ERG. Focal ERGS are extracted from each location using a cross-correlation technique [51]. This is made possible as each hexagon follows the same exact sequence of black and white presentation, although each location begins at a different point in the cycle. From our experience, using a DTL electrode and a matrix of 61 hexagons, good results are achieved with a 4-min protocol. However, because the patient cannot blink during the stimulation and has to maintain a perfect fixation at a cross in the center of

Figure 9 Representative example of a mfERG recording obtained with a matrix of 61 hexagons, which subtended a field of about 40° (VERISTM, Science 4, ElectroDiagnostic Imaging, Inc, San Mateo, CA). In (A), a 3-D topographical map (scalar plot) of local response density. Observe the central peak (fovea), which has the highest density of photoreceptors as well as a noticeable depression caused by the blind spot seen on the far left-hand side. In (B), a trace arrays of all 61 electroretinograms are presented. Observe that the overall amplitude at each location is maintained because the hexagons are scaled with eccentricity (larger in the periphery). Having similar amplitudes allows the detection of area of abnormalities just by looking at the trace arrays. Responses can also be averaged into rings, quadrants, or hemiretinal areas, to name a few (not presented). The total recording was 3 min 38 s, achieved using 16 slightly overlapping segments 15 s long (signal amplification: 100,000; bandwidth: 10–100 Hz). The mfERG recording was obtained from the left eye of a male (35 years old) dilated with Tropicamide 0.8% using a DTL fiber electrode.

the screen, the 4-min protocol is split into 16 segments of 15 s duration (30 s if using a contact lens). Depending on how the patient handles that task, the 4-min protocol can be completed in 5 to 15 min. In contrast to the full-field ERG, which necessitates minimal cooperation from the patients, cooperation is crucial to achieve the mfERG. Therefore, it would be difficult to perform this type of recording in infants. As with the pattern ERG, refraction is also important.

Recently, ISCEV introduced the first guidelines for the mfERG. In these guidelines it is recommended that when the hexagons are in the white state they should produce a luminance of 200 cd/m2, whereas when they are in the black state they should be close to a luminance of 0 cd/m2. The background should be

set at 100 cd/m2, yielding to an overall mean luminance of 100 cd/m2. Amplification is usually 100,000 times and the bandwidth set at 1–300 Hz or 10–300 Hz.

Typical views of the recording are presented in Figure 9. Looking at Figure 9 (right-hand side), it is clear that mfERG response shares some resemblance with the full-field cone ERG as it is composed of a negative deflection (N1) followed by a positive deflection (P1). However, in the mfERG, the peak of the response occurs at an earlier time and it is devoid of the fast oscillation observed in the full-field ERG, namely, the oscillatory potentials. This difference in the waveforms is likely due to the stimulation paradigm that is employed and how the data analysis is performed. In contrast to the full-field ERG in which the final response is composed of consecutive responses that are simply averaged together, in the mfERG the final response is composed of additions and subtractions of preselected responses that depend on the analysis selected. For instance, with the first-order analysis (most commonly used), only white hexagons that were followed by a black hexagon are averaged together at each location. With the second-order analysis, we are measuring the impact of successive flashes on the mfERG response [51]. Similar to the full-field ERG, the generators of the firstorder analysis of the mfERG appear to be the photoreceptors and bipolar cells. Higher order analyses that take into account the effect of consecutive flashes are believed to be indicators of ganglions cell activity as well as optic nerve function [51]. Studies are still needed to confirm the exact components and generators of the mfERG. The research interest in mfERG is that it provides a spatial resolution of the macula that cannot be achieved with the full-field ERG. Of note, animal research can be performed with mfERG, as shown by reports on monkeys, cats, rats, and pigs.

IX. CONCLUDING REMARKS

The purpose of this chapter was to provide the reader with standardized means to assess the retinal function in clinical or experimental conditions. Assessing the functional status of the retina using noninvasive procedures is relatively simple. The methods described can yield a wealth of extemely valuable information that can be used to make a diagnosis whether the retinal disorder was acquired through experimentation or otherwise, or inherited. However, in order to achieve the highest level of diagnostic accuracy possible, users are urged to create their own sets of normative data for all the species and age ranges under analysis.

ACKNOWLEDGMENTS

This work was supported by grants-in-aid from McGill University-Montreal Children’s Hospital Research Institute, the Canadian Institutes of Health Research