Ocular electrophysiology provides an objective assessment of the integrity of the visual pathway from the eye to the brain. By recording the electrical signals produced by visual stimulation via the electroretinogram, electro-oculogram or visual evoked potential, the location and extent of the disruption in the visual pathway can be determined. These assessments measure evoked potentials that are transient electrical responses produced by neurones in response to a stimulus.
1.5.4.1 Electroretinogram
The electroretinogram (ERG) is the objective recording of the summed electrical activity produced by the retina in response to a visual stimulus. This can be measured under photopic or scotopic conditions and with a flash or pattern stimulus. Although the ERG is a summed response, different groups of cells become active at different times after the onset of a stimulus. Hence, each part of the waveform produced gives information about different cells within the retina. Recording the ERG requires three electrodes, an active on the cornea, reference on the skin at the outer canthus, or in the contralateral eye and earth on the forehead.
Since 1989 The International Society for Clinical Electrophysiology of Vision (ISCEV) has provided regularly reviewed standards for the recording parameters of full field clinical ERG techniques (McCulloch et al. 2014). The actual form of the wave produced depends on:
state of adaptation stimulus wavelength stimulus intensity
stimulus temporal frequency
Low temporal frequencies, shorter wavelengths, lower luminance stimuli and dark adaptation all help to isolate the rod response (Perlman 2001). The stimulus can also be patterned, global or focal to isolate different retinal locations.
1.5.4.2 Flash electroretinogram
Figure 1-12 shows the waveform produced in response to a single flash full field stimulus that records summed potentials from all the retinal cells. The resultant ERG has a
repeatable waveform.
Figure 1-12. Flash electroretinogram waveform from a normal subject showing the amplitude of the a wave (a) and b-wave (b), and the implicit time of the a wave (t) (McCulloch et al. 2014).
In 1933, Granit determined that the scotopic transient ERG of the cat was attributable to the summation of 3 physiological processes, PI, PII and PIII (Granit 1933). These three processes were numbered depending on the order of disappearance under progressive anaesthesia. PI is a positive wave of long latency, PII is a reasonably fast positive wave and PIII, which develops faster than PI or PII, is a negative wave. PIII has been further divided into the ‘fast’ and ‘slow’ components that have also been attributed to different layers of the retina. However, these processes cannot be isolated in the standard ERG. Figure 1-13 shows that the three processes sum together to form the overall ERG
Vo lt ag e µ V Time (ms)
waveform that is shown in black. The clinically recordable peaks and troughs of the ERG are referred to as the sub-components. The following section will discuss certain sub- components of the ERG (those most relevant to this thesis), the recording parameters required to elicit them, and their proposed retinal origins.
Figure 1-13. Granit’s three processes of the ERG.
1.5.4.3 The a-wave
Appearing first in the waveform, the a-wave is a negative potential, and has been attributed to the leading edge of Granit’s PIII component seen in Figure 1-13. The origin of this wave has been has been determined by intra-retinal electrode studies which found signals from the photoreceptors contributed to PIII (Brown et al. 1957). Current source density analysis suggests that the rod outer segments are the main source of the scotopic ERG a-wave, as the photocurrent they produce is of a similar sign and waveform to the currents that underlie the a-wave (Penn and Hagins 1969). Studies using the glutamate agonist sodium aspartate, which prevents the transmission of glutamate from the photoreceptors to bipolar cells, have found that the a-wave is not extinguished by its
administration which is also suggestive of a photreceptor source (Wakabayashi et al. 1988). In photopic conditions, the origins of the a-wave are more complicated and may involve postreceptoral activity from hyperpolarising bipolar cells or horizontal cells (Bush and Sieving 1994).
1.5.4.4 The b-wave
Evidence for b-wave origins come from single cell intracellular recordings from the inner nuclear layer. It is this layer and the bipolar cells within it which contribute to the positive b-wave (Brown et al. 1957; Brown and Wiesel 1961), which is attributed to Granit’s
process PII (Granit 1933). In particular, it is thought to be the ON bipolar cells that contribute to the b-wave. 2-amino-4-phosphonobutyrate, which abolishes the response of ON- bipolar cells to light has been found to diminish the b-wave (Tian and Slaughter 1995; Stockton and Slaughter 1989). Studies of the mud puppy retina have also suggested that there may be an involvement of the Müller cells (Miller and Dowling 1970), although studies have found that blocking the Müller cell activity using Barium ions does not reduce the b-wave amplitude (Lei and Perlman 1999).
1.5.4.5 The focal ERG
The focal ERG allows the identification of localised defects by using a uniform flickering or flashing stimulus with a desensitising surround to record responses from the central retina. When centred on the macula, the combination of a small stimulus and the suppression of signals from the peripheral retina by the surround allow a focal cone
driven response to be recorded. The flicker stimulus results in a sinusoidal waveform, and a flicker of greater than 30Hz will result in a cone driven response (McCulloch et al. 2014). A postreptoral contribution to the flicker ERG has also been proposed due to the effect of glutamate on the waveform (Kondo and Sieving 2001).
This recording is useful in conditions where localised functional changes occur such as in AMD and cone dystrophies, where small changes to the waveform may be swamped by the summed signals of the peripheral, healthy retina. However there are no international standards set for the recordings and good fixation is required (Berrow et al. 2010) which may make recording from individuals with nAMD and GA difficult.
1.5.4.6 The ERG in AMD
Different types of ERG recording have been used to investigate retinal function in AMD. Using the full field ERG to study peripheral visual function Sunness et al. (1985) concluded that peripheral visual function was not affected by AMD severity. Jackson et al. (2004) later agreed, finding no difference in the response onset or amplitude of the a-wave of the full field scotopic ERG, between participants with early and late AMD and age matched and younger controls. This is probably due to the full field ERG summing the overall retinal response, which would include peripheral rods not affected by macular degeneration and may mask macular abnormalities. In contrast, significant reductions in the a and b-wave amplitudes of the full field rod ERG and prolonged b-wave implicit times have been found in another study, suggesting that there is a global impairment in AMD,
possibly due to vascular changes (Walter et al. 1999). The differences between these studies can be explained in part by the small sample (n=7) of late stage AMD participants used in the Jackson et al. (2004) study, whereas there was no breakdown of disease classification of the 66 participants studied by Walter et al. (1999), which included both participants with nAMD and GA, which may have skewed results.
Using new multifocal ERG (mfERG) techniques which target the rod pathway, rod function has been found to be preferentially affected by AMD (Chen et al. 2004; Feigl et al. 2006). When assessing both cone and rod mfERGs, the rod mfERG response was found to be delayed in participants with early AMD, but there were no significant changes in the cone response (Feigl et al. 2005a). However, Li et al. (2001) did find changes in the foveal amplitude and latency of the cone dominated mfERG in participants with AMD and in asymptomatic eyes of participants with early AMD or AMD in the other eye, suggesting that the mfERG may be useful in diagnosing and monitoring at risk groups. Differences in the recordings have also been found between participants with nAMD and GA, with the response densities of N1 and P1 decreasing dramatically for the nAMD group compared to the participants with GA, suggesting that the mfERG may be used for AMD disease quantification (Huang et al. 2000). The mfERG may also be used for monitoring the effectiveness of treatments (Berrow et al. 2010).
The focal cone ERG has been found to have delayed implicit times in eyes with early AMD (Binns and Margrain 2007) and in the fellow eyes of participants with unilateral nAMD
present before changes in VA. A reduction in mean amplitude has been found in patients with GA, which was correlated with GA severity (Falsini et al. 1999). There is also evidence that focal ERG abnormalities could precede the morphological changes typical of more advanced disease (Falsini et al. 1999). Implicit time delays of 1ms have been found in eyes with a choroidal perfusion defect (Remulla et al. 1995), which may underlie hypoxia driven progression of AMD as discussed in section 2.2.5.