In our study, we examined the long-term changes of the mfERG responses in patients with retinitis pigmentosa. We also observed the changes of the visual field parameters and compared them with earlier findings by other authors.
17 18 19 20 21 22 23 24 25 26 27 28 29 30 In our group of patients, the progression of visual field loss followed an exponential decay, which was similar to the change described by Iannaccone et al.
16 (Table 2) . Our estimate for the yearly loss of visual field is 14.5%. The slope data (−0.156, respectively) is close to the −0.136 for target V4e and −0.172 for target I4e found by Iannaccone et al.
16 or −0.112 for target V4e found by Holopigian et al.
20 and thus confirm their data. Other studies, in which more patients were included found either somewhat higher
19 or lower slopes
21 22 (for details, see
Table 2 ).
Using a similar regression model, we found a well-defined reduction of Ganzfeld ERG and mfERG amplitudes of 6% to 10% per year of disease duration (see
Table 3 ). To overcome or reduce possible problems arising from changes introduced by the development of the VERIS versions 1 to 4.8, we reanalyzed all data with VERIS version 4.8. Other factors like the onset of disease, which was only determined in case history, still affect our estimates. However, our estimates of amplitude loss per disease year for the mfERG ring averages are similar to those of the visual field with comparable error estimates and confidence limits. Because of averaging, the response waveforms of the hexagons of a given eccentricity the signal-to-noise ratio improves. Other factors like electrode placement or target fixation limit the test–retest reliability and may hide slight disease progression. One should keep in mind that two successive recordings can show 10% to 20% variation in amplitude.
31 32 33 34
We further looked at the relationship between remaining visual field area and mfERG response amplitudes. Even though the loss of mfERG amplitude and the visual field show a similar decay, we found only a weak correlation between both (
r S = 0.61). Similarly, we found a weak (
r S = 0.58) correlation of the scotopic mixed cone–rod response with the visual field loss. Some studies emphasize the existence of a substantial correlation between visual field and Ganzfeld ERG response amplitude.
25 26 27 28 29 30 Iannaccone et al.
26 found a strong correlation between ERG mixed cone-rod response b-wave amplitude and visual field area determined for Goldmann I4e and III4e isopters (
r = 0.89 and
r = 0.87, respectively) but similar correlations to ours for the V4e isopter (
r = 0.69). Others found only weak correlations between visual field diameter and ERG amplitudes.
28 29 Using a Naka-Rushton equation to estimate maximum amplitude of rod ERG-responses, Birch et al.
30 found a significant correlation of this amplitude to the size of the dark-adapted visual field, whereas Massof et al.
27 did not find a correlation using similar methods but light-adapted visual field testing.
Of note, Sandberg et al.
25 found weak correlations of the 0.5- and 30-Hz ERG amplitudes with visual field area (
r S = 0.54 and 0.60, respectively), but observed higher correlations (
r S up to 0.85) in a subgroup analysis, suggesting that the relationship of visual field size to ERG amplitude depends on genetic type, the altered protein, and/or the specific mutation. Because we found a significant correlation between the outermost ring average amplitude with the Ganzfeld scotopic mixed cone–rod response amplitude (
r S = 0.87) and the cone response amplitude (
r S = 0.85), this may be true of the mfERG results as well.
In conclusion, the mfERG provides a useful measure of the retinal function; it does not replace, but complements psychophysical methods including visual field and color vision testing. The response amplitude follows a similar exponential decay as the visual field and may provide reproducible responses, even if the Ganzfeld ERG is nearly extinguished. This issue is becoming more important, now that essential steps toward possible therapies for retinal degenerations are being made and reliable and objective testing methods are needed. The mfERG is well-suited for observation and long-term follow-up in disease development and—in addition to other psychophysical methods—it could be used as an objective outcome measure in upcoming treatment studies involving patients with advanced retinal diseases.