advertisement

---

Editors Selection IGR 18-3

Examination methods: PERG and OCT

Comment by Brad Fortune on:

14037 The relationship between retinal ganglion cell function and retinal nerve fiber thickness in early glaucoma, Ventura LM; Sorokac N; Santos Rde L et al., Investigative Ophthalmology and Visual Science, 2006; 47: 3904-3911

Find related abstracts

---

The goal of a recent study by Ventura et al. (638) was "to compare relative reduction of retinal ganglion cell (RGC) function and retinal nerve fiber layer (RNFL) thickness in early glaucoma by means of steady-state pattern electroretinogram (PERG) and optical coherence tomography (OCT), respectively." The rationale behind the study was based on the idea that deviations from a proportional decline between measures of structure and function, over a range of (early) glaucomatous damage, may enable one to test the hypothesis that RGCs exhibit a degree of dysfunction prior to their ultimate demise and disappearance. The authors used the PERG as a measure of RGC function (over the central 12.5 degrees of the visual field), and RNFL thickness as a surrogate measure of RGC number, matched for the corresponding area of the retina. The general result was that there was a slightly greater 'relative' loss of PERG amplitude as compared with RNFL thickness, on average, in their population of glaucoma suspects (GS) and patients with early manifest glaucoma (EMG). The frequency of eyes with abnormally small PERG was also slightly greater than the frequency of eyes with abnormally thin temporal RNFL. These findings led the authors to conclude that there is a population of viable RGCs (within the central retina) that are dysfunctional in early glaucoma.

There is a population of viable RGCs (within the central retina) that are dysfunctional in early glaucoma

The authors also present a very thorough critique of their own study within the Discussion section, which I recommend as essential reading to those interested in the conclusions of this study. There was, however, another limitation inherent in the study design, not explicitly considered by the authors. Because this was a cross-sectional study in which 'relative' losses of functional capacity and structural integrity were compared, it is important that the normative databases differed critically. The RNFL values derived from OCT scans were compared to a normative database collected (and published) by another laboratory (under presumably different conditions). In contrast, the PERG normative database was collected in the same laboratory during a previous study (under presumably identical conditions). Thus, deviations from the expected normal value (the primary outcome measure used to compare PERG and RNFL losses) were actually 'relative' to two different normal populations. Any systematic difference between these two normative populations would introduce a bias such that either the structural measure (RNFL) or the functional measure (PERG) might appear to have a larger 'relative' loss and/or be more frequently abnormal. Although the populations defining the two normative databases were similar in size, age and ethnic mix, this potential problem could have easily been averted by including a control group in the current study, in which both OCT and PERG measures were obtained and used as a common normative database.

Nevertheless, to the extent that the underlying assumptions are valid, particularly, that PERG and RNFL values are both linearly related to the stage of glaucomatous damage, and more specifically, to the number of RGCs, and that the slopes of both functions versus the number of RGCs are equal, the observed divergence above a slope of 1 in the function relating relative PERG amplitude to relative RNFL thickness suggests that RGC dysfunction exists prior to loss of RNFL thickness. Further studies are needed to test these underlying assumptions, as well as the robustness of the interesting findings presented by Ventura et al.

---

• ← Previous
• Next →

---