Published in:
01-02-2019 | Zonisamide | Original Research Article
The Topographical Relationship between Visual Field Loss and Peripapillary Retinal Nerve Fibre Layer Thinning Arising from Long-Term Exposure to Vigabatrin
Published in: CNS Drugs | Issue 2/2019
Login to get accessAbstract
Background
The antiepileptic drug vigabatrin is associated with characteristic visual field loss (VAVFL) and thinning of the peripapillary retinal nerve fibre layer (PPRNFL); however, the relationship is equivocal.
Objective
The aim of this study was to determine the function–structure relationship associated with long-term exposure to vigabatrin, thereby improving the risk/benefit analysis of the drug.
Methods
A cross-sectional observational design identified 40 adults who had received long-term vigabatrin for refractory seizures, who had no evidence of co-existing retino-geniculo-cortical visual pathway abnormality, and who had undergone a standardized protocol of perimetry and of optical coherence tomography (OCT) of the PPRNFL. Vigabatrin toxicity was defined as the presence of VAVFL. The function–structure relationship for the superior and inferior retinal quadrants was evaluated by two established models applicable to other optic neuropathies.
Results
The function–structure relationship for each model was consistent with an optic neuropathy. PPRNFL thinning, expressed in micrometres, asymptoted at an equivalent visual field loss of worse than approximately − 10.0 dB, thereby preventing assessment of more substantial thinning. Transformation of the outcomes to retinal ganglion cell soma and axon estimates, respectively, resulted in a linear relationship.
Conclusions
Functional and structural abnormality is strongly related in individuals with vigabatrin toxicity and no evidence of visual pathway comorbidity, thereby implicating retinal ganglion cell dysfunction. OCT affords a limited measurement range compared with perimetry: severity cannot be directly assessed when the PPRNFL quadrant thickness is less than approximately 65 µm, depending on the tomographer. This limitation can be overcome by transformation of thickness to remaining axons, an outcome requiring input from perimetry.