Attenuation of retinal nerve fibre layer in people with epilepsy receiving valproate

2019 
Abstract Objectives Valproate (VPA) is one of the most frequently used anti-epileptic drugs (AEDs) worldwide. Its effects in decreasing the retinal nerve fibre layer (RNFL) thickness remain debatable. We aimed to evaluate the effect of VPA usage on the RNFL in comparison with other AEDs and no AED usage in people with epilepsy (PWE). Methods In this observational case-control study, PWE were enrolled and divided into three groups: PWE 1) receiving VPA monotherapy throughout their clinical course; 2) receiving an AED other than VPA as monotherapy; and 3) who never took any AED. RNFL thickness of the right eye was measured by optical coherence tomography (OCT). In each individual, disease-related information was recorded. Results A total of 86 individuals (51 males; median age, 25 years) with an average epilepsy duration of 6.88 years were enrolled. No difference in the demographics except for sex was noted between the groups. The average RNFL thickness in 26 individuals who had received VPA (group I) was 93.73 ± 9.24 μm, which was significantly lower than the corresponding values for the 31 individuals who received other single AED regimens (group II; 99.71 ± 8.50 μm; p = 0.031) or the 29 individuals who never used any AED (group III; 102.79 ± 8.05 μm; p = 5.67 × 10-4), especially in the superior and inferior quadrants. The RNFL attenuation was significantly correlated with the epilepsy duration in groups II and III (r = 0.351, p = 0.006). However, no correlation between epilepsy duration, cumulative dosage of VPA, duration of treatment with VPA and RNFL thickness was found in group Ⅰ. Conclusion These preliminary findings suggest an association between VPA usage and reduction of retinal thickness in PWE, especially in the superior and inferior quadrants. Epilepsy itself might also be another risk factor for RNFL attenuation. Further studies need to confirm this finding and to unravel the underlying mechanism.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    31
    References
    3
    Citations
    NaN
    KQI
    []