Author response: Longitudinal optic neuritis-unrelated visual evoked potential changes in NMO spectrum disorders

We thank Dr. Chen for his comment. We included neuromyelitis optica spectrum disorder (NMOSD) patients with longitudinal visual-evoked potential (VEP) data covering ≥3 months. We calculated the changes per year using regression analysis. For eyes with <12 months longitudinal OCT follow-up available, the change rate was interpolated. However, as interpolation of data with short intervals can lead to amplification of measurement variability, we also performed a more conservative subanalysis including only eyes with ≥12 months interval between VEP measurements. For this subanalysis, only 58% and 69% of eyes were available for the investigation of the rates of change of latencies and amplitudes, respectively. Of note, it has previously been reported that the amplitudes show a stronger variability of measurement compared to latencies.^1,2 Therefore, we believe that the sample size was not sufficient to achieve statistical significance in the subanalysis. Moreover, the prominent outliers of amplitude changes of eyes with <12 months of follow-up are likely to result from amplification of measurement variability and should be interpreted with caution. On the other hand, the latency findings show consistent and reliable increases also in the conservative ≥12 months subanalysis, thus representing the main finding of our study and a better parameter to investigate chronic changes.