Longitudinal Assessment of Increased Ellipsoid Zone Attenuation Associated with Hydroxychloroquine Use: Possible Role for Automated Toxicity Detection

Abstract

Purpose : To evaluate longitudinal OCT threshold changes in ellipsoid zone (EZ) attenuation that may precede overt hydroxychloroquine (HCQ) toxicity using an automated EZ mapping platform.

Methods : All participants underwent macular cube scan at two time points using the Cirrus HD-OCT, which were imported and analyzed in an automated retinal layer and EZ mapping segmentation platform. Clinical characteristics were collected. Longitudinal change in en face percentage of en face EZ attenuation (EZ thickness < 20μm). Clinical factors associated with an absolute longitudinal en face EZ partial attenuation increase of 4% were analyzed.

Results : Four hundred one eyes of 401 patients were included. Mean age was 57.6±13.0 years, daily mean HCQ dose was 367.1±72.6 mg, mean actual body weight (ABW) dose was 4.9±1.6 mg/kg, and mean ideal body weight dose was 6.6±1.6 mg/kg. Mean drug duration was 5.8±3.9 years and mean interval between OCTs was 3.1±0.9 years. There was a significant increase in en face EZ attenuation from the first (4.7±18.2%) to the second OCT (5.8±19.1%; p=0.02). An increase in en face EZ attenuation of 4% or more was significantly associated with age of HCQ initiation (p<0.001), age at time of the first (p<0.001) and second OCT (p<0.001), VA at the first (p<0.001) and second (p<0.001) OCT, cumulative HCQ dose at the second OCT (p=0.02), actual body weight HCQ dose (p=0.02), and ideal body weight HCQ dose (p=0.008).

Conclusions : Longitudinal OCT assessment revealed a significant increase in EZ attenuation which correlated with age, daily dose, and dose based on actual as well as ideal body weight. Additional research is needed to further validate these subclinical progressive changes, to further characterize the long-term significance of these changes, and their impact on identification of HCQ toxicity.

This is a 2020 ARVO Annual Meeting abstract.