Assessment of parafoveal cone density in patients taking hydroxychloroquine in the absence of clinically documented retinal toxicity.

Purpose To measure cone density in patients taking hydroxychloroquine (HCQ), with no clinical evidence of maculopathy. Methods Patients visiting for HCQ macular toxicity screening in the Besancon University Hospital Ophthalmology Department (France) were studied. They underwent routine examination including spectral-domain optical coherence tomography, fundus autofluorescence and multifocal electroretinogram to detect HCQ-induced retinal toxicity. Cone metrics (density, spacing and percentage of cones with six neighbours) were obtained using an adaptive optics camera (RTX1, Imagine Eyes, Orsay, France). The region of interest corresponded to a 0.3° × 0.3° square placed nasally and temporally at 2° of eccentricity from the fovea. Results Forty eyes of 23 patients were studied. The majority of the patients (21/23) were female. They were aged from 25 to 60 years (mean age ± SD: 40.1 years ± 10). The cumulative dose for HCQ ranged from 24 to 2160 g (777 ± 558 g). None of them displayed HCQ toxicity on screening tests. Bivariate analysis showed moderate cone loss with escalating doses of HCQ (linear regression, r² = 0.23, p = 0.018). Cone spacing also increased with increasing cumulative dose (r² = 0.17, p = 0.008). Cone packing remained unchanged (p > 0.05). Multivariate analysis showed that age and cumulative dose were additive and independent factors of cone dropout. Conclusions In this pilot study, we observed moderate cone loss as HCQ cumulative doses increased. The early detection of parafoveal cone metric changes may represent the earliest sign of HCQ macular toxicity during screening.