The prevalence of, and risk factors for, hydroxychloroquine retinal toxicity seem to be more common than previously recognized.
Melles RB, Marmor MF. The Risk of Toxic Retinopathy in Patients on Long-term Hydroxychloroquine Therapy. JAMA Ophthalmol. (2014) Online October 2, 2014. doi:10.1001/jamaophthalmol.2014.3459
Scholl HPN, Shah SMA. We Need to Be Better Prepared for Hydroxychloroquine.JAMA Ophthalmol. (2014) Online October 2, 2014. doi:10.1001/jamaophthalmol.2014.4090
Hydroxychloroquine-related retinopathy occurs in 7.5% of patients, about three times higher than previous estimates, researchers report in JAMA Ophthalmology.
Daily dose, cumulative dose, duration of use, kidney disease and lower weight correlate with hydroxychloroquine-related retinopathy. There also may be an association between the disease and the concurrent use of tamoxifen for breast cancer treatment.
Irreversible retinopathy, sometimes leading to blindness, is a rare adverse effect of hydroxychloroquine (HCQ). Earlier studies have focused on short-term users or severe retinal toxicity. More sensitive diagnostic methods find less severe retinopathy earlier, and retinopathy increases significantly after 10 years.
The investigators reviewed electronic health records from 2,361 patients in California's Kaiser Permanente system. Patients had taken HCQ for at least 5 years, and underwent a central visual field examination or spectral-domain optical coherence imaging (SD-OCT) study.
The investigators pinpoint long-term use as a significant risk factor. At a daily dose of HCQ at 5.0 mg/kg of real body weight, the prevalence of retinal toxicity remains at less than 2% within the first decade of use, but then jumps to almost 20% after 20 years of use.
Current screening guidelines suggest calculating dose by ideal body weight to reduce the theoretical risk of overdosing obese patients, but these authors prefer dose calculation based on real body weight, saying that it predicts renal toxicity better.
However, an accompanying commentary says that that this does not have support from randomized studies.