Background/Purpose:

Hydroxychloroquine (HCQ) is widely used in the treatment of systemic lupus erythematosus (SLE), rheumatoid arthritis (RA), and other rheumatic diseases. While generally well-tolerated and regarded as one of the safest treatment options for these conditions, the major long-term risk is vision-threatening retinopathy. The risk of retinopathy has been shown to vary by amount of HCQ exposure. However, the true risk of HCQ retinopathy remains unclear, and clinical consensus may be lacking between US and European rheumatology and ophthalmology societies regarding the safe dosing of HCQ. Our objective was to systematically review and appraise current evidence of the risk of HCQ retinopathy and its purported risk factors.

Methods:

We conducted a systematic search of MEDLINE and EMBASE databases through June 2017. Inclusion criteria were: 1) study sample including individuals treated with HCQ; 2) study design of either retrospective or prospective study with a comparator group; and 3) retinopathy outcomes reported. The primary outcome was HCQ retinopathy incidence and/or prevalence, and other data of interest included measures of HCQ exposure (e.g., mean daily dose, mean cumulative dose, and mean duration of HCQ use), method of HCQ retinopathy ascertainment, risk factors for HCQ retinopathy, and rheumatologic diagnosis.

Results:

We identified 3,236 unique citations. Of these, 28 studies were ultimately included in the systematic review. Ten studies (all published in 2014 or later) reported current standard-of-care screening modalities (e.g., Humphreys visual fields (VF) and spectral domain-optical coherence tomography (SD-OCT)). Eleven studies reported the use of only older screening modalities (e.g., funduscopic, visual acuity, and color vision assessments), 5 studies used ophthalmologist reports of toxicity, and 2 studies did not describe a method of retinopathy ascertainment. Of the 10 recent studies, HCQ retinopathy risk estimates ranged from 2.9-30.5%, with 2 studies reporting a risk of 10% or higher. Of the 18 older studies, HCQ retinopathy risk estimates ranged from 0.5%-21%, with 2 studies reporting a risk over 5%. Twelve studies examined predictors of retinopathy. Of these, 3 found age and 1 found CKD to be predictors of retinopathy. Duration of HCQ use, cumulative dose of HCQ, and daily dose of HCQ were associated with HCQ retinopathy risk in 5, 4, and 2 studies, respectively. Underlying rheumatologic disease (e.g., SLE versus RA) did not impact retinopathy risk. The mean daily HCQ doses varied from 3.1mg/kg/day to 8.4mg/kg/day for the overall cohorts exposed to HCQ, with some studies reporting dose per ideal body weight and others reporting actual body weight.

Conclusion:

This is the first systematic review on the risk of toxic retinopathy secondary to HCQ. The risk estimate of HCQ retinopathy varies considerably in the published literature, with increased risk estimates reported in recent studies. More recent ophthalmologic screening methods including SD-OCT have increased sensitivity to detect earlier stages of retinopathy. Prospective studies are urgently needed to accurately characterize the risk and risk factors of HCQ retinopathy.