Hydroxychloroquine and chloroquine retinopathy: a systematic review evaluating the multifocal electroretinogram as a screening test.

PURPOSE: To determine the validity of multifocal electroretinography (mfERG) as a screening tool for detecting chloroquine (Aralen, Sanofi Aventis, Bridgewater, NJ) (CQ) and hydroxychloroquine (Plaquenil, Covis Pharmaceuticals, Inc, Zug, Switzerland) (HCQ) retinal toxicity in patients using these medications. To evaluate the sensitivity and specificity of mfERG when compared with automated visual fields (AVFs), fundus autofluorescence (FAF), and optical coherence tomography (OCT).

CLINICAL RELEVANCE: The 2011 American Academy of Ophthalmology recommendations on screening for CQ/HCQ retinopathy recommended a shift toward more objective testing modalities. Multifocal electroretinography may be effective in detecting functional change before irreversible structural damage from CQ/HCQ toxicity.

METHODS: We performed a search for records reporting the use of mfERG for screening CQ/HCQ retinopathy in MEDLINE (PubMed and OVID), EMBASE, and Web of Science, and assessed these using the QUADAS-2 risk of bias tool. We conducted an analysis of 23 individual studies and their reported individual patient data (449 eyes of 243 patients) published from January 2000 to December 2014.

RESULTS: Multifocal electroretinography had the greatest proportion of positive test results, followed by AVF. The pooled sensitivity and specificity of mfERG were 90% (95% confidence interval [CI], 0.62-0.98) and 52% (CI, 0.29-0.74), respectively, with AVF as reference standard (13 studies). Sensitivity was high, but specificity was variable when OCT, FAF, and the positivity of 2 of 3 tests was used as the reference standard. When verified against AVF as the reference test, patients with a false-positive mfERG result received higher HCQ cumulative doses (1068 g) than patients with true-negative (658 g, P < 0.01) and false-negative (482 g, P < 0.01) results.

CONCLUSIONS: Multifocal electroretinography was shown to have a high sensitivity but variable specificity when verified against AVF, OCT, FAF, and a combination of tests. The greater average cumulative dose in the false-positive group compared with the true-negative group when mfERG was verified against AVF suggests that mfERG may have the ability to detect cases of toxicity earlier than other modalities. There is an unclear risk of bias in the available evidence, and future studies should adhere to Standards for Reporting of Diagnostic Accuracy reporting guidelines.