A comparison of structural and functional changes in patients screened for hydroxychloroquine retinopathy

Abstract

Purpose: The aims of this retrospective study were to compare the results of recommended screening tests for hydroxychloroquine-related retinal toxicity and analyze disparities between the structural and functional findings.

Methods: Thirty-four patients (31 women and 3 men) were included in the study. All were evaluated with standard automated perimetry using the 10-2 and/or 24-2 visual field program (Zeiss, Meditec), multifocal electroretinography (mfERG), spectral-domain optical coherence tomography (SD-OCT), and short-wavelength fundus autofluorescent imaging (SW-FAF). The results for the right eye from each patient were analyzed. Visual fields were classified as normal or abnormal based on pattern deviation plots, and mfERGs based on a comparison of R5 ring ratios to values from 20 controls. The SW-FAF images were examined for areas/rings of abnormal hypo- and/or hyperautofluorescence, and the SD-OCT line scans were classified as abnormal based on visual inspection and thickness measurements of the outer segment plus retinal pigment epithelial layer and total receptor layers compared to mean thicknesses from 35 controls.

Results: Fifteen patients had abnormal results on at least one test; however, only two patients had abnormal results on all four tests. Excluding SW-FAF, seven of the 15 had abnormal visual fields, mfERG ring ratios, and SD-OCTs. The remaining eight had either abnormal mfERGs and/or visual fields and normal SD-OCTs. We found no evidence of abnormal SD-OCTs in the presence of normal mfERG and visual field results.

Conclusions: The findings suggest that functional deficits precede structural changes seen on SD-OCT in these patients.

Figure 1

Ring averaged waveforms R₁–R₆ for a normal observer obtained after grouping the 103 mfERG responses in the trace array (shown in field view) in the form of 6 rings. The locations of the first negative peak N1 and first positive peak P1 for the average waveforms are shown. Response densities are calculated as P1-N1 expressed in nV/deg². The R₅ ring ratio values are presented in the column on the left hand side.

Figure 2

Image of a SD-OCT scan through the horizontal meridian for a normal observer. The colored lines indicate the borders that were segmented. The thicknesses of the OS+ and REC+ layers were measured.

Figure 3

HVF 24-2 and 10-2 pattern deviation plots and multifocal ERG trace arrays (field view) for P22 and P6.

Figure 4

R₅ ring ratio values for 34 eyes of 34 patients. Horizontal lines indicate the 5th and 95th percentiles of the values for the group of normals. The filled red circles are the values for 20 eyes of 20 controls.

Figure 5

SW-FAF and corresponding SD-OCT images of P22, P10 and P33. The vertical arrows indicate the location of discontinuity of the EZ band.

Figure 6

Thickness values along the horizontal meridian of the REC+ (A) and OS+ (B) layers for 34 eyes of patients on HCQ, The dashed and continuous black lines represent the mean +/− 2SDs thickness values for 35 normal control eyes.

Figure 6

Thickness values along the horizontal meridian of the REC+ (A) and OS+ (B) layers for 34 eyes of patients on HCQ, The dashed and continuous black lines represent the mean +/− 2SDs thickness values for 35 normal control eyes.

Figure 7

HVF 10-2 gray scale, total deviation and pattern deviation plots, multifocal ERG trace arrays (field view) and R₅ ring ratio values for P25 (represented by inverted green triangles in A) and P13 (represented by filled blue circles in B).

Figure 7

HVF 10-2 gray scale, total deviation and pattern deviation plots, multifocal ERG trace arrays (field view) and R₅ ring ratio values for P25 (represented by inverted green triangles in A) and P13 (represented by filled blue circles in B).