PurposeTo describe the findings in a case of hydroxychloroquine-induced maculopathy using optical coherence tomography (OCT) Spectralis (Heidelberg Engeneering) with en-face transversal section.
CaseA 77-year-old woman on hydroxichloroquine for treatment of erythematous systemic lupus was referred from to screen for hydroxichloroquine-induced toxicity. The daily dosage was 50 mg for 8 years (approximated accumulate dose: 146g). Clinical examination was normal with a best-corrected visual acuity of 20/20 on both eyes (OU). Humphrey visual field 10-2 white showed significant persistent paracentral defect on right eye (OR) and a subtle superior paracentral defect on pattern deviation map on left eye (OS) without correlation on grey scale map (Figure 1). A multifocal electroretinogram (mERG) confirmed the diagnosis of premaculopathy showing a decrease in retinal response in OU with an abnormal R1/R2 relation being 2.62 on OD and 5.22 on OS, assuming the normal value is 2.6 according to the age of the patient [1,2] (Figure 2). The Fanswoth-Munsell test revealed a moderate decrease of sensitivity to colors on OD with a possible protanopia, being normal OI. The Spectral Domain OCT (Heidelberg Engineering) showed aberration of ellipsoid layer and retinal pigment epithelium (RPE) defects using Retina Fast protocol but those findings where nonspecific and difficult to correlate with the HFA results on OD, although OS appears to be unaffected (Figures 3 and 4). Using the enface protocol on OD hypo/hyperreflective lesions where described as well as in OS which had been previously informed as normal OCT using sectional protocols (Figures 3 and 4).
DiscussionHydroxychloroquine is a wide range used drug for rheumatologic or dermatological disorders. It may induce retinal toxicity due to the deposit of the drug at the RPE causing alterations at that level that induce cellular death and secondary to the metabolically disturbance, photoreceptor loss. Since 2011, the recommendations for surveillance of these patients include spectral domain OCT (SD-OCT), ophthalmological exploration and 10-2 perimeter. If available, it is recommended to perform an auto fluorescence and mERG among others tests [3]. There is not a gold standard method, so the use of a combination of tests is recommended [4].
SD-OCT examination interpretation can be done in either a qualitative or quantitative method. The qualitative methods take into consideration the alterations on RPE or the disruption of the inner segment (IS) layer due to cellular loss [5-7]. On the other hand, the quantitative analysis uses the thinning of inner layers to measure the thickness from inner limiting membrane (ILM) to RPE [8]. Johson, et al. found a significant relationship between the thinning of inner nasal and temporal quadrants related to the accumulated dose and, according to its reproducibility, it could be used as an early detector. However, it presents a low sensitivity, being very specific (97%), so in case of presenting a pathological exploration, the thinning of the ILM - RPE measurement would be pathognomic for toxicity [9], but being useless as an screening method due to the low sensitivity. The qualitative analysis showed even lower sensibility than the quantitative, and also less reproducibility because of the difficulty for surveillance, so it won't be recommended as a screening method [10]. As Browning, et al. suggests, mERG may be affected on an early stage being very sensible, nevertheless, associating a high rate of false positives so its role as an screening method is also questioned [11].
In our case, the right eye is affected with a subtle scotoma that maintained persistent on nasal inferior quadrant, using autofluorescence (HFA-AF) a hypoautofluorescent lesion could be observed at temporal superior quadrant. The topographical correspondence of the lesion correlates with the lesions showed with transversal reconstruction en-face presenting an hyperreflective patron at inner layers and hyporreflective at outter layers, corresponding to the region affected among other tests such as HFA-AF or 10-2 visual field. Nevertheless, on the left eye, it was noticed a subtle decrease of sensibility using the 10-2 perimetry pattern map with at superior and paracentral area. The OCT B-Scan sectional protocol appears unaffected as well as the autoflourescence, however, using en-face protocol, hyporreflective lesions were shown at the inferior and paracentral area of ellipsoid layer.
Recently, Itoth Y, et al. [12] obtained using ellipsoid mapping with en-face OCT a volumetric scale of this layer showing a decrease in volume of those patients with hydroxicloroquine intake among other maculopathies such as geographical atrophy and after ocriplasmin inyection. Also there are described morphologic changes that could be associated, in pattern-based analysis, with bull-eye maculopathy in advanced cases and with the disruption of ellipsoid layer in cases of moderate affection. Using en-face protocol, other cases of paramaculopathies have been studied, such as retinal artery obstruction. On those patients, were found hyperreflective lesions at ischemic areas which coincide with the areas of hypoperfusion using angioOCT. In addition, en-face protocol produce images with enough reproducibility to perform an accurate follow-up [13]. In the case of diabetic patients, the reconstruction showed the decrease of photoreceptors at IS layer, being unremarkable at the B mode imaging [14]. Similar findings are described affecting outter retina using different methoths such as adaptive optics, microperimetry, HFA... with good topographical correlation with the images produced by enface OCT scans [15-17].
This evidence suggests that there is a good feasibility between the OCT en-face imaging and the demonstration of cellular lost. The findings of the en-face OCT scan in this case would correlate the lesions attributed to cellular lost with the evidence of a subtle sensibility decrease by the 10-2 perimetry pattern deviation map. The early detection of this lesions using enface instead of B-scan sectional protocol may be due to a more precise scanning avoiding the loss of defects in non-obtained cuts and, on the other hand, due to the layer reconstruction, making more evident the defects that affect a specific region or cellular population [12,18].
These findings have not been reported earlier in other cases of hydroxychloroquine maculopathy could be implemented as an early detector on the screening of these patients. Nevertheless, it is need to perform larger studies to determine the diagnostic feasibility of this test.