Evaluation of retinotoxicity of COVID-19 treatment: Hydroxychloroquine and lopinavir/ritonavir.

To the Editor,

Despite no established treatment for coronavirus disease 2019 (COVID‐19), hydroxychloroquine (HCQ) and lopinavir/ritonavir are being used. We read with interest the recent article of Chong et al who suggested that potentially significant cardiac side effects are common in patients treated with HCQ add on therapy to lopinavir/ritonavir.

The risk of cardiotoxic effects with antimalarial drugs has been highlighted by some authors, but retinotoxicity should not be overlooked. HCQ may produce characteristic changes in the outer retina (Figure 1A). Lopinavir does not appear to cause retinotoxicity, but there are cases of retinal pigment epitheliopathy associated with ritonavir. Because both HCQ and ritonavir can affect the outer retina, possible synergic toxicity has been hypothesized.

Figure 1

A, Characteristic macular optical coherence tomography showing hydroxychloroquine retinotoxicity with alteration of the outer retina. B, Macular optical coherence tomography from one of the patients of the study, where the outer retina presents no alterations

We performed a study to assess the appearance of characteristic changes associated with these drugs in patients with COVID‐19 using optical coherence tomography (OCT) and fundus autofluorescence (FAF).

This is a cross‐sectional study conducted at the Hospital Clinico San Carlos in Madrid (Spain). The study included patients between 18 and 70 years of age with a positive nasopharyngeal swab for severe acute respiratory syndrome coronavirus‐2 that were attended in the Emergency Department. Those patients, on quarantine, unable to attend the hospital, unable or unwilling to give consent, as well as those with concomitant neurological or ophthalmological disorders were excluded.

Sociodemographic data, medical history, clinical severity, laboratory workup, and COVID‐19 treatment were collected. All patients underwent slit‐lamp biomicroscopy, fundus examination, macular OCT, and FAF 1 and 3 months after COVID‐19 diagnosis, and, therefore, from the beginning of the therapy. Images were obtained using the Spectralis‐OCT (Heidelberg Engineering, Heidelberg, Germany) and afterward, two ophthalmologists blindly reviewed the images. Poor‐quality images were excluded.

The study included 79 patients (142 eyes) with laboratory‐confirmed COVID‐19 (Table 1). HCQ regimen was 400 mg/12 hours the first day and 200 mg/12 hours for 4 days, while lopinavir/ritonavir dosage was 400/100 mg every 12 hours for 7 days.

Table 1

Clinical characteristics

Sociodemographic data
Sex
Male, No. (%)	36 (46)
Female, No. (%)	43 (54)
Age, y (SD)	55.6 (9.1)
Race
Caucasic, No. (%)	53 (67)
Hispanic, No. (%)	26 (33)
Medical history
Arterial hypertension, No. (%)	23 (29)
Diabetes mellitus, No. (%)	8 (10)
Dyslipidemia, No. (%)	22 (28)
Renal disease, No. (%)	1 (1)
Liver disease, No. (%)	2 (3)
Tamoxifen use, No. (%)	0 (0)
Laboratory findings
Creatinine, mean (SD)	0.9 (0.3)
Clinical severity
Mild, No. (%)	26 (33)
Moderate, No. (%)	21 (27)
Severe, No. (%)	32 (41)
COVID‐19 treatment
None, No. (%)	16 (20)
Hydroxychloroquine, No. (%)	15 (19)
Lopinavir/ritonavir, No. (%)	2 (3)
Hydroxychloroquine + lopinavir/ritonavir, No. (%)	46 (58)

Abbreviations: COVID‐19, coronavirus disease 2019; SD, standard deviation.

None of the patients reported a visual loss during infection, nor after resolution. Funduscopic examination did not show visible macular alterations. In OCT, no thinning of the outer retinal layers or disruption of the ellipsoid zone was noted (Figure 1B). In FAF, no abnormal areas of hyper‐ or hypofluorescence were observed.

HCQ and lopinavir/ritonavir are being used in COVID‐19 and could produce retinotoxicity. In a large series of patients, most of them treated with both drugs, we found no macular alterations in OCT and FAF.

HCQ retinotoxicity is asymptomatic, dose‐dependent, and increases in patients suffering from renal disease, with maculopathy or who use tamoxifen. Characteristic alterations in HCQ toxicity include thinning of the photoreceptor layers, disruption of the interdigitation zone, and outer nuclear layer thinning. FAF can reveal early damage as an area of increased autofluorescence. None of our patients presented these findings.

Regarding ritonavir, macular alterations described in the literature include macular retinal pigment epitheliopathy (granular appearance), intraretinal crystalline deposits, and parafoveal telangiectasia. Patients with abnormal liver function are at a higher risk of retinotoxicity, the former being only present in two of our patients. Toxicity appears after chronic use and is usually associated with a decrease in visual acuity, which none of our patients referred. ,

Although most patients were on both drugs, the doses used in COVID‐19 are shorter in time than those used in other diseases, normally chronic. This supports the idea that there are is a low expected risk of retinal effects in the future, although long‐term follow‐up is needed for confirmation.

The study has several limitations. First, it is carried out on a relatively low sample considering the low incidence of retinotoxicity described. In addition, the changes were evaluated qualitatively and there were no previous examinations of the patients to compare.

Nevertheless, this is the first study to analyze the changes in OCT and FAF with COVID‐19 treatment. These are short‐term results, although we have not observed any qualitative alterations characteristic of these drugs. Despite it being a long latency period toxicity, it seems unlikely that with the doses used, future changes will occur. Currently, the World Health Organization does not recommend the use of HCQ in combination with antivirals. Undoubtedly, further research and longer follow‐up is needed to determine long‐term consequences of these drugs.

CONFLICT OF INTERESTS

The authors declare that there are no conflict of interests.