Effects of COVID-19 on Retinal and Choroidal Thickness by Optical Coherence Tomography

Précis:

The aim of our study was to evaluate the subclinical changes in the macula, retinal nerve fiber layer (RNFL), and choroidal thickness after coronavirus disease 2019 (COVID-19) infection using spectral domain optical coherence tomography.

Methods:

Our study was prospectively designed and involved 170 eyes of 85 patients. Patients with polymerase chain reaction (PCR)-positive COVID-19 infection were examined in the ophthalmology clinic before and after infection were included. All included patients had mild COVID-19 with no hospitalization and no need for intubation. Control ophthalmic examination was repeated at least 6 months after PCR positivity. Macular and choroidal thickness and RNFL parameters were compared before and at least 6 months after PCR-positive COVID-19 infection using optical coherence tomography.

Results:

When the mean macular thickness data were evaluated, a significant decrease was detected in the inner (mean difference, −3.37 µm; 95% CI: −6.09 to −0.65, P = 0.021) and outer (mean difference, −6.56 µm; 95% CI: −9.26 to −3.86, P < 0.001) temporal segments and the inner (mean difference, −3.39 µm; 95% CI: −5.46 to −1.32, P = 0.002) and outer (mean difference, −2.01 µm; 95% CI, −3.70 to −0.31, P = 0.018) superior segments in the post-COVID-19 measurements compared with pre–COVID-19 measurements. Similarly, on RNFL evaluation, some thinning was evident in the temporal superior (mean = 1.14 µm, P = 0.004) and temporal inferior (mean = 1.30 µm, P = 0.032) regions. All choroidal regions, including central, nasal 500 µm and 1500 µm and temporal 500 µm and 1500 µm, exhibited significant thinning (P < 0.001).

Conclusion:

At least 6 months after mild COVID-19 infection, significant thinning was seen in the temporal and superior quadrants of the macula, the temporal superior and temporal inferior regions of the RNFL, and all measured areas of choroidal regions.

Coronavirus disease 2019 (COVID-19), caused by a coronavirus designated severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has grown into a pandemic with enormous global impact since it was first diagnosed in early December 2019.¹ The disease shows a wide range of symptoms, from minimal symptoms to acute respiratory distress syndrome and multiple organ damage. Microvascular damage and coagulopathy caused by a severe inflammatory response, impaired endothelial cell function by binding to angiotensin-converting enzyme 2 (ACE2) receptors, and endothelial cell damage due to triggered inflammation are among the reported mechanisms in this condition, progressing with diffuse vascular damage.^2,3

Although its destructive effects on the respiratory and gastrointestinal systems have been shown, many effects on the ocular system have also been published and continue to be investigated.^4–7. Regarding ocular involvement, there are reports of 10% of patients affected by conjunctivitis and ocular surface diseases, as well as virus detection in the tear film, and retinal and optic nerve involvement.^6–10. Research is continuing into the effect of the virus, which enters human cells by binding to ACE receptors, on the tissues where these receptors are located. Some of the structures known to have ACE receptors present include retinal components, such as ganglion cells, Müller cells, and retinal photoreceptor cells.^10,11 There have been a number of controlled studies evaluating the macula and optic nerve head with the hypothesis that the retina and optic nerve may be affected by COVID-19. There are a few studies of choroidal thickness after COVID-19 but inconsistent findings have been reported, depending on the study endpoint. In a study conducted in the first month after COVID-19, it was reported that there was choroidal thickening and the macular layer did not change, whereas, in another study conducted 3 months after COVID-19, it was reported that the choroidal thickness was reduced in all measured areas and these changes were due to the developing inflammation.^12,13 These differences in results were attributed to differences in study designs, endpoints, number of patients included, differences in systemic disease between patients and controls, and differences in the severity of COVID-19 disease. Optic nerve head examination is another area of focus. Although many studies on optic nerve head evaluation have detected an increase in retinal nerve fiber layer (RNFL) thickness in the early stages of COVID-19, it has been shown that there is thinning in the RNFL 1 year after the disease.^14–16

Spectral-domain optical coherence tomography (SD-OCT) is a noninvasive method for imaging retinal tissues and the optic nerve head, with the ability to perform cross-sectional scanning with high resolution. It is the most important imaging tool used in ophthalmology in the follow-up and treatment of retinal diseases and neurological diseases involving the optic nerve head.¹⁷ Although there are many studies evaluating the macula and optic nerve head in the early period after COVID-19, especially the first 1–2 months, the late effects (≥3 mo) need to be clarified.^14–16 In an attempt to resolve this, we aimed to compare macular thickness, choroidal thickness, and RNFL parameters before and ~6 months after COVID-19 using SD-OCT

METHODS

This study was conducted with a prospective design in accordance with the Helsinki Declaration. It was approved by the human research Ethics Committee of Recep Tayyip Erdogan University, Rize, Turkey (July 8, 2021–2021/143). Written informed consent was obtained from all participants, consistent with the Turkish National Research Ethics Committee resolution for research conducted during the COVID-19 pandemic.

The study included 170 eyes of 85 patients who were diagnosed with polymerase chain reaction (PCR) positive COVID-19 infection and had an ocular examination in our ophthalmology clinic before infection with SARS-Cov-2. Subjects were recruited from hospital personnel, who applied to our ophthalmology clinic for refraction examination, routine eye examinations, or medical board report within the last 6 months, and whose examination data were recorded. These recorded data were used as pre–COVID-19 data. The patients were questioned about any possible symptoms of COVID-19 during this examination and none were PCR-positive at that time. Volunteers were instructed to inform the authors about their condition when they had COVID-19 infection. If the volunteers who were eligible to participate in the study had mild COVID-19 infection, control ophthalmological examinations were performed ~6 months after their PCR positivity. All included patients had mild COVID-19 with no hospitalization and no need for intubation.

Those with any systemic chronic disease, any ocular pathology, and those with refractive errors >2D were not included in the study. Participation in the study was on a voluntary basis and the majority of the participants were hospital personnel. Patients with any retinal and/or choroidal pathologies, glaucoma, uveitis, neuro-ophthalmological pathology, history of any ocular surgery, any chronic systemic disease, and pregnant and breastfeeding women were excluded. Patients who developed complications after COVID-19 and required either prolonged treatment or chronic medication were excluded from the study.

Ophthalmological Examinations and Optical Coherence Tomography Imaging

All patients underwent a thorough ophthalmic examination including: best-corrected visual acuity; slit-lamp anterior segment and fundus examination; macular and coroidal thickness; and RNFL measurements using a single device (Spectral domain OCT, software version 5.6.3.0). In macular SD-OCT, the central and mean layer thickness was analyzed by scanning an area of 6×6 mm with the Early Treatment Diabetic Retinopathy Study (ETDRS) macular scan, comprising 9 ETDRS macular areas (including the central 1 mm circle and the inner and outer rings 3 and 6 mm in diameter), which were defined automatically within circular areas at 3 different diameters (Fig. 1). Only images with an artifact-free quality index >15 were used, as recommended.¹⁸ All OCT images were received and segmentation errors were manually corrected using the OCT machine’s built-in software. Peripapillary RNFL thickness measurements were recorded in 6 quadrants (superior temporal, temporal, inferior temporal, superior nasal, nasal, and inferior nasal) using an optic nerve-centered circular scanning pattern. The thicknesses of these regions were automatically segmented using OCT software. Enhanced depth imaging OCT captured the image of the full thickness of the choroid. Choroidal thickness was measured as the vertical distance between the pigment epithelium and the inner surface of the sclera in these enhanced depths imaging-OCT images. Choroidal thickness was measured from the center of the fovea and at 500 μm and 1500 μm nasally and temporally to the central fovea. All measurements were performed by the same experienced blinded individual (F.S.), as reported.¹⁹

FIGURE 1.

Representative image of macular segmental assessment by methods including Early Treatment Diabetic Retinopathy Study (ETDRS) scan of a patients right eye at pre-coronavirus disease 2019 measurement. ETDRS, indicates early treatment diabetic retinopathy.

Statistical analyses

All statistical analyses were performed using IBM SPSS for Windows, version 20.0 (SPSS). The Kolmogorov-Smirnov test was used to assess the assumption of normality. Numeric variables are presented as mean ± SD, together with the 95% CI for the mean difference. Comparisons of dependent samples were carried out using the Wilcoxon Signed Ranks test. A P value <0.05 was considered statistically significant.

RESULTS

This study included 170 eyes of 85 PCR-positive patients treated for mild COVID-19 infection. The mean age of subjects was 30.08 ± 3.96 years. Pre–COVID-19 macular, choroidal thickness, and RNFL parameters were compared with their measurements at least 6 months after COVID-19 PCR positivity. The mean ± SD time interval between pre–COVID-19 ophthalmological examination and COVID-19 PCR positivity was 30.3 ± 9.77 (12–52) days. The mean ± SD time interval between COVID-19 positivity and post–COVID-19 ophthalmological examination was 203.78 ± 20.02 (156–240) days. All patients had 1.0 Snellen visual acuity, their anterior and posterior segment examinations were normal, and their refractive errors were below 2D at both examination time points.

The mean macular thickness measurements before and after COVID-19 disease are shown in Table 1. The temporal inner (mean difference, −3.37 µm; 95% CI: −6.09 to −0.65, P = 0.021) and temporal outer macular thickness (mean difference, −6.56 µm; 95% CI: −9.26 to −3.86, P < 0.001) and the superior inner (mean difference, −3.39 µm; 95% CI: −5.46 to −1.32, P = 0.002) and superior outer macular thickness (mean difference, −2.01 µm; 95% CI −3.70 to −0.31, P = 0.018) parameters were significantly decreased after COVID-19. Differences in macular thickness parameters in all regions are shown in Figure 2.

TABLE 1.

Comparison of Mean Macular Thickness Measurements Before and After COVID-19 Disease

	Before COVID-19	After COVID-19
Macular thickness (µm)	Mean (SD)	Mean (SD)	P	Mean differences	95% CI
Central macular thickness	267.91 (5.71)	267.96 (5.65)	0.882	0.05	−1.15 to 1.26
Temporal inner macula	336.23 (16.52)	332.85 (15.20)	0.021	−3.37	−6.09 to −0.65
Superior inner macula	340.51 (16.15)	337.11 (15.56)	0.002	−3.39	−5.46 to −1.32
Nasal inner macula	332.87 (14.22)	330.99 (13.47)	0.064	−1.87	−4.11 to 0.36
Inferior inner macula	337.28 (14.90)	335.74 (14.17)	0.057	−1.54	−3.47 to 0.38
Temporal outer macula	309.48 (21.27)	302.91 (16.85)	<0.001	−6.56	−9.26 to −3.86
Superior outer macula	303.66 (13.60)	301.65 (10.58)	0.018	−2.01	−3.70 to −0.31
Nasal outer macula	300.99 (18.64)	300.20 (14.35)	0.230	−0.78	−3.65 to 2.07
Inferior outer macula	296.38 (16.59)	294.36 (15.52)	0.092	−2.01	−4.27 to 0.24
Coroidal thickness	319.05 (10.15)	314.37 (13.06)	0.002	−4.67	−7.34 to −2.00

COVID-19 indicates coronavirus disease 2019.

Statistically significant values are in bold.

FIGURE 2.

Differences in macular thickness (A), choroidal thickness (B) and retinal nerve fiber layer thickness (C) parameters in all regions.

The mean macular volume parameters before and after COVID-19 disease are shown in Table 2. There was no statistically significant difference.

TABLE 2.

Comparison of the Mean Macular Volume Parameters Before and After COVID-19 Disesase

	Before COVID-19	After COVID-19
Macular volume (mm³)	Mean (SD)	Mean (SD)	P	Mean differences	95% CI
Fovea volume	0.217 (0.023)	0.219 (0.023)	0.599	0.001	−0.003 to 0.006
Temporal inner macula	0.534 (0.021)	0.536 (0.020)	0.532	0.001	−0.003 to 0.006
Superior inner macula	0.540 (0.022)	0.541 (0.022)	0.573	0.001	−0.003 to 0.005
Nasal inner macula	0.530 (0.026)	0.532 (0.027)	0.388	0.002	−0.003 to 0.008
Inferior inner macula	0.538 (0.020)	0.539 (0.020)	0.566	0.001	−0.003 to 0.005
Temporal outer macula	1.629 (0.094)	1.629 (0.093)	0.935	−0.0002	−0.022 to 0.022
Superior outer macula	1.612 (0.084)	1.614 (0.085)	0.637	0.002	−0.016 to 0.020
Nasal outer macula	1.618 (0.080)	1.618 (0.079)	0.743	−0.0002	−0.016 to 0.015
Inferior outer macula	1.561 (0.080)	1.564 (0.081)	0.852	0.003	−0.015 to 0.021
Total volume	8.767 (0.315)	8.773 (0.317)	0.663	0.005	−0.062 to 0.074

COVID-19 indicates coronavirus disease 2019.

The mean RNFL thickness measurements in all sections before and after COVID-19 are shown in Table 3. There were significant decreases in the temporal superior (mean difference, −1.14; 95% CI: −1.88 to −0.40, P = 0.004) and temporal inferior (mean difference, −1.30; 95% CI: −2.38 to −0.21, P = 0.032) RNFL segments. Other RNFL areas did not exhibit significant differences after COVID-19. Differences in RNFL thickness parameters in all regions are shown in Figure 2.

TABLE 3.

Comparison of the Mean RNFL Measurements Before and After COVID-19 Disesase

	Before COVID-19	After COVID-19
Optic nerve OCT (µm)	Mean (SD)	Mean (SD)	P	Mean differences	95% CI
RNFL-G	112.32 (4.66)	112.30 (4.62)	0.780	−0.02	−0.78 to 0.74
RNFL-N	93.74 (8.24)	94.36 (7.64)	0.424	0.61	−0.49 to 1.72
RNFL-NS	122.65 (9.84)	122.64 (8.91)	0.897	−0.00	−1.45 to 1.44
RNFL-NI	118.78 (8.63)	118.92 (7.85)	0.823	0.14	−1.18 to 1.48
RNFL-T	83.56 (6.47)	83.91 (6.30)	0.493	0.34	−0.54 to 1.22
RNFL-TS	131.87 (4.47)	130.72 (4.54)	0.004	−1.14	−1.88 to −0.40
RNFL-TI	134.45 (6.65)	133.15 (6.41)	0.032	−1.30	−2.38 to −0.21

COVID-19 indicates coronavirus disease 2019; OCT, optical coherence tomography; RNFL, retinal nerve fiber layer.

Choroidal thickness evaluations were made in 5 locations twice and compared between the pre and post–COVID-19 measurements (Table 4). There were significant decreases in all measured choroidal regions (P < 0.001), including the central (mean difference, −14.41; 95% CI −15.95 to −12.86), nasal 500 µm (mean difference, −9.60; 95% CI −10.02 to −9.18), nasal 1500 µm (mean difference, −8.39; 95% CI −8.73 to −8.05), temporal 500 µm (mean difference, −10.18; 95% CI −10.78 to −9.59), and temporal 1500 µm (mean difference, −8.37; 95% CI: −8.71 to −8.04). All the intraindividual reproducibility correlation coefficients for choroidal thickness parameters were obtained >0.99, which means there was a strong agreement between the two measurements (P <0.001 for all). Differences in choroidal thickness parameters in all measured regions are shown in Figure 2.

TABLE 4.

Comparison of Mean Choroidal Thickness Measurements Before and After COVID-19 Disease

	Before COVID-19	After COVID-19
Choroidal thickness (µm)	Mean (SD)	Mean (SD)	P	Mean differences	95% CI
Central	322.28 (8.07)	307.86 (9.93)	<0.001	−14.41	−15.95 to −12.86
Nasal 500	270.38 (5.03)	260.77 (4.66)	<0.001	−9.60	−10.02 to −9.18
Nasal 1500	257.06 (5.28)	248.67 (4.82)	<0.001	−8.39	−8.73 to −8.05
Temporal 500	283.88 (6.77)	273.69 (6.54)	<0.001	−10.18	−10.78 to 9.59
Temporal 1500	272.43 (8.16)	264.05 (7.42)	<0.001	−8.37	−8.71 to −8.04

COVID-19 indicates coronavirus disease 2019.

DISCUSSION

The coronavirus pandemic, which continues to affect the whole world, has been shown to have had detectable effects on many organ systems. Pathologic involvement of eye tissues, which have an important place among the affected structures, are presented as case reports and controlled studies are continuing. Although COVID-19 conjunctivitis is one of the most common ocular manifestations,²⁰ many pathologic retinal conditions (retinal vascular occlusions, many neuro-ophthalmic manifestations such as optic neuritis, papillophilebitis, and Adie tonic pupil) have been added to ocular findings.^21–25 In light of these reports, our aim was to investigate whether there were detectable changes in the structures of the macula, choroid, and RNFL in patients who had mild COVID-19 infection and did not develop any systemic or ocular complication, by comparing them with the preinfection data of the individuals and data from post–COVID-19 measurements. Our results show a significant decrease in the thickness of both some regions of the macula (inner and outer segments of temporal and superior quadrants) and in all of the 5 points along the horizontal scan passing the foveal center of choroidal regions and a decrease in some regions of RNFL thickness, specifically the temporal superior and temporal inferior regions in this cohort. Considering the contribution of our study to the literature, we think that the results will be more striking because it is a study with the same participant group without the effects of systemic disease, age, sex, and COVID-19 disease severity.

Studies have begun to report the effects of COVID-19 on the retina. It has been shown by optic coherence tomography angiography (OCTA) studies that there are changes in retinal microcirculation in the early postinfection period and vascular densities in both deep and superficial capillary plexus decrease, and this effect is evident if the disease severity is moderate or severe.^26,27 The changes in the thickness of retinal layers can be examined with Spectral Domain Optical Coherence Tomography ( SD-OCT), which is used to noninvasively detect retinal changes at both clinical and subclinical levels. In many neurodegenerative diseases that affect the retinal layers, the thickening that occurs in the tissues as a result of the edema that develops in the early period is evident in the form of thinning, with atrophy in the late period. SARS-Cov-2 is known to be a viral pathogen that affects the nervous system together with producing neurological symptoms (ageusia, anosmia, headache, and dizziness) and there are reported neuro-ophthalmological pathologies in COVID-19.^28,29 In experimental animal models of coronavirus retinopathy, thickening of the tissues due to inflammation in the early period and thinning due to retinal degeneration in the late stage have been shown.^9,30 In a study comparing macular thicknesses in the early period after COVID-19 infection with healthy individuals, thickening was found in the macular layers. This was interpreted as being due to inflammation, such as edema, in the acute phase and the authors predicted that there may be thinning due to atrophy in the chronic period.¹⁴ Mavi Yildiz et al³¹ reported that there was an increase in central foveal thickness in the macular evaluation performed in the period between 2–8 weeks after COVID-19, and this was associated with thickening of the outer nuclear layer, which was positively correlated with disease severity. In our study, the macular thickness was measured in 9 Early Treatment Diabetic Retinopathy Study (ETDRS) areas and significant thinning was detected in both temporal and superior quadrants ~6 months after COVID-19 infection. These findings suggest that, despite there being no ophthalmic pathology, subclinical macular thickness decreases, this decrease seems to begin in the temporal and superior quadrants in patients with COVID-19 infection, and this is clinically similar to the late period atrophy evident in neurodegenerative diseases. Macular volume did not exhibit significant change, which may be due to atrophy in mild COVID-19 patients that has not yet become evident and is not reflected in volume parameters. Determining which layer of the macula is affected by this change remains to be resolved. Future studies may clarify whether the atrophy arises in the outer layers of the retina due to a microcirculation disorder caused by thinning in the choroid or arises in the inner layers of the retina due to inflammation at the retinal microglial level.

An increase in RNFL values was reported 3 months after COVID-19 and a global decrease in RNFL values was reported 1 year after the disease in studies investigating the effects of COVID-19 on the head of an optic nerve.^15,16 In contrast, Oren et al³² found no significant differences in RNFL thickness between 14 and 30 days after COVID-19 disease. Microvascular endothelial damage and cytokine release have been shown to be key factors in the reduction of peripapillary vessel density in COVID-19.^26,27 A linear correlation was found between RNFL thickness and peripapillary vessel densities and these differences were found in the superior, inferior, and temporal vessel densities in mild-stage COVID-19.¹⁶ In our study, the optic nerve head was evaluated by the thicknesses in the RNFL quadrants, and there were significant decreases in the temporal superior (mean difference, 1.14 µm) and temporal inferior (mean difference, 1.30 µm) regions. We hypothesize that this is due to the patients having had mild COVID-19 and were in the earlier stages of the atrophy process. The subclinical thinning of the optic nerve head seems to begin in the temporal inferior and temporal superior regions in mild COVID-19 according to our study data.

The choroid is the layer of the eye with the highest vascularization and is also affected by many other systemic viral diseases, such as Ebstein-Bar Virus (EBV), hepatitis B virus (HBV),Human Immunodeficiency Virus (HIV).^33–35 Two mechanisms are proposed in the hematogenous spread of COVID-19 infection to ocular tissues. One is direct spread through ACE2-expressing vascular endothelial cells and the other is through infected leukocytes that cross the blood-retinal barrier. This can result in infection in the retina and choroid by either or both mechanisms.³⁶ In a study examining the period from 14 to 60 days after the onset of infection, a minimal and nonsignificant decrease in choroidal thickness was reported.³⁷ In addition, Yildiz et al³¹ found no change in choroidal thickness in the central, temporal, and nasal quadrants between 2 and 8 weeks after COVID-19. In a study examining 6 months, post–COVID-19 retinal findings with OCTA and SD-OCT, a comparison was made with healthy individuals, and a decrease in superficial capillary plexus and radial capillary plexus vessel densities was detected but only in the whole-image measurements, whereas a decrease in deep capillary plexus vessel density was reported in all macular regions.³⁸ Another study evaluating the early and late period findings after COVID-19 with OCTA showed that the value of the choriocapillaris flow area decreased significantly in the late period.³⁹ In our study, the thickness of all 5 measured choroidal regions exhibited a remarkable and significant decrease after COVID-19. We suggest that this may be associated with the significant reduction in the deep capillary plexus and choriocapillaris flow area shown in previous studies. The choroidal layer, which has a rich blood vessel network, and the outer retinal layers, which also provide blood supply, are likely to be affected by a viral disease, such as COVID-19, that results in microcirculation disruption. Taken together, these mechanisms may result in thinning in the choroid layer and retinal layer in the late period. To confirm this relationship, there is a need for additional studies with large sample sizes that examine the choroid in terms of structural and vascular density multimodally. Furthermore, the thinning, which may be due to this microcirculation disorder in the choroid layer, may also cause thinning in the outer retinal layers, which obtain some of their nutrition from the choroid, and this may be reflected in the macular thickness. Further studies examining the thicknesses of the macular layers separately are needed to elucidate these mechanisms.

The limitations of our study include a short follow-up period, lack of multimodal imaging, and including images from both eyes. The lack of comprehensive assessment of the choroid, retina, and their multimodal imaging with OCT angiography, measurement of all macular layers separately, using additional image interpretation software and using measurements from both eyes without adjusting for the correlated nature of the data may affect the findings. In addition, the measurements of the choroidal thickness were made manually, which may have further limitations. However, the strengths of this study are that it is the first to examine changes in the macula, choroid, and RNFL thickness before and after mild COVID-19 infection using the same group of individuals by SD-OCT.

CONCLUSION

Significant thinning was detected in macular thickness in the temporal and superior quadrants, temporal superior and temporal inferior RNFL and all measured choroidal regions. These changes suggest that even after mild COVID-19 disease, optic nerve head, macula, and choroid structures should be carefully evaluated and followed up in the ophthalmic evaluation of patients.