New findings on retinal microvascular changes in patients with primary COVID-19 infection: a longitudinal study

Abstract

Purpose: To investigate the longitudinal alterations of retinal microvasculature in patients with primary coronavirus disease 2019 (COVID-19) infection.

Methods: A cohort of participants, who had never been infected with COVID-19, was recruited between December 2022 and May 2023 at Peking Union Medical College Hospital in Beijing, China. Participants underwent comprehensive ophthalmologic examinations and fundus imaging, which included color fundus photography, autofluorescence photography, swept-source optical coherence tomography (SS-OCT) and SS-OCT angiography (SS-OCTA). If participants were infected with COVID-19 during the study, follow-ups with consistent imaging modality were conducted within one week and two months after recovery from the infection.

Results: 31 patients (61 eyes), with a mean age of 31.0 ± 7.2 years old, were eligible for this study. All participants contracted mild COVID-19 infection within one month of baseline data collection. The average period was 10.9 ± 2.0 days post-infection for the first follow-up and 61.0 ± 3.5 days for the second follow-up. No clinical retinal microvasculopathy features were observed during the follow-ups. However, SS-OCTA analysis showed a significant increase in macular vessel density (MVD) from 60.76 ± 2.88% at baseline to 61.59 ± 3.72%(p=0.015) at the first follow-up, which subsequently returned to the baseline level of 60.23 ± 3.33% (p=0.162) at the two-month follow-up. The foveal avascular zone (FAZ) remained stable during the follow-ups with areas of 0.339 ± 0.097mm², 0.342 ± 0.093mm², and 0.344 ± 0.098mm² at the baseline, first follow-up (p=0.09) and second follow-up (p=0.052), respectively. Central macular thickness, cube volume and ganglion cell-inner plexiform layer showed a transient decrease at the first follow-up(p<0.001, p=0.039, p=0.002, respectively), and increased to baseline level at the two-month follow-up(p=0.401, p=0.368, p=0.438, respectively).

Conclusion: Mild COVID-19 infection may temporarily and reversibly impact retinal microvasculature, characterized by a transient increase in retinal blood flow during the early recovery phase, which returns to the pre-infection level two months post-infection.

Keywords: coronavirus disease 2019; foveal avascular zone; optical coherence tomography angiography; retinal microvasculature; vessel density.

Figure 1

Analysis of macular vessel density on swept-source optical coherence tomography angiography. The macular region were further divided into (A) superficial capillary plexus(SCP), (B) intermediate capillary plexus(ICP), and (C) deep capillary plexus(DCP). The changes of mean vessel density(%) at 0–1mm(central fovea), 1–3mm(inner ring), and 3–6mm(outer ring) are shown and compared between pre-infection and post-infection follow-ups. (*p <0.05, **p <0.01, ***p <0.001).

Figure 2

Longitudinal swept-source optical coherence tomography angiography(SS-OCTA) vessel density(VD) changes in the right eye of a 28-year-old male. Top three rows display SS-OCTA VD heatmaps of the superficial (SCP), intermediate (ICP), and deep capillary plexuses (DCP) across three time points: COVID-19 pre-infection, first follow-up (10-day post-infection), and second follow-up (2-month post-infection), which are overlaid by Early Treatment Diabetic Retinopathy Study grid. Compared to pre-infection, an increase in VD was observed at first follow-up for SCP (from 33.49% to 35.75%) and ICP (from 22.23% to 25.28%), whereas VD approached pre-infection levels with SCP at 33.46% and ICP at 22.89% at the second follow-up. The DCP exhibited a marginal decline at the first follow-up (from 10.16% to 9.21%) and further decreased at the second follow-up (8.92%). Bottom row shows the foveal avascular zone (FAZ) at COVID-19 pre-infection, 10-day post-infection, and 2-month post-infection. The FAZ measurements remained relatively unchanged pre- and post-COVID-19 infection, with values of 0.195 mm² at baseline, 0.182 mm² at the first follow-up, and 0.197 mm² at the second follow-up.