Uveitis following COVID-19 Infection: A Case Series including Vogt-Koyanagi-Harada Disease and Birdshot Chorioretinopathy

Abstract

Introduction

Uveitis in the context of COVID-19 is rare, yet a relatively higher frequency of uncommon posterior variants has been described both following COVID-19 infection and after COVID-19 vaccination. In this series, we report 4 cases of new-onset uveitis following a PCR-confirmed COVID-19 infection: 2 cases of incomplete Vogt-Koyanagi-Harada (VKH) disease and 2 cases of birdshot chorioretinopathy (BSCR).

Case Presentations

Four patients presented with ocular symptoms and characteristic clinical findings consistent with either VKH disease or BSCR shortly after PCR-confirmed COVID-19 infection. VKH disease and BSCR occurred within 2 weeks of infection and 1 month after infection, respectively. Human leukocyte antigen (HLA) genotyping revealed HLA-A29 in both BSCR cases. One VKH case carried the HLA-DRB1*04:04 allele, indicating possible genetic susceptibility. No relevant medical history was reported, and systemic work-up was negative in all cases.

Conclusion

We report the first 2 cases of BSCR in the setting of PCR-confirmed COVID-19 infection, and add 2 cases of COVID-19-associated VKH disease to the existing literature. These observations further support the hypothesis that viral infections may trigger autoimmune uveitic entities. While causality remains to be fully established, COVID-19 should be considered a potential immunological trigger for BSCR, in line with its documented role in VKH disease. Further studies are needed to elucidate the potential role of viral infections in the pathogenesis of BSCR.

Introduction

Ocular manifestations of COVID-19 triggered by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) are diverse, ranging from mild surface disease to more severe posterior involvement including retinal vascular occlusions and rarely uveitis [1]. In this case series, we report the clinical findings of 4 patients with new-onset uveitis following a PCR-confirmed COVID-19 infection. These include 2 cases of incomplete Vogt-Koyanagi-Harada (VKH) disease according to the 2001 Revised Diagnostic Criteria for VKH, and 2 cases of birdshot chorioretinopathy (BSCR) according to the 2021 Standardization of Uveitis Nomenclature Working Group criteria, the latter of which has not previously been described in this context. This case series was reported according to the CARE guidelines with the completed CARE checklist included as online supplementary material (for all online suppl. material, see https://doi.org/10.1159/cop/acaag006).

Case Series

Case 1

A 36-year-old woman presented with acute blurred vision of the right eye, accompanied by mild headache, fever, cough, and tinnitus. Medical history revealed no relevant prior disease. She previously received 3 doses of a COVID-19 vaccine with the last administration (Comirnaty, BNT162b2 mRNA) more than 2 months prior to presentation.

Best-corrected visual acuity (BCVA) was Snellen 0.5 in the right eye and 1.0 in the left eye. Biomicroscopic examination showed no signs of anterior inflammation. Fundus examination combined with optical coherence tomography (OCT) revealed bilateral posterior vitreous cells, a large collection of papillomacular and peripapillary subretinal fluid in the right eye and a smaller collection of subretinal fluid in the left eye, as well as bilateral choroidal thickening (shown in Fig. 1a–d). Infectious work-up including testing for syphilis and interferon-gamma release assay (IGRA) was negative. Chest radiography showed no abnormalities. In the context of viral symptoms, PCR testing using a combined nose-throat swab was positive for COVID-19. A pan-PCR test for COVID-19, influenza A, influenza B, and respiratory syncytial virus, performed approximately 1 month earlier in accordance with institutional COVID-19 screening policies, had been negative.

Fig. 1.

VKH disease, case 1. Horizontal (a) and vertical (b) optical coherence tomography (OCT) at initial presentation, showing a large collection of papillomacular and peripapillary subretinal fluid (a) and a smaller collection of subretinal fluid (b) corresponding with hyperautofluorescent regions on right-eye (c) and left-eye (d) autofluorescence (AF), respectively (dotted arrows indicate scan location and direction), as well as choroidal thickening (double-headed arrow with measured choroidal thickness). It should be noted that the AF images were acquired 1 week after the OCT scan depicted, already showing mild regression of the subretinal fluid. e, f Late-phase fluorescein angiography (FA) at 1-week follow-up showing mild pinpoint leakage in the papillomacular region with mild disk staining. g, h Indocyanine green angiography (ICGA) at 1-week follow-up showing widespread choroidal infiltration with diffuse scattered small hyporeflective dots throughout the whole posterior pole.

Oral corticosteroid therapy was initiated at a dose of 1 mg/kg/day, with the addition of cyclosporine after 1 month. At 1-week follow-up, BCVA of the right eye improved to Snellen 0.9. Fundus fluorescein angiography (FA) showed mild pinpoint leakage in the late phase in the papillomacular region and mild disk staining in both eyes (shown in Fig. 1e–f), but indocyanine green angiography (ICGA) showed widespread choroidal infiltration with diffuse scattered small hyporeflective dots throughout the whole posterior pole (shown in Fig. 1g–h). A diagnosis of COVID-19-associated incomplete VKH disease was made. OCT monitoring at 1-month follow-up revealed significant reduction in subretinal fluid in the right eye and complete resolution in the left eye, with stable bilateral resolution at the 8-week follow-up. Adalimumab was administered every 2 weeks between 3 months and 2 years after diagnosis, primarily because of intolerance to cyclosporine and concerns regarding the less favorable long-term safety profile of cyclosporine rather than due to insufficient clinical response. There was no recurrence of the subretinal fluid on OCT up to 1 year after termination of adalimumab therapy, which marks her most recent follow-up visit.

Case 2

A 10-year-old boy was referred by his pediatrician with a 2-week history of painful red eyes, unresponsive to antiallergic treatment, now accompanied by subacute vision loss. Systemic complaints included tinnitus and mild abdominal pain. Respiratory symptoms were noted less than 2 weeks earlier, following a suspected high-risk COVID-19 exposure. The patient had not received any COVID-19 vaccination. Further medical history revealed no prior disease, and there was no family history of autoimmune disease.

BCVA was Snellen 0.1 in the right eye and limited to counting fingers in the left eye. Biomicroscopic examination revealed bilateral granulomatous anterior uveitis including diffuse endothelial precipitates and dense posterior synechiae. Ocular ultrasound showed bilateral dense vitritis, serous retinal detachment, optic disk edema, and diffuse choroidal thickening without evidence of posterior scleritis. Initial serological testing revealed microcytic anemia and thrombocytosis. Extensive testing for underlying autoimmune causes was within normal limits. Screening for syphilis, tuberculosis (IGRA) and brucellosis (Wright test) was negative. Chest and abdominal radiography as well as abdominal ultrasound were normal. Urine analysis showed no detectable levels of alpha-1-microglobulin or beta-2-microglobulin, arguing against tubulointerstitial nephritis. COVID-19 PCR testing at the time of the uveitis diagnosis was positive, suggesting COVID-19-associated incomplete VKH disease.

Treatment was initiated with prednisolone and cyclopentolate eye drops, alongside oral corticosteroids at a dose of 1 mg/kg/day. OCT of the right eye at the 2-day follow-up and the left eye at the 7-day follow-up showed diffuse choroidal thickening associated with marked secondary subretinal fluid in the papillomacular region (shown in Fig. 2a). Over time, a significant decrease in subretinal fluid with subjective improvement of symptoms was seen. A progressive decrease in vitreous haze at the 4-week follow-up allowed for fundus examination, fundus autofluorescence, FA, and ICGA, revealing bilateral disk leakage, macular and peripheral retinal pigment epithelial alterations due to chronic inflammation, and peripheral retinal capillaritis (shown in Fig. 2b). At the 2-month follow-up, methotrexate was initiated due to recurrent inflammation occurring during tapering of oral corticosteroids. Adalimumab was added at the 3-month follow-up due to insufficient control of inflammation.

Fig. 2.

VKH disease, case 2. a OCT of the right eye at the 2-day follow-up and the left eye at the 7-day follow-up, showing diffuse choroidal thickening associated with marked secondary subretinal fluid in the papillomacular region. b FA 4 weeks after initial presentation, revealing bilateral disk leakage, macular and peripheral retinal pigment epithelial alterations due to chronic inflammation, and peripheral retinal capillaritis.

Case 3

A 68-year-old woman was referred for uveitis work-up of bilateral optic disk swelling and retinal vasculitis. She reported mild visual decline after her third COVID-19 vaccination (Comirnaty, BNT162b2 mRNA), but symptoms markedly worsened 8 months later, with the onset of bilateral photopsia 1 month after a PCR-confirmed COVID-19 infection. Her medical history revealed no prior disease.

At presentation at our tertiary center, BCVA was Snellen 0.7 in the right eye and 0.8 in the left eye. Biomicroscopic examination revealed subtle vitreous cells in the left eye. Fundus examination combined with FA demonstrated bilateral multifocal cream spots, disk leakage, and perivascular hyperfluorescence consistent with retinal venous vasculitis, most pronounced in the right eye. Early-phase ICGA showed multiple hypofluorescent spots distributed along the choroidal vessels, persisting into the late phase (shown in Fig. 3a–c). OCT revealed discrete macular edema, most marked in the right eye. On the macular thickness map, diffuse thickening was observed along the course of the retinal venous arcade (shown in Fig. 3d–e). Infectious work-up was negative for syphilis and tuberculosis (IGRA). Extensive autoimmune serology was unremarkable. Chest radiography was normal. HLA genotyping revealed positivity for the HLA-A29 allele. A diagnosis of BSCR was made.

Fig. 3.

BSCR, case 3. a Color fundus image showing multifocal cream spots (white arrowheads indicate the most prominent spots). b Early-phase ICGA at referral, prior to treatment, showing multiple hypofluorescent spots distributed along the choroidal vessels, persisting into the late phase. c Mid-phase fluorescein angiography (FA) 1 month into treatment, showing bilateral disk leakage and residual mild perivascular hyperfluorescence consistent with retinal venous vasculitis. d OCT at referral showing discrete macular edema, most marked in the right eye. e On the macular thickness map, diffuse thickening is observed along the course of the retinal venous arcade; case 4. f Mid-phase FA showing bilateral disk leakage and perivascular hyperfluorescence consistent with retinal venous vasculitis, most pronounced in the left eye with additional peripheral retinal capillaritis.

Oral corticosteroids at a dose of 1 mg/kg/day and cyclosporine were initiated, the latter of which was switched to adalimumab due to intolerance. A gradual reduction in inflammation was observed over the following 6 months. Discontinuation of adalimumab therapy due to abdominal discomfort was followed by a recurrence of optic disk leakage and vasculitis. Adalimumab was reintroduced, resulting in control of inflammation, and was ongoing at the last follow-up after 6 months.

Case 4

A 54-year-old woman was referred for a second opinion regarding bilateral chronic posterior uveitis, insufficiently responsive to oral corticosteroid therapy. She initially reported a subacute onset of floaters and blurred vision in the right eye 1 month following a positive COVID-19 PCR test, with subsequent involvement of the left eye. She had previously received three doses of a COVID-19 vaccine with the last administration (Comirnaty, BNT162b2 mRNA) more than 1 year prior to symptom onset. Medical history revealed no prior relevant disease, and there were no systemic complaints. Biomicroscopic examination at initial presentation revealed bilateral vitreous cells, most pronounced in the right eye. An extensive systemic work-up including infectious serology (syphilis, toxoplasmosis, Borrelia, CMV, VZV), IGRA, chest radiography, brain MRI, and vitreous sampling during vitrectomy of the right eye did not demonstrate systemic disease or malignancy. HLA-B27 testing was negative. Oral corticosteroids were initiated at a dose of 1 mg/kg/day but were eventually discontinued by the patient.

At presentation at our tertiary center, 1 year after the initial ophthalmological work-up, bilateral subtle cream choroidal lesions were detected on fundus examination. Mid-phase FA demonstrated bilateral disk leakage and perivascular hyperfluorescence consistent with retinal vasculitis, most marked in the left eye with additional peripheral retinal capillaritis (shown in Fig. 3f). Since the clinical picture was highly suggestive of BSCR, HLA genotyping was carried out and revealed positivity for the HLA-A29 allele. Oral corticosteroids were reintroduced, with methotrexate added due to recurrent inflammation occurring during tapering. Eventually, adalimumab was added after corticosteroid tapering, given that methotrexate alone proved insufficient to control inflammation.

Discussion

Uveitis in the context of COVID-19 is rare, yet among such cases, uncommon posterior variants such as multiple evanescent white dot syndrome, tubulointerstitial nephritis with uveitis, and VKH disease have been reported in the literature with a higher relative frequency both following COVID-19 infection and after COVID-19 vaccination. This pattern suggests a shared immunopathogenic mechanism, potentially mediated by molecular mimicry between viral proteins and host antigens [1].

In this case series, we report the first two cases of BSCR in the context of a PCR-confirmed COVID-19 infection in the literature to date. BSCR is characterized by a bilateral chronic inflammation of the choroid and retina with a distinct clinical phenotype. The etiology of BSCR remains unclear but is thought to reflect an immune-related process with a central role of HLA-A29. Since not all individuals carrying the HLA-A29 allele develop BSCR, additional triggers are likely required for disease expression, with viral peptides postulated as potential initiators. Kuiper et al. [2] suggested that hypothesized structural similarities between viral proteins and retinal or choroidal antigens may provoke autoreactive CD8+ T cells, initiating the immune response that manifests as BSCR. Both BSCR cases in our series developed ocular symptoms approximately 1 month following COVID-19 infection, with a considerably longer interval of 8 to 16 months since their third COVID-19 vaccination (Comirnaty, BNT162b2 mRNA).

In line with prior reports suggesting COVID-19 infection or vaccination may precipitate VKH disease, we provide 2 additional cases of incomplete VKH following COVID-19 infection. VKH disease is a T cell-mediated immune-related multisystem disorder that targets melanocyte-rich tissues such as the eyes, meninges, inner ear, and skin. Ocular involvement is characterized by bilateral granulomatous panuveitis, primarily affecting choroidal tissue. In VKH disease, a genetic predisposition has been linked to HLA-DR4-restricted peptides derived from melanocyte proteins [3]. In our series, HLA-DQB1/DRB1 typing was performed using next-generation sequencing. This analysis demonstrated that our second patient, a 10-year-old boy, carried the HLA-DRB1*04:04 allele, further supporting the role of HLA-DR4 alleles in disease susceptibility. Both VKH cases showed a clear temporal link to infection, with symptoms emerging 2 to 3 weeks post-COVID-19 infection, consistent with six previously reported cases [4–10]. Only one of them received three doses of a COVID-19 vaccine approximately 2 months earlier.

We acknowledge that COVID-19 vaccination has also been associated with new-onset uveitis and VKH disease in rare cases. However, in our patients, the temporal relationship between the onset of ocular inflammation and the PCR-confirmed COVID-19 infection was considerably stronger, with uveitis developing shortly after acute infection, whereas vaccination had occurred several months earlier (between 2 and 16 months before presentation). This temporal pattern makes acute infection the more likely triggering event, although a contributory role of prior vaccination cannot be completely excluded. None of the patients experienced systemic complications of COVID-19 beyond the reported symptoms. Importantly, we do not consider either COVID-19 infection or vaccination to be a direct cause of uveitis, but rather potential immunological triggers capable of inducing immune dysregulation in genetically predisposed individuals.

With regard to disease course and treatment requirements, all 4 patients in our series eventually required escalation to adalimumab. Both VKH disease and BSCR are chronic immune-mediated disorders characterized by autoreactive T-cell responses and persistent immune dysregulation, irrespective of the inciting trigger. In clinical practice, these entities frequently require prolonged systemic immunosuppressive therapy and, in a subset of patients, escalation to biologic agents due to recurrent inflammation, incomplete disease control, or intolerance to conventional immunosuppressive agents such as methotrexate or cyclosporine. Current therapeutic strategies aim to suppress T-cell activation and the downstream inflammatory cascade rather than to cure the underlying disease process. Accordingly, based on our limited case series, no firm conclusions can be drawn regarding a more refractory clinical course in COVID-19-associated uveitis compared with non-COVID-associated disease.

An overview of the presented VKH and BSCR cases can be found in Tables 1 and 2, respectively. While causality remains to be fully established, COVID-19 should be considered a potential immunological trigger for BSCR, in line with its documented role in VKH disease. Further studies are needed to elucidate the potential role of viral infections in the pathogenesis of BSCR.

Table 1.

Overview of new-onset VKH cases following recent COVID-19 PCR+

Patient	Age, years, sex	Most recent COVID-19 vaccination	COVID-19 PCR+	Onset (ocular) symptoms	Medical history	HLA-DQB1/DRB1
Case 1	36 F	02/2022: BNT162b2 mRNA (Comirnaty)	04/2022	Acute blurred vision	None	HLA-DRB1*07:01, HLA-DQB1*02:02
				Accompanied by: fever/cough/tinnitus (PCR confirmation at time of diagnosis)
Case 2	10 M	No vaccination	01/2022	±1–2 weeks following respiratory symptoms: painful red eyes, followed by bilateral subacute vision loss	None	HLA-DRB1*04:04, HLA-DRB1*11:04, HLA-DQB1*03:01, HLA-DQB1*03:02
				Accompanied by: tinnitus (PCR confirmation at time of diagnosis)

HLA, human leukocyte antigen.

Table 2.

Overview of new-onset BSCR cases following recent COVID-19 PCR+

Patient	Age, years, sex	Most recent COVID-19 vaccination	COVID-19 PCR+	Onset (ocular) symptoms	Medical history	HLA-A29
Case 3	68 F	10/2021: BNT162b2 mRNA (Comirnaty)	05/2022	Initial mild visual decline following third COVID-19 vaccination	None	Positive
				±1 month following COVID-19 PCR+: subacute worsening of vision, onset of bilateral photopsia
Case 4	54 F	12/2021: BNT162b2 mRNA (Comirnaty)	03/2023	±1 month following COVID-19 PCR+: subacute onset of floaters and blurred vision right eye, subsequent involvement left eye	None	Positive

HLA, human leukocyte antigen.

Statement of Ethics

This study protocol was reviewed and approved by EC Research UZ Leuven, Approval No. S71452. Written informed consent was obtained from the patients in cases 1, 3, and 4 and from the legal guardians in case 2 for publication of the details of their medical case and any accompanying images.

Conflict of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

This study was not supported by any sponsor or funder.

Author Contributions

All authors (M.V., C.D., M.W., J.V.C., P.P.S.) attest that they meet the current ICMJE Criteria for Authorship. M.V.: conceptualization, writing – original draft, review and editing, and visualization. C.D.: conceptualization, writing – review and editing, and supervision. M.W.: admitting physician and writing – review and editing. J.V.C.: admitting physician and supervision. P.P.S.: admitting physician, conceptualization, writing – review and editing, supervision, and validation. All authors contributed to the revision and approved the submitted version.

Funding Statement

This study was not supported by any sponsor or funder.

Data Availability Statement

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.