Introduction

The COVID-19 pandemic caused by the SARS-CoV-2 has had devastating impacts worldwide. COVID-19 vaccines have offered much needed hope in tackling the pandemic, remaining the mainstay of disease prevention efforts. In the local setting of the Hong Kong Special Administrative Region (HKSAR), the government approved the emergency use of COVID-19 vaccines in early 2021, including the Comirnaty (BNT162b2) messenger RNA (mRNA) vaccine supplied by Fosun Pharma and BioNTech (equivalent to the Pfizer-BioNTech vaccine distributed elsewhere) [1], alongside the CoronaVac inactivated virus vaccine developed by Sinovac Biotech [2].

Vaccination has previously been observed to associate with immune-related ocular phenomena. Before COVID-19, uveitis was reported to associate with various antiviral vaccines, including the hepatitis B vaccine [3, 4], the varicella vaccine [5], the herpes zoster vaccine [6], the bacille Calmette–Guérin (BCG) vaccine [7], and the human papillomavirus (HPV) vaccine [8]. Given public concerns regarding the safety of COVID-19 vaccines, vaccination data are much needed for the monitoring and surveillance of potential adverse ocular effects.

The most common ocular manifestations of COVID-19 include conjunctivitis, keratitis, keratoconjunctivitis, episcleritis, uveitis, posterior ischemic optic neuropathy, and other conditions involving the retinal vasculature [9,10,11]. In particular, intraocular inflammation and retinal vascular involvement have been described among patients infected with COVID-19 [11, 12]. It remains uncertain whether vaccination against COVID-19 can produce potentially vision-threatening inflammatory responses similar to COVID-19 infection. This study reports a case series describing the clinical spectrum of ocular inflammatory events related to COVID-19 vaccination observed locally in Hong Kong.

Methods

Case series data were obtained from the medical records for patients at the Eye Clinic for Prince of Wales Hospital who were consecutively diagnosed with uveitis shortly after receiving COVID-19 vaccines in Hong Kong (either the Comirnaty or CoronaVac vaccine) between March and October 2021. All patients were diagnosed with uveitis resulting from COVID-19 vaccination after ruling out alternative causes for uveitis. The causality between vaccination events and the onset of uveitis was evaluated using the World Health Organization Adverse Reaction Terminology (WHO-ART) [13]. Uveitis presentations were described using the classification developed by the Standardization of Uveitis Nomenclature (SUN) Working Group. Additional demographic and clinical data were also obtained from medical records, including patient age, gender, past medical and ocular history, dates of receiving the first and second doses of COVID-19 vaccines, vaccine type, presenting symptoms, visual acuity, other ocular findings, treatment, and clinical outcomes.

Serology and viral polymerase chain reaction (PCR) and real-time polymerase chain reaction (RT-PCR (SARS-CoV-2)) testing were performed for patients on an individual basis. Administered serology tests included those for erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), antineutrophil cytoplasmic antibodies (ANCA), antinuclear antibodies (ANA), C3 protein, C4 protein, rheumatoid factors, syphilis, and human leukocyte antigen B27 (HLA-B27). PCR testing methods included anterior chamber tapping and vitreous fluid tapping.

This study was conducted in accordance with the tenets of the Declaration of Helsinki, with the approval from the joint Chinese University of Hong Kong—New Territories East Cluster Clinical Research Ethics Committee (CREC NTEC 2020.349).

Results

Our case series includes 16 eyes from ten female patients. Their mean age was 49.4 ± 17.4 years (range: 15–79 years). Eight patients (80%) had received the Comirnaty vaccine, while the other two (20%) had received the CoronaVac vaccine. Nine (90%) developed postvaccination uveitis after their second vaccine dose, while one (10%) developed it after her first dose. Among the occurrences of postvaccination uveitis observed in the case series, anterior uveitis was the most prevalent type (five patients; 50%), followed by intermediate uveitis (three patients; 30%) and posterior uveitis (two patients; 20%), respectively. The median time from vaccination to uveitis onset was 15.2 days (range: 0–6 weeks). Regarding treatment outcomes, topical steroids resulted in complete resolution of uveitis for 11 out of 16 eyes (68.75%). Table 1 summarizes the clinical information for the case series.

Table 1 Case summaries
Full size table

Anterior uveitis

Among the five cases (1, 2, 5, 6, and 8) presenting with anterior uveitis, the first four had a previous history of anterior uveitis. The disease had been quiescent for a period between 4 months and 10 years before these four patients received COVID-19 vaccines. Severity of anterior uveitis and associated features including the presence of keratic precipitates, posterior synechiae, and Busacca and Koeppe’s nodules are reported in Table 1. The presence of keratic precipitates and posterior synechiae were commonly observed in our cases of AU; however, associated granulomatous inflammatory features such as Busacca or Koeppe’s nodules were not observed. Moreover, two cases (1 and 5) had ankylosing spondylitis (AS) as a comorbidity, testing positive for HLA-B27. In particular, Case 5, who developed a right anterior uveitis flare-up after their first dose of the Comirnaty vaccine, had a strong family history of AS. Only she had been receiving long-term immunomodulating therapy (methotrexate) to control their AS; the other four patients were not taking any immunosuppressive medications for systemic diseases at diagnosis. Case 5 was shown in Fig. 1.

All patients in this group presented with redness in the eye and blurred vision within 1–14 days of receiving a dose of the Comirnaty vaccine. The median duration between vaccination administration and uveitis flare-up was 15.2 ± 11.4 days. No case exhibited sustained elevation in intraocular pressure. All were treated using a course each of topical steroids and cycloplegic eye drops, which sufficed to resolve all signs and symptoms of anterior uveitis.

Intermediate uveitis

Among the three cases (3, 4, and 10) presenting intermediate uveitis flare-ups, one (Case 3) had a background history of intermediate uveitis; the disease had been quiescent for 9 months before she received the first dose of the Comirnaty vaccine. Case 4 saw a relapse of unilateral uveitis after the second dose of Comirnaty, while Case 10 saw new-onset intermediate uveitis develop bilaterally 3 weeks after the second dose of Comirnaty. For the latter two patients, their ocular inflammation and visual symptoms were completely resolved after being administered a course each of topical steroids and cycloplegic eye drops.

Retinal vasculitis

The two cases of postvaccination posterior uveitis include one presenting sinister ocular inflammation (Case 7) and another presenting severe retinal vasculitis (Case 9). For Case 7, the patient presented with fever, headaches, and blurred vision 2 weeks after receiving the second dose of the CoronaVac vaccine. Visual examinations discovered bilateral frosted branch angiitis with widespread vascular sheathing and multiple dots, the latter indicating retinitis (Fig. 2). Optical coherence tomography (OCT) revealed multiple pockets of subretinal fluid with evidence of retinal pigment epitheliitis and outer retinitis. Moreover, fluorescein angiography uncovered evidence of perivascular leakage with limited capillary dropout. However, indocyanine green angiography (ICG) observed no hypercyanescence.

Fig. 1
figure 1

Slit-lamp photographs showing diffuse anterior scleritis and anterior uveitis. This patient presented shortly after receiving the second dose of the Comirnaty mRNA vaccine. Interval improvements were observed at 3 and 9 days after the initiation of treatments using topical corticosteroids and cycloplegic eye drops

Full size image

Prior to PCR and serological testing, the patient of Case 7 received an intravitreal injection of empirical foscarnet (2.4 mg). Following anterior chamber paracentesis, she tested negative on PCR tests for the varicella zoster, cytomegalovirus, and herpes simplex viruses. She also received negative results for her serology workup, including tests for autoimmune markers, complete blood count, toxoplasmosis, HIV, syphilis, and tuberculosis. After all infectious causes were excluded, the patient was treated using an infusion of intravenous pulse methylprednisolone, followed by oral steroids. Ocular inflammation subsided quickly following treatment.

Fig. 2
figure 2

From top, a fundus photographs of a female (Case 7) who developed a mixed picture of combined frosted branch angiitis, choroiditis, and outer retinal inflammation at 2 weeks after receiving the second dose of the CoronaVac inactivated COVID-19 vaccination. b Fundus autofluorescence images with extensive multifocal gray–white chorioretinal hyperautofluorescence spots. c Red-free images at the acute phase. d Optical coherence tomography (OCT) scans of the images in (c). These two sets of images indicate a good response to intravenous pulse steroids after excluding infectious causes, with visual acuity recovering to 1.0 three months after disease the onset (presenting bilateral visual acuity was 0.1). e Resolution of subretinal fluids after systemic steroid treatment, with areas of remaining inner segment/outer segment disruption observed bilaterally

Full size image
Fig. 3
figure 3

An elderly female (Case 9) presented with unilateral multiple evanescent white dot syndrome (MEWDS) in her left eye a after receiving the second dose of the Comirnaty mRNA vaccine. Her case represents an occurrence of MEWDS in a patient older than the typical age range (15–50 years). Fundus autofluorescence images b, c show pattern hyperautofluorescence typical of MEWDS in her left eye. The patient responded well to combined treatments with oral steroids and intravitreal Ozurdex injections

Full size image

Discussion

Despite reports on the ocular effects of COVID-19 vaccination being limited until now, certain studies have observed diverse manifestations of ocular phenomena following COVID-19 vaccination. These studies cover a wide range of presentations, including anterior uveitis [14, 15], posterior uveitis [16], multifocal choroiditis [17], panuveitis, Vogt–Koyanagi–Harada syndrome (VKH) [18], and white dot syndromes[19] (Table 2). Nevertheless, most reports have insufficient evidence other than the temporal association to prove the causality between the described ocular inflammation and COVID-19 vaccination. Presumed cases of uveitis related to COVID-19 vaccination reported in the literature include new presentations of onset bilateral juvenile idiopathic arthritis (JIA)-associated anterior uveitis, unilateral anterior uveitis, bilateral choroiditis, and bilateral panuveitis, in addition to the recurrence of VKH syndrome [14,15,16,17].

Table 2 Summary of the literature studies
Full size table

Clinical presentations of vaccine-associated uveitis

Among the literature describing ocular inflammatory events following COVID-19 vaccination, most reports noted the occurrence of anterior segment inflammation. Recently, a large-scale study including 1094 cases of vaccine-associated uveitis from different countries was reported, summarizing the results from adverse event reporting system; anterior uveitis remained as the most common observed vaccine-associated uveitis (VAU) [23]. Similar to our report, a female predominance was observed in the reported VAU cases. Another large case series for anterior segment inflammation was reported by Testi et al. in a multicenter study of 70 patients. In this series, the most common postvaccination inflammatory events were anterior uveitis (58.6%), followed by posterior uveitis (12.9%) and scleritis (10%). Most patients were managed with topical corticosteroids, with vision remaining the same for 92.9% of cases [24].

Posterior segment involvements in vaccine-associated uveitis

On the other hand, a few reports revealed potential retinal layer or posterior segment involvement following COVID-19 vaccination, suggesting neurological tissue inflammation. In Israel, Rabinovitch et al. reported 21 cases of uveitis following COVID-19 vaccination, of which two cases also developed multiple evanescent white dot syndrome (MEWDS) affecting the retina [22]. Pichi et al. described seven patients presenting with uveitis, along with two instances of rare clinical entities, namely, acute macular neuroretinopathy (AMN) and paracentral middle maculopathy (PAMM) [25]. Similar to the multicenter report by Testi et al. [24], most of the ocular inflammatory events observed by Pichi et al. occurred after patients received the Pfizer-BioNTech vaccine.

Several mechanisms have been suggested to explain ocular inflammatory responses following COVID-19 vaccination. Among them, the most generally accepted theories include the activation of antigens generated secondary to molecular mimicry resulting from similarities between vaccine and uveal peptides, type III hypersensitivity reactions, and other innate immune reactions induced by vaccination [26, 27]. In sum, our study reported the temporality of COVID-19 vaccination and the onset of ocular inflammatory events, observing relatively short intervals between these two events. Causality is possible given current understanding of uveitis events following other vaccinations, as well as generally accepted proposals for biological responses following vaccination.

Newly observed uveitis entities in our series

In our case series, we also report a new entity of frosted branch angiitis, choroiditis, and outer retinal inflammation, which mimics a mixed picture of widespread retinal vasculitis and choroiditis, following the administration the CoronaVac inactivated whole-virus vaccine. Formulating from harvested, inactivated, and purified whole particles of the SARS-CoV-2, the CoronaVac vaccine is generally expected to induce a broader immune response compared to mRNA vaccines that only target spike proteins [28, 29]. Neutralizing antibodies and/or activated T-helper cells may cross-react with proteins and antigens in multiple ocular tissue layers, including parts of the retinal vessels, outer retinal layers, retinal pigment epithelial cells, and the choroid. So far, retinal vasculitis has only been reported following COVID-19 infection [30]. This case in our series illustrated the potential for COVID-19 vaccines to induce similar vascular inflammatory events, albeit to a milder extent than infection.

In our series, most patients presenting with anterior uveitis (six out of seven patients) experienced uveitis attacks after receiving the Comirnaty vaccine, whereas only one patient developed ocular inflammation involving multiple layers of chorioretinal tissue after receiving the CoronaVac vaccine. The differences in the technologies used by both vaccines may imply distinct immunogenicity mechanisms specific to each vaccine platform. Considering the low occurrence rates of vaccine-related uveitis, continued data collection and future research are warranted to evaluate the safety profiles of each vaccine, including ocular side effects. Nevertheless, given that nearly all uveitis cases following vaccination with the Comirnaty vaccine could be resolved rapidly and completely following a short course of topical steroids, the benefits of vaccination still far outweigh the potential risk of ocular adverse events.

Clinical course of the available VAU case series

Within the current literature, anterior uveitis is the most prevalent presentation of uveitis flare-ups after various vaccination regimens, followed by intermediate uveitis. The literature also observes that most reported cases resolve rapidly following one course of topical steroids. In the most comprehensive review of vaccine-induced uveitis yet, Benage and Fraufelder identify a median time range from vaccination to uveitis onset of 16 days (range: 1 day–6 years) [4]. For our case series, the median date of anterior and intermediate uveitis presentations was 15 days (range: 0–6 weeks) following vaccination with the Comirnaty vaccine. Our series did not include specific antibody tests and titrations for the SARS-CoV-2. Nevertheless, factors including the temporal sequence of vaccination and disease events, the absence of identifiable alternative causes from accessory examinations, the transience of severe ocular inflammation symptoms, and good treatment responses to steroids all strongly suggest that uveitis can be a vaccine-induced immune response.

Potential mechanisms for vaccine-associated uveitis

Of the two available vaccines in Hong Kong, the Comirnaty mRNA vaccine uses lipid nanoparticle encapsulated mRNA to encode a full-length spike protein of the SARS-CoV-2, which is also locked down by two proline mutations to avoid integration into the host cell genome [31]. This vaccine induces strong activation of the cellular and humoral immune responses, exhibiting dose-dependent side effects [32]. Because of the similarities between the SARS-CoV-2 spike protein and proteins in uveal tissues, molecular mimicry may cause immune cross-reactivity, triggering autoimmune diseases such as anterior uveitis [31, 33]. In accordance with available reports, most cases in our series developed uveitis following the second dose of COVID-19 vaccines, potentially because of greater, and possibly dose-dependent, reactogenicity.

Several mechanisms have been suggested to explain ocular inflammatory responses following COVID-19 vaccination. Among them, the most generally accepted theories include the activation of antigens generated secondary to molecular mimicry resulting from similarities between vaccine and uveal peptides, type III hypersensitivity reactions, and other innate immune reactions induced by vaccination [26, 27]. In sum, our study reported the temporality of COVID-19 vaccination and the onset of ocular inflammatory events, observing relatively short intervals between these two events. Causality is possible given current understanding of uveitis events following other vaccinations, as well as generally accepted proposals for biological responses following vaccination.

Potential role of vaccines in creating inflammatory reaction less than the extent of cytokine storms by SARS-CoV-2 may provide us an insight into the potential mechanisms of vaccine-associated uveitis. Coronavirus-2 (SARS-CoV-2) causes acute respiratory distress syndrome (ARDS) in 15% of COVID-19 cases. ARDS is mainly triggered by elevated levels of pro-inflammatory cytokines, referred to as cytokine storm which is induced by an excessive immune response rather than the viral load. Cytokine storm is defined as acute overproduction and uncontrolled release of pro-inflammatory markers systemically, as well as reduced macrophages functions and peripheral lymphopenia [34, 35].

Besides, recent studies suggest that excessive production of some cytokines, such as interleukin-6, interleukin-1, interleukin-17, and tumor necrosis factor-alpha, may be the leading cause of inflammatory response in COVID-19-related cytokine storms [36]. We understand that these interleukins and inflammatory factors play a crucial role in many ocular inflammatory events. Also, SARS-CoV-2 has a predilection target at organs expressing ACE2 receptor, which is abundantly expressed in the pulmonary system, and to a lesser extent, in the eye and brain. Thus, ocular tissue is also highly suspectable to SARS-CoV-2 entry and replication. These two factors may shed a light into the phenomenon of vaccine-associated uveitis.

Conclusion

This study reported a case series of intraocular inflammation events following COVID-19 vaccination, covering a broad disease spectrum. Anterior uveitis is most prevalent presentation in the series, followed by intermediate uveitis. Because early treatment of uveitis often results in rapid and complete resolution, uveitis flare-ups should not deter people from receiving COVID-19 vaccines. Nevertheless, the general public should be made aware about the potential risk of uveitis attacks and possible ocular symptoms following vaccination.