Article Text
PDF
Abstract
Purpose To evaluate the cases of corneal graft rejection following SARS-CoV-2 vaccination reported to Centers for Disease Control and Prevention Vaccine Adverse Event Reporting System.
Methods A descriptive analysis of the demographics, clinical history and presentation was performed. We evaluated the correlation between the vaccines and duration of vaccine-associated graft rejection (VAR) onset following vaccination using a one-way analysis of variance test. A post hoc analysis was performed between VAR onset-interval following vaccination dose and vaccine type. Finally, a 30-day cumulative incidence analysis was performed to assess the risk of VAR in short term following different doses, vaccines and type of corneal transplantation.
Results A total of 55 eyes of 46 patients were diagnosed with VAR following vaccination with BNT162b2 (73.91%) and mRNA-1273 (26.09%). The mean age of the patients was 62.76±15.83 years, and 28 (60.87%) were female. The patients diagnosed with VAR had undergone penetrating keratoplasty (61.82%), Descemet membrane endothelial keratoplasty (12.73%), descemet stripping endothelial keratoplasty (18.18%), anterior lamellar keratoplasty (3.64%) and corneal limbal allograft transplantation (1.82%). The mean time for VAR since penetrating and endothelial keratoplasty was 8.42±9.23 years and 4.18±4.40 years, respectively. 45.65% of the cases of VAR were reported after the second dose of vaccine. The duration of VAR onset was significantly shorter after the second dose compared with the first and booster doses (p=0.0165) and in patients who underwent endothelial keratoplasty compared with penetrating keratoplasty (p=0.041).
Conclusions This study outlines a possible temporal relationship between corneal graft rejection and SARS-CoV-2 vaccination. An earlier onset of VAR was observed in patients who had a history of endothelial keratoplasty and following the second dose of vaccination.
- Covid-19
- Cornea
Data availability statement
Data are available in a public, open access repository. The data are publicly available via the Centers for Disease Control and Prevention (CDC) Vaccines Adverse Event Response System at https://wonder.cdc.gov/vaers.html.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Several case reports and series have reported a potential temporal relationship between corneal graft rejection and SARS-CoV-2 vaccination.
WHAT THIS STUDY ADDS
This study evaluates the largest cohort of corneal graft rejection cases reported to the Centers for Disease Control and Prevention Vaccine Adverse Event Reporting System following the US Food and Drugs Administration authorised SARS-CoV-2 vaccination from all over the world. Among patients who developed vaccine-associated graft rejection, an earlier onset was observed in those who had undergone endothelial keratoplasty compared with penetrating keratoplasty and after the second dose of vaccination compared with the first and booster doses.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
The cornea specialists are recommended to closely monitor these patients for graft rejection following SARS-CoV-2 vaccination.
Introduction
The global response to counter the rapidly spreading SARS-CoV-2 infection primarily focused on the swift development of vaccines. The collective efforts of various research groups and commercial pharmacological entities have resulted in the development of 336 vaccine candidates.1 As of July 2022, 32 vaccines have received emergency use authorisation (EUA) and are in use worldwide. In the USA, the first SARS-CoV-2 vaccine received EUA on 11 December 2020.2 Since then, three vaccines—BNT162b2 (Pfizer Inc./BioNTech SE, Mainz, Germany), mRNA-1273 (Moderna Therapeutic, Cambridge, Massachusetts, USA) and Ad26.COV2.S (Janssen Pharmaceuticals, Beerse, Belgium) have been approved for use in the USA and several other countries.3–5 As these vaccines were approved through EUA, the Centers for Disease Control and Prevention (CDC) expanded its passive surveillance system—Vaccine Adverse Event Reporting System (VAERS), to evaluate any potential vaccine-associated adverse events.6 Vaccine-associated graft rejection (VAR) was also included as an adverse event of interest to the VAERS.
Several studies have highlighted a temporal association between VAR and SARS-CoV-2 vaccination; however, the evidence is limited to case reports and review articles. Moreover, the immunological mechanisms that lead to VAR are yet to be understood. The corneal tissue has a unique immune privilege attributed to several mechanisms that have been extensively studied over the past few decades. These mechanisms include anterior chamber-associated immune deviation, lack of vascularisation in a quiescent cornea, atypically low levels of major histocompatibility antigen expression and high levels of immunosuppressive cytokines in the aqueous humour, presence of characteristic T-cell apoptotic factors and dormant antigen-presenting cells.7 However, these immunoregulatory mechanisms may be disrupted in patients following corneal transplantation.8 It has been reported that the immune response on exposure to the vaccine can lead to induction of major histocompatibility compex-II (MHC-II) antigens in the grafted corneal tissue and has been speculated to cause VAR due to allorecognition or allogeneic response generated by humoral immunity.9 10 Several studies have highlighted the potential immune response generated by the primary components (viral messenger RNA in BNT162b2 and mRNA-1273 vaccines and recombinant replication-incompetent adenovirus type 26 vector in Ad26.COV2.S vaccine) of the SARS-CoV-2 vaccines.11–13 As a part of these immune responses, the cross-reactivity of viral antigen with corneal endothelial cells and generation of IFN-γ expressing CD4+ helper T-cells may cause endothelial cell apoptosis and graft rejection. The underlying mechanisms of VAR following SARS-CoV-2 mechanisms are purely speculative and can only be established after extensive investigations.
Since the commencement of the most extensive global vaccination efforts, few studies have evaluated the safety concern of ophthalmic adverse events following SARS-CoV2 vaccination using the VAERS database.14–16 Here, we assess the cases of VAR following the administration of the three United States Food and Drugs Administration (FDA) authorized COVID-19 vaccines (BNT162b2, mRNA-1273 and Ad26.COV2.S) reported to the VAERS from all over the world. In addition, we evaluate the demographics, association with the type of corneal transplantation, average VAR onset interval post-vaccination and clinical characteristics observed in patients.
Methods
Data source
This retrospective analysis was conducted using the CDC-VAERS database (CDC, Atlanta, Georgia, USA). In 1990, the CDC established VAERS as a national early warning system for unforeseen adverse events associated with the vaccines authorised or licensed by the US FDA. It is used to detect new, atypical or rare vaccine-related adverse events, monitor increases in known adverse events, identify potential patient risk factors for post-vaccination adverse events and assess the safety of FDA licensed vaccines reported from all over the world. The data in VAERS are publicly available, deidentified, anonymous data of vaccine adverse events reported by clinicians, vaccine manufacturers, patients and regulatory bodies.
The database is accessible through the Wide-ranging Online Data for Epidemiologic Research (WONDER) platform, developed and operated by the CDC, and includes demographic information, date of vaccination and adverse event onset, brief medical and surgical history, current comorbidities and medications, history of adverse events, and a detailed report of the clinical signs and symptoms and the diagnoses of the adverse event after vaccination.17 All the reports submitted to VAERS that appear to be false or fabricated to mislead CDC and FDA are reviewed before being added to the VAERS database and are punishable by a fine and imprisonment in violation of the Federal law (18 US Code § 1001). These reports are then evaluated and categorised by third-party professional coders using Medical Dictionary for Regulatory Activities preferred terms.18 On requesting explicit permission to analyse and publish these data, the authors were informed that CDC WONDER allows access to the information freely and could use, copy, distribute or publish this information without additional or explicit permission.19
Study population
This study included patients diagnosed with VAR following administration of BNT162b2, mRNA-1273 and Ad26.COV2.S between 11 December 2020 and 1 July 2022. The VAR cases were reported from 12 countries, including the USA. The data query to the VAERS included cases coded as ‘corneal graft rejection’ in the database including patients of all ages and genders. The generated data were grouped by symptoms, age, sex, location and onset interval. The additional measures included in the data were adverse event description, lab data, current illness, adverse events after prior vaccination, medications at the time of vaccination and history/allergies. The data points of interest were manually extracted (by RBS and JL) from the unstructured adverse event descriptions for analysis.
Statistical analysis
The descriptive statistical analysis was performed using R Studio (R Foundation for Statistical Computing, Vienna, Austria). A descriptive analysis of the social demographic characteristics and vaccination data was performed. We assessed the association between the onset interval of VAR and vaccine type, age, sex, dosage and type of corneal transplant surgery using the one-way analysis of variance test (ANOVA). To reduce the risk of type 1 error, Bonferroni correction was applied for multiple comparisons. A post hoc analysis was performed to assess the time interval in the age groups, dosage and vaccine type. A 30-day cumulative incidence analysis was performed to assess the risk of VAR in short-term following different doses, vaccines and type of corneal transplantation performed in the patients. The value of p<0.05 was considered statistically significant.
Results
During the study period, VAR was reported in 55 eyes of 46 patients following administration of SARS-CoV-2 vaccines, including BNT162b2 (n=34, 73.91%) and mRNA-1273 (n=12, 26.09%) vaccines. None of the patients had received Ad26.COV2.S vaccine. The mean age of the patients was 62.76±15.83 years, and 60.87% of the patients were female. The demographic and vaccination data of the patients in the cohort are summarised in table 1.
The demographics of patients diagnosed with vaccine-associated rejection post-SARS-CoV-2 vaccination
Most of the patients with VAR had undergone penetrating keratoplasty (n=34, 61.82%). The remaining patients had received Descemet membrane endothelial keratoplasty (n=7, 12.73%), Descemet stripping endothelial keratoplasty (n=10, 18.18%), anterior lamellar keratoplasty (n=2, 3.64%) and corneal limbal allograft transplantation (n=1, 1.82%). The forty-six patients included in this study had undergone corneal transplantation procedures for Fuchs endothelial corneal dystrophy (n=8, 17.39%), keratoconus (n=6, 13.04%), pseudophakic bullous keratopathy (n=4, 8.70%), and Salzmann nodular degeneration, exfoliation syndrome, congenital hereditary endothelial dystrophy, and infectious keratitis (one case each, 2.17%). The underlying aetiology for corneal transplantation was not known in 24 patients. The mean time for VAR since penetrating and endothelial keratoplasty was 8.42±9.23 years and 4.18±4.40 years, respectively (p=0.17). VAR was reported following the first dose in 17 patients (36.96%), second dose in 21 patients (45.65%) and third dose in 2 patients (4.35%). The vaccine dosage was unknown in four patients who received the BNT162b2 vaccine and two who received the mRNA-1273 vaccine. The mean VAR onset duration following BNT162b2, and mRNA-1273 vaccines was 15.10±12.76 and 19.27±31.54 days, respectively (table 2). The duration of VAR onset was significantly shorter following the second dose (10.37±9.32 days, p=0.0165; adjusted cut-off following Bonferroni’s correction p=0.017) compared with the first (16.26±12.97 days) or booster dose (34.5±3.53 days). In the cohort, 11 patients (23.91%) were diagnosed with VAR within the first week of vaccination. The VAR onset was significantly shorter in patients who had endothelial keratoplasty (9.55±5.54 days, p=0.041) compared with those who had penetrating keratoplasty (20.43±22.13 days). The onset duration following vaccination was comparable for unilateral and bilateral corneal graft rejection (p=0.91). Additionally, the 30-day cumulative incidence analysis showed that the onset of VAR was significantly earlier in patients who underwent endothelial keratoplasty compared with penetrating keratoplasty (p<0.0001) in the short term (figure 1). We did not observe a significant difference in VAR onset in patients who received either BNT162b2 or mRNA-1273 vaccines (p=0.7) (figure 2). The analysis also showed earlier onset of VAR following the second dose than the first dose (0.02) in the short term(figure 3).
The 30-day cumulative incidence analysis comparing VAR onset in patients with penetrating and endothelial keratoplasty. VAR, vaccine-associated graft rejection.
The 30-day cumulative incidence analysis comparing VAR onset in patients who received BNT162b2 and mRNA-1273. VAR, vaccine-associated graft rejection.
The 30-day cumulative incidence analysis comparing VAR onset in patients following first and second dose. VAR, vaccine-associated graft rejection.
Analysis to evaluate the factors associated with onset interval of VAR in 55 eyes of 46 patients following SARS-CoV-2 vaccination
The patients presented with reduced vision (n=18, 39.13%), corneal oedema (9, 19.57%), anterior chamber cells (n=8, 17.39%), Descemet folds (n=8, 17.39%) and conjunctival hyperaemia (n=7, 15.22%) and keratic precipitates (n=7, 15.22%). The patients were prescribed topical (n=24, 52.17%), oral (n=4, 8.70%) and subconjunctivally injected (n=5, 10.87%) corticosteroids for managing VAR. At presentation, the patients were on corticosteroids (n=11, 23.91%), other immunosuppressants (n=5, 10.87%), glaucoma (n=10, 21.74%) and antimicrobial (n=4, 8.70%) drugs. The clinical presentations and therapeutic management are detailed in table 3. The post hoc analysis for the VAR onset with different vaccine types and dosages are detailed in table 4A,B. The corneal graft rejection resolved in 17 patients and 2 patients required repeat corneal transplantation. In the remaining patients, the corneal graft rejection did not resolve in 5 and the status was unknown for 15 patients at the time of conducting this study.
Ocular presentation, therapeutic management and previously prescribed drugs in 46 patients with vaccine-associated rejection
(A) Post hoc analysis comparing onset interval with different vaccine doses
Discussion
The global vaccination efforts to protect the general population against SARS-CoV-2 were a critical step in mitigating the effects of this rapidly spreading virus. In the USA, the three FDA authorised vaccines—BNT162b2, mRNA-1273 and Ad26.COV2.S have helped in significantly reducing the severity of the disease, hospitalisations and prevented long-term effects of SARS-CoV-2.20–22 The evidence of VAR following SARS-CoV2 vaccination has been restricted to a few case reports and series. In this descriptive analysis, we report the largest cohort of 55 eyes of 46 patients diagnosed with corneal graft rejection following vaccination.
The literature lacks evidence regarding the causal association between VAR and SARS-CoV-2 vaccines, and this database analysis only provides evidence of a temporal association. However, graft rejection in patients who had undergone corneal transplantation more than 4-5 years ago provide evidence of the possible disruption of immune quiescence post-vaccination in these patients. Steinemann et al were the first to report five VAR cases following influenza, hepatitis B and tetanus booster vaccinations.23 Since the initiation of SARS-CoV-2 vaccination, more than 20 cases of corneal graft rejection have been reported following vaccination with BNT162b2 (Pfizer), mRNA-1273 (Moderna), ChAdOx1 nCoV-19 (Oxford-AstraZeneca) and CoronaVac (Sinopharm).9 10 24–39 The BNT162b2, mRNA-1273 and ChAdOx1 nCoV-19 are based on lipid nanoparticle encapsulated messenger RNA, whereas CoronaVac uses the inactivated DNA containing virus vector which encodes for SARS-CoV-2 spike protein, triggering an immune response in the recipients.21 40 41
Several cases of VAR following mRNA-based influenza vaccines, which have similar immune activation mechanisms to the vaccines included in this study, have been reported in the literature.23 42–44 The corneal graft rejection following vaccination with mRNA-based vaccines is speculated to be caused by a cross-reaction between viral antigen-specific T-cells with the human leucocyte antigens of the corneal endothelial cells.25 The vaccine may induce an expression and recognition of antigens of the major histocompatibility complex in all the donor corneal tissue following vaccination, thereby triggering VAR.10 Moreover, the neutralising antibodies generated in response to SARS-CoV-2 vaccines have been shown to elicit a robust CD4+Th1 response.37 These CD4+ Th1 cells are considered one of the primary mediators of corneal graft rejection. More of the reported VAR cases were after the second dose than after the first dose, which may be attributed to an accelerated immune response after antigenic sensitisation following the first dose of the vaccine. Moreover, the VAR onset duration was shorter following the second dose than the first and booster doses. Interestingly, no VAR cases have been reported following vaccination with Adenovirus vector-based vaccines. It can be speculated that the immune response discussed above may be exclusive to mRNA-based vaccines. However, we also need to consider that substantially fewer doses of Adenoviral vector-based vaccines (~18 million) have been administered compared with mRNA-based vaccines (~579 million).
Limitations
This study reporting vaccination-associated corneal graft rejections following SARS-CoV-2 vaccines has several limitations. VAERS is a passive surveillance system established by CDC that records adverse event reports. Despite the mandatory requirement to report certain vaccine-associated adverse events, under-reporting and delayed reporting are very common. Due to the lack of a control group, a relative risk analysis could not be performed. In some cases, the reports submitted to VAERS are incomplete and lack uniformity in data reporting. The 30-day cut-off for post hoc risk analysis was set at the initiation of the study to exclude cases of corneal graft rejection due to other confounders. However, on data analysis, we found that only one patient had a corneal graft rejection time of more than 30 days. The lack of dosage, frequency and duration data for ophthalmic medications prescribed prior to the presentation limited our ability to delineate their possible confounding effect on accelerating or delaying VAR. The adverse event reporting associated with SARS-CoV-2 vaccines is limited to a few countries in the European Union, North America, Australasia and a few Asian countries. Therefore, there is no reporting from the countries where most vaccine recipients reside. Additionally, there are no data for several other vaccines, such as ChAdOx1 nCoV-19, ZyCoV-D, Sputnik, Covidecia, Sputnik, Sinopharm, Abdala, Soberna, Zifivax and Novavax.
In conclusion, the analysis of the largest adverse event global database suggests earlier onset of VAR in patients who underwent endothelial keratoplasty compared with penetrating keratoplasty and following the second dose of vaccination. Therefore, it is recommended that cornea specialists should closely monitor these patients for graft rejection following vaccination.
Data availability statement
Data are available in a public, open access repository. The data are publicly available via the Centers for Disease Control and Prevention (CDC) Vaccines Adverse Event Response System at https://wonder.cdc.gov/vaers.html.
Ethics statements
Patient consent for publication
Ethics approval
This study was conducted in compliance with the tenets of the Declaration of Helsinki and National Statement on Ethical Conduct in Human Research 2007. Since the study includes publicly available, deidentified, anonymous data, the University of Adelaide Human Research Ethics Committee exempted it from ethical review (Ref ID: 36025).
References
Footnotes
Twitter @rbs_md
Contributors VJ accepts full responsibility for the work and/or the conduct of the study, had access to the data, and controlled the decision to publish. Conceptualisation: VJ and RBS; Methodology: VJ and BHJ. Software: UPSP and RBS; Validation: VJ and BHJ; Formal analysis: RBS and UPSP; Investigation: RBS and JL; Resources: VJ; Data curation: JL snd RBS; Writing-original draft : RBS and VJ; Writing-review and editing : VJ and BHJ; Visualisation: RBS, UPSP and VJ; Supervision: VJ; Project administration: VJ; Funding acquisition: NA.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Linked Articles
- Highlights from this issue