Adverse Events Related to SARS-CoV-2 Vaccine in a Nationwide Cohort of Patients With Inflammatory Bowel Disease

INTRODUCTION: There are limited data on the safety profile of the severe acute respiratory syndrome coronavirus-2 vaccine among patients taking immunosuppressive medications. Our aim was to evaluate the adverse events related to the vaccines in a nationwide cohort of patients with inflammatory bowel disease on diverse immunosuppressive medications. METHODS: This was a retrospective cohort study using data from the Veterans Health Administration. The primary outcome was any adverse event of special interest (cerebrovascular accident, venous thromboembolism, acute myocardial infarction, Bell palsy) within 90 days of vaccination. RESULTS: A total of 17,201 patients were included, and 12,351 patients (71.8%) received at least 1 vaccine dose. The most common adverse events were acute myocardial infarction and venous thromboembolism. In inverse probability treatment weighting-adjusted logistic regression, full vaccination was not significantly associated with increased adverse events through 90 days, relative to unvaccinated patients. DISCUSSION: Full severe acute respiratory syndrome coronavirus-2 vaccination was not associated with an increased rate of key adverse events relative to unvaccinated individuals among patients with inflammatory bowel disease.


INTRODUCTION
The severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) has affected over 596 million people worldwide as of August 24, 2022 (1). To prevent this, 2 vaccines (Pfizer BioNTech vaccine and Moderna) were initially approved by the US Food and Drug Administration.
Inflammatory bowel disease (IBD), comprising ulcerative colitis (UC), and Crohn's disease (CD), is a chronic inflammatory disorder of the gastrointestinal tract. Vaccination is strongly recommended to prevent the development of SARS-CoV-2 among patients with IBD. The vaccines have been shown to be effective among patients with IBD (2). However, despite their efficacy, a recent study in US Veteran Health Administration (VHA) cohort of patients with IBD found that only 61.8% of patients with IBD were vaccinated (3). A possible reason for this number may be the fear of adverse events reported after receipt of the SARS-CoV-2 vaccine. Our aim was to evaluate the adverse events related to the vaccines in a nationwide cohort of patients with IBD.

Study design and cohort creation
This was a retrospective cohort study using data from an established national IBD cohort in the VHA. We used a previously validated algorithm based on administrative codes and pharmacy data to identify patients with UC or CD before December 18, 2020, the start date of the VHA COVID-19 vaccination campaign (index date) (3,4). Patients 18 years or older with at least 2 outpatient appointments before the index date were included. Patients were excluded if they received the Janssen vaccine during the study window given the small sample size.

Exposures
The primary exposure in this study was SARS-CoV-2 vaccination (first dose, second dose, unvaccinated). Vaccine administration events (first and second dose, as applicable) were identified using Current Procedural Terminology codes. For each patient, we additionally obtained baseline demographics (age, sex, race), alcohol use history, tobacco use history, geographic region, IBD type (UC or CD), and Charlson Comorbidity Index. IBD medication groups in the 3 months before the index date were categorized as 5-ASA alone, thiopurines (azathioprine or mercaptopurine, i.e., thiopurines (TPs), with or without 5-aminosalicylic acid medications), antitumor necrosis factor (anti-TNF) agents alone, anti-TNF 1 TPs, vedolizumab, ustekinumab, and tofacitinib. Steroid use was also ascertained in a 3-month window before the index date using prescriptions for budesonide, methylprednisolone, prednisolone, or prednisone.

Outcomes
The primary outcome was development of any adverse event of interest after SARS-CoV-2 vaccination, with follow-up data obtained through February 2, 2022. Based on observed adverse events in Clinical and Translational Gastroenterology VOLUME 14 | APRIL 2023 www.clintranslgastro.com clinical trials, we used incident International Classification of Diseases 9/10 codes to ascertain the first occurrence of the following composite events identified as adverse events of special interest (AESIs), which were commonly reported in the Pfizer and Moderna trials: cerebrovascular accident, venous thromboembolism (VTE), acute myocardial infarction, and Bell palsy. AESIs were identified within 90 days of first or second vaccination doses. As a reference group, AESIs for unvaccinated patients were evaluated relative to an index date set to the median time to first vaccination in patients who were vaccinated (February 14, 2021). Importantly, every adverse event was manually adjudicated for confirmation through detailed chart review. For statistical analysis, refer supplementary material.

Cohort and vaccination characteristics
After application of selection criteria and exclusion of 387 patients who received the Janssen vaccine, we identified a total of 17,201 patients with IBD. Of these, 12,351 (71.8%) received at least 1 vaccine dose and 4,850 (28.2%) remained unvaccinated (Table 1).

Postvaccination adverse events
A summary of postvaccination AESIs is provided in Table 2.

Association between vaccination and adverse events
After creation of propensity scores and application of inverse probability weights, excellent covariate balance was achieved between unvaccinated and vaccinated groups, demonstrated by standardized mean differences reduced to within 6 0.1 for each exposure variable (see Supplemental Figure 1

DISCUSSION
In the Pfizer vaccine trial, the serious adverse events which were higher among those who received the vaccine vs placebo were acute myocardial infarction and cerebrovascular accident (5). For the Moderna vaccine, these were Bell palsy, myocardial infarction, and nephrolithiasis (6). Because patients with IBD have an inherently increased risk of developing myocardial infarction (7) as well as VTE (8), it was reassuring that fully vaccinated state was not associated with an increased risk of these events. Major strengths of this work include the use of a large, national cohort of patients with IBD taking diverse medication regimens and an inverse probability weighted design to account for variation in baseline characteristics between vaccinated and unvaccinated individuals. Each adverse event was confirmed by individual review of the chart. The primary limitation is that the cohort comprised an older predominantly male population, which may hinder broad generalizability of the results.
In conclusion, our study showed that among patients with IBD, full vaccination state is not associated with an increased risk of adverse events attributable to vaccination. Furthermore, the overall rate of adverse events was very low. This is of relevance given the large proportion of patients who remain unvaccinated and hopefully will lead to increased vaccine adoption.