key: cord-0704444-vwhkz22k
authors: Zevit, Noam; Chehade, Mirna; Leung, John; Marderfeld, Luba; Dellon, Evan S.
title: Eosinophilic Esophagitis Patients Are Not At Increased Risk of Severe COVID-19: A Report from a Global Registry
date: 2021-10-22
journal: J Allergy Clin Immunol Pract
DOI: 10.1016/j.jaip.2021.10.019
sha: 616f4a98ca8f149db286016da89c504a0c7b72a6
doc_id: 704444
cord_uid: vwhkz22k

Background The impact of COVID-19 on eosinophilic esophagitis (EoE) and eosinophilic gastrointestinal diseases (EGID) is unknown. Objective We aimed to characterize EoE and EGID patients who had COVID-19, assess severity of COVID-19 in the EoE/EGID population, and evaluate for COVID-19-induced EoE/EGID flares. Methods We established an online global registry collecting physician entered, de-identified data related to patient demographics, EoE/EGID disease features, comorbidities, and treatments, COVID-19 source of exposure, symptoms, illness severity, hospitalizations, and deaths. Results Ninety four cases were reported between March 2020 and April 2021 (median age 21 years (range 1.5-53); 73% male). Most had atopy (73%), and 80% had isolated EoE. Prior to COVID-19, the EoE/EGID activity was reported as clinical remission in 51 (54%) and moderate in 20 (21%). EoE/EGID treatments at the time of COVID-19 included PPIs 49(52%), swallowed/topical steroids 48(51%), and dietary elimination 34(36%). COVID-19 symptoms included cough (56%), fever (49%), anosmia (21%), and ageusia (22%). Most patients with COVID-19 had a mild course (70%), with 15% asymptomatic, 12% moderate, and 2% severe. Three patients were hospitalized, and no ICU admissions or deaths were reported. Mean time from first symptoms to resolution in symptomatic patients was 10 days (range 1-90). A single EGID flare was reported during COVID-19. Conclusions In a global EoE/EGID registry, relatively few COVID-19 cases have been reported. COVID-19 severity was comparable to the general population. Based on this registry it does not appear that EoE patients are at increased risk for severe COVID-19 infection or that COVID-19 leads to EGID flares.

CoV-2, ACE2, is expressed in multiple human tissues, including the esophagus(2) and intestines(3). 77

Binding of the SARS-CoV-2 spike protein to ACE2 and proteolytic cleavage by TMPRSS2, a cellular serine 78 protease, facilitate viral entry into cells(4). While the infection remains asymptomatic or mild in most 79 patients, a minority may progress to moderate/severe disease or death. Risk factors for severe 80 outcomes in COVID-19 include older age, obesity, chronic lung and cardiovascular disease, diabetes, 81 immune suppression and others(5, 6). 82

Early attempts to assess the COVID-19 burden in patients with eosinophilic esophagitis (EoE) and 83 eosinophilic gastrointestinal diseases (EGIDs), whose main disease pathology is concentrated in tissues 84 expressing the proteins necessary for viral invasion, failed to identify SARS-CoV-2 infected EoE/EGID 85 patients(7-9). Not only do EoE/EGIDs involve organs involved in SARS-CoV-2 infection, but they are also 86 a form of immune dysregulation, and their mainstay treatments involve medications (proton pump 87 inhibitors (PPI) and steroids) which have been implicated in the risk for more severe COVID-19(10-13). 88 Furthermore, while normal and elevated peripheral eosinophil counts have been found to be associated 89 with better outcomes in COVID-19(14, 15), it is theoretically possible that eosinophil depleting drugs, 90 which are sometimes used for treatment of EoE/EGID, may induce more severe disease(16). 91

To date, there have been no published series of COVID-19 or SARS-COV-2 infection in patients with 92 either EoE or EGIDs distal to the esophagus, and only a single case report has been published(17). A 93 clear understanding of the impact that pre-existing EoE/EGID may have on COVID-19 severity, and 94 conversely, how COVID-19 may impact EoE/EGID disease activity, is essential for guidance to both 95 patients and practitioners. Using an international registry of EoE and EGID patients with COVID-19, we 96 aimed to determine the characteristics of patients with EoE/EGID who were diagnosed with COVID-19, 97 assess the severity of disease in this cohort, and evaluate for COVID-19 induced EoE/EGID flares. with a diagnosis of EoE or EGID. SECURE EoE/EGID was established as a broad collaborative international 104 effort due to the rarity of eosinophilic gastrointestinal diseases, and was based on similar work in 105 inflammatory bowel disease(13). The project was endorsed and promoted by Gastroenterology Societies 106 in Europe, North America, and Latin America, Allergy and Immunology societies, and patient advocacy 107 groups (Supplemental Table E1 ). Following the initial announcement of the database, several reminders 108 were sent throughout 2020 and early 2021 and the link to the website was added to other disease 109 specific COVID-19 websites. 110

Medical providers were encouraged to report all EoE/EGID patients diagnosed with COVID-19 by a 111 laboratory test (or when testing was unavailable, suspected based on close contact with a test-112 confirmed patient and typical symptoms) regardless of the disease severity. To enable practitioners to 113 easily access data entry, a dedicated website was established (covideoeegid.org). Practitioners were 114 instructed to enter data after the patient had COVID-19 for a long enough duration to experience 115 complete or partial recovery, hospitalization, or death. For cases reported while patient was still 116 symptomatic, the physician was contacted to assess for final outcome prior to data analysis. We utilized 117 Figure E2 ). Within the United States, patients resided in 17 different States (Supplemental Figure E2) , 142 the majority from the Northeastern and Southwestern US. The median age was 21years (IQR 16-31; 143 range 1.5-53), 73% were male, and 82% reported as white. Atopy was present in 71% of patients; 39% 144 had asthma, of which 43% had active disease based on either use of beta-agonists twice weekly, 145 prophylactic inhaler use, or an asthma related hospitalization in the previous year. The vast majority 146 (98%) were non-smokers. ( Table 1 ). The median (IQR) BMI of patients over 16 years of age was 23.5 147 (21.8-26.9). 148

Insofar as the segment of the GI tract involved in the eosinophilic disease, while the majority 90 (96%) 149 had esophageal involvement, only 75 (79.8%) had isolated EoE, 2 (2.1%) had isolated eosinophilic colitis, 150 and one had combined eosinophilic enteritis and colitis (1.0%). The remainder had eosinophilic 151 esophageal involvement (EoE) of their EGID combined with other segments of the GI tract including: 152 eosinophilic gastritis in 6 (6.4%), eosinophilic enteritis in 5 (5.3%), eosinophilic gastroenteritis in 3 153

(3.2%), and pan-GI involvement in 2 (2.1%). (Table 1 ). The majority were in clinical remission (54%) at 154 the time of the last assessment prior to COVID-19, though only 34% reported combined clinical and 155 histological remission at the time of the last gastroenterologist follow-up prior to COVID-19. Seventeen 156 patients (18%) had mild and 20 (21%) had moderate disease activity, while only 6 (6%) were assessed to 157 have severe EoE/EGID activity. Other than atopic conditions, there were 22 comorbidities reported in 24 158 patients including anxiety disorder (4), attention deficit and hyperactivity disorder (3), celiac disease (2), 159 inflammatory bowel disease (2), mast cell activation syndrome (2) and others (Supplemental Table E2 ). receiving systemic steroids or enteric budesonide, one received azathioprine/6-mercaptopurine, and 4 165 patients were receiving a biologic treatment ( Table 1) . Feeding tubes were reported by three patients; 166 five patients had undergone esophageal dilation in the three months prior to COVID-19. 167

The diagnosis of COVID-19 was made by naso-pharyngeal swab PCR in 79%, an antigen-based test in 7%, 168 and serology in 3%. There were 9% of patients who were diagnosed as "suspected" cases on the basis of close contact with a confirmed case and typical symptoms. These were all from the first months of the 170 pandemic when diagnostic testing was very limited in many countries. The most common sources of 171 exposure were a household contact in 28%, a work contact in 10%, travel in 3% and unknown in 30%. 172

The mean time from diagnosis of EoE/EGID to COVID-19 was 5.0±4.2 years. 173

Symptomatic patients (n=80) most often presented with cough, low grade fever, dyspnea, ageusia, 174 rhinorrhea, and anosmia (Figure 1) . Only a minority of patients reported gastrointestinal complaints 175 during the duration of their COVID-19. These included diarrhea (8%), nausea/vomiting (5%), and 176 abdominal pain (4%), with only three patients having an overlap between different GI complaints. 177 COVID-19 severity (Figure 1 ) was reported as mild (70%) or asymptomatic (15%) in most. Only 12% 178 reported moderate and 2% severe disease. 179

There was no significant association between COVID-19 severity (asymptomatic/mild vs. 180 moderate/severe) and any of the standard EoE/EGID medications used by the patient ( Table 2 Three patients had been hospitalized for COVID-19. All three were over 40 years old, and none needed 193 mechanical ventilation or extracorporeal membrane oxygenation (ECMO). The first was a 40 year old 194 non-atopic, non-obese female eosinophilic colitis patient who was hospitalized with fever, cough, 195 dyspnea, headaches and myalgia. She received lopinavir and hydroxychloroquine, and was discharged 196 after 7 days. Symptoms resolved after a total of 17 days. The second patient was a 41 year old, non-197 asthmatic, overweight (BMI 26) male with EoE treated with topical fluticasone (0.8mg/day). He was 198 hospitalized with a low fever, cough, dyspnea and diarrhea, and then developed a pulmonary 199 thromboembolism. The patient received lopinavir and ritonavir. His symptoms resolved after 28 days.

The final hospitalization was of an obese (BMI 32) 43 year old male with mild intermittent asthma, food 201 allergy and EoE. He was treated with an elimination diet and PPIs, and had prior esophageal dilation. He 202 was hospitalized for three days with high fever, cough, dyspnea, rhinorrhea and diarrhea. It is unknown 203 if he received anti-viral treatment. His symptoms fully resolved after 60 days. 204

In order to make an estimate of the minimal value for the year-long incidence of COVID-19 in EoE/EGID, 205 centers including patients were asked the number of EoE/EGID patients they follow. Of seven US centers 206 who responded, there were 39 COVID-19 cases out of 4,391 patients giving an incidence of 1:113 207 (8,850/1 million). On a world-wide basis, there were 57 COVID-19 cases out of 5,549 patients, giving an 208 incidence of 1:97 (10,309/ 1 million). This is compared to the total US incidence since the beginning of 209 the pandemic of 104,050 cases/1 million population, or a worldwide incidence of 23,848 cases/1 million 210 population(18). (14, 15, 22) . These studies further 252 support the notion that individuals whom harbor a TH2 predominant milieu have better outcomes, 253 possibly because this profile protects from the damaging effects of a TH1 associated cytokine storm(9, 254 21). This theory, however, has been challenged by others who note the decreased lymphocyte counts in 255 severe COVID-19 and that Th17 hyper-response may actually be responsible for the cytokine storm, 256 suggesting that augmenting a TH1 response in severe patients may actually be beneficial to them(23, 257 24). Because only few of the patients in our cohort had blood tests performed soon before or during 258 their COVID-19 (none of which included T-cell subtyping), the fact that the vast majority had relatively 259 mild disease, and our study construct, we could not assess for this association in EoE/EGID patients. 260

In our study, neither the use of PPI nor of swallowed topical steroids were associated with moderate or 261 severe COVID-19. Chronic PPI use, one of the main EoE/EGID associated treatments, has been 262 implicated in risk of SARS-CoV-2 infection(10) as well as poorer COVID-19 outcomes(10-12), though 263 recent data have challenged this(25). Similarly, patients using chronic systemic steroids in inflammatory 264 bowel diseases(13) and rheumatic disease(26, 27) have had more severe COVID-19 outcomes. However, 265 inhaled steroids used for asthma and COPD maintenance treatment appear to be safe(28, 29). It is 266 possible that the general health and relatively young age of our patients mitigated the effects of PPI. As 267 to swallowed topical steroids, debate exists as to the systemic exposure of patients using these 268 formulations of budesonide or fluticasone(30-32). If systemic exposure exists, it is minor in most cases -269 presenting a situation akin to that of inhaled corticosteroids for asthma or COPD, which may explain the 270 lack of an association with deleterious effects in EoE patients. 271

This study has limitations which must be acknowledged. First, while this is the first and largest series of 272 patients with EoE/EGIDs and COVID-19, there were still too few patients with EGID distal to the esophagus to fully assess the effect of either the EGID or its treatments on the natural history of COVID-274 19 in such patients. Furthermore, even for EoE, the size of the cohort still makes conclusive statements 275 difficult. In order to increase the numbers of patients entered, we contacted physicians several times 276 throughout the year to promote the database including the addition of the URL on society dedicated 277 COVID-19 sites, emails to dedicated EoE/EGID research groups and list-serves, and added requests 278 through patient advocacy groups to notify their physicians in the event that they, as patients, were 279 infected with SARS-CoV-2 regardless of severity. In contrast to other chronic diseases, such as IBD or 280 chronic liver disease, patients with EoE/EGID may not be in as close contact with their treating physician, 281 and clinic visits are less frequent, especially during clinical remission. However, it is not implausible that 282

there were simply few COVID-19 cases in this patient group as reflected by the minimal incidences rates 283 calculated from our cases. However, because patients were neither actively contacted nor universally 284 tested for COVID-19, actual or maximal incidence rates are unknown, and these values are minimum 285 rates. The study was not intended as a formal epidemiological study, and as most SARS-CoV-2 infections 286 are asymptomatic or mild, it may be that many cases were not reported to the treating physicians or 287 that visibility of the database was not high enough, introducing a reporting bias or recruitment bias from 288 highly motivated centers. In addition, severe cases -including either death or disability, may not have 289 been reported because patients were not in contact with their gastroenterologist/immunologist and 290 were thus regarded as lost to follow-up. Another limitation is that we relied on physician reports which 291 were sometimes secondhand reports of data passed on from the patients who had been treated 292 elsewhere. Disease severity was recorded as assessed by the physician entering the data. Since formal 293 guidelines on the nomenclature of disease severity were only developed as the pandemic advanced, and 294 differ between time periods, reporting countries and guidelines(33), these were not specifically queried 295 nor reported. Furthermore, we do not have data on the need for supplemental oxygen or COVID19 296 associated imaging which are used by some guidelines to assess severity. Other disease specific 297 databases similarly did not assess severity by specific metrics during the initial data collections (34, 35) . 298

Due to the rarity of COVID-19 in our population, we chose to include the few cases which were not 299 confirmed by testing rather diagnosed epidemiologically based on close contact with a testing-300 confirmed patient and having disease appropriate symptoms. These cases were all from high prevalence 301 areas at times when confirmatory testing was not yet available. These may also have introduced bias 302 into the data. The main strength of this study is the robust, worldwide collaboration which enabled us 303 to collect data on a rare incidence taking place in rare diseases. 

Digestive system is a potential route of COVID-19: 326 an analysis of single-cell coexpression pattern of key proteins in viral entry process

Intestinal Inflammation 329 Modulates the Expression of ACE2 and TMPRSS2 and Potentially Overlaps With the Pathogenesis of 330 SARS-CoV-2-related Disease

SARS-CoV-2 Cell 332 Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor

Risk factors for severe and critically 335 ill COVID-19 patients: A review

A Systematic Review of Systematic Reviews on the COVID-19 Pandemic

Clinical and Psychological 339 Impact of COVID-19 Infection in Adult Patients with Eosinophilic Gastrointestinal Disorders during the 340 SARS-CoV-2 Outbreak

Lack of complications in 342 patients with eosinophilic gastrointestinal diseases during SARS-CoV-2 outbreak

Eosinophilic esophagitis: is the Th2 345 inflammation protective against the severe form of COVID-19?

Increased Risk of COVID-19 Among Users of Proton Pump 348 Inhibitors

Proton pump inhibitors and the risk of 350 severe COVID-19: a post-hoc analysis from the Korean nationwide cohort

Do proton pump inhibitors influence SARS-CoV-2 352 related outcomes? A meta-analysis. Gut

But 354 Not TNF Antagonists, Are Associated With Adverse COVID-19 Outcomes in Patients With 355 Inflammatory Bowel Diseases: Results From an International Registry

Patients Is Protective Against Severe COVID-19 Illness

Longitudinal hematologic and immunologic 361 variations associated with the progression of COVID-19 patients in China

Anti-IL5 Drugs in COVID-19 Patients: Role of Eosinophils in SARS-CoV-2-Induced 364

SARS-COV2 and eosinophilic esophagitis: a 366 first case

Worldometer COVID-19 coronavirus pandemic

COVID-19) Outbreak in China: Summary of a Report of 72 314 Cases From the Chinese Center for 371 Disease Control and Prevention

Asymptomatic 373 SARS-CoV-2 infection: A systematic review and meta-analysis

Type 2 Immunity and Age Modify 376

Gene Expression of Coronavirus-induced Disease 2019 Receptors in Eosinophilic Gastrointestinal 377 Disorders

Eosinophil responses during COVID-19 infections and 379 coronavirus vaccination

The cytokine storm-An appropriate, over-reactive response to SARS-CoV-2 or 381 the wrong immune pathway?

COVID-19: more than a cytokine storm

Proton Pump Inhibitor Use Is 385 Not Strongly Associated With SARS-CoV-2 Related Outcomes: A Nationwide Study and Meta-analysis