Incidence of persistent SARS-CoV-2 gut infection in patients with a history of COVID-19: Insights from endoscopic examination

Background and study aims Gut infection is common during acute COVID-19, and persistent SARS-CoV-2 gut infection has been reported months after the initial infection, potentially linked to long-COVID syndrome. This study tested the incidence of persistent gut infection in patients with a history of COVID-19 undergoing endoscopic examination. Patients and methods Endoscopic biopsies were prospectively collected from patients with previous COVID-19 infection undergoing upper or lower gastrointestinal endoscopy (UGE or LGE). Immunohistochemistry was used to detect the presence of persistent SARS-CoV-2 nucleocapsid proteins. Results A total of 166 UGEs and 83 LGE were analyzed. No significant differences were observed between patients with positive and negative immunostaining regarding the number of previous COVID-19 infections, time since the last infection, symptoms, or vaccination status. The incidence of positive immunostaining was significantly higher in UGE biopsies than in LGE biopsies (37.34% vs. 16.87%, P =0.002). Smokers showed a significantly higher incidence of positive immunostaining in the overall cohort and UGE and LGE subgroups ( P <0.001). Diabetic patients exhibited a significantly higher incidence in the overall cohort ( P =0.002) and UGE subgroup ( P =0.022), with a similar trend observed in the LGE subgroup ( P =0.055). Conclusions Gut mucosal tissues can act as a long-term reservoir for SARS-CoV-2, retaining viral particles for months following the primary COVID-19 infection. Smokers and individuals with diabetes may be at an increased risk of persistent viral gut infection. These findings provide insights into the dynamics of SARS-CoV-2 infection in the gut and have implications for further research.


Introduction
The COVID-19 pandemic, caused by the novel coronavirus "severe acute respiratory syndrome coronavirus 2" (SARS-CoV-2), began as an outbreak of pneumonia in December 2019.By March 2020, the World Health Organization had declared it a global pandemic [1].The implications have been profound, resulting in substantial life loss and significant economic crises worldwide.New cases and deaths related to COVID-19 continue to be reported [2].COVID-19 extends beyond the respira-Incidence of persistent SARS-CoV-2 gut infection in patients with a history of COVID-19: Insights from endoscopic examination Background and study aims Gut infection is common during acute COVID-19, and persistent SARS-CoV-2 gut infection has been reported months after the initial infection, potentially linked to long-COVID syndrome.This study tested the incidence of persistent gut infection in patients with a history of COVID-19 undergoing endoscopic examination.
Patients and methods Endoscopic biopsies were prospectively collected from patients with previous COVID-19 infection undergoing upper or lower gastrointestinal endoscopy (UGE or LGE).Immunohistochemistry was used to detect the presence of persistent SARS-CoV-2 nucleocapsid proteins.
Results A total of 166 UGEs and 83 LGE were analyzed.No significant differences were observed between patients with positive and negative immunostaining regarding the number of previous COVID-19 infections, time since the last infection, symptoms, or vaccination status.The incidence of positive immunostaining was significantly higher in UGE biopsies than in LGE biopsies (37.34% vs. 16.87%,P = 0.002).Smokers showed a significantly higher incidence of positive immunostaining in the overall cohort and UGE and LGE subgroups (P <0.001).Diabetic patients exhibited a significantly higher incidence in the overall cohort (P = 0.002) and UGE subgroup (P = 0.022), with a similar trend observed in the LGE subgroup (P = 0.055).
Conclusions Gut mucosal tissues can act as a long-term reservoir for SARS-CoV-2, retaining viral particles for months following the primary COVID-19 infection.Smokers and individuals with diabetes may be at an increased risk of persistent viral gut infection.These findings provide insights into the dynamics of SARS-CoV-2 infection in the gut and have implications for further research.
tory system, with documented effects on other systems, notably the gastrointestinal tract.More than 10% of patients have reported symptoms such as nausea, vomiting, and diarrhea during acute illness, occasionally preceding the onset of respiratory symptoms [3,4,5,6,7,8,9].In addition, viral RNA has been detected in the feces of approximately 50% of COV-ID-19 patients, regardless of gastrointestinal symptoms.In vitro models have shown that human-entered cells can host productive SARS-CoV-2 infections [10,11,12].
Immunohistochemistry has been effectively utilized to detect SARS-CoV-2-infected cells by identifying viral proteins within the cytoplasm of these cells [5,9,13,16,17,19].Despite this, there are limited data regarding persistent SARS-CoV-2 gut infection and the patient characteristics associated with this persistence or the development of long COVID-19 syndrome.Consequently, this study aimed to assess the incidence of persistent SARS-CoV-2 infection in the gut cells of patients with the history of COVID-19.These assessments were performed using immunohistochemistry during endoscopic examinations at a high-volume endoscopic unit, and the results were used to compare the characteristics of patients with positive and negative immunostaining.

Patients and methods
This was a random patient selection in a cross-sectional study design conducted between October 2022 and February 2023.Patients undergoing upper gastrointestinal endoscopy (UGE) or colonoscopy (lower gastrointestinal endoscopy [LGE]) during the study period were asked about their history of COVID-19.Those with polymerase chain reaction (PCR)-confirmed COV-ID-19 were invited to participate in the study.Participants provided informed consent for the use of their data in future research.The study was approved by our institutional Ethics Committee and adhered to the 1964 Declaration of Helsinki.

Endpoints
The primary endpoint was detection of Persistent SARS-CoV-2 gastrointestinal infection gut infection using immunohistochemistry in endoscopic biopsies from patients with a history of COVID-19 [9,20].The secondary endpoints were comparison of immunostaining results between UGE and colonoscopy; assessment of the association between positive immunostaining and clinical factors, including number of previous COVID-19, time since last infection, symptoms of initial COVID-19, and vaccination status; and analysis of risk factors associated with positive immunostaining.

Inclusion criteria
Patients were eligible if they had a history of PCR-confirmed COVID-19, with or without gastrointestinal symptoms, at least 1 month before endoscopy.They also had to exhibitno active illness indicative of COVID-19, including respiratory and gastrointestinal symptoms.All patients exhibiting only clinical symptoms or suspicions with mild or high fever were subjected to PCR testing and were advised to return home to await the results, ensuring the absence of COVID-19 during endoscopic biopsy.

Data collection
Data included demographic information, associated medical problems, original pathology in the endoscopic biopsies, COV-ID-19 serology tests at the time of endoscopy, vaccination information (type, number of doses, side effects), time elapsed since last vaccine dose, breakthrough infection and details about previous COVID-19 (symptoms, number of episodes, and the duration between the last illness and the time of endoscopy).

Tissue collection and examination
Endoscopic punch biopsies for viral particle detection were randomly taken from healthy-appearing mucosa in the stomach, colon, and terminal ileum using cold biopsy forceps.Biopsies were fixed in 10% buffered formalin for 24 hours, embedded in paraffin, and sectioned into 5-µm slices with a Leica RM2235 rotary microtome.The slices were mounted on positively charged slides and underwent 3,3'-diaminobenzidine (DAB) immunohistochemistry using an automated Dako autostainer (LINK 48).After antigen retrieval using EDTA solution for 15 minutes at pH 9, the slides were incubated with an anti-SARS-CoV-2 nucleocapsid monoclonal mouse immunoglobulin-G (IgG) antibody (Bio-techne, USA, #MAB10474-100) at a 1:1000 concentration for 30 minutes.After adding freshly prepared chromogen (DAB) for 2 minutes and counterstaining with hematoxylin, the stained slides were assessed for positivity [13].

Quantification of SARS-CoV-2 antibody response
The serological response to SARS-CoV-2 was measured using an enzyme-linked immunosorbent assay (ELISA_-based SARS-CoV-2 IgG assay (AnshLabs, Webster, Texas, United States) that quantifies antibodies to spike and nucleocapsid proteins.The Dynex automated analyzer was used to calculate the antibody concentration in arbitrary concentration units (AU/mL).A concentration of >12 AU/mL was considered positive, <10 AU/mL negative, and 10 to 12 AU/mL was considered an indeterminate [21].

Statistical analysis
For the analyses, we used descriptive and inferential statistics.All data were first tested for normality using the Kolmogorov-Smirnov, Q-Q plot, and Levene's tests.Categorical variables were expressed as n (%).Continuous normally distributed variables were described with their means and standard deviations, while non-normally distributed variables were expressed with their medians and interquartile ranges.When appropriate, categorical variables were tested using Pearson's Chi-squared or Fisher's exact test.Continuous normally distributed data were tested with the student's t-test for independent samples.For non-normally distributed data, the Mann-Whitney U-test was used for independent samples.
All independent variables counting more than 10 events and showing P <0.1 were eligible for multivariable analysis, achieved through backward selection.The estimated effects from the model were presented as adjusted odds ratios (AORs) and 95% confidence intervals (95% CI) to represent the likelihood of positivity associated with each variable.An AOR greater than 1 indicates a higher risk of positivity, whereas an AOR less than 1 suggests a lower risk.Statistical significance was set at P ≤0.05.Data were analyzed using R software version 4.2.2(The R Project for Statistical Computing, Vienna, Austria).

Results
This cross-sectional study included 249 patients with a median age of 53 (range 14 to 88 years).Males represented 52% of the cohort.Hypertension and diabetes mellitus (DM) were the most commonly observed associated medical conditions at 14.5% and 12.4%, respectively.Of the biopsies undertaken, 67% were from the upper gastrointestinal tract and 33% were from the lower gastrointestinal tract.The lower gastrointestinal tract samples comprised both colon and ileum biopsies per patient, accounting for 95%.
Regarding SARS-CoV-2 findings, nucleocapsid proteins of the virus were identified in 76 biopsies, accounting for 31%.The stomach was the predominant site of these findings, with 82% of the identified proteins.Of the biopsies, 89% showed no general pathological abnormalities, but gastritis was evident in 12 patients, which is 4.8%.Antibodies specific to SARS-CoV-2 (IgG) were detected in 32.5% of the patients, translating to 81 individuals.The median antibody titer among these was 36.8AU/mL.SARS-CoV-2 nucleocapsid tissue staining was discernible in gastric biopsies from 37.34% of patients for cytoplasmic staining.
In addition, colonic and ileal biopsies from 16.87% of patients displayed this staining.In gastric biopsies, the staining at the crypt base manifested as fine brownish cytoplasmic granules.In contrast, the colonic and ileal biopsies presented focal staining within the epithelial cells.Despite the positive staining, the histology was predominantly normal, though there was a mild increase in the lymphoplasmacytic infiltrate (▶ Table 1, ▶ Fig. 1, ▶ Fig. 2).

History of previous COVID-19
The median duration since the last infection was reported as 17 months, with a range spanning 7 to 30 months.Of the patients, 79% had experienced a single COVID-19 episode.One hundred eight (43%) reported gastrointestinal symptoms related to their infection.The primary symptoms were loss of smell/taste (41%), followed by nausea (26%) and diarrhea (18%).Regarding vaccination, 172 patients (69%) had received an anti-COVID-19 vaccine.Sinovac was the most common vaccine administered, accounting for 51%, with Pfizer being the second most common at 32%.Among those vaccinated, 72% had completed their vaccination with at least two doses.Notably, 127 subjects (91.5%) reported no side effects post-vaccination (▶ Table 2).

Original article
Characteristics of patients with positive and negative SARS-CoV-2 In the entire cohort, biopsies that tested positive were significantly more frequent among UGE patients (37.3%) compared to LGE patients (16.8%;P = 0.002).The distribution of negative SARS-CoV-2 results was 62.6% in UGE and 83.1% in LGE, respectively.Among patients with positive staining, smoking (36.8% vs. 7.5%) and DM (22.3% vs. 8.1%) were significantly associated when compared to those with negative staining (P <0.001 and P = 0.002, respectively) (▶ Table 3).

Multiple logistic regression analysis
In the logistic regression analysis assessing factors influencing the likelihood of a positive stain test result, several predictors were examined: age, sex, DM, smoking status, type of endoscopy, number of previous COVID-19 infections, time since the last infection, and vaccination status.Smoking status showed a pronounced effect, with an odds ratio (OR) of 7.68 (95% CI: 3.56, 17.46, P <0.001), signifying an increased likelihood of a positive test among smokers.The type of endoscopy was another significant predictor; patients undergoing upper endoscopy were more likely to test positive than those with lower endoscopy, with an OR of 2.74 (95% CI: 1.38, 5.72, P = 0.005).DM was significantly associated with a positive test result, demonstrated by an OR of 2.84 (95% CI: 1.17, 6.99, P = 0.021).In the segmented analysis focusing on patients with prior vaccination, we examined additional factors: the type of vaccine, number of doses received, time since the last vaccine dose, and history of breakthrough COVID-19 post-vaccination.None of these factors significantly influenced the likelihood of a positive stain test result (▶ Table 6).

Discussion
This study had 249 cross-sectional patients with UGE or colonoscopy (LGE), whereby samples for SARS-CoV-2 infection were tested.

Gastrointestinal manifestations during acute COVID-19
A multicenter study by Vanella et al. elucidated endoscopic findings in patients with COVID-19, with 44.4% experiencing gastrointestinal symptoms.Predominantly, ischemic-like colitis was observed in colonoscopies, whereas gastro-duodenal ulcers and erosions were frequent in UGE [3].Similarly, Mauro et al. identified ulcers as the most common finding in UGE for upper gastrointestinal bleeding in patients with acute COVID-19 [23].Moreover, Massironi et al. reported bulbar ulcers, gastric erosions, esophagitis, and ischemic colitis as principal findings in endoscopic examinations of acute COVID-19 patients [24].Neuberger et al. reported severe erosive duodenitis in 8% of critically ill, intensive care unit-admitted patients with COVID-19 who had pneumonia and were experiencing severe gastrointestinal symptoms, such as bleeding or feeding intolerance.These symptoms were directly attributed to enterocyte infection by SARS-CoV-2, as confirmed by positive immunostaining for the virus's spike protein in endoscopic biopsies [5].The virus's direct invasion of the endothelial epithelium, which expresses ACE-2 receptors, could elucidate the ischemic gastrointestinal lesions and general thrombotic events observed in acute COV-ID-19 [3].

Potential pathophysiological pathways
SARS-CoV-2, responsible for COVID-19, uses the angiotensinconverting enzyme 2 (ACE2) receptor to enter cells.These ACE2 receptors are abundant not just in the respiratory tract, but also in the small intestine's enterocytes, making the gastrointestinal system a target for infection [25,26].The virus can directly damage the gastrointestinal epithelium, manifesting in symptoms like diarrhea.Notably, viral particles have been detected in feces even when negative respiratory tests indicate potential prolonged gastrointestinal infection or viral shedding [6,27].The immune response to the virus can also cause local inflammation in the gut, increasing its barrier permeability and potentially leading to or worsening diarrhea [28].In addition, there is a notable connection between the gut and lungs, known as the "gut-lung axis."Any disruption in the gut can influence respiratory health and vice versa [29].Systemically, the virus's inflammatory response, especially releasing cytokines like interleukin-6, can affect gastrointestinal physiology [30].
Lastly, there are reports from some "long-haulers" who experience extended gastrointestinal symptoms after the primary infection phase has passed, indicating potential lasting effects on the gastrointestinal system.However, the exact causes are still under study [31].

Persistent gut infection by SARS-CoV-2
Persistent gut infection by SARS-CoV-2 has been reported, with viral RNA and nucleocapsid proteins detected in surgical samples from the stomach, colon, intestine, gallbladder, bile, and peritoneal fluid months post-infection [13,14,15].These viral components have also been identified in endoscopic biopsies Original article ▶ from the small bowel and colon long after COVID-19 resolution [5,9,16,17].This study detected SARS-CoV-2 nucleocapsid proteins in 30.5% of patients, primarily in the UGE subgroup (37.34%).Only 43.4% of these patients had prior gut involvement in their COVID-19.Immunohistochemistry confirmed persistent infection of the gut.Despite biopsies being from seemingly healthy mucosa, an increased lymphoplasmacytic infiltrate, indicative of a potent immune response, was seen in samples with positive immunostaining.
This study underscores the relatively common occurrence of persistent SARS-CoV-2 in the gut post-COVID-19, even without acute gastrointestinal symptoms, aligning with prior research identifying the gut as a long-term SARS-CoV-2 reservoir [9].Arostegui et al. identified a case of enduring SARS-CoV-2 in the colon 3 months post-infection, linked with chronic abdominal pain and elevated fecal calprotectin.This persistent viral colonization, evidenced by positive immunohistochemical staining for SARS-CoV-2 nucleocapsid proteins and dense lymphocytic infiltrations in colonic biopsies, may suggest a sustained immune response and potential etiology for gastrointestinaldominant long COVID-19 [17].
Zollner et al. detected SARS-CoV-2 nucleocapsid proteins and RNA in endoscopic biopsies from patients with inflammatory bowel disease several months post-acute COVID-19 resolution [16].Similarly, Cherne et al. studied SARS-CoV-2 persistence in endoscopic gastrointestinal biopsies and fluid samples, employing PCR, immunohistochemical staining, and virus isolation assays.The study included 100 patients with unknown prior COVID-19 infection and 12 with confirmed infection.Results showed a low incidence of persistent SARS-CoV-2 gut infection-only one biopsy from the unknown infection status group and three from the confirmed group tested positive.Moreover, no viable SARS-CoV-2 virions were isolated from the samples, and the virus and its RNA were found to be inactivated entirely within 24 hours of exposure to colonic fluids [9].▶ number of doses 0.524 Original article ▶ The inability to isolate infectious virions from gut tissues and fluids and their rapid inactivation suggest these persistent particles pose a minimal infection risk to healthcare professionals performing endoscopic procedures.However, persistent viral particles have been associated with "long COVID-19" syndrome [15,16,18], characterized by lingering symptoms post-primary COVID-19 [32].Affecting 30% to 87% of COVID-19 patients [17], long COVID-19 can manifest various gastrointestinal symptoms, including anorexia, dysphagia, bowel motility changes, abdominal pain, and weight loss [18].

Long COVID-19 syndrome
The gastrointestinal variant of long COVID-19 syndrome, characterized by persistent viral colonization leading to ongoing inflammation and cellular abnormalities, has been extensively documented [15,16,17,18,33,34,35,36]. Our study identified a higher incidence of persistent SARS-CoV-2 infection, as indicated by positive immunostaining, among smokers and individuals with DM.This difference was significant in the overall cohort and the subgroup undergoing UGE, suggesting smoking and DM could be a potential risk factor for developing long COVID-19 syndrome.
Although research on long COVID-19 risk factors is still limited, studies by Bai et al. [36].Barthélémy et al. [37], and Su et al. [38] have identified associations between long COVID-19 ▶ Our study employed antibodies targeting nucleocapsid proteins to detect persistent SARS-CoV-2 virions using a well-established approach reported in previous research [5,9,13,16,17,19].Immunohistochemistry staining for nucleocapsid and spike proteins exhibited higher sensitivity compared to RNA detection via reverse transcription polymerase chain reaction (RT-PCR), likely due to the increased stability of proteins.However, it should be noted that immunostaining was positive in only 75% of RT-PCR-positive gut mucosa samples in another study [16].
In our cohort, 32.5% of patients displayed SARS-CoV-2 antispike IgG antibodies, with a median titer of 36.8AU/mL.This antibody response could have stemmed from either prior infections or vaccinations.The median duration between the most recent infection and endoscopy was 7 months (range 3 to 30 months).In addition, the vaccination rate stood at 69%.There were no notable disparities in antibody prevalence, past infections, or vaccination status when comparing patients with positive and negative tissue staining results.
Our study employed an enzyme-linked immunosorbent assay (ELISA), which typically yields higher seroprevalence estimates than random access immunoassay (RAIA).Existing literature corroborates the persistence of a robust anti-spike IgG response for over a year post-infection [32,39].
Furthermore, it is important to emphasize that our study did not utilize validated diagnostic criteria to test for long COVID-19 syndrome.There need to be more universally accepted and validated testing methods for long COVID-19.As such, our study did not specifically assess the identification and characterization of long COVID-19 in our cohort.Future research using validated diagnostic criteria for long COVID-19 syndrome will offer a more thorough understanding of its prevalence and related risk factors.

Limitations
The absence of PCR testing in individuals for whom there was no suspicion of COVID-19 or who were symptom-free could have indicated missed infections during endoscopy; therefore, we cannot rule out the possibility of active subclinical COVID-19 in these cases.Another limitation is the need for follow-up data to monitor changes in persistent gut infection or disease progression over time in a patient.This study, being a crosssectional design, has inherent limitations.
Follow-up has dual implications in this context.First, there is the duration over which we monitored for the development of persistent SARS-CoV-2 gastrointestinal infections.These results indicated that this duration was sufficient.Nevertheless, cross-sectional studies provide a snapshot of data at a specific time, restricting our ability to establish causality or capture changes over time.Consequently, the findings should be interpreted within the context of this study design.Despite these limitations, the relatively large cohort from a single endoscopic center offers valuable insights.However, future multicenter studies with larger cohorts are warranted to validate the results further and minimize potential biases and confounding factors.Randomly selecting patients in this cohort reduces selection bias but limits the available variables, making complex modeling challenging.In addition, including longitudinal follow-up would provide a better understanding of the dynamics of persistent gut infection and its implications.

Conclusions
This study underscores the relatively prevalent persistence of SARS-CoV-2 infection in gut cells after an initial COVID-19 episode, even in the absence of gastrointestinal symptoms.Smokers and patients with diabetes seem to be at an elevated risk of continuous viral gut infection and the subsequent onset of long COVID-19 syndrome.These observations emphasize the need for more in-depth research to understand better the mechanisms and clinical consequences of enduring gut infection and its correlation with protracted COVID-19 symptoms.

Table 1
Patient characteristics.

Table 2
History of COVID-19 infection and vaccination.

Table 3
COVID-19Ag-positive and -negative cases in the whole cohort.

Table 6
Logistic regression analysis for predicting the probability of positive stain results.RNAemia, Epstein-Barr virus viremia, and auto-antibodies during the initial COVID-19.Patients with these risk factors, especially those with DM, should receive close monitoring due to their increased susceptibility to longterm COVID-19 and potential exacerbation of acute illness.