COVID-19 has become a worldwide pandemic causing widespread illness and mortality. SARS-CoV-2 mainly infects the respiratory tract through attachment to angiotensin-converting enzyme 2 (ACE2) receptors (Lan et al., 2020). ACE2 is also highly expressed on intestinal epithelial cells, allowing SARS-CoV-2 to infect the intestinal tract (Xiao et al., 2020a). Recent meta-analyses show that 48%–54% of fecal samples from COVID-19 patients have tested positive for viral RNA, and 15%–17% of patients have gastrointestinal (GI) symptoms (Cheung et al., 2020; Mao et al., 2020; Sultan et al., 2020). Additionally, live virus can be isolated from fecal samples of COVID-19 patients (Wang et al., 2020). Some studies have proposed fecal–oral transmission as the cause of intestinal infection (Guo et al., 2021). However, direct evidence for fecal–oral transmission is still lacking. Meanwhile, the virus has been detected in the blood of both symptomatic and asymptomatic patients (Chang et al., 2020), and disseminated virus could infect extrapulmonary organs (Jacobs and Mellors, 2020). Thus, the potential that intestinal infection occurs via blood transmission should be carefully considered.

Pulmonary infection triggers cytokine storm and induces a prothrombotic state (McFadyen et al., 2020; Moore and June, 2020). Venous and arterial thrombosis are common in COVID-19 (Moore and June, 2020). Systematic reviews estimate that 14%–31% of in-hospital patients develop a clinically apparent thrombotic event (Suh et al., 2021; Tan et al., 2021), while autopsy reports show a high prevalence of microthrombi in multiple organs, including lung, heart, liver, kidney, and gastrointestinal tract (Bradley et al., 2020; Polak et al., 2020). A cohort study showed that COVID-19 patients with intestinal ischemia had markedly elevated D-dimer levels and poor outcomes (Norsa et al., 2020). Additionally, recent studies have shown that mesenteric thrombosis often results in intestinal resection and significantly increases mortality (Bhayana et al., 2020; El Moheb et al., 2020). Therefore, it is essential to outline the mechanisms of intestinal thrombosis and its contribution to intestinal damage and disease progression.

In this review, we discuss blood transmission as a potential route for intestinal infection. We then summarize the characteristics and mechanism of intestinal thrombosis formation in COVID-19. Next, we focus on the effects of intestinal infection and thrombosis on intestinal damage and disease severity. Finally, we discuss therapeutic strategies to prevent intestinal damage.

Multiple studies have reported GI symptoms in COVID-19 patients, including diarrhea, nausea, vomiting, anorexia, and abdominal pain (Cheung et al., 2020; Mao et al., 2020; Sultan et al., 2020). According to a meta-analysis comprising 10,890 COVID-19 patients, the pooled prevalence estimates of GI symptoms were: diarrhea (7.7%), nausea or vomiting (7.8%), and abdominal pain (2.7%; Sultan et al., 2020) with 10% of these patients reporting GI symptoms as being their initial symptoms (Cheung et al., 2020). These data indicate potential gastrointestinal infection by SARS-CoV-2, which is reported to infect and replicate in epithelial cells of human small intestinal organoids (Zang et al., 2020). Both viral nucleocapsid proteins and viral particles have been detected in infected patient intestinal biopsies (Livanos et al., 2021). Additionally, SARS-CoV-2 RNA and live virus can be found in the stool of patients (Wang et al., 2020). More importantly, SARS-CoV-2 subgenomic mRNA is transcribed in actively replicating cells and has been detected in fecal samples (Wölfel et al., 2020). Further, rectal viral shedding persists for longer than that of the respiratory system (Zhao et al., 2020). All these data demonstrate that SARS-CoV-2 directly infects and replicates in intestinal epithelial cells of patients.

With the deepening understanding of COVID-19, GI symptoms have been recognized as early signs of the disease. The high expression of ACE2 in the GI tract, isolation of live virus from fecal samples, and a subset of patients presenting with only GI symptoms seem to suggest fecal–oral transmission. However, problems with the feasibility of this mode of transmission remain. First, studies have shown that SARS-CoV-2 loses infectivity in simulated gastric acid within 10 min (Chan et al., 2020; Zang et al., 2020; Zhong et al., 2020). Secondly, SARS-CoV-2, as an enveloped virus, is largely unable to withstand the detergent effect of bile salts and the activity of digestive enzymes in the duodenum (Figure 1). Although some studies have suggested that highly viscous mucus in the gastrointestinal tract protects SARS-CoV-2, allowing the virus to retain its infectivity (Guo et al., 2021; Zhang H. et al., 2021), there is still a lack of direct evidence. Bushman et al. (2019) had previously investigated the links between the structures of viruses and routes of transmission and found a strong association between fecal–oral transmission and the absence of a lipid envelope. Lastly, although some studies have isolated intact viruses from feces (Wang et al., 2020; Zhang Y. et al., 2020; Zhou et al., 2020; Xiao et al., 2020b), most of them have not further confirmed the infectivity of these viruses (Wang et al., 2020; Zhang Y. et al., 2020; Xiao et al., 2020b). Zhou et al. (2020) confirmed viral propagation by RT-PCR, but only in a single fecal sample. Previous research has shown that SARS-CoV-2 is completely inactivated in simulated human colonic fluid over the course of 24 h, which may explain the sporadic detection of infection-active SARS-CoV-2 from feces samples.

Several lines of evidence suggest that SARS-CoV-2 may infect the intestinal tract via the bloodstream. Deng et al. (2020) detected SARS-CoV-2 RNA in anal swabs from intratracheally but not intragastrically infected rhesus macaques, suggesting blood transmission. Indeed, SARS-CoV-2 RNA has been detected in blood and urine samples of patients (Wang et al., 2020). The virus can also be detected in multiple organs (including heart, brain, and kidney) and is associated with organ injury, indicating that the virus can reach and infect extrapulmonary organs (Puelles et al., 2020). Another study showed that SARS-CoV-2 viremia was associated with intestinal damage, independent of disease severity (Li Y. et al., 2021). Thus, blood transmission could be the cause of intestinal infection. Specifically, SARS-CoV-2 replicating in alveolar epithelial cells and capillary ECs is released into the bloodstream and infects new vascular ECs. The capillary network is then the main route by which the virus enters and infects extrapulmonary organs. The extensive surface area of intestinal capillaries makes intestinal epithelial cells more susceptible to infection than other extrapulmonary organs. Following infection of intestinal capillaries, SARS-CoV-2 is released into the gut and infects surrounding intestinal epithelial cells along the intestinal tract (Figure 1). Once established in the gut, SARS-CoV-2 can also reenter the capillaries, potentially leading to recurrence of disease. Consistent with this, in patients who experienced recurrence, the phylogenetic analysis of infection samples has shown that recurrent virus evolves from the original parent virus (Hu et al., 2020).

Additionally, SARS-CoV-2 RNA can also be detected in the blood and urine of asymptomatic patients, suggesting a second pathway to viremia through the nasal cavity (Chang et al., 2020; Hasanoglu et al., 2021). The abundant blood vessels, thin mucous membrane, and higher levels of ACE2 (Huang et al., 2021) make it possible for the virus to initiate viremia from the nasal cavity. Specifically, SARS-CoV-2 is released from infected ciliary cells of the nasal cavity and breaks through the basement membrane, infecting the vascular ECs and eventually entering circulation (Figure 1). Blood transmission after nasal infection is therefore another potential route of intestinal infection.

A recent study has shown that a fecal sample positive for SARS-CoV-2 RNA at any time during hospitalization was associated with higher mortality [HR: 3.4 (1.2–9.9); Das Adhikari et al., 2021]. Similarly, another study showed that small-bowel thickening on CT was strongly associated with ICU admission (Wölfel et al., 2020). This relationship did not hold for colon or rectal thickening. These data indicates that small-bowel damage contributes to poor outcomes. As the main organ for nutrient absorption, damage to the small intestine will result in malabsorption and malnutrition, both of which commonly occur in COVID-19 patients (Di Filippo et al., 2021; Lv et al., 2021) and are associated with disease severity (Luo et al., 2020; Zhang P. et al., 2021). A fecal metabolome study showed that feces of COVID-19 patients were enriched with important nutrients that should be metabolized or absorbed, consistent with malabsorption (Lv et al., 2021). A prospective study showed that 29% of COVID-19 patients (31% of hospitalization patients and 21% of patients quarantined at home) had lost >5% of body weight [median weight loss, 6.5 (5.0–9.0) kg or 8.1 (6.1–10.9) %; Di Filippo et al., 2021]. Those patients with weight loss had greater systemic inflammation, impaired renal function and longer disease duration. A large, multicenter study (including 3,229 patients with GI symptoms) showed that 23% of patients had malnutrition, of whom 56.4% were unable to gain weight after 6 months follow-up (Rizvi et al., 2021). Studies also showed that malnutrition was associated with higher incidences of acute respiratory distress syndrome, acute myocardial injury, secondary infection, shock, and 28-day ICU mortality (Luo et al., 2020; Zhang P. et al., 2021). Overall, malabsorption and malnutrition due to damaged small intestine increased disease severity and mortality.

Nutrient absorption in the small intestine is mainly through ATP-dependent active transport. Intestinal infection, hypoxemia, and intestinal ischemia contribute to malabsorption. SARS-CoV-2 adhesion depletes ACE2 levels on intestinal epithelial cells, which alters the expression of the neutral amino acid transporter B0AT1, reducing the intake of tryptophan and the production of nicotinamide (D’Amico et al., 2020). Meanwhile, uncontrolled viral replication consumes large amounts of ATP and nutrients, resulting in decreased nutrients entering the bloodstream. More importantly, anaerobic glycolysis caused by hypoxemia and intestinal ischemia significantly decreases ATP and active transport, leading to malabsorption. Additionally, hypoxemia and intestinal ischemia can also cause anorexia, nausea, vomiting, and enteral nutrition intolerance, reducing food intake. A prospective multicenter study showed that reduced food intake was associated with higher ICU admission and mortality (Caccialanza et al., 2021).

Intestinal ischemia is a common manifestation in COVID-19 patients. Autopsy results have shown that 31.6% of deceased patients had focal ischemic intestinal changes (Chiu et al., 2020). In a separate imaging study, bowel wall thickening and pneumatosis intestinalis, which indicate intestinal ischemia, were found on 38.1% (16 of 42) of abdominal CT images (Bhayana et al., 2020). Of these, 4 (9.5%) patients with pneumatosis intestinalis developed severe intestinal necrosis and needed resection. In another cohort study, 55.8% (58/104) of ICU patients developed an ileus (Kaafarani et al., 2020). Although mechanical factors cannot be ruled out, insufficient intestinal motility due to intestinal ischemia was more likely to be the cause of ileus in COVID-19 patients. In these patients with ileus, 4 (3.8%) developed severe intestinal ischemia and require emergency surgery. Both studies found microthrombi in these resected intestinal samples, which were the main cause of intestinal ischemia and increased mortality.

Additional intestinal ischemia and necrosis follows the formation of mesenteric thrombosis. However, there is currently relatively little data of mesenteric thrombus in COVID-19. Therefore, we have summarized the characteristics of 40 patients in 39 case reports published on PubMed (Supplementary Table 1). The median age of these patients was 50 (20–82) years, 26 (65%) were male, 38 (95%) developed bowel ischemia or necrosis, 30 (75%) needed bowel resection, 7 (17.5%) required no surgery, at least 3 (7.5%) developed sepsis, and 13 (32.5%) died. Other abdominal thrombotic events (such as celiac aortic thrombosis) leading to mesenteric ischemia can also result in severe intestinal necrosis and require intestinal resection (Zamboni et al., 2021).

Mild intestinal ischemia can lead to reduced diet and malabsorption. Severe intestinal ischemia or necrosis leads to the dissemination of gut bacteria, endotoxins, and microbial metabolites into the blood (Figure 2 bottom), aggravating hyperinflammation and the hypercoagulability state. Such patients need emergency excision of the necrotic bowel, which significantly increases mortality.

Long-term GI complications are common in recovering COVID-19 patients. In one systematic review of post-acute COVID-19 manifestations, diarrhea was among the top 10 most common complaints, with a prevalence of 6%. Other long-term GI symptoms include nausea, vomiting, abdominal pain, loss of appetite, and weight loss (Aiyegbusi et al., 2021; Huang et al., 2021). The exact mechanisms of the GI sequelae remain unclear. Recently, persistent endotheliopathy, higher levels of thrombin (Fogarty et al., 2021), and residual SARS-CoV-2 viral antigens in the GI tract (Cheung et al., 2022) were described in convalescent COVID-19 patients. These data suggest that prolonged intestinal infection, persistent endothelial injury (abnormal intestinal–blood barrier), and microthrombi could be causes of the persistent GI symptoms.

Vaccines and antithrombotic therapy are effective measures to reduce intestinal damage and fight against the COVID-19 pandemic (Baden et al., 2021; Chalmers et al., 2021). Vaccines induce adaptive immunity to clear the virus, reducing intestinal infection and intestinal damage. However, the usefulness of vaccines is limited by incomplete vaccine acceptance and viral mutations (Hacisuleyman et al., 2021; Wang et al., 2021). Vaccines are also ineffective for already infected patients. Therefore, more attention should be paid to antithrombotic therapy. Studies had shown that thrombotic events mainly occurred within 7 days of COVID-19 diagnosis (both inpatients and outpatients; Mouhat et al., 2020; Ho et al., 2021). Meanwhile, two large randomized controlled trials (RCTs) from the same platform showed that therapeutic anticoagulation reduced mortality in moderate cases but not in severe ones, suggesting that delayed anticoagulant therapy may lead to treatment failure (REMAP-CAP Investigators et al., 2021a,b). More importantly, a recent study reported three asymptomatic COVID-19 patients who developed abdominal (or intestinal) thrombosis leading to intestinal necrosis (Zamboni et al., 2021). All these data suggest that antithrombotic therapy should be initiated once COVID-19 is diagnosed (excluding patients with contraindications). Early prophylactic antithrombotic therapy can reduce the activation of vascular ECs and blood cells, preventing intestinal thrombosis, ensuring sufficient intestinal perfusion, maintaining the normal gut–blood barrier, avoiding malabsorption, malnutrition, and intestinal flora entering the bloodstream. Further, attenuated injury and decreased microthrombi in convalescent patients may lower the risk of long-term GI sequelae. Meanwhile, unobstructed systemic circulation can also accelerate the removal of SARS-CoV-2, inflammatory cytokines and damaged blood cells by the mononuclear phagocyte system.

During COVID-19 disease progression, SARS-CoV-2 infiltrates the blood stream from the initial respiratory tract infection, causing viremia, hyperactivated platelets and PS storm. The virus settles into the vascular beds of extrapulmonary organs, ultimately causing infection of intestinal epithelial cell. Damaged ECs, combined with hyperactivated platelets and PS storm, promote intestinal thrombosis, resulting in intestinal ischemia or necrosis. The damaged gut–blood barrier leads to malabsorption, malnutrition and intestinal flora entering the bloodstream, which significantly increase disease severity and mortality. Prolonged intestinal infection, persistent endothelial injury and microthrombi contribute to the long-term GI sequelae after discharge. Early prophylactic antithrombotic therapy can prevent microthrombi, ensuring sufficient intestinal perfusion, maintaining the normal intestinal function, and reducing the risk of long-term GI sequelae. More active antithrombotic therapy should be adopted in patients with other thrombotic risk factors or co-morbidities. Even in vaccinated COVID-19 patients, antithrombotic therapy is also important to decrease (intestinal) thrombosis, mortality and the risk of long-term GI sequelae.

With the Omicron pandemic, patients requiring hospitalization and ICU treatment decline rapidly. However, people are increasingly concerned about Long Covid. In terms of long-term GI sequelae, the detailed mechanisms of prolonged intestinal infection and persistent microthrombi remain unclear. And whether anticoagulant therapy can decrease GI symptoms in patients with long-term GI sequelae deserves further study. Finally, the impact of vaccines on long-term GI sequelae remains unclear in previously infected and breakthrough infected patients.

XW, TJ, and JS designed the research. XW and JS analyzed the data and wrote the primary manuscript. HJ, CW, YW, NZ, and TJ contributed to literature retrieval, data acquisition, and analysis of data. VN analyzed the data and revised the primary manuscript. JS and VN discussed and confirmed the final manuscript. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The reviewer (UDA) declared a shared affiliation with the authors (VN and JS) to the handling editor at the time of review.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.