If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Hepatic steatosis, gallbladder distension and sludge are common imaging findings in COVID-19 hospitalized patients.
Acute kidney injury is seen in about one third of hospitalized COVID-19 patients with increased or heterogeneous renal echogenicity seen on ultrasound and preserved cortical thickness.
Splenic and renal infarcts are reported in up to 5% on CT scans of COVID-19 patients.
Ileus is the most commonly reported bowel abnormality in hospitalized COVID-19 patients.
Children with multisystem inflammatory syndrome most commonly present with gastrointestinal symptoms and less frequently with respiratory symptoms.
The coronavirus disease 2019 (COVID -19) pandemic caused by the novel severe acute respiratory syndrome coronavirus (SARS-CoV-2) has affected almost every country in the world, resulting in severe morbidity, mortality and economic hardship, and altering the landscape of healthcare forever. Although primarily a pulmonary illness, it can affect multiple organ systems throughout the body, sometimes with devastating complications and long-term sequelae. As we move into the second year of this pandemic, a better understanding of the pathophysiology of the virus and the varied imaging findings of COVID-19 in the involved organs is crucial to better manage this complex multi-organ disease and to help improve overall survival. This manuscript provides a comprehensive overview of the pathophysiology of the virus along with a detailed and systematic imaging review of the extra-thoracic manifestation of COVID-19 with the exception of unique cardiothoracic features associated with multisystem inflammatory syndrome in children (MIS-C). In Part I, extra-thoracic manifestations of COVID-19 in the abdomen in adults and features of MIS-C will be reviewed. In Part II, manifestations of COVID-19 in the musculoskeletal, central nervous and vascular systems will be reviewed.
The coronavirus 2019 disease (COVID-19), which originated in Wuhan, China, has quickly become a global pandemic, bringing normal life to a standstill in almost all countries around the world. The severe acute respiratory syndrome coronavirus (SARS-CoV-2) is a novel virus preceded by two other recent coronavirus infections, the severe acute respiratory syndrome coronavirus (SARS-CoV-1) and the Middle Eastern respiratory syndrome coronavirus (MERS–CoV), but it has more far-reaching and devastating consequences. As of March 2021, the COVID-19 pandemic has resulted in over 29 million cases in the United States and over 121 million cases globally. As of April 2021, it is responsible for the deaths of over half a million people in the United States and more than 2 ½ million worldwide [
]. As the disease has evolved over the past year, so has our understanding of the virus, including its pathophysiology, clinical presentation and imaging manifestations. Although COVID-19 is predominately a pulmonary illness, it is now established to have widespread extra-pulmonary involvement affecting multiple organ systems. The SARS-CoV-2 has a highly virulent spike protein which binds efficiently to the angiotensin converting enzyme 2 (ACE2) receptors which are expressed in many organs, including the airways, lung parenchyma, several organs in the abdomen, particularly the kidneys and GI system, central nervous system and the smooth and skeletal muscles of the body [
This article provides a comprehensive review of the pathophysiology and imaging findings of the extra-thoracic manifestations of COVID-19 with the exception of unique cardiothoracic features associated with multisystem inflammatory syndrome in children (MIS-C). In Part I, extra-thoracic manifestations of COVID-19 in the abdomen in adults and the varying features of multisystem inflammatory syndrome in children will be reviewed, with imaging findings summarized in Table 1, Table 2. In Part II, manifestations of COVID-19 in the musculoskeletal system, the central nervous system and central and peripheral vascular systems will be reviewed.
Table 1Summary of abdominal imaging findings in COVID-19 in adults.
Increased or coarsened echogenicity on US
Hypoattenuation on non-contrast or contrast enhanced CT
Periportal edema and heterogeneous enhancement on CT
Loss of signal on opposed-phase sequences on MRI
Portal vein thrombus
Features of acute interstitial pancreatitis
Biliary ductal dilatation
Increased or heterogeneous parenchymal echogenicity on US
Loss of corticomedullary differentiation on US
Preserved cortical thickness
Perinephric fat stranding and thickening of Gerota's fascia on CT
Wedge shaped perfusion defects on CT or MRI
Thrombus in the renal artery or vein
Bladder wall thickening
Portal vein gas
Acute mesenteric ischemia
Vascular occlusion (superior mesenteric artery, superior mesenteric vein, or portal vein)
Mesenteric fat stranding, ascites
Active gastrointestinal bleeding (duodenal or gastric ulcer) on CTA
Varying derangements of the liver, biliary system, gallbladder, portal vein and pancreas may occur in COVID-19 with hepatic parenchymal injury and biliary stasis reported with highest frequency. The mechanism of involvement of these structures appears to be multifactorial. The most direct form of injury results from SARS CoV-2 entry into host cells by binding to ACE2 receptors detected in several locations in the hepatobiliary system, including biliary epithelial cells (cholangiocytes), gallbladder endothelial cells and both pancreatic islet cells and exocrine glands [
Relevance of SARS-CoV-2 related factors ACE2 and TMPRSS2 expressions in gastrointestinal tissue with pathogenesis of digestive symptoms, diabetes-associated mortality, and disease recurrence in COVID-19 patients.
]. One study excluding COVID-19 patients receiving hepatotoxic drugs, still found patients with liver injury. Therefore, liver damage in COVID-19 patients is likely not entirely drug-induced but may also be due to acute infection [
] and including raised levels of alanine aminotransferase, aspartate aminotransferase, and γ-glutamyl transferase with mild elevation of bilirubin. The majority of cases are mild and self-limited, with severe liver damage rare [
]. Based on a meta-analysis of hepatic autopsy findings of deceased COVID-19 patients in 7 countries, hepatic steatosis (55%), hepatic sinus congestion (35%) and vascular thrombosis (29%) were the most common [
]. In a retrospective study of abdominal imaging findings of 37 COVID-19 patients, 27% who underwent ultrasound had increased hepatic echogenicity considered to represent fatty liver with elevated liver enzymes being the most frequent indication for ultrasound [
]. It should be noted that since obesity is a major risk factor for severe COVID-19 infection, it might contribute to the frequency of steatosis identified on imaging. In another retrospective abdominal sonographic study of 30 ICU patients with COVID-19, the most common finding was hepatomegaly (56%), with most cases having increased hepatic echogenicity and elevated liver function tests [
]. Steatosis was based on a single ROI measurement in the right lobe with an attenuation value ≤ 40 HU. However, underlying risk factors for steatosis such as diabetes, obesity, hypertension and abnormal lipid profile, were not available to exclude preexisting conditions leading to steatosis. Finally, unlike in the spleen and kidney where infarcts are reported in COVID-19, hepatic infarction is not a distinct feature. This is likely due to the liver's unique dual blood supply.
On imaging the liver may be enlarged. On ultrasound, the liver of patients with abnormal liver function tests may be coarsened and/or increased in echogenicity (Fig. 1, Fig. 2). On CT scan, the liver may be hypoattenuated on non-contrast or contrast-enhanced exam due to steatosis (Fig. 3). Periportal edema and heterogeneity of hepatic enhancement may be seen on contrast-enhanced CT or MRI due to parenchymal inflammation. On MRI, loss of signal on opposed-phase sequences (Fig. 4) may be seen due to steatosis and periportal edema may be conspicuous on T2-weighted images or on contrast-enhanced images [
On ultrasound, portal vein thrombosis appears as echogenic material within the vein, although it may be anechoic and undetectable in the early acute stages. Color Doppler images may show total or partial absence of flow in the vein. On contrast-enhanced CT or MRI, portal vein thrombosis appears as non-enhancing filling defect within the main or intrahepatic branches (Fig. 6). Clots may be occlusive or non-occlusive and may expand the involved vein.
2.1.3 Biliary stasis
Biliary stasis is quite common in COVID-19, manifesting as gallbladder distention, gallbladder sludge and/or biliary ductal dilatation on imaging. In a retrospective study of 37 COVID-19 patients who underwent abdominal imaging, gallbladder sludge and distension were reported in 54% of patients who underwent right upper quadrant ultrasound [
]. Acalculous cholecystitis has been reported in patients with COVID-19, although it is unclear if this is due direct viral injury or due to the multiple confounding factors of general sepsis, prolonged parenteral nutrition, or systemic cytokine release [
]. In a large retrospective study of 80 COVID-19 patients with positive findings on abdominal imaging, abnormalities of the gallbladder and biliary system were found in 25% of patients, including gallbladder findings in 15% and biliary ductal dilatation in 10% [
]. Gallbladder findings included distention (6.2%), mural edema (2.5%) and possible or definite acute cholecystitis (5%).
On imaging, biliary stasis may manifest as increased gallbladder distention (typically defined as >/=5 cm) and sludge, and acalculous cholecystitis is suggested when there is superimposed gallbladder mural thickening and pericholecystic fluid [
] (Fig. 7). Prolonged biliary stasis may obstruct the biliary tree and result in cholangitis. This is demonstrated on imaging as extrahepatic or intrahepatic biliary ductal dilatation, and increased biliary ductal mural thickening and enhancement (Fig. 8). Gallstones should be excluded on ultrasound, as choledocholithiasis is a more common cause of biliary infection.
The management of acute cholecystitis in COVID-19 patients is controversial. Intravenous antibiotics alone or with percutaneous cholecystostomy are advocated by many as a bridging agent to cholecystectomy [
Imaging plays a role in the diagnosis of cholecystitis as well as in the in detection of complications of percutaneous cholecystostomy. These complications include placement in the wrong site, catheter dislodgement, intraabdominal hemorrhage or bile leak. Fistulae to the liver can develop when catheters are left in for extended periods of time. In order to avoid potential cholecystostomy risks, some institutions offer surgery in COVID-19 patients as first-line therapy [
Pancreatic involvement in COVID-19 is much less common than hepatic or biliary involvement, and may occur from direct viral injury or secondary to general inflammation and systemic illness. Pancreatic islet cells contain ACE2 receptors that may permit direct viral entry, while the host immune-mediated cytokine response can also incite pancreatic damage. In a study of 52 patients hospitalized for COVID-19 pneumonia, 9 (17%) had pancreatic injury defined as elevation of amylase or lipase with only one requiring mechanical ventilation. None had clinically severe pancreatitis [
]. Only 1–2% of patients with mild viral illness had elevated pancreatic enzymes while almost 18% of patients with severe illness had elevated pancreatic enzymes. On CT scan, no patients had evidence of necrosis and only 7.5% (5 of 67) patients with severe illness had imaging evidence of pancreatitis described mainly as focal pancreatic enlargement or dilatation of the pancreatic duct. Although reported cases of pancreatic injury in COVID-19 tend to be mild, the role of pancreatitis in aggravating systemic inflammation, contributing to ARDS or leading to potential chronic pancreatitis later on remains unknown [
Imaging findings of pancreatitis in COVID-19 patients mimic those in other settings, with CT demonstrating diffuse or focal pancreatic enlargement, decreased pancreatic attenuation due to edema, surrounding fat stranding and indistinct gland margins. MRI shows similar manifestations, with the gland appearing enlarged with poorly defined margins, and with edema and free fluid appearing bright on T2 weighted images (Fig. 10, Fig. 11).
2.2 Genitourinary derangement
2.2.1 Acute kidney injury
Acute kidney injury (AKI) is a common complication of SARS-CoV-2 infection, seen in about one third of hospitalized COVID-19 patients [
]. Patients with AKI have a higher associated morbidity and mortality with one meta-analysis showing 77% of patients developing a more severe infection, 5% requiring renal replacement therapy and having an overall increased mortality rate of 50% [
]. Among the patients who developed AKI and survived, only a third recovered renal function at the time of discharge, and about a third of patients failed to recover to baseline even on subsequent follow-up [
]. In addition, patients with chronic renal disease are at a higher risk of developing severe upper respiratory infection and pneumonia following exposure to COVID-19, secondary to their inherent proinflammatory state and diminished immunity [
The mechanism of kidney injury in COVID-19 is likely multifactorial. Both the increased expression of ACE2 receptors in the kidneys and certain genetic traits and polymorphism of renal ACE2 receptors enable SARS-CoV-2 binding and viral cell entry [
]. The pathogenesis of AKI involves both direct and indirect effects. The direct effect of the virus causes endothelial damage, inflammatory response, coagulopathy and complement activation while the indirect effects are sequelae of the systemic effects of multisystem involvement including hypovolemia, critical care interventions and organ crosstalk [
]. Organ crosstalk is a complex communication network between distant organs that communicate by signaling factors such as cytokines, growth factors and release of damage associated molecular patterns (DAMPs) from injured tissue. Histologic evaluation of COVID-19 patients has revealed varying types of injury including collapsing glomerulopathy, proximal tubular injury, microthrombi and microangiopathy [
] [Fig. 12]. Involved kidneys typically have preserved cortical thickness, which helps differentiate them from those with chronic kidney disease. Additionally, a combination of nonspecific findings including increased renal parenchymal echogenicity, loss of corticomedullary differentiation, increased resistive indices and decreased color Doppler flow can be seen in COVID-19 patients with collapsing focal segmental glomerulosclerosis [
]. On non-contrast CT, perinephric fat stranding and thickening of Gerota's fascia correlated with higher serum creatinine levels in patients with AKI, suggesting more severe inflammation and parenchymal injury [
Increased urinary frequency has been reported in COVID-19 patients. It is believed that expression of ACE2 receptors in the urinary bladder, although not as high as the kidneys, predisposes patients to interstitial or hemorrhagic cystitis. It is unclear whether the receptors are expressed on the basal or luminal bladder surface. However, infection arises both from urine along the luminal surface and from blood via the basal surface [
]. The mechanism of organ infarction in COVID-19 is likely multifactorial. There is a direct cytopathic effect of SARS-CoV-2 which results in a viral mediated platelet-dependent endothelial inflammation causing endothelial dysfunction and rise in proinflammatory cytokines, leading to an increase in tissue factor expression [
]. Additionally, hypercoagulability from increased levels of coagulation factors and antiphospholipid antibodies along with decreased levels of anticoagulation proteins contribute to the increased incidence of thrombotic events in COVID-19 patients [
]. Three splenic and one renal infarct were identified with patent vasculature on contrast-enhanced CT. One additional case of renal vein thrombus was reported without reported renal infarct. This supports disseminated microvascular thrombosis in the etiology of solid organ infarction [
]. On CT and MRI, renal infarcts may appear as geographic focal or multifocal areas of diminished or absent parenchymal enhancement in one or both kidneys. The renal defects are typically wedge-shaped spanning the cortex and medulla with their apex towards the hilum and extending to the renal capsule. Several days after acute onset they typically have overlying thin capsular enhancement (“rim sign”) due to preserved collateral capsular perfusion. Segmental infarcts have a characteristic geographic appearance that follows the distribution of the anterior and posterior branch vessels of the renal artery [
]. On MRI, infarcts vary with age and are usually hypointense on T1-weighted and T2-weighted images acutely, becoming higher in signal on T2-weighted images over the next few days as coagulation necrosis occurs. They may appear hyperintense on T1-weighted images if hemorrhage occurs. Healed or chronic infarcts have progressively decreased signal due to fibrosis and atrophy. There is often loss of normal corticomedullary differentiation in the infarcted segment and corresponding wedge-shaped areas of hypoennhancement on post contrast images [
]. Less commonly, a filling defect may be seen in the renal artery on post-contrast images [Fig. 15] or vein [Fig. 16].
2.3.2 Splenic infarction
Although relatively uncommon, splenic infarcts have been reported in COVID-19 patients. Based on a retrospective study of 141 COVID-19 patients with abdominal symptoms, the incidence of splenic infarct was 5% on CT scans [
]. One patient had both multifocal splenic infarcts and unilateral focal renal infarct and the other had small bowel necrosis and massive splenic infarction requiring surgery. Asymptomatic splenic infarct incidentally noted on CT scan performed for suspected pulmonary embolism has also been reported [
On CT and MRI, splenic infarcts appear as single or multiple peripheral wedge-shaped hypoenhancing defects, with their apex pointed towards the hilum [Fig. 17]. They can also be rounded or linear in shape. Their appearance varies with age, being vague hypodense areas on CT with mottled enhancement in the first 24 h, and becoming progressively better defined non-enhancing hypodense defects over the next week [
]. On MRI, splenic infarcts are typically hypointense on T1- and T2-weighted images without enhancement on the post contrast images. However, their appearance varies based on the presence of hemorrhage and with their age. Acute and hemorrhagic infarcts are hyperintense on T1-weighted images and hypointense on T2-weighted images. Subacute infarcts have fluid signal. Chronic infarcts may fully resolve or become hypointense on T2 and T2-weighted images [
]. Less commonly, a filling defect may be seen in the splenic vein or artery on post-contrast images [Fig. 18].
2.4 Gastrointestinal derangement
The mechanism of SARS-CoV-2 induced gastrointestinal symptoms is not fully understood and may in part be due to direct viral entry into enterocytes. A single-cell transcriptomic study of enteric and lung cells determined that co-expression of ACE2 and TMPRSS2 cell receptors is necessary to permit SARS-CoV-2 entry. Co-expression was found not only in the lung but also in the esophageal upper epithelial and gland cells, the ileum and colon [
Gastrointestinal symptoms in COVID-19 are relatively frequent. In a large meta-analysis of 43 studies of gastrointestinal manifestations of COVID-19 including over 18,000 patients, diarrhea was the most common GI symptom (11.5%), followed by nausea and vomiting (6.3%), and abdominal pain (2.3%) [
]. Among symptomatic patients, 17.5% had severe COVID-19 illness and 9.8% had a milder illness. In a retrospective study of 141 patients with severe COVID-19 in the ICU, 45% had gastrointestinal symptoms on admission including abdominal pain, diarrhea and vomiting [
Gastrointestinal symptoms in COVID-19 may be due to hepatobiliary or enteric involvement and, in some, may be due to side effects of drug therapy or advanced sepsis. In addition, when enteric involvement occurs, the mechanism of bowel disease remains unclear with postulations including cytokine storm and inflammation, edema, or ischemia [
]. Further studies are needed to determine if there is a direct correlation between viral entry into enterocytes and imaging findings of enteritis or colitis, as mild cases are managed without operative intervention or definitive tissue diagnosis. The origin of bowel involvement in mild cases is therefore unknown and may be infectious, ischemic or due to an alternative cause.
Finally, it has also been reported that many patients with abdominal complaints lack abdominal imaging findings. In a study of abdominal CT scans of 23 patients whose lung bases had findings typical of COVID-19 and presented with abdominal complaints, there were no bowel abnormalities detected [
]. Of those who lacked abdominal findings in the study, 64% had basilar lung findings on CT. In another study of 43 hospitalized COVID-19 patients who underwent CT for abdominal complaints, 63% had no abdominal abnormality [
]. Therefore, it is not clear how often COVID-19 patients have relevant bowel involvement when symptomatic. Since abdominal symptoms are common, it can be postulated that some cases are due to nonspecific viral-induced pain or referred symptoms from basilar pneumonia.
2.4.1 Bowel imaging patterns of COVID-19
The most common and consistently reported imaging manifestations of COVID-19 affecting the bowel include mural thickening, non-specific ileus, fluid-filled colon, and pneumatosis intestinalis. Portal vein gas and pneumoperitoneum are less commonly seen [
]. In a retrospective review of 141 COVID-19 patients in the ICU, 56% had clinically or radiologically diagnosed ileus, 2% had Ogilvie-like syndrome, 4% had bowel ischemia (3% small bowel, 1% small and large bowel and 1% cecal), 11% had GI bleeding and 4% had Clostridium difficile colitis [
]. Out of 81 hospitalized COVID-19 patients who underwent abdominal CT, 24% had intestinal findings of which colorectal thickening (5%), small bowel thickening (12%), and ileus (18%) were the most common, while pneumatosis (1%) and perforation (1%) were relatively rare [
]. Small bowel thickening was reported in 12%, all of who were in the ICU. Colonic or rectal thickening was reported in 17%, of which nearly half were in the ICU. In another larger retrospective study of hospitalized COVID-19 patients who underwent CT for abdominal symptoms, 57% had positive abdominal imaging findings [
]. Bowel wall thickening was found at a lower rate of 15%, including mostly colonic and small bowel, with rare cases of esophageal and gastric thickening. An even smaller rate of bowel abnormalities was reported in a retrospective study of 32 hospitalized COVID-19 patients who underwent CT, where only 7% had nonspecific enteritis, with pneumatosis intestinalis in one patient [
Finally, Horvat et al. found the presence of intestinal findings on CT in hospitalized COVID-19 patients positively correlated with a higher risk of worse clinical outcome including death or assisted ventilation [
Ileus can manifest as gaseous distention of the small bowel, colon or both and is relatively common in hospitalized COVID-19 patients (Fig. 19). Kaafarani et al. reported that 56% of patients in their case series of 141 patients with severe COVID-19 requiring ICU care had a clinical or radiographic diagnosis of ileus [
]. The pathophysiology of ileus in COVID-19 is likely multifactorial. In severely ill COVID-19 patients, ileus is commonly caused by metabolic and/or electrolyte derangements but may also be reactive to other complications of COVID-19 [
Acute mesenteric ischemia is a serious and not uncommon complication of COVID-19 infection with a high mortality. It is theorized to occur secondary to viral-induced hypercoagulability or thromboinflammation leading to thrombosis, as opposed to direct viral infection of the gastrointenstinal tract [
]. If COVID-19 patients present with suspicious symptoms, there should be a low threshold to initiate a search for acute mesenteric ischemia with CT angiogram of the abdomen and pelvis.
Acute mesenteric ischemia can be caused by thromboembolic arterial or venous occlusion of medium or large mesenteric vessels or by microthrombosis of small vessels. A systematic literature review of studies reporting mesenteric ischemia in COVID-19 patients who had at least one gastrointestinal finding on imaging, defined as mesenteric arterial or venous thromboembolism or signs of small bowel ischemia, yielded 22 studies and 31 patients [
]. Nine (29%) had superior mesenteric arterial thromboembolism, 6 (19%) had occlusive thrombosis of the portal system or superior mesenteric vein and 16 (52%) had no visible vascular occlusion. More than two-thirds (64%) of all patients required laparotomy and bowel resection. The overall mortality was 39%. In this meta-analysis, almost half of patients with bowel ischemia had macrovascular arterial/venous thrombosis and half did not.
Imaging findings of COVID-19 related bowel ischemia on CT angiogram of the abdomen and pelvis are not unique and include mural thickening, a targetoid mural enhancement pattern, absent or diminished mural enhancement, mural hyperenhancement, ileus [Fig. 20], mesenteric vascular engorgement, mesenteric fat stranding and ascites. Pneumatosis and/or portomesenteric venous gas may reflect ischemia or infarction, but when seen together typically indicate infarction [
] [Fig. 21]. Arterial filling defects or zones of vascular narrowing may occur in arterial ischemia with preservation or thinning of bowel wall thickness. Venous filling defects may occur in venous ischemia typically with associated mesenteric venous engorgement, mural thickening and ascites. In nonocclusive mesenteric ischemia, bowel wall may be normal or thickened due to reperfusion. Mesenteric fat stranding and ascites are also common [
A few CT imaging reports of mesenteric ischemia in COVID-19 with confirmed bowel necrosis on surgical intervention have been reported. These include one case of non-enhancing distal ileal loops with adjacent free air and patent mesenteric vessels [
Finally, pneumatosis or portal venous gas is reported in COVID-19 patients with bowel ischemia in up to 20% (4 of 20) of ICU patients, and is typically considered a marker for possible mesenteric ischemia [
]. However, in the absence of clinical symptomatology, like in non-COVID-19 patients, it may not on its own indicate ischemia or mandate surgical intervention. Meini et al. report a patient hospitalized for COVID-19 pneumonia that was successfully managed conservatively for asymptomatic intraperitoneal air bubbles and pneumatosis of the cecum and ascending colon noticed on chest CT performed for pulmonary embolism [
]. The etiology was theorized to be bowel wall injury and gut flora impairment during SARS-CoV-2 infection.
2.4.4 Ischemic colitis
Ischemic colitis has been reported rarely in patients hospitalized with COVID-19 illness with bloody diarrhea. In one case it was attributed to shock and hemodynamic compromise triggered by small vessel vasoconstriction and mesenteric hypoperfusion [
Pneumoperitoneum in hospitalized COVID-19 patients has many causes including recent surgery, iatrogenic procedural trauma, bowel perforation, ischemia, and benign causes such as barotrauma in ventilated patients. Benign pneumomediastinum and pneumoperitoneum can be due to pulmonary interstitial emphysema, known as the Macklin effect [
Malignant interstitial emphysema of the lungs and mediastinum as an important occult complication in many respiratory diseases and other con- ditions: interpretation of the clinical literature in the light of laboratory experiment.
]. In the setting of barotrauma, typically due to invasive mechanical ventilation, alveoli can rupture leading to small pockets of air tracking along the peribronchovascular sheaths into the mediastinum. Once air collects within the mediastinum, it can dissect into the peritoneal cavity, causing pneumoperitoneum via major diaphragmatic portals [Fig. 22]. Alternatively, even if pneumomediastinum is absent, air can track directly through pleural or diaphragmatic defects leading to isolated pneumoperitoneum [
]. Duarte et al. report a case of pneumoperitoneum in a hospitalized COVID-19 patient who developed pulmonary interstitial edema from nasal cannula oxygen supplementation leading to benign pneumoperitoneum [
When unexpected pneumoperitoneum is detected in a COVID-19 patient on chest radiograph or abdominal imaging, it is essential to rule out potential bowel perforation or ischemia. If no cause is determined on abdominal CT, the patient should be observed closely to ensure bowel perforation or ischemia was not missed, and that the pneumoperitoneum is regressing. Given the high morbidity and mortality associated with exploratory laparotomy in hospitalized COVID-19 patients, there should be a high threshold for surgical intervention [
]. In certain situations, if a chest CT is performed, the Macklin effect can be confirmed by visualizing thin linear collections of air tracking along the bronchovascular sheath.
2.4.6 GI bleeding
Although the exact incidence of GI bleeding in COVID-19 is unknown, GI bleeding is not uncommonly observed in hospitalized COVID-19 patients, and in some cases may be the primary indication for abdominopelvic CT scan. In one retrospective study, 5% of hospitalized COVID-19 patients had GI bleeding as the indication for abdominal CT [
]. Patients may also present with symptoms of perforated gastric or duodenal ulcers, which has been attributed to stress-related mucosal damage in the setting of severe illness or anticoagulation, rather than direct infection or ischemia of the gastrointestinal tract [
]. In a large case series of 4871 patients in Northern Italy, 23 patients developed upper GI bleeds (0.5%), 78% of whom were on anticoagulation at the time bleeding, including 44% on full dose anticoagulation [
]. While most patients in the case series developed bleeding during their hospital course, 26% of patients initially presented to the emergency room for GI bleeding and then tested positive for COVID-19 infection, suggesting that in some cases bleeding may be directly due to COVID-19 infection rather than stress-related mucosal damage or anticoagulation [
A multicenter international study of 114 endoscopies performed on COVID-19 patients for acute GI bleeds, found that 25% of bleeds were caused by ulcers, 16% by erosive/ulcerative gastro-duodenopathy and 9% by petechial or hemorrhagic gastropathy [
Dedicated CT angiogram with a GI bleeding protocol, including non-contrast, arterial and portal venous phase images can detect foci of active bleeding at a rate of 0.3–0.5 ml/min which is less sensitive than RBC scanning (0.1–0.5 ml/min) but more sensitive than fluoroscopic angiography (0.5–1.0 ml/min) [
] (Artigas). In patients with intermittent bleeding or in stable patients with elevated creatinine, a technetium 99 m (99mTc) nuclear medicine scan with a sulfur colloid or red blood cell pharmaceutical agent may be used as an alternative imaging modality. 99mTc scintigraphy has a high sensitivity but poor anatomic localization of acute GI bleed [
]. CT angiogram has the added benefit of detecting free air prompting a search for perforated gastric or duodenal ulcer as a source of GI bleed [Fig. 23, Fig. 24].
2.4.7 Pseudomembranous colitis
Clostridium difficile colitis is frequently seen in severely ill COVID-19 patients after treatment with broad-spectrum antibiotics, which were given frequently early on in the pandemic and are still routinely given to COVID-19 patients with severe sepsis. In addition to antibiotic use, direct infection of enterocytes by the SARS-CoV-2 virus may disrupt the gut microbiome, increasing a patient's susceptibility to gastrointestinal infection, including Clostridium difficile [
]. Elderly patients, especially those living in long-term care facilities, have both an increased risk of severe COVID-19 illness and baseline elevated rates of Clostridum difficile colonization. This confounds our understanding of the true impact of COVID-19 infection on the development of Clostridium difficile colitis [
]. In a retrospective clinical laboratory review of hospitalized patients with Clostridium difficile infection, 9 patients co-infected with COVID-19 were found, all of whom had severe COVID-19 illness and developed symptomatic Clostridium difficile colitis during their hospital course, all related to antibiotic administration [
Pseudomembranous colitis should be suggested whenever abdominopelvic CT scan shows a pattern of long segment colitis [Fig. 25] most commonly involving the rectosigmoid but which can be also be segmental involving the right or transverse colon or pan-colonic in extent. It is characterized by a more extensive pattern of circumferential mural thickening than other causes of colitis, with low attenuation intramural edema, and associated features including the “accordion sign,” the “target sign,” pericolonic stranding and ascites. Mural thickening is marked with a reported mean of 15 mm (range 3–32 mm), and is often irregular and shaggy [
]. Although the majority of symptomatic children with COVID-19 have a respiratory illness, gastrointestinal symptoms have been reported in 33–35% of cases, with vomiting, diarrhea and abdominal most common [
]. This review will focus on the multifaceted post-infectious clinical and imaging manifestations of COVID-19 unique to children, coined multisystem inflammatory syndrome in children (MIS-C).
Although infected children tended to be only mildly symptomatic during the early months of the pandemic, by April and May of 2020, several medical centers in Europe and North America reported children with a multiorgan Kawasaki disease-like syndrome, consisting of fever and mucocutaneous rash, sometimes with features of toxic shock syndrome [
]. This syndrome was previously given various overlapping definitions, however as more cases subsequently arose around the globe, the Centers for Disease Control and Prevention (CDC) and the World Health Organization (WHO) have given this entity the preferred label of “Multisystem Inflammatory Syndrome in Children” (MIS-C) [
]. Although SARS-CoV-2 is not definitively known to cause MIS-C, temporal occurrences to outbreaks in Europe and America are highly suggestive of a link, and some cases even occur in the presence of an acute respiratory infection [
]. Some MIS-C patients test positive for SARS-CoV-2 IgM and IgG antibodies, rather than testing positive for the virus on RT-PCR, and a minority of patients have recorded COVID-19 symptoms a few weeks prior to onset of MIS-C [
]. Kawasaki disease has lower rates of myocardial involvement than MIS-C, and is more frequently associated with coronary artery aneurysms. Kawasaki disease typically presents before 5 years of age, with the highest incidence of disease in Asia. In contrast, MIS-C usually occurs in older children and adolescents, and has not been reported in China or Japan.
Treatment of MIS-C is both supportive and directed, consisting of fluid resuscitation, vasopressors and ionotropic support, anticoagulation, intravenous immune globulin (IVIG), glucocorticoids, and other immunomodulatory and antiviral drugs [
]. Critical cases may require mechanical ventilation or extracorporeal membrane oxygenation (ECMO). Despite the severity of MIS-C at initial presentation, short-term outcomes are favorable, with the majority of patients recovering between a few days up to two weeks [
The diagnosis of MIS-C is made on clinical grounds, and imaging in not routinely indicated. However, imaging may be requested to exclude other acute pathologies, and imaging abnormalities are associated with fulminant illness [
]. Although this review is focused on the extra-pulmonary manifestations of COVID-19, the cardiothoracic imaging findings of MIS-C will be included because the majority of MIS-C patients have pulmonary findings that are slightly different than those of routine COVID-19 pneumonia and may have associated cardiac sequelae.
3.1 Thoracic manifestations of MIS-C
The most common respiratory complaint in MIS-C is tachypnea, although some patients may progress to respiratory failure requiring mechanical ventilation, likely secondary to shock and cardiac dysfunction [
]. While initial chest radiographs can be normal, as the disease progresses, common radiographic findings include bilateral diffuse airspace opacities with basilar predominance, peribronchial thickening or interstitial opacities, bilateral pleural effusions, and cardiomegaly [
] (Fig. 26, Fig. 27). The pulmonary opacities tend to be hazy and symmetric, suggesting that they reflect pulmonary edema, ARDS, and/or third spacing. These phenomena may be secondary to cardiac dysfunction, the hyperinflammatory status of the patient, or possibly even the sequelae of aggressive fluid resuscitation [
]. Although chest radiograph findings in MIS-C can overlap with acute COVID-19 pneumonia, COVID usually presents as non-diffuse peripheral or subpleural opacities without pleural effusions and MIS-C usually presents as diffuse opacities with lower lobe predominance with pleural effusions [
]. The most common chest CT findings are bibasilar consolidation with atelectasis and bilateral pleural effusions, with diffuse ground-glass opacity, septal thickening, and mild hilar lymphadenopathy less common [
Echocardiography is routinely performed at baseline and in follow-up of patients with MIS-C. The most common finding is depressed left ventricular ejection fraction, with other findings including valvular regurgitation and small pericardial effusion [
]. Data on thrombotic complications in pediatric COVID-19 and MIS-C is limited, likely due to its low incidence, and currently published guidelines for thromboprophylaxis in children are extrapolated from the adult population. The highest reported rate of symptomatic imaging-confirmed thromboembolic disease (deep venous thrombosis or pulmonary embolism) in one multicenter registry of MIS-C was 7% [
]. It is thought that children with COVID-19 and/or MIS-C who experience hospital-associated thrombotic events have markedly elevated plasma D-dimer levels, or have one or more superimposed risk factors (e.g. prior history of venous thrombosis, indwelling central line, active malignancy, etc) [
]. Imaging is not necessarily required prior to initiation of therapy if a thromboembolic event is suspected, but can be obtained at the time of discharge, in order to guide the length and intensity of the anticoagulation regimen [
Since patients with MIS-C most commonly present with gastrointestinal symptoms, often mimicking acute appendicitis, abdominal imaging is sometimes obtained even prior to recognition of the MIS-C diagnosis [
Gastrointestinal symptoms as a major presentation component of a novel multisystem inflammatory syndrome in children that is related to Coronavirus Disease 2019: a single center experience of 44 cases.
]. Acute hepatitis, defined as elevated alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels, was found in 43% of subjects in a single-center report, and may account for some of the abdominal ultrasound findings in MIS-C such as hepatomegaly, gallbladder wall thickening, and ascites [
The SARS-CoV-2 virus has many extrathoracic facets in adults with a host of abdominal manifestations increasingly recognized. MIS-C is a unique postinfectious manifestation in children that encompasses gastrointestinal symptoms most commonly, but also includes unique pulmonary and cardiac manifestation different than those of typical COVID-19 respiratory infection. It is important for radiologists to have a better understanding of the pathophysiology of this disease and incidence of its involvement in various organ systems, along with improved familiarity with the broad spectrum of imaging manifestations of COVID-19. This will improve our ability to more accurately search for and diagnose COVID-19 sequelae and hopefully better assist clinicians in managing this complex disease. As the long-term effects of COVID-19 continue to emerge, we must stay vigilant and closely watch and monitor patients for any future unexpected manifestations of this novel COVID-19 disease.
Relevance of SARS-CoV-2 related factors ACE2 and TMPRSS2 expressions in gastrointestinal tissue with pathogenesis of digestive symptoms, diabetes-associated mortality, and disease recurrence in COVID-19 patients.