The coronavirus SARS-CoV-2 (severe acute respiratory syndrome coronavirus 2), which is responsible for the disease COVID-19 (coronavirus disease 2019), has infected over 9.5 million people and has caused more than 480,000 deaths globally, as of 24 June 2020 (ref. 1). While SARS-CoV-2 is known to cause substantial pulmonary disease, including pneumonia and acute respiratory distress syndrome (ARDS), clinicians have observed many extrapulmonary manifestations of COVID-19. Our clinical experience and the emerging literature suggest that the hematologic, cardiovascular, renal, gastrointestinal and hepatobiliary, endocrinologic, neurologic, ophthalmologic, and dermatologic systems can all be affected (Supplementary Table)2,3,4,5,6. This pathology may reflect either extrapulmonary dissemination and replication of SARS-CoV-2, as has been observed for other zoonotic coronaviruses7, or widespread immunopathological sequelae of the disease. To provide a perspective on these extrapulmonary manifestations, we discuss the pathophysiology and clinical impact of COVID-19 on various organ systems, accompanied by insights from our experience at the Columbia University Irving Medical Center in New York City at the epicenter of the pandemic.


SARS-CoV-2 seems to employ mechanisms for receptor recognition similar to those used by prior virulent coronaviruses such as SARS-CoV, the pathogen responsible for the SARS epidemic of 2003 (refs. 8,9,10,11). The coronavirus spike protein facilitates entry of the virus into target cells. The spike subunit of SARS-CoV and that of SARS CoV-2 engage ACE2 (angiotensin-converting enzyme 2) as an entry receptor (Fig. 1). In addition, cell entry requires priming of the spike protein by the cellular serine protease TMPRSS2 or other proteases12. Co-expression on the cell surface of ACE2 and TMPRSS2 is required for the completion of this entry process. In addition, the efficiency with which the virus binds to ACE2 is a key determinant of transmissibility, as shown in studies of SARS-CoV13. Recent studies have demonstrated higher affinity of binding of SARS-CoV-2 to ACE2 than of SARS-CoV to ACE2, which may partially explain the increased transmissibility of SARS-CoV-214,15,16.

Fig. 1: Pathophysiology of COVID-19.
figure 1

SARS-CoV-2 enters host cells through interaction of its spike protein with the entry receptor ACE2 in the presence of TMPRSS2 (far left). Proposed mechanisms for COVID-19 caused by infection with SARS-CoV-2 include (1) direct virus-mediated cell damage; (2) dysregulation of the RAAS as a consequence of downregulation of ACE2 related to viral entry, which leads to decreased cleavage of angiotensin I and angiotensin II; (3) endothelial cell damage and thromboinflammation; and (4) dysregulation of the immune response and hyperinflammation caused by inhibition of interferon signaling by the virus, T cell lymphodepletion, and the production of proinflammatory cytokines, particularly IL-6 and TNFα.

Key mechanisms that may have a role in the pathophysiology of multi-organ injury secondary to infection with SARS-CoV-2 include direct viral toxicity, endothelial cell damage and thromboinflammation, dysregulation of the immune response, and dysregulation of the renin–angiotensin–aldosterone system (RAAS) (Fig. 1). The relative importance of these mechanisms in the pathophysiology of COVID-19 is currently not fully understood. While some of these mechanisms, including ACE2-mediated viral entry and tissue damage, and dysregulation of the RAAS, may be unique to COVID-19, the immune pathogenesis caused by the systemic release of cytokines and the microcirculation dysfunctions may also occur secondary to sepsis17.

Direct viral toxicity

SARS-CoV-2 is transmitted mainly through direct or indirect respiratory-tract exposure. It has tropism for the respiratory tract, given the high expression of ACE2, its entry receptor, in multiple epithelial cell types of the airway, including alveolar epithelial type II cells in the lung parenchyma18,19. Live SARS-CoV-2 virus and viral subgenomic mRNA isolated from the upper airway can successfully be detected by RT-PCR. Later in the disease course, viral replication may occur in the lower respiratory tract20, which manifests in severe cases as pneumonia and ARDS.

Studies evaluating body-site-specific viral replication of SARS-CoV-2 have isolated viral RNA from fecal samples at high titers2,20 and, less commonly, from urine and blood21,22. Histopathological studies have reported organotropism of SARS-CoV-2 beyond the respiratory tract, including tropism to renal21,23, myocardial21,24, neurologic21, pharyngeal21, and gastrointestinal25 tissues. In addition, single-cell RNA-sequencing studies have confirmed expression of ACE2 and TMPRSS2 in lung alveolar epithelial type II cells, nasal goblet secretory cells, cholangiocytes, colonocytes, esophageal keratinocytes, gastrointestinal epithelial cells, pancreatic β-cells, and renal proximal tubules and podocytes21,26,27,28. These findings suggest that multiple-organ injury may occur at least in part due to direct viral tissue damage. The mechanism of extrapulmonary spread of SARS-CoV-2, whether hematogenous or otherwise, remains elusive.

Endothelial cell damage and thromboinflammation

Endothelial cell damage by virtue of ACE2-mediated entry of SARS-CoV-2 and subsequent inflammation and the generation of a prothrombotic milieu are other proposed pathophysiological mechanisms of COVID-1929,30,31. ACE2 expression has been demonstrated in arterial and venous endothelium of several organs29,32, and histopathological studies have found microscopic evidence of SARS-CoV-2 viral particles in endothelial cells of the kidneys31 and lungs29. Infection-mediated endothelial injury (characterized by elevated levels of von Willebrand factor) and endothelialitis (marked by the presence of activated neutrophils and macrophages), found in multiple vascular beds (including the lungs, kidney, heart, small intestine, and liver) in patients with COVID-19, can trigger excessive thrombin production, inhibit fibrinolysis, and activate complement pathways, initiating thromboinflammation and ultimately leading to microthrombi deposition and microvascular dysfunction31,33,34,35,36. Platelet–neutrophil cross-communication and activation of macrophages in this setting can facilitate a variety of proinflammatory effects, such as cytokine release, the formation of neutrophil extracellular traps (NETs), and fibrin and/or microthrombus formation37,38,39,40. NETs further damage the endothelium and activate both extrinsic coagulation pathways and intrinsic coagulation pathways. They were detected at higher levels in patients hospitalized with COVID-19 in a study from a large academic center in the USA (50 patients and 30 control participants), with a ‘pro-NETotic state’ positively correlating with severe illness41. Hypoxia-mediated hyperviscosity and upregulation of the HIF-1 (hypoxia-inducible factor 1) signaling pathway subsequent to acute lung injury may also contribute to the prothrombotic state42. Finally, direct coronavirus-mediated effects may also lead to an imbalance of pro- and anti-coagulant pathways43,44. Small case reports and case series have demonstrated the presence of fibrinous exudates and microthrombi in histopathological examinations in patients with COVID-1944,45,46,47,48.

Dysregulation of the immune response

Dysregulated immune response and cytokine-release syndrome, due to overactivation of innate immunity in the setting of T cell lymphodepletion, characterize the presentations of severe COVID-1949. Prior preclinical and human studies with pathogenic human coronaviruses have proposed rapid viral replication, antagonism of interferon signaling, and activation of neutrophils and monocyte-macrophages as mediators of hyperinflammation50,51. Elevation of serum inflammatory markers such as C-reactive protein, ferritin, erythrocyte sedimentation rate, D-dimer, fibrinogen, and lactate dehydrogenase is predictive of subsequent critical illness and mortality in patients with COVID-194,5,52,53,54. These patterns of laboratory abnormalities have been compared with secondary hemophagocytic lymphohistiocytosis–macrophage-activation syndrome, previously demonstrated in pathological samples from patients who died from infection with SARS-CoV55,56. Higher levels of the cytokine IL-6 in the serum have also been linked to a worse prognosis4,5,52,54,57 and have been found to correlate with fibrinogen levels in patients with COVID-1958,59,60. Clinical trials for treating COVID-19 by targeting the IL-6 signaling pathway are underway and hope to mitigate the deleterious effects of the activation of this pathway55. Immune system–related manifestations found in patients with COVID-19, including those of cytokine-release syndrome, are presented in Box 1.

Dysregulation of the RAAS

Maladaptive functions of the RAAS constitute another plausible pathophysiological mechanism of SARS-CoV-2 infection–related tissue damage. The RAAS is composed of a cascade of regulatory peptides that participate in key physiological processes of the body, including fluid and electrolyte balance, blood-pressure regulation, vascular permeability, and tissue growth61. ACE2, a membrane-bound aminopeptidase, has emerged as a potent counter-regulator of the RAAS pathway. ACE2 cleaves angiotensin I into inactive angiotensin 1-9 and cleaves angiotensin II into angiotensin 1-7, which has vasodilator, anti-proliferative, and antifibrotic properties62,63,64. While the pathophysiology of SARS-CoV-2 may not be limited exclusively to ACE2-related pathways, these findings may have implications for the organ-specific clinical manifestations of COVID-19 (Fig. 2).

Fig. 2: Extrapulmonary manifestations of COVID-19.
figure 2

The pulmonary manifestation of COVID-19 caused by infection with SARS-CoV-2, including pneumonia and ARDS, are well recognized. In addition, COVID-19 is associated with deleterious effects on many other organ systems. Common extrapulmonary manifestations of COVID-19 are summarized here.

Hematologic manifestations

Patients with COVID-19 may present with several laboratory abnormalities and thromboembolic complications. The hematologic manifestations and management considerations of COVID-19 are presented in Box 1.

Epidemiology and clinical presentation

Lymphopenia, a marker of impaired cellular immunity, is a cardinal laboratory finding reported in 67–90% of patients with COVID-19, with prognostic association in the vast majority of studies published so far2,4,5,57,65,66,67,68,69. Studies examining specific lymphocyte subsets have revealed decreases in both CD4+ T cells70 and CD8+ T cells5 to be associated with severe COVID-1971. In addition, leukocytosis (especially neutrophilia), seen less commonly, is also a negative prognostic marker4,5,66. Thrombocytopenia, although often mild (in 5–36% of admissions), is associated with worse patient outcomes2,4,68,69,72. COVID-19-associated coagulopathy is marked by elevated levels of D-dimer and fibrinogen, with minor abnormalities in prothrombin time, activated partial thromboplastin time, and platelet counts in the initial stage of infection73. Elevated levels of D-dimer at admission (reported in up to 46% of hospitalized patients) and a longitudinal increase during hospitalization have been linked with worse mortality in COVID-192,4,5,53,54,74,75.

Thrombotic complications were first reported from intensive care units (ICUs) in China76 and the Netherlands77 in up to 30% of patients. There is also emerging evidence of thrombosis in intravenous catheters and extracorporeal circuits, and arterial vascular occlusive events, including acute myocardial infarction (MI), acute limb ischemia, and stroke, in severely affected people in studies from the USA, Italy and France78,79,80,81,82. Subsequent studies from France and Italy have also reported high rates of thromboembolic events in critically ill patients with COVID-19 (17–22%) despite their having received prophylactic anticoagulation80,83,84,85. Indeed, in a cohort of 107 patients admitted to a single-center ICU with COVID-19, their rates of pulmonary emboli were notably higher than those of patients admitted to the same ICU during the same time interval in 2019 (20.6% versus 6.1%, respectively)86. Furthermore, multiple small studies in which critically ill patients with COVID-19 were routinely screened for thrombotic disease demonstrated high rates of thrombotic complications in these patients ranging from 69% to 85% despite thromboprophylaxis83,87,88. Variability in thromboprophylaxis regimens and screening schedules can help explain this variation in event rates across published studies.


The potential proposed mechanisms by which lymphopenia occurs include direct cytotoxic action of the virus related to ACE2-dependent or ACE2-independent entry into lymphocytes29,89,90, apoptosis-mediated lymphocyte depletion50,91,92, and inhibitory effects of lactic acid on lymphocyte proliferation93. In addition, atrophy of the spleen and widespread destruction of lymphoid tissues have been described for both SARS and COVID-1989,94. Leukocytosis (especially neutrophilia) is thought to be a consequence of a hyperinflammatory response to infection with SARS-CoV-2 and/or secondary bacterial infections67. The abnormally high levels of D-dimer and fibrinogen in the blood during the early stages of infection are reflective of excessive inflammation, rather than overt disseminated intravascular coagulation, which is commonly seen only in later stages of COVID-1977,80.

The untempered inflammation, along with hypoxia and direct viral mediated effects, probably contributes to the high rates of thrombotic complications in COVID-19. The increased expression of ACE2 in endothelial cells after infection with SARS-CoV-2 may perpetuate a vicious cycle of endothelialitis that promotes thromboinflammation29. Collectively, hemostatic and inflammatory changes, which reflect endothelial damage and activation as well as critical illness, constitute a prothrombotic milieu, at least similar to and possibly more severe than that of other viral illnesses86,95,96.

In addition to the macrothrombotic events, the development of in situ thrombosis in small vessels of the pulmonary vasculature (pulmonary intravascular coagulopathy) is an area that requires further study97. Autopsy studies of patients who died due to COVID-19 have shown high rates of microvascular and macrovascular thromboses, especially in the pulmonary circulation29,98,99,100. A post-mortem series of seven patients from Germany showed that alveolar capillary microthrombi were nine times more common in people who died of COVID-19 than in those who died of influenza29. Microthrombi and microangiopathic pathology, associated with foci of hemorrhage, were also noted on autopsies of ten African-American patients with severe COVID-19 from New Orleans, Louisiana, USA100.

Management considerations

Longitudinal evaluation of a complete blood count, with white-blood-cell differential, D-dimer, prothrombin time, and fibrinogen, is recommended during the hospitalization of patients with COVID-19, in accordance with interim guidelines from the International Society of Hemostasis and Thrombosis101. Trending inflammatory indices may help in the prediction of clinical outcomes and response to therapy in hospitalized patients. In addition, recently published interim consensus-based guidelines for the prevention and management of thrombotic disease in patients with COVID-19102 recommend routine risk assessment for venous thromboembolism for all hospitalized patients with COVID-19. Standard-dose pharmacological prophylaxis should be considered in the absence of absolute contraindications in such patients. Empiric use of higher-than-routine prophylactic-dose or therapeutic-dose anticoagulation in patients admitted to the ICU in absence of proven thromboses has also been implemented in some institutions. This is an area of ongoing intense discussions among experts, particularly for those patients who exhibit marked COVID-19-associated coagulopathy73,103.

A retrospective analysis found lower in-hospital mortality rates in patients with COVID-19 who received therapeutic anticoagulation104, although this study was carried out in a single center, and there is currently not sufficient evidence to recommend such a strategy. Randomized clinical trials investigating these questions are currently underway and will be crucial to establishing effective and safe strategies. Parenteral anticoagulants (such as low-molecular-weight or unfractionated heparin) are preferred to oral anticoagulants in the inpatient setting, given their short half-life and the ready availability of reversal agents, due to the possibility of drug–drug interactions when they are taken with antiviral treatments (such as ritonavir) and antibacterial treatments (such as azithromycin)102.

Cardiovascular manifestations

Several cardiovascular presentations of COVID-19 have been reported. Clinical manifestations and management considerations pertaining to the cardiovascular system are presented in Box 2.

Epidemiology and clinical presentation

SARS-CoV-2 can cause both direct cardiovascular sequelae and indirect cardiovascular sequelae, including myocardial injury, acute coronary syndromes (ACS), cardiomyopathy, acute cor pulmonale, arrhythmias, and cardiogenic shock, as well as the aforementioned thrombotic complications105,106. Myocardial injury, with elevation of cardiac biomarkers above the 99th percentile of the upper reference limit, occurred in 20–30% of hospitalized patients with COVID-19, with higher rates (55%) among those with pre-existing cardiovascular disease3,107. A greater frequency and magnitude of troponin elevations in hospitalized patients has been associated with more-severe disease and worse outcomes3,107. Biventricular cardiomyopathy has been reported in 7–33% of critically ill patients with COVID-1952,65. Isolated right ventricular failure with and without confirmed pulmonary embolism has also been reported108,109. Cardiac arrhythmias, including new-onset atrial fibrillation, heart block, and ventricular arrhythmias, are also prevalent, occurring in 17% of hospitalized patients and 44% of patients in the ICU setting in a study of 138 patients from Wuhan, China110. In a multicenter New York City cohort, 6% of 4,250 patients with COVID-19 had prolonged QTc (corrected QT; >500 ms) at the time of admission111. Among 393 patients with COVID-19 from a separate cohort from New York City, atrial arrhythmias were more common among patients who required mechanical ventilation than among those who did not (17.7% versus 1.9%)68. Reports from Lombardi, Italy, show an increase of nearly 60% in the rate of out-of-hospital cardiac arrest during the 2020 COVID-19 pandemic relative to a similar time period in 2019, which suggests the etiology to be either COVID-19 or other untreated pathology due to patients’ reluctance to seek care112.


The pathophysiology underlying cardiovascular manifestations is probably multifactorial. ACE2 has high expression in cardiovascular tissue, including cardiac myocytes, fibroblasts, endothelial cells, and smooth-muscle cells32,113, in support of a possible mechanism of direct viral injury. Myocarditis is a presumed etiology of cardiac dysfunction, and the development of myocarditis may relate to viral load. While isolation of the virus from myocardial tissue has been reported in a few autopsy studies21,24,99, other pathological reports have described inflammatory infiltrates without myocardial evidence of SARS-CoV-2114,115. Additionally, the finding of direct viral infection of the endothelium and accompanying inflammation, as reported in a patient with circulatory failure and MI, lends credence to the possibility of virus-mediated endothelial-cell damage as an underlying mechanism31. Systemic inflammatory response syndrome (cytokine storm) is another putative mechanism of myocardial injury17. Furthermore, patients with pre-existing cardiovascular disease may have higher levels of ACE2, which would potentially predispose them to more-severe COVID-19116,117. Moreover, isolated right ventricular dysfunction may occur as a result of elevated pulmonary vascular pressures secondary to ARDS118, pulmonary thromboembolism108,109, or potentially virus-mediated injury to vascular endothelial and smooth-muscle tissue31.

Other potential etiologies of myocardial damage not specific to COVID-19 include severe ischemia or MI in patients with pre-existing coronary artery disease, stress-mediated myocardial dysfunction, tachycardia-induced cardiomyopathy, and myocardial stunning after resuscitation or prolonged hypotension.

While patients with viral infections are at risk for MI in general119, this risk may be exaggerated in patients with COVID-19, given reports of disproportionately increased hypercoagulability in affected people, which would lead to a possible increase in thrombotically mediated MI. Moreover, distinguishing the presentation of atherosclerotic plaque-rupture MI from myonecrosis due to supply–demand mismatch (type 2 MI) in the setting of severe hypoxia and hemodynamic instability and myocarditis can be challenging120. This was especially evident in a recent case series of 18 patients with COVID-19 who developed ST-segment elevation on electrocardiogram, 10 of whom were diagnosed with noncoronary myocardial injury78.

Management considerations

Whether upregulation of ACE2 by ACE inhibitors or angiotensin-receptor blockers (ARBs) is lung protective121,122 or increases susceptibility to infection with SARS-CoV-2123 has been intensely debated within the cardiovascular community61,124. This has implications for patients with hypertension, heart failure, and/or diabetes, who are overrepresented among critically ill patients with COVID-19125. There is no evidence to support an association between the use of ACE inhibitors and ARBs and more-severe disease; some large studies indicate no relationship between the use of these agents and the severity of COVID-19126,127,128, whereas other data suggest that they may attenuate the severity of disease129,130,131. Routine discontinuation of these medications is not recommended, as endorsed by the guidelines of several international cardiology societies132,133.

As for the management of ACS, there is guidance available from multiple specialty societies134,135. Although primary percutaneous coronary intervention remains the preferred approach for most patients with ST-segment-elevation MI, fibrinolytic therapy may be appropriate in select patients, especially if personal protective equipment is not available136. Additionally, point-of-care echocardiography may be used to assess regional wall-motion abnormalities to guide decisions about cardiac catheterization. Less-urgent or elective procedures should be deferred in an effort to minimize the risk of viral transmission105,134,135,137. The patient’s baseline QTc interval should be obtained before the administration of any drugs that may lead to prolongation of this interval138. Diagnostic workup of myocardial dysfunction in patients with COVID-19 is challenging, given the sparing use of cardiac imaging, invasive angiography and hemodynamic assessments, and endomyocardial biopsies in consideration of the serious risk of viral infection of patients and healthcare workers and contamination of facilities139,140.

Renal manifestations

A substantial proportion of patients with severe COVID-19 may show signs of kidney damage. Clinical manifestations and management considerations pertaining to the renal system are presented in Box 3.

Epidemiology and clinical presentation

Acute kidney injury (AKI) is a frequent complication of COVID-19 and is associated with mortality141,142. In China, the reported incidence of AKI in hospitalized patients with COVID-19 ranged from 0.5% to 29% (refs. 2,4,5,142) and occurred within a median of 7–14 days after admission5,142. Studies from the USA have reported much higher rates of AKI. In a study of nearly 5,500 patients admitted with COVID-19 in a New York City hospital system, AKI occurred in 37%, with 14% of the patients requiring dialysis143. About one third were diagnosed with AKI within 24 hours of admission in this study. Of note, these rates are much higher than those reported during the SARS-CoV epidemic144. AKI occurred at much higher rates in critically ill patients admitted to New York City hospitals, ranging from 78% to 90% (refs. 53,54,111,143,145). Of 257 patients admitted to ICUs in a study from New York City, 31% received renal replacement therapy (RRT)54. Furthermore, hematuria has been reported in nearly half of patients with COVID-19143, and proteinuria has been reported in up to 87% of critically ill patients with COVID-1954. Hyperkalemia and acidosis are common electrolyte abnormalities associated with the high cell turnover seen in patients with COVID-19, even among patients without AKI. COVID-19 is also increasingly reported among patients with end-stage renal disease and kidney transplant recipients, with higher mortality rates than those seen in the general population146,147,148.


Several possible mechanisms specific to SARS-CoV-2 that distinguish this renal abnormality from the more general AKI that accompanies severe illness are noteworthy. First, SARS-CoV-2 may directly infect renal cells, a possibility supported by histopathology findings and the presence of ACE2 receptors23,27,32. Histopathological findings include prominent acute tubular injury and diffuse erythrocyte aggregation and obstruction in peritubular and glomerular capillary loops21,23. Viral inclusion particles with distinctive spikes in the tubular epithelium and podocytes, and endothelial cells of the glomerular capillary loops, have been visualized by electron microscopy21,23,31. Second, the demonstration of lymphocytic endothelialitis in the kidney, in addition to viral inclusion particles in glomerular capillary endothelial cells, suggests that microvascular dysfunction is secondary to endothelial damage31. Third, similar to severe infection with influenza virus, cytokine storm may have an important role in the immunopathology of AKI149. In fact, it has been speculated that this is an underlying mechanism of the clinical ‘viral sepsis’ and multiple-organ dysfunction, including AKI, in patients with COVID-1917. Glomerular injury mediated by immunocomplexes of viral antigen or virus-induced specific immunological effector mechanisms is also plausible, and this is reflected in the development of collapsing focal segmental glomerulosclerosis in people infected with SARS-CoV-2 who have two high-risk variants of APOL1 (the gene that encodes apolipoprotein L1)150,151,152. Finally, while proteinuria is not a typical manifestation of AKI, transient heavy albuminuria might occur secondary to endothelial dysfunction or direct podocyte injury. It is also possible that the pattern of severe proximal-tubular injury leads to a defect in receptor-mediated endocytosis, which results in the observed instances of proteinuria23. Other potential etiologies of AKI common to critical illness presentations, including ARDS, rhabdomyolysis, volume depletion, and interstitial nephritis, all remain relevant in patients with COVID-19153.

Management considerations

Urine analysis and protein-to-creatinine ratio may be obtained at admission for patients with COVID-19, as proteinuria and hematuria seem to be associated with a more severe clinical course and higher mortality, and this would provide an opportunity for early risk stratification142,154. In patients with suspected or confirmed COVID-19, an emphasis should be placed on optimization of volume status to prevent prerenal AKI, particularly given the high prevalence of AKI at presentation, while avoiding hypervolemia, which may worsen the patient’s respiratory status. The Surviving Sepsis guidelines for critical illness in COVID-19 recommend a conservative fluid-resuscitation strategy while acknowledging that the supporting evidence base is weak155. A dramatic increase in the need for RRT in critically ill patients may require judicious resource planning, including the use of shared continuous RRT protocols, co-localization of patients, and the utilization of acute peritoneal dialysis in select patients156. The prothrombotic state poses additional challenges in the initiation and maintenance of the extracorporeal circuits needed for RRT. In a multicenter prospective cohort study from four ICUs in France, 97% (28 of 29) patients receiving RRT experienced circuit clotting80. In the absence of contraindications, patients with COVID-19 may require systemic anticoagulation during RRT157.

Gastrointestinal manifestations

COVID-19 may cause gastrointestinal symptoms in some patients. Clinical manifestations and management considerations pertaining to the gastrointestinal system are presented in Box 4.

Epidemiology and clinical presentation

The incidence of gastrointestinal manifestations has ranged from 12% to 61% in patients with COVID-195,158,159,160,161. Gastrointestinal symptoms may be associated with a longer duration of illness but have not been associated with increased mortality158,160. In a recent meta-analysis of 29 studies (of which the majority are from China), the pooled prevalence of individual symptoms was reported, including that of anorexia (21%), nausea and/or vomiting (7%), diarrhea (9%), and abdominal pain (3%)160. In a study from the USA, a higher prevalence of these symptoms was reported (anorexia, 34.8%; diarrhea, 33.7%; and nausea, 26.4%)161. In addition, the presence of gastrointestinal symptoms at presentation was associated with a 70% increased risk of detection of SARS-CoV-2 in a study from a New York City hospital162. Gastrointestinal bleeding was rarely observed in this study, despite presence of traditional risk factors, including prolonged mechanical ventilation, thrombocytopenia, or systemic anticoagulation162.


The pathophysiology of gastrointestinal damage in COVID-19 is probably multifactorial. Virus-mediated direct tissue damage is plausible, given the presence of ACE2 in intestinal glandular cells20,25,163,164, as well as the visualization of viral nucleocapsid protein in gastric, duodenal, and rectal epithelial cells, and glandular enterocytes25. Viral RNA has been isolated from stool, with a positivity rate of 54% (ref. 160). Live viral shedding of infectious virions in fecal matter has been reported even after the resolution of symptoms, and this needs further evaluation as a potential source of transmission25. In addition, histopathological evidence of diffuse endothelial inflammation in the submucosal vessels of the small intestine from patients with COVID-19 and mesenteric ischemia suggests microvascular small-bowel injury31. Support for inflammation-mediated tissue damage is provided by the presence of infiltrating plasma cells and lymphocytes and of interstitial edema in the lamina propria of the stomach, duodenum, and rectum of patients25. It has also been hypothesized that alteration of the intestinal flora by the virus may contribute to gastrointestinal symptoms and severe disease progression165.

Management considerations

Current multi-society guidelines emphasize the avoidance of diagnostic endoscopy for non-urgent reasons during the COVID-19 pandemic166,167. Most practitioners reserve procedures for patients with COVID-19 to those for whom therapeutic intervention is necessary due to large-volume upper gastrointestinal bleeding or biliary obstruction. During the COVID-19 pandemic at a New York City hospital, upper endoscopy was performed at lower-than-usual hemoglobin levels and after transfusion of larger volumes of packed red blood cells168. Interestingly, this was seen both in patients testing positive for SARS-CoV-2 and in those testing negative, probably reflective of a reluctance to perform endoscopy for the former group and reluctance of the latter group to come to the hospital. Patients who do present with gastrointestinal symptoms and later test positive for SARS-CoV-2 have been reported to experience delays in diagnosis160. When feasible, COVID-19 should be considered as a differential diagnosis in these patients even in absence of respiratory symptoms. In resource-limited settings, patients with symptoms of diarrhea or nausea and/or vomiting in addition to respiratory symptoms should be prioritized for testing162.

Hepatobiliary manifestations

Signs of hepatobiliary damage may be observed in patients with severe presentations of COVID-19. Clinical manifestations and management considerations pertaining to the hepatobiliary system are presented in Box 4.

Epidemiology and clinical presentation

In critically ill patients with COVID-19, a hepatocellular injury pattern is seen in 14–53% of hospitalized patients2,4,5,65,66. Aminotransferases are typically elevated but remain less than five times the upper limit of normal. Rarely, severe acute hepatitis has been reported169,170. A recent systematic review incorporating 12 studies reported a pooled prevalence of liver function abnormalities at 19% (95% confidence interval, 9–32%) with an association with disease severity160. Elevated bilirubin at hospital admission has also been linked to disease severity and progression to critical illness in a few studies171,172, although the association of longitudinal changes in bilirubin with prolonged ARDS is currently unclear.


SARS-CoV-2 may directly damage the biliary ducts by binding to ACE2 on cholangiocytes170. Hyperinflammation seen with cytokine storm and hypoxia-associated metabolic derangements are other potential mechanisms of liver damage17. Drug-induced liver injury, particularly secondary to investigational agents such as remdesivir, lopinavir, and tocilizumab, may also occur17,159,173. In a prospective clinicopathologic series of 11 patients from Austria, Kupffer cell proliferation was seen in all patients, and chronic hepatic congestion was seen in 8 patients. Other histopathologic changes in the liver included hepatic steatosis, portal fibrosis, lymphocytic infiltrates and ductular proliferation, lobular cholestasis, and acute liver-cell necrosis, together with central-vein thrombosis98.

Management considerations

In accordance with guidelines issued by an expert panel of the American Association for the Study of Liver Diseases, additional diagnostic tests are not recommended for aminotransferase elevations unless additional features raise the probability of findings that require further measures (such as hyperbilirubinemia, right-upper-quadrant pain, and hepatomegaly)174. Other COVID-19-related etiologies of elevated liver biochemistries, including myositis, cardiac injury (in conjunction with troponin elevation), ischemia, cytokine-release syndrome, and co-infection with other viruses, should be considered. Longitudinal monitoring of hepatic transaminases is recommended, particularly in patients receiving investigational treatments, including remdesivir, lopinavir, and tocilizumab, although low-level elevations should not necessarily be considered a contraindication to treatment with these agents174.

Endocrinologic manifestations

While patients with pre-existing endocrinologic disorders may be predisposed to more-severe presentations of COVID-19, observations of a range of endocrinologic manifestations in patients without pre-existing disease have also been made. Clinical findings and management considerations relating to the endocrinologic system are presented in Box 5.

Epidemiology and clinical manifestations

Patients with diabetes mellitus and/or obesity are at risk of developing more-severe COVID-19 illness125. In a report from the US Centers for Disease Control, 24% of hospitalized patients and 32% of patients admitted to the ICU had underlying diabetes125. In an initial experience with 257 critically ill patients hospitalized in a tertiary-care hospital in New York City, 36% had diabetes and 46% were obese54. Similar observations were made in studies from China and Italy that demonstrated an association of underlying diabetes with severe illness and death2,175,176. Moreover, patients hospitalized with COVID-19 have exhibited a range of abnormalities of glucose metabolism, including worsened hyperglycemia, euglycemic ketosis, and classic diabetic ketoacidosis. In a retrospective study from China, among a group of 658 patients hospitalized with COVID-19, 6.4% presented with ketosis in the absence of fever or diarrhea177. Of these, 64% did not have underlying diabetes (with an average hemoglobin A1c level of 5.6% in this group).


Several mechanisms may account for the more severe disease course, including worsened hyperglycemia and ketosis, observed in patients with COVID-19 and diabetes. Factors related to SARS-CoV-2 include substantially elevated cytokine levels, which may lead to impairments in pancreatic β-cell function and apoptosis178 and, consequently, decreased insulin production and ketosis. In addition, ACE2 expression has been reported in the endocrine pancreas179,180, albeit inconsistently181. This raises the possibility that direct binding of SARS-CoV-2 to ACE2 on β-cells might contribute to insulin deficiency and hyperglycemia, as has been shown previously for infection with SARS-CoV179. Accelerated fat breakdown in patients with COVID-19 has also been proposed as a possible mechanism, but this requires further investigation177. Factors not specific to COVID-19 in patients with diabetes and infections include an altered immune response182,183,184 and an increase in counter-regulatory hormones that promotes hepatic glucose production, decreased insulin secretion, ketogenesis, and insulin resistance185,186. Key extrapulmonary manifestations of COVID-19 can also be interlinked with diabetic complications, from reduced renal function to pro-thrombotic and coagulopathic states, to cardiac dysfunction and hepatocyte injury187.

Obesity is another risk factor for more-severe illness in COVID-19188. This may be related to its effects on pulmonary function, such as reduced lung volumes and compliance, and an increase in airway resistance, as well as an association with diabetes189,190. In addition, increased adiposity has been linked with alterations in multiple cytokines, chemokines, and adipokines, including increased pro-inflammatory cytokines such as TNFα, IL-6, IL-8, leptin, and adiponectin189,190,191, which all potentially exacerbate the exuberant inflammatory response seen in this disease.

Management considerations

Hemoglobin A1C should be assessed in patients with COVID-19 who present with hyperglycemia and/or ketoacidosis, to identify possibly undiagnosed diabetes. Logistically, the management of diabetic ketoacidosis poses an increased risk to medical personnel, due to the need for hourly glucose checks while patients are on an insulin drip. There may be a role for remote glucose monitoring via continuous glucose monitors to alleviate this problem and reduce demands on nursing staff. Alternative protocols for subcutaneous insulin in selected groups of patients with mild to moderate diabetic ketoacidosis may be considered on an individual patient-level basis192.

Neurologic and ophthalmologic manifestations

There is growing evidence of neurologic complications of COVID-19. Neurologic and ophthalmologic manifestations of COVID-19 and management considerations are summarized in Box 6.

Epidemiology and clinical presentation

Similar to SARS and Middle East respiratory syndrome193,194, multiple neurological manifestations of COVID-19 have been described. An analysis of 214 patients with severe COVID-19 found that neurologic symptoms occurred in 36% of the patients195. A number of non-specific mild neurological symptoms are notable in hospitalized patients with COVID-19, including headache (8–42%), dizziness (12%), myalgia and/or fatigue (11–44%), anorexia (40%), anosmia (5%), and ageusia (5%)69,75,110, although the epidemiology may be different in milder outpatient presentations2,196,197. More-severe presentations of COVID-19 manifest with acute stroke of varying arterial and venous mechanisms (in up to 6% of those with severe illness)81,198, and confusion or impaired consciousness (8–9%)195,199. Acute inflammatory demyelinating polyneuropathy (Guillain-Barré syndrome) has also been reported in some patients200,201. In addition, meningoencephalitis79, hemorrhagic posterior reversible encephalopathy syndrome202, and acute necrotizing encephalopathy, including the brainstem and basal ganglia, have been described in case reports203,204. Ocular manifestations, such as conjunctival congestion alone, conjunctivitis, and retinal changes205, have also been reported in patients with COVID-192,206,207.


SARS-CoV and the coronavirus that causes Middle East respiratory syndrome have known neuroinvasive and neurotropic abilities193,194. Direct viral invasion of neural parenchyma is a possibility; SARS-CoV-2 may access the central nervous system via the nasal mucosa, lamina cribrosa, and olfactory bulb or via retrograde axonal transport. Nasal epithelial cells display the highest expression of ACE2 (the receptor for SARS-CoV-2) in the respiratory tree18,28; this may account for the symptoms of altered sense of taste or smell frequently reported retrospectively in the majority of outpatients with COVID-19196,197,208. Other neurological manifestations support, at least, the neurovirulence of COVID-19, perhaps reflecting the proinflammatory and prothrombotic cascade in the wake of cytokine storm30,55 as it affects brain vasculature and the blood–brain barrier, especially in the setting of toxic-metabolic sequelae of multi-organ dysfunction often seen in COVID-19-associated critical illness.

Management considerations

Provisional guidelines during the COVID-19 crisis call for continued adherence to established guidelines for acute ischemic stroke, including providing access to thrombolysis and thrombectomy, while recognizing the need to minimize the use of personal protective equipment209. Post-acute-care monitoring guidelines may be adjusted for pandemic constraints210. The liberal use of ‘telestroke’, or remote video evaluation of patients who have had a stroke, which has been a mainstay of stroke care for hospitals without access to specialists for many years, is also recommended and may have a role even within large hospitals. Potent baseline immunomodulatory therapies may be considered for extended-interval or delayed dosing in conditions such as multiple sclerosis during COVID-19211. Long-term considerations, such as post-infectious neurodegenerative and neuroinflammatory involvement, as well as the efficacy of an eventual vaccine in some immunosuppressed populations, are under investigation.

Dermatologic manifestations

Dermatologic manifestations have occasionally been described in patients suffering from COVID-19.

Epidemiology and clinical presentation

The dermatologic manifestations of COVID-19 were first reported in a single-center observational study in Italy, with a frequency of 20% in hospitalized patients with no history of drug exposure in the previous 2 weeks212. Approximately 44% of the patients had cutaneous findings at disease onset, while the remaining patients developed these during the course of their illness. No correlation with disease severity was noted in this small study. The cutaneous manifestations included erythematous rash, urticaria, and chickenpox-like vesicles. A preliminary systematic review of 46 studies (including case reports and series) found acro-cutaneous (pernio or chilblain-like) lesions to be the most commonly reported skin manifestation213. Other cutaneous findings included maculopapular rash, vesicular lesions, and livedoid and/or necrotic lesions. A transverse study included in this systematic review found chilblain-like lesions to be associated with less-severe COVID-19, while livedoid and/or necrotic lesions were seen in more-severe COVID-19214. Case reports have also described exanthematous rashes and petechiae in COVID-19215,216,217.


Drug exposure and temporal association with hydroxychloroquine, remdesivir, tocilizumab, and other experimental drugs should always be evaluated before any skin lesion is attributed to the viral infection218. Potential mechanisms for COVID-19-related cutaneous manifestations include an immune hypersensitivity response to SARS-CoV-2 RNA, cytokine-release syndrome, deposition of microthrombi, and vasculitis219. Superficial perivascular dermatitis and dyskeratotic keratinocytes have been most commonly described from histopathological examination of skin rashes215,220. Biopsy of acrocutaneous lesions has shown diffuse and dense lymphoid infiltrates, along with signs of endothelial inflammation220. Small thrombi in vessels of the dermis have occasionally been seen215.

Management considerations

Most cutaneous manifestations of COVID-19 have been self-resolving. It is not clearly understood whether patients with dermatologic diseases who receive biologic therapies are at increased risk of complications from COVID-19221. In their interim guidelines, the American Academy of Dermatology recommends discontinuation of biologic therapy in COVID-19-positive patients (similar to recommendation for patients with other active infections) and recommends case-by-case discussion about the continuation of these drugs for at-risk patients222.

Special considerations in COVID-19

Here we review the epidemiological and clinical features of pediatric and pregnant patients with COVID-19.


Epidemiology and clinical presentation

In a review of 72,314 patients with COVID-19 reported by the Chinese Center for Disease Control and Prevention, less than 1% of the patients were younger than 10 years of age176. In two retrospective studies from China, of >1,000 pediatric patients with COVID-19, the majority of the patients had mild or moderate disease, and only 1.8% required ICU admission, with two reported deaths223,224. In a recent study from a large group of North American pediatric ICUs, however, 38% of 48 critically ill children required invasive ventilation, with an in-hospital mortality rate of 4.2% (ref. 225).

In the later phase of the pandemic, reports of healthy children presenting with a severe inflammatory shock that had features similar to those of atypical Kawasaki disease or toxic-shock syndrome were highlighted by physicians in Europe and the USA226,227. This syndrome was recognized as multisystem inflammatory syndrome in children (MIS-C) and was defined as follows: a person <21 years of age presenting with fever, laboratory evidence of inflammation, and evidence of clinically severe illness requiring hospitalization, with multisystem (two or more) organ involvement in the setting of current or recent infection with SARS-CoV-2. If untreated, Kawasaki disease can result in coronary aneurysms in 25% of patients228.


Potential reasons for less-severe manifestations of COVID-19 in children include the evolution of ACE2 expression229 and T cell immunity and a pro-inflammatory cytokine milieu230 with age. The possibility of competition between SARS-CoV-2 and other pre-existing viruses that are common in the respiratory mucosa of young children has been hypothesized as well230. In contrast, increased ACE2 expression around birth, before its decrease, and a reduced ability of T cells to fight viral infections at birth may be responsible for the susceptibility of infants to severe COVID-19230. Interestingly, many children with MIS-C tested negative for SARS-CoV-2 but positive for antibodies, suggestive of a likely trigger from the development of acquired immunity rather than direct viral injury as the underlying mechanism.

Management considerations

The majority of patients with MIS-C required vasopressor support, and few required mechanical circulatory support, in experience from New York hospitals. Due to some similarities with Kawasaki disease, treatment strategies have included intravenous immunoglobulin and corticosteroids, and occasionally an IL-1 antagonist (anakinra)227,231. Treatment with aspirin or lovenox has also been debated, given the hypercoagulable state and concern for coronary involvement similar to that of Kawasaki disease. Other medical therapies that have been extrapolated from studies of adults include compassionate use of the anti-viral drug remdesivir and the IL-6 receptor antagonist tocilizumab232,233.

Pregnant women


Pregnancy and childbirth have not been shown to substantially alter susceptibility to or the clinical course of infection with SARS-CoV-2234,235,236. Preliminary data indicate that rates of ICU admission for pregnant women are similar to those of the nonpregnant population176,237. Pregnant women with COVID-19 have not been reported to have severe maternal complications but were noted to be at increased risk of preterm and cesarean delivery in a few studies235,238,239. Maternal deaths from cardiopulmonary complications and multi-organ failure in previously healthy women have also been reported240,241. Evidence for vertical transmission to neonates has been mixed so far, which suggests that vertical transmission is possible but is probably not a common occurence242,243,244.


It is unknown whether the normal immunological changes of pregnancy affect the severity of COVID-19 illness, a disease marked by hyperinflammation in its severe forms. Histopathological evidence of infection of placental and fetal membrane samples with SARS-CoV-2 has been reported in a few cases245,246, but so far, vaginal and amniotic samples have tested negative in pregnant women infected with SARS-CoV-2247.


The management of hospitalized pregnant women is not substantially different from that of non-pregnant people. Changes to the route of delivery or management of labor are not routinely recommended for pregnant patients with COVID-19248.

Conclusions and future directions

Beyond the life-threatening pulmonary complications of SARS-CoV-2, the widespread organ-specific manifestations of COVID-19 are increasingly being appreciated. As clinicians around the world brace themselves to care for patients with COVID-19 for the foreseeable future, the development of a comprehensive understanding of the common and organ-specific pathophysiologies and clinical manifestations of this multi-system disease is imperative. It is also important that scientists identify and pursue clear research priorities that will help elucidate several aspects of what remains a poorly understood disease. Some examples of areas that require further attention include elucidation of the mechanism by which SARS-CoV-2 is disseminated to extrapulmonary tissues, understanding of the viral properties that may enhance extrapulmonary spread, the contribution of immunopathology and effect of anti-inflammatory therapies, anticipation of the long-term effects of multi-organ injury, the identification of factors that account for the variability in presentation and severity of illness, and the biological and social mechanisms that underlie disparities in outcomes. A number of organ-system-specific research questions are summarized in Table 1. There is also a need for common definitions and data standards for research relating to COVID-19. Regional, national, and international collaborations of clinicians and scientists focused on high-quality, transparent, ethical, and evidence-based research practices would help propel the global community toward achieving success against this pandemic.

Table 1 Future directions for research pertaining to various organ systems affected by COVID-19