Lethal gastrointestinal manifestations of COVID-19 infection at autopsy

Introduction

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was first identified in December 2019 and has since caused considerable morbidity and mortality globally. Although first thought to be primarily a disease focussed on the respiratory tract, with the release of significant amounts of pro-inflammatory cytokines inducing interstitial pneumonia and acute respiratory distress syndrome (ARDS), it is now recognised that it can have quite serious effects on multiple organ systems of infected individuals.^1,2 One of striking features of this infection is the predisposition to thromboembolic phenomena involving the heart, brain and limbs, and rarely the gastrointestinal tract. ³ However, although reported in the clinical literature, there has been little written on this latter phenomenon in forensic journals. For this reason the following two cases are reported to demonstrate the significant and potentially lethal effects of intestinal involvement that may be encountered in a mediocolegal setting, with a forensic approach to such cases.

Case reports

Case 1: A 74-year-old woman presented to hospital with a 3 day history of watery diarrhoea with increasing abdominal pain, dehydration and ataxia. Her medical history included hypertension, hypercholesterolaemia, type 2 diabetes mellitus, kidney disease, hypothyroidism, diverticulitis, gastro-oesophageal reflux disease, gout, alcohol misuse and cognitive decline over the past 3 to 4 years. Admission blood samples showed acute kidney injury and raised inflammatory markers and a computed tomography (CT) scan of the abdomen showed distension of the small and large intestine suggestive of an ileus. She became hypotensive, hypoxic and confused with a rise in serum potassium, a fall in haemoglobin (to 66 g/L), raised creatine kinase and a coagulopathy. Cardiac arrest occurred that was unresponsive to resuscitation.

At autopsy there were extensive ischaemic changes of the distal two-thirds of the small intestine, the caecum, appendix, ascending colon and a portion of the transverse colon associated with traces of fibrinopurulent exudate in the peritoneal cavity with no perforation (Figure 1(a)). Opening of the intestine revealed ischaemic changes (Figure 1(b)). Other findings included marked diffuse pulmonary oedema with bilateral pleural effusions, an essentially normal heart, patency of the superior mesenteric artery and its major branches (Figure 1(c)), a focus of focal nodular hyperplasia in the liver, moderate diffuse cortical thinning of both kidneys, and enlargement and distortion of the uterine fundus by multiple fibroids.

OPEN IN VIEWER

Figure 1.

At autopsy in case 1 there were extensive ischaemic changes of the distal two-thirds of the small intestine and a portion of the transverse colon associated with traces of fibrinopurulent exudate in the peritoneal cavity with no perforation (a). Opening of the small intestine demonstrated ischaemic changes (b) with no luminal obstructive lesions within the superior mesenteric artery and its major branches (c).

Histology of multiple sections from the ischaemic areas of small intestine showed marked congestion and dilatation of submucosal vessels with focal acute inflammation and necrosis of the mucosa and intramucosal haemorrhage. Fibrin thrombi were noted in some of the intramucosal and submucosal vessels (Figure 2(a)). The muscularis propria generally appeared well preserved although there were focal areas of acute inflammation extending from the submucosa into the muscularis. Sections from the caecum and ascending colon showed similar congestion of submucosal and intramucosal vessels with focal fibrin thrombi and areas of intramucosal haemorrhage. There was no evidence of vasculitis. No frank mucosal necrosis was seen but this could not be completely excluded due to the confounding effects of mucosal autolysis. The lungs showed congestion and oedema but no inflammation.

OPEN IN VIEWER

Figure 2.

Histology in case 1 showed marked congestion and dilatation of submucosal vessels. The mucosa showed variable changes with acute inflammation and necrosis in some areas and, in other areas, congestion of intramucosal vessels with foci of intramucosal haemorrhage. Fibrin thrombi were present in some of the intramucosal and submucosal vessels (a). Prominent fibrin thrombi were present throughout the ischaemic intestinal wall in case 2 (b).

Nucleic acid tests undertaken on specimens of small and large intestinal contents obtained at autopsy were negative for common enteric pathogens. Respiratory viral nucleic acid testing performed on a retronasal swab obtained at autopsy was positive for COVID-19 (SARS-CoV-2).

There were no other underlying natural diseases which could have caused or contributed to death and there was no trauma. Death was therefore attributed to multi-organ failure due to mesenteric ischaemia associated with COVID-19 infection.

Case 2: A 94-year-old woman who was resident in an aged care facility contracted COVID-19. She became drowsy and lethargic and subsequently developed abdominal pain. A sacral pressure sore had been noted, which appeared to be progressing, and she suffered an episode of melaena and bright red rectal bleeding followed by death.

Her medical history included osteoporosis, recurrent urinary tract infections, congestive cardiac failure, gastro-oesophageal reflux, atrial fibrillation, reduced cognition, hypercholesterolaemia and hypertension.

At autopsy there was serosal erythema of the distal small intestine suggestive of ischaemic change with superficial ulceration of the descending colon. Histological examination showed autolytic changes of the small intestine with mucosal, and in places transmural, acute inflammation, focal mucosal congestion, serosal exudate and multiple microthrombi (Figure 2(b)). There was no evidence of vasculitis. There was also no evidence of pseudo-membranous colitis, or of pneumonia or diffuse alveolar damage within the lungs.

Nucleic acid tests undertaken on intestinal contents obtained at autopsy revealed Clostridioides difficile. Respiratory viral nucleic acid testing performed on a retronasal swab obtained at autopsy was positive for COVID-19 (SARS-CoV-2).

There were no other underlying natural diseases which could have caused or contributed to death and there was no trauma. Death was therefore attributed to multi-organ failure due to mesenteric ischaemia associated with COVID-19 infection. Although C. difficile was also identified within intestinal contents no pseudo-membranes were identified suggesting that the enteritis was primarily ischaemic, rather than infective.

Discussion

SARS-CoV-2 is a coronavirus responsible for COVID-19 that caused a recent global health emergency with deaths due to pneumonia, severe acute respiratory distress and multi-organ failure. While the most common symptoms are fever, myalgia and cough, the possibility of COVID-19 should be considered in cases with a history of gastrointestinal symptoms as 39% to 61% of cases may present with abdominal pain, anorexia, nausea, diarrhoea and gastrointestinal haemorrhage with abnormal liver function tests.^4,5

COVID-19 has also been associated with acute mesenteric ischaemia, a rare cause of abdominal pathology occurring in <1% of patients presenting with an acute abdomen. Acute mesenteric ischaemia generally has a high mortality rate of 50% to 90% associated with delays in diagnosis from its non-specific manifestations ⁶ and underlying comorbidities. Anatomically the intestine is supplied by the coeliac artery, and the superior and inferior mesenteric arteries, with the superior mesenteric artery supplying most of the small intestine and the colon to the splenic flexure. Acute mesenteric ischaemia may be caused by arterial or venous occlusion or it may be non-occlusive due to hypoperfusion.^7,8 The most common causes of occlusive acute mesenteric ischaemia are embolism (40%–50% of cases) and thrombosis (20%–35%) with a small number of cases arising from dissection usually involving the superior mesenteric artery. Venous thrombosis is often associated with hypercoagulable states. Secondary mesenteric ischaemia may occur with twisting of the mesentery associated with small intestinal obstruction. ⁶ The striking feature with cases of COVID-19 is the presence of non-occlusive ischaemia due to microvascular involvement by fibrin microthrombi. ³

COVID-associated acute mesenteric ischaemia is a rare event with only 2.1% of 18,185 patients with acute mesenteric ischaemia found on review of the National Inpatient Sample (NIS) database in the United States in 2020 to have COVID-19. The mortality for this group was higher (up to 62.5%) despite medical and surgical intervention, with risk factors including age, male sex, diabetes mellitus, cardiovascular disease, chronic kidney disease, obesity and hypertension.^4,9 The mortality has approached 100% in patients aged over 65 years. ¹⁰

The aetiology of mesenteric ischaemia in COVID infections is incompletely understood and may involve a predisposition to thrombosis, endothelial damage, hypoperfusion (which may be generalised or local) or direct infection of enterocytes by SARS-CoV-2 due to binding to angiotensin–converting enzyme 2 (ACE2) receptors in the gut lining.^4,11 As was demonstrated in the reported cases there is usually no evidence of large vessel mesenteric thrombosis, with 100% of patients in one study with intestinal ischaemia due to COVID-19 having patent vessels on abdominopelvic CT scans. ¹⁰ It is a condition that may affect children as well as adults. ¹²

The most characteristic finding in cases of COVID-19-associated ischaemic enterocolitis is patchy mucosal ulceration with sloughing of the surface epithelium, partial or transmural involvement of the muscularis propria and small vessel fibrin thrombi (96% of cases), ¹³ as was clearly demonstrated in the reported cases. Pneumatosis was not observed, although it has been previously documented. ¹⁰

In conclusion, mesenteric ischaemia with gut infarction may be caused by a range of pathological conditions including arterial and venous thromboses/thromboemboli. ⁷ For this reason cases should have the relevant major intestinal branches of the aorta (the coeliac, superior and inferior mesenteric arteries) dissected at the time of autopsy to check for occlusive lesions. In the absence of major vessel narrowing the possibility of COVID-19 infection should be considered with careful histological sampling of the intestine and mesentery to check for microvascular fibrin thrombi, and retronasal swabbing for SARS-CoV-2.