Cerebrovascular disease in patients with COVID-19: neuroimaging, histological and clinical description

Abstract

Abstract Since the appearance of the first case of coronavirus disease 2019 (COVID-19) a pandemic has emerged affecting millions of individuals worldwide. Although the main clinical manifestations are respiratory, an increase in neurological conditions, specifically acute cerebrovascular disease, has been detected. We present cerebrovascular disease case incidence in hospitalized patients with SARS-CoV-2 infection. Patients were confirmed by microbiological/serological testing, or on chest CT semiology. Available data on co-morbidity, laboratory parameters, treatment administered, neuroimaging, neuropathological studies and clinical evolution during hospitalization, measured by the modified Rankin scale, were analysed. A bivariate study was also designed to identify differences between ischaemic and haemorrhagic subtypes. A statistical model of binary logistic regression and sensitivity analysis was designed to study the influence of independent variables over prognosis. In our centre, there were 1683 admissions of patients with COVID-19 over 50 days, of which 23 (1.4%) developed cerebrovascular disease. Within this group of patients, cerebral and chest CT scans were performed in all cases, and MRI in six (26.1%). Histological samples were obtained in 6/23 cases (two brain biopsies, and four arterial thrombi). Seventeen patients were classified as cerebral ischaemia (73.9%, with two arterial dissections), five as intracerebral haemorrhage (21.7%), and one leukoencephalopathy of posterior reversible encephalopathy type. Haemorrhagic patients had higher ferritin levels at the time of stroke (1554.3 versus 519.2, P = 0.004). Ischaemic strokes were unexpectedly frequent in the vertebrobasilar territory (6/17, 35.3%). In the haemorrhagic group, a characteristic radiological pattern was identified showing subarachnoid haemorrhage, parieto-occipital leukoencephalopathy, microbleeds and single or multiple focal haematomas. Brain biopsies performed showed signs of thrombotic microangiopathy and endothelial injury, with no evidence of vasculitis or necrotizing encephalitis. The functional prognosis during the hospital period was unfavourable in 73.9% (17/23 modified Rankin scale 4–6), and age was the main predictive variable (odds ratio = 1.5; 95% confidence interval 1.012–2.225; P = 0.043). Our series shows cerebrovascular disease incidence of 1.4% in patients with COVID-19 with high morbidity and mortality. We describe pathological and radiological data consistent with thrombotic microangiopathy caused by endotheliopathy with a haemorrhagic predisposition.