Microvascular brain injury was seen in COVID-19 patients who died, but no evidence of a direct viral attack on the brain was detected, a pathology report showed.
Damage caused by thinning and leaky brain blood vessels consistently appeared on high-resolution MRI, but there were no signs of SARS-CoV-2 infection in tissue samples, reported Avindra Nath, MD, clinical director of the National Institute of Neurological Disorders and Stroke (NINDS), and co-authors in a New England Journal of Medicine letter.
“We found there were many foci of small blood vessel damage from which there was leakage of blood products into the brain tissue,” Nath said. “The cause of this was not clear but is most likely due to damage from the immune cells or lymphocytes. We found some lymphocytes attached to the endothelial cells in the blood vessels and in the perivascular regions,” he told MedPage Today.
“The inflammatory response is key to the neuropathogenesis of this syndrome, since we were unable to find virus in the brain,” Nath added. “The study potentially has important implications for long-term damage to several structures in the brain, particularly the olfactory bulb and the brainstem.”
Other researchers have found SARS-CoV-2 RNA and protein in the brain and nasopharynx of patients who died with COVID-19, with highest levels of viral RNA found in the olfactory mucous membrane.
In this study, Nath and colleagues looked at brain tissue samples of COVID-19 patients — 16 cases from the New York City chief medical examiner’s office and three from the University of Iowa in Iowa City — who died between March and July 2020. Their analysis included 11.7T MRI images of 13 patients at resolutions of 25 μm for the olfactory bulb and 100 μm for the brain, plus conventional histopathological brain exams of 18 patients.
Ages ranged from 5 to 73, with a median of 50. Fourteen patients had chronic illnesses, including diabetes and hypertension, and 11 had been found dead or had died suddenly. Of 16 patients with available medical histories, one had delirium, five had mild respiratory symptoms, four had acute respiratory distress syndrome, two had pulmonary embolism, and symptoms were unknown in three.
Magnetic resonance microscopy showed punctate hyperintensities representing areas of microvascular injury and fibrinogen leakage in nine patients that correlated with histopathological exam, which showed thinning of the basal lamina of endothelial cells.
In contrast, punctate hypointensities in 10 patients corresponded to congested blood vessels with areas of fibrinogen leakage and “relatively intact vasculature,” the researchers said. They also observed minimal perivascular inflammation but no vascular occlusion, consistent with other studies.
Perivascular-activated microglia, macrophage infiltrates, and hypertrophic astrocytes were seen in 13 patients. In eight patients, T cells were observed in perivascular spaces and in lumens adjacent to endothelial cells.
“We were completely surprised,” Nath said. “Originally, we expected to see damage that is caused by a lack of oxygen. Instead, we saw multifocal areas of damage that is usually associated with stroke and neuroinflammatory disease.”
SARS-CoV-2 was not detected in brain tissue, but it’s possible the virus was cleared by the time of death or viral copy numbers were below levels of detection by the assays used, the researchers noted.
So far, the findings suggest the damage was not caused by direct brain infection, Nath observed. With limited clinical information available in this study, no conclusions can be drawn about how these findings relate to neurologic features of COVID-19.
But studies are underway to “further characterize the inflammatory infiltrates and the pattern of neuronal injury in the autopsy material,” Nath said. In addition, NINDS researchers are studying a cohort of COVID-19 patients with neurologic sequelae to determine whether they also have microvascular injury and whether inflammatory infiltrates are linked to persistent neurologic symptoms.
This study was supported by the NIH Intramural Research Program at NINDS.
The researchers had no disclosures.