Mapping of SARS-CoV-2 Brain Invasion and Histopathology in COVID-19 Disease

Abstract

The coronavirus SARS-CoV-2 (SCV2) causes acute respiratory distress, termed COVID-19 disease, with substantial morbidity and mortality. As SCV2 is related to previously-studied coronaviruses that have been shown to have the capability for brain invasion, it seems likely that SCV2 may be able to do so as well. To date, although there have been many clinical and autopsy-based reports that describe a broad range of SCV2-associated neurological conditions, it is unclear what fraction of these have been due to direct CNS invasion versus indirect effects caused by systemic reactions to critical illness. Still critically lacking is a comprehensive tissue-based survey of the CNS presence and specific neuropathology of SCV2 in humans. We conducted an extensive neuroanatomical survey of RT-PCR-detected SCV2 in 16 brain regions from 20 subjects who died of COVID-19 disease. Targeted areas were those with cranial nerve nuclei, including the olfactory bulb, medullary dorsal motor nucleus of the vagus nerve and the pontine trigeminal nerve nuclei, as well as areas possibly exposed to hematogenous entry, including the choroid plexus, leptomeninges, median eminence of the hypothalamus and area postrema of the medulla. Subjects ranged in age from 38 to 97 (mean 77) with 9 females and 11 males. Most subjects had typical age-related neuropathological findings. Two subjects had severe neuropathology, one with a large acute cerebral infarction and one with hemorrhagic encephalitis, that was unequivocally related to their COVID-19 disease while most of the 18 other subjects had non-specific histopathology including focal β-amyloid precursor protein white matter immunoreactivity and sparse perivascular mononuclear cell cuffing. Four subjects (20%) had SCV2 RNA in one or more brain regions including the olfactory bulb, amygdala, entorhinal area, temporal and frontal neocortex, dorsal medulla and leptomeninges. The subject with encephalitis was SCV2-positive in a histopathologically-affected area, the entorhinal cortex, while the subject with the large acute cerebral infarct was SCV2-negative in all brain regions. Like other human coronaviruses, SCV2 can inflict acute neuropathology in susceptible patients. Much remains to be understood, including what viral and host factors influence SCV2 brain invasion and whether it is cleared from the brain subsequent to the acute illness.

Keywords: RT-PCR; coronavirus; encephalitis; hemorrhage; hypoxia; infarction; neuropathology.

Figure 1.

Mayo Clinic Case M10 (see Tables), a male in his 30s. Neuropathological gross examination was consistent with encephalitis-associated brain swelling causing transtentorial uncal herniation (A) and associated with acute hemorrhages (arrows A inset, and in B and C). Microscopic examination of semi-adjacent temporal lobe sections show microscopic hemorrhage (D; H & E stain) and neuropil infiltration with T- and B-lymphocytes and macrophages (E-G).

Figure 2.

Mayo Clinic Case M10 (continued from Figure 1). Sections of pons adjacent to area of gross acute hemorrhage (Figure 1A inset). Acute pontine hemorrhage and fibrinoid necrosis of a blood vessel in a section stained with H & E (A). Semi-adjacent sections are immunohistochemically stained for B-lymphocytes (B, CD20), macrophages (C, CD68) and T-lymphocytes (D, CD3).

Figure 3.

Cerebral white matter sections from Mayo Clinic Case M10 (continued from Figures 1 and 2, A–D) and BSHRI Case B5 (E, F), a male in his 90s. Sections of cerebral white matter from Case M10 are unremarkable on H & E stains, at low (A) and medium (B) magnifications while semi-adjacent sections stained with an immunohistochemical method for β-amyloid precursor protein (APP) show patchy staining at low magnification and intense staining of axons at higher magnification (C and D, respectively). Lower (E) and higher (F) magnifications of APP immunostaining in Case B5b show a patchy quality at lower magnification, with axonal staining visible at higher magnifications. Axons frequently have periodic axonal swellings (arrow in F shows example).

Figure 4.

Banner Sun Health Research Institute (BSHRI) Case B3, a male in his 70s. Clinical findings indicated a massive acute left middle cerebral artery territory ischemic infarct, as shown in the MRI image (A, arrows). Gross examination of the brain at autopsy showed widespread hemorrhagic areas, especially in cerebral cortex (B). The left middle cerebral artery within the Sylvian fissure was completely occluded by firm thrombus, confirmed on microscopic examination (C). There were multiple thrombi within parenchymal arterioles (D, arrows). Sections were stained with H & E, on 6 μm paraffin sections

Figure 5.

Case B3 (continued from Figure 4). Microscopic examination confirmed the presence of widespread acute cortical hemorrhages (A). Other areas within the left middle cerebral artery distribution showed acute ischemic infarction, with microvacuolation of the neuropil and loss of normal tissue eosinophilia (B), perivascular neuropil infiltration by polymorphonuclear leukocytes (C), and acute hypoxic-ischemic changes, including perikaryal cytoplasmic eosinophilia and nuclear pyknosis of cortical pyramidal neurons (D). Sections are all stained with H & E, on 6 μm paraffin sections (A, B, D) or 80 μm thick sections (C)

Figure 6.

BSHRI Case B1 (A and B), a male in his 90s. Microscopic examination showed focal white matter rarefaction in the temporal lobe white matter (A) with increased numbers of microglial nuclei within area of rarefaction (B). BSRHI Case B9, a malein his 70s, showed acute microhemorrhages, seen here in the cortex of the superior frontal gyrus (C). BSHRI Case B8 (D-F), a male in his 80s, had an acute microscopic infarct in cortex of the middle frontal gyrus (D) and laminar mineralization of pyramidal neurons (E, F) in lateral occipital association cortex. All images are from H & E-stained 80 μm thick sections.

Figure 7.

Several cases had occasional perivascular mononuclear cell aggregates, in both gray and white matter. Shown are examples from BSHRI Case B4, a male in his 80s(A), Case B2, a female in her 70s (B), Case B6, a male in his 70s (C), Case B8, a male in his 80s(D), and Case B9, a male in his 70s (E). Also shown is a microglial nodule in the posterior medulla in the region of the nucleus gracilis, in BSHRI Case B6 (F). Images are from H & E-stained 80 μm thick sections (B-E) or 6 μm paraffin sections (A, F).

Figure 8.

Standard curve for obtaining estimated SARS-CoV-2 sample copy numbers, constructed with serial dilutions of commercially-obtained synthetic SCV2 RNA. Ct values for patient samples that were positive for threshold amplification (colored symbols, see legend; some symbols are superimposed on others on graph) of sample SCV2 RNA are plotted on the same graph. The highest copy numbers were obtained from lung (black arrows show lung samples with higher Ct values), olfactory bulb (OB, lavender arrows show both of the positive OB samples) and temporal cortex (TC, orange arrow shows the only positive TC sample). Ent = entorhinal area; Leptos = leptomeninges; AMYG = amygdala; FC = frontal cortex.