RNA SARS-CoV-2 Persistence in the Lung of Severe COVID-19 Patients: A Case Series of Autopsies

Abstract

The exact role of viral replication in patients with severe COVID-19 has not been extensively studied, and it has only been possible to demonstrate the presence of replicative virus for more than 3 months in a few cases using different techniques. Our objective was to study the presence of RNA SARS-CoV-2 in autopsy samples of patients who died from COVID-19 long after the onset of symptoms. Secondary superimposed pulmonary infections present in these patients were also studied. We present an autopsy series of 27 COVID-19 patients with long disease duration, where pulmonary and extrapulmonary samples were obtained. In addition to histopathological analysis, viral genomic RNA (gRNA) and viral subgenomic RNA (sgRNA) were detected using RT-PCR and in situ hybridization, and viral protein was detected using immunohistochemistry. This series includes 26 adults with a median duration of 39 days from onset of symptoms to death (ranging 9-108 days), 92% of them subjected to immunomodulatory therapy, and an infant patient. We detected gRNA in the lung of all but one patient, including those with longer disease duration. SgRNA was detected in 11 out of 17 patients (64.7%) with illness duration up to 6 weeks and in 3 out of 9 patients (33.3%) with more than 6 weeks of disease progression. Viral protein was detected using immunohistochemistry and viral mRNA was detected using in situ hybridization in 3 out of 4 adult patients with illness duration of <2 weeks, but in none of the 23 adult patients with an illness duration of >2 weeks. A remarkable result was the detection of viral protein, gRNA and sgRNA in the lung cells of the pediatric patient after 95 days of illness. Additional pulmonary infections included: 9 acute bronchopneumonia, 2 aspergillosis, 2 cytomegalovirus, and 1 BK virus infection. These results suggest that in severe COVID-19, SARS-CoV-2 could persist for longer periods than expected, especially in immunocompromised populations, contributing to the persistence of chronic lung lesions. Additional infections contribute to the fatal course of the disease.

Keywords: COVID-19; RNA; autopsy; immunosuprression; viral persistance.

FIGURE 1

(A) Hematoxylin and eosin. Diffuse alveolar damage (DAD), exudative phase with hyaline membranes formation. Scale bar: 100 μm. (B) Hematoxylin and eosin. DAD, organizing phase. Scale bar: 100 μm. (C) Hematoxylin and eosin. DAD, fibrotic phase. Scale bar: 100 μm. (D) Hematoxylin and eosin. Numerous filamentous fungi are seen in the lung. Scale bar: 100 μm. Inset: Periodic acid–Schiff staining, highlighting the fungi. (E) Hematoxylin and eosin. Filamentous fungi invading the arteriolar wall in angioinvasive aspergillosis. Septate hyphae with dichotomous branching at acute angles of around 45° are apparent. Scale bar: 100 μm. (F) Hematoxylin and eosin. Endothelial pulmonary cells with viral cytopathic changes. Scale bar: 100 μm. Inset: Immunohistochemistry for cytomegalovirus, showing positive staining.

FIGURE 2

(A) Hematoxylin and eosin. Ciliated bronchial epithelial cells with amphophilic nuclear inclusions (arrowheads). Scale bar: 100 μm. (B) Polyomavirus BK immunohistochemistry. Same area as panel (A), showing positive staining in the nuclear inclusions (arrowheads). Scale bar: 100 μm. (C) Electron micrograph of an infected cell containing intranuclear inclusions (between asterisks). Scale bar: 2 μm. (D) Intranuclear icosahedral inclusions. Scale bar: 1 μm. Inset: Note the 40 nm particles characteristic of the polyoma virus group. Scale bar: 200 nm.

FIGURE 3

SARS-CoV-2 cycle threshold values heatmap visualization. Upper blue bar indicates sample type (swab or formalin fixed paraffin embedded tissue –FFPE) of nasopharyngeal or lung locations. Right red bar indicates the weeks from onset of symptoms and right yellow bar shows the presence of sub-genomics regions. For antemortem nasopharyngeal swabs, three positive patients did not have the Ct values recorded (NA). One additional patient had a negative result in our hospital, but a positive result in another center. For post-mortem samples, NA indicates no sample available. The infant corresponds to Patient 26.

FIGURE 4

Box plot of SARS-CoV-2 cycle threshold values of the different types of samples. Median is shown as a black line, mean as a green circle and the outliers as red points. ANOVA: Pr(>F) = 4.04e–12***. Ct, cycle threshold; NPS, nasopharyngeal; RL, right lobe; LL, left lobe; RUL, right upper lobe; RML, right medium lobe; RIL, right inferior lobe; LUL, left upper lobe; LIL, left inferior lobe; IQR, interquartile range. *Three positive patients did not have the Ct values recorded. One additional patient had a negative result in our hospital, but a positive result in another Center.

FIGURE 5

Spearman correlation between days from onset of symptoms and SARS-CoV-2 cycle threshold values of FFPE samples, where regression line is shown in blue and confident interval in gray.

FIGURE 6

(A) Ciliated bronchial epithelial cells positive for SARS-CoV-2 spike protein in patient 26 (immunohistochemistry). (B) In situ hybridization for SARS-CoV-2 RNA revealed positivity in the cytoplasm of several alveolar cells. (C) From left to right: Tape Station report of sgRNA amplification in infant patient. Electropherogram obtained by Sanger sequencing of the amplified product of gRNA where the leader sequence is observed close to the N gene.