This study demonstrated that it is possible to detect SARS-CoV-2 in several FFPE tissues and cytology specimens, mainly with qPCR techniques. While IHC was only useful in placental tissue with acute inflammation, qPCR also showed a high viral load. In this regard, qPCR techniques were more sensitive than IHC, with qPCR method A showing the highest percentage of positive cases (20.93%) and IHC using the spike primary antibody showing only 10.34% positivity.
The better sensitivity of qPCR method A may be due to both a better nucleic acid extraction method and a better qPCR process; its extraction process was fully automatic, and the qPCR method B extraction process was manual. Furthermore, method A was based on primers and probes from IDT technologies designed to detect the SARS-CoV-2 target gene encoding nucleocapsid 1 (N1)7, while qPCR method B was based on the “LightMix® Modular Sarbecovirus E-gene” from TIB MOLBIOL (Berlin, Germany), distributed by Roche Diagnostics, and included primers to detect 76 bp fragments from the E gene from SARS and SARS-CoV-2 viruses8.
In this study, we identified SARS-CoV-2 in a greater variety of tissues than those previously described2,3,4. Positive FFPE tissues using qPCR techniques included the oesophagus, large intestine, kidney, placenta, lung, and brain. Additionally, with IHC, we detected weak positivity in isolated cells in kidney, lung, brain, and histologically normal placentas. However, only one placenta sample showed strong diffuse IHC positivity.
SARS-CoV-2 RNA was previously detected in fresh gastrointestinal specimens obtained by endoscopy from oesophageal lesions and from stomach, duodenum and rectum samples in 3 of 6 patients, although no histopathological study was described9. The ACE2 receptor is highly expressed in oesophageal epithelial cells and absorptive enterocytes from the ileum and colon10. SARS-CoV-2 RNA was also previously detected in FFPE specimens from the oesophagus in autopsies, but it was not detected in the large intestine11.
In postmortem FFPE core biopsies, Tian et al. detected SARS-CoV-2 RNA in the heart, lung, and liver12. In one autopsy, we were able to detect the virus in the lung and brain. Sekulic et al. (2020) found the highest SARS-CoV-2 RNA levels in FFPE tissues from the lung, bronchi, lymph nodes, and spleen in two limited autopsy cases (brain was not studied) using 2019-nCoV N1 and N2 primer/probe sets from IDT11. As in our autopsy case, in patients dying from diffuse alveolar damage, SARS-CoV-2 RNA may not be found in the kidney.
SARS-CoV-2 virus has also been successfully detected in FFPE tissue blocks from lung samples obtained in autopsies using a one-step RT–qPCR assay specific for the amplification of the SARS-CoV-2 E gene13 or using immunofluorescence techniques with an antibody directed against the Rp3 NP protein14.
In renal biopsies in COVID-19 patients, SARS-CoV-2 was detected using in situ hybridization in renal tubules and endothelial cells in six of nine (67%) kidney specimens, but IHC for SARS-CoV-2 spike protein was positive in only one case (11%)15. Using a primary antibody against the 2019-nCov N-protein, positive cases were found in 9/16 (56%) renal biopsies, with the virus mainly detected in proximal tubule epithelial cells and isolated distal tubule cells, with only one patient showing IHC positivity in more than 10% of the tubules. In this series, in situ hybridization was positive in 2/9 cases (22%), and real-time reverse-transcriptase polymerase chain reaction (RT–PCR) performed in FFPE tissue to detect the E and N1/N2 genes of SARS-CoV-2 in kidney samples was able to detect viral RNA in only 1/16 cases (6.25%)16. Most authors agree that SARS-CoV-2 can infect the kidney, at least in severe cases17. This confirms our findings suggesting that in COVID-19 patients, SARS-CoV-2 can occasionally be found in FFPE renal tissue using qPCR techniques, as in only one of 5 patients with a low viral load in our series, and IHC usually shows only a few positive epithelial tubular cells.
Most available studies have found that placentas from SARS-CoV-2-positive women do not show any specific histopathology pattern18, although some authors have described a higher frequency of changes associated with maternal–foetal vascular malperfusion19,20.
In our series, the highest SARS-CoV-2 virus load was detected in one placental FFPE specimen with inflammatory changes. With IHC, other authors confirmed the presence of the nucleocapsid protein of the virus in syncytiotrophoblasts, and no virus was identified in Hofbauer cells, but as in our series, isolated IHC-positive inflammatory cells for SARS-CoV-2 were also found in placental tissue19,21,22. In a series of 51 SARS-CoV-2-positive women, both IHC (using spike antibody) and in situ hybridization were negative in placental FFPE tissue20.
We detected SARS-CoV-2 RNA using qPCR techniques in 3/5 placentas (60%). In a review of 19 studies that tested for SARS-CoV-2 RNA in the placenta, only 4 studies reported positive results19. Placental inflammatory changes in COVID-19 patients have been frequently described, but acute inflammatory changes are less frequent19.
Positive cytology specimens in our series for qPCR included ascitic fluid, pleural fluid and urine. In ascitic fluid, the presence of SARS-CoV-2 RNA was first described by Culver et al. (2020) with a RT–PCR technique targeting the gene encoding the envelope (E) protein23. SARS-CoV-2 RT–PCR was also positive on nasopharyngeal swabs, bronchial aspirates and blood samples. Passarelli et al. (2020) also described SARS-CoV-2 RNA in ascitic fluid in a male patient with kidney transplantation, peritoneal dialysis, and liver cirrhosis, with ascitic fluid showing a significant number of macrophages24. Other authors were not able to detect SARS-CoV-2 in peritoneal or intra-abdominal samples in patients undergoing abdominal surgery25.
SARS-CoV-2 has also been previously detected by RT–PCR in pleural fluid, showing reactive mesothelial cells and lymphocytes in cytology examination, in patients with lung infiltrates26, and in children27. Viral RNA can also be detected in pleural fluid in patients without lung parenchyma involvement28. We detected SARS-CoV-2 RNA in one of the two pleural effusion samples examined in a patient with chronic heart failure, with no evidence of lung infiltrates.
Bennett et al. described a cytology of pleural fluid showing mesothelial cells with large multiple nuclei, consistent with the viral cytopathic effect in a COVID-19 patient28. We could not find any cytopathic effect of the virus in any of the tissue or cytology samples examined.
Urine samples have been reported to be positive for SARS-CoV-2 RNA in less than 1–19% of COVID-19 patients with a low or moderate viral load (approximately 300 copies/mL)29,30. We also detected SARS-CoV-2 RNA in urine using qPCR with the N1 target gene in a patient with persistent COVID-19 and a previous history of urothelial carcinoma, in which positive SARS-CoV-2 RNA in urine was detected long after RT–PCR from a nasopharyngeal swab became negative. A higher mortality rate has been described in patients with SARS-CoV-2 RNA positivity in urine30. Our patient was alive and well after 20 months of follow-up.
Liquid-based cytology has also been used for cytological and immunocytochemical studies of samples obtained with nasopharyngeal swabs. In this series, it was also confirmed that no viral cytopathic effect was found in epithelial cells, and granular cytoplasmic immunocytochemical positivity was observed only in nasopharyngeal squamous cells of SARS-CoV-2 RT–PCR-positive patients. Macrophages and neutrophils did not show immunoreactivity. In this study, RT–PCR was not performed in liquid-based cytology medium31.
In our series, the qPCR technique identified SARS-CoV-2 in specimens within an interval between nasopharyngeal PCR positivity and tissue/cytology sample collection ranging from 0 to 170 days. Considering the results of the two qPCR techniques, SARS-CoV-2 RNA could be identified in 25.58% of the selected tissue/cytology samples in COVID-19 patients.
Correlation analysis established that the shorter the interval and persistence time, the higher the viral load of the sample. This analysis is relevant to take into consideration, as samples obtained and submitted to different departments within short periods of time should be treated with additional precautions, since they will probably contain a high viral load.
We conclude that IHC cannot be used as a screening method, but it can offer useful data in patients with previous qPCR positivity in tissue or cytology. Strong IHC positivity in placental samples also showing acute villitis must be considered during histopathological examination of the placenta in COVID-19-positive women, especially when a lower Cp value (high viral load) is detected in qPCR techniques in placental FFPE tissue. In selected cases, both S and N protein antibodies can be reliably used in IHC to detect SARS-CoV-26.
We recommend the use of qPCR in FFPE tissue and liquid cytology samples, where SARS-CoV-2 RNA can also be detected. The selection of adequate nucleic acid extraction and qPCR techniques used can be very important to obtain reliable results. In these samples, we found higher sensitivity when using the automated nucleic acid extraction method from Qiagen and SARS-CoV-2 nucleocapsid 1 (N1) primers and probes from IDT technologies.
SARS-CoV-2 can be detected in multiple organs, and in some cases (e.g., lung or placenta), it may be associated with known histopathological findings, but we need larger cohorts to understand the long-term role of the presence of this virus in some organs.
Liquid cytology and FFPE tissue blocks are the most valuable sources of biological samples in scientific research, although even in well-processed specimens, RNA is increasingly degraded in FFPE archival samples with time32. However, with adequate methodology, SARS-CoV-2 can be detected in FFPE tissue blocks more than one year after sample collection.