To determine whether severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) can establish a faecal–oral transmission route, it is essential to confirm that infectious virus particles are shed in faeces from patients with COVID-19. We would like to thank Pedersen et al. for their Correspondence on our Perspective (Guo, M. et al. Potential intestinal infection and faecal–oral transmission of SARS-CoV-2. Nat. Rev. Gastroenterol. Hepatol. 18, 269–283 (2021))1, which raises some important issues (Pedersen et al. Rectally shed SARS-CoV-2 lacks infectivity: time to rethink faecal–oral transmission? Nat. Rev. Gastroenterol. Hepatol. https://doi.org/10.1038/s41575-021-00501-w (2021))2.

Our Perspective refers to four studies that discuss isolation of infectious virus particles from patient faeces3,4,5,6. Two of these studies5,6 only reported data obtained by electron microscopy, which are indeed insufficient to determine infectivity, as argued by Pedersen et al.2. However, Zhou and colleagues reported the extraction of infectious virus particles from a patient with COVID-19 based on increases in viral load in human intestinal organoids4. Also, Xiao et al. reported isolation of SARS-CoV-2 particles from two of three stool samples3; the viral isolates, first observed by electron microscopy, could successfully infect Vero cells (confirmed by immunofluorescence staining)3. More recently, when three viral isolates were extracted from stool samples of patients with COVID-19 and added to Vero cells7, the culture supernatants tested positive for SARS-CoV-2 by quantitative reverse transcription PCR. The infectious viral particles were collected from culture supernatant viral stock, then frozen and subsequently thawed stock were able to infect Vero and Calu-3 (human lung adenocarcinoma) cells. Two of the three viral faecal isolates were more infectious in cells than viral isolates extracted from nasopharyngeal swabs and sputum in vitro7. These results strongly suggested that infectious virus could be shed in faecal samples from patients with COVID-19.

Notably, three of five studies mentioned here also included patient information3,6,7, and all of the viral isolates in these studies were extracted from patients with severe disease. Without more information, we cannot exclude the possibility that one study mentioned by Pederson et al. that failed to isolate infectious SARS-CoV-2 from stool could be attributable to a mild disease course in those patients8. Moreover, Xiao et al. failed to isolate virus from stool at later time points, indicating the absence of infectious virus, with only viral fragments apparently shed in faeces during late stages of the disease3. Thus, disease severity and time course for sampling could be determining factors in the isolation of infectious viral particles.

Research in animals might also provide valuable insights. One study investigating SARS-CoV-2 infection in rhesus macaques reported the successful isolation of infectious viruses from faeces, and confirmed this finding by electron microscopy and TCID50 (median tissue culture infective dose) assays (105/ml)9. Similarly, a study of SARS-CoV-2 infection in ferrets showed that oral inoculation with faecal supernatants of infected specimens could establish infection in naive ferrets10. These animal studies therefore demonstrated the presence of infectious virus in the faeces of COVID-19 animal models. Although research has suggested that infectious viral particles can be isolated from both humans and animals, based on the available evidence (Table 1), a clear faecal–oral transmission route has not been established for SARS-CoV-2.

Table 1 Research on the infectivity of virus shedding in faecal samples in human and animal studies