Detection of cytomegalovirus (CMV) by digital PCR in stool samples for the non-invasive diagnosis of CMV gastroenteritis.

Background: CMV gastroenteritis is common in patients receiving allogeneic hematopoietic stem cell transplantation and it is difficult to distinguish from acute graft-versus-host disease (aGvHD), which has very similar symptoms but needs quite different treatment. CMV gastroenteritis is caused by local infection or reactivation of CMV in the gastrointestinal tract while aGvHD is due to immune rejection. The gold standard of diagnosis of CMV gastroenteritis and aGvHD is gastrointestinal biopsy under endoscopy, which is invasive and can potentially lead to severe side effects. Stool samples testing with quantitative polymerase chain reaction (qPCR) may be an alternative, while the application in trace level measurements and precision are not all satisfactory enough in reported research.
Methods: In this study, we designed a novel method that extracted the cell free DNA (cfDNA) from the fecal supernatant to perform digital PCR (dPCR) for the detection of CMV, analyzed the performance and compared it with the total DNA extracted by the current procedure.
Results: Twenty-two paired stool samples using two DNA extraction methods proved that the cfDNA extraction method had markedly higher DNA concentrations and control gene copy number, suggesting that cfDNA may be more informative and more useful for the detection of CMV DNA segment. The dPCR approach in detecting CMV DNA segment also exhibit good linearity (R ²  = 0.997) and higher sensitivity (limit of detection at 50% was 3.534 copies/μL). Eighty-two stool samples from 44 immunocompromised patients were analyzed, CMV-positive rate was 28%, indicating that more than one-quarter of the gastrointestinal symptoms within these patients may be caused by CMV infection or reactivation.
Conclusion: The combined results suggest that detection of CMV by dPCR in cfDNA of stool supernatant is a powerful method to identify CMV gastroenteritis and helps in clinical treatment decision making.
(© 2022. The Author(s).)