RR:C19 Evidence Scale rating by reviewer:
Reliable. The main study claims are generally justified by its methods and data. The results and conclusions are likely to be similar to the hypothetical ideal study. There are some minor caveats or limitations, but they would/do not change the major claims of the study. The study provides sufficient strength of evidence on its own that its main claims should be considered actionable, with some room for future revision.
***************************************
Review: The van Weyenbergh and colleagues at KU Leuven analyzed 48 patients enrolled from a single hospital, for which a long-COVID diagnostic (WHO criteria) was made. Patient-reported and clinician-reported outcomes, as well as comorbidities, were documented and analyzed in conjunction with blood transcriptome data assessed digitally (nCounter, Nanostring) in a customized gene panel, including transcripts related to Myeloid and Innate Immunity, SARS-CoV-2 and ACE2/TMPRSS2 receptors.
The authors identified 212 genes differentially expressed in long-COVID patients, including SARS-CoV-2 genes (upregulated Nucleocapsid, ORF7A, ORF3A, Mpro, and ORF1ab, indicating active, ongoing viral replication and downregulated Spike level), SARS-CoV-2 related host genes (ACE2/TMPRSS2 receptor, DPP4/FURIN protease), memory B-cells (CD27/IGHE/BMP8A) and platelets (CD99/PBX1/PDZK1IP1). Neither age nor sex is associated with the altered blood transcript status; the number of comorbidities is associated with increased odds of blood viral load, while the number of vaccines is associated with lower odds of high transcript status. Pathway analysis showed a significantly decreased lymphocyte activation and immunometabolism, negatively correlating with the blood viral load.
Overall, the methodologies employed to generate and analyze transcriptome data are appropriate, and the integration with the patients’ clinical status is judicious. Nevertheless, I have several minor suggestions that might improve the quality of the manuscript. They are as follows:
It is not clear how the sample size was calculated. Is the number of 48 patients enrolled enough to provide adequate power for the statistical analysis?
A nomogram and a decision curve analysis based on the variables of the proposed prediction model would be useful for assessing the full clinical usefulness of their results.
Given that the patients were recruited from a single center, the low number of patients included, and the assumption that the cohort is representative of the entire long-COVID population, I would suggest using a bootstrapping method to estimate the intrinsic reliability of the authors’ statistical approach/model and to assess for overfitting.
The discussion should place the results in the context of the already published data on long-COVID (at least blood) transcriptome changes.
How do the authors explain the low level of Spike RNA, given the presumed ongoing, active viral replication suggested by the high levels of ORF1ab?
Creative Commons Attribution 4.0 International License (CC-BY 4.0)
