Profiling of immune dysfunction in COVID-19 patients allows early prediction of disease progression

  1. André F Rendeiro
  2. Joseph Casano
  3. Charles Kyriakos Vorkas
  4. Harjot Singh
  5. Ayana Morales
  6. Robert A DeSimone
  7. Grant B Ellsworth
  8. Rosemary Soave
  9. Shashi N Kapadia
  10. Kohta Saito
  11. Christopher D Brown
  12. JingMei Hsu
  13. Christopher Kyriakides
  14. Steven Chiu
  15. Luca Vincenzo Cappelli
  16. Maria Teresa Cacciapuoti
  17. Wayne Tam
  18. Lorenzo Galluzzi
  19. Paul D Simonson
  20. Olivier Elemento
  21. Mirella Salvatore
  22. Giorgio Inghirami

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

With a rising incidence of COVID-19–associated morbidity and mortality worldwide, it is critical to elucidate the innate and adaptive immune responses that drive disease severity. We performed longitudinal immune profiling of peripheral blood mononuclear cells from 45 patients and healthy donors. We observed a dynamic immune landscape of innate and adaptive immune cells in disease progression and absolute changes of lymphocyte and myeloid cells in severe versus mild cases or healthy controls. Intubation and death were coupled with selected natural killer cell KIR receptor usage and IgM+ B cells and associated with profound CD4 and CD8 T-cell exhaustion. Pseudo-temporal reconstruction of the hierarchy of disease progression revealed dynamic time changes in the global population recapitulating individual patients and the development of an eight-marker classifier of disease severity. Estimating the effect of clinical progression on the immune response and early assessment of disease progression risks may allow implementation of tailored therapies.

Article activity feed

  1. Version published to 10.26508/lsa.202000955
  2. ScreenIT

    SciScore for 10.1101/2020.09.08.20189092: (What is this?)

    Please note, not all rigor criteria are appropriate for all manuscripts.

    Table 1: Rigor

    Institutional Review Board StatementIRB: Study design, sample acquisition, and clinical data: The study was approved by the Institutional Review Board of Weill Cornell Medicine.
    Consent: Informed consent was obtained from all participants.
    RandomizationThe cross-validation loop was repeated 100 times and models were fit with real or randomized labels.
    Blindingnot detected.
    Power Analysisnot detected.
    Sex as a biological variablenot detected.

    Table 2: Resources

    Antibodies
    SentencesResources
    CD56+, CD16 bright, mature NK cells were then interrogated for their reactivity with individual anti-KIR (CD158) and anti-NKG2A (CD159a) antibodies with gates on histogram plots.
    anti-KIR
    suggested: None
    CD158
    suggested: None
    anti-NKG2A ( CD159a
    suggested: (Abcam Cat# ab24849, RRID:AB_448437)
    Software and Algorithms
    SentencesResources
    Supervised quantification of immune cell populations (gating): Immune populations were quantified by manual analysis with BD FACSDiva.
    BD FACSDiva
    suggested: (BD FACSDiva Software, RRID:SCR_001456)
    Absolute counts of populations were exported to CSV and relative population sizes were calculated in Microsoft Excel.
    Microsoft Excel
    suggested: (Microsoft Excel, RRID:SCR_016137)
    Compensation was applied using FlowKit60 version 0.5.0, and an inverse hyperbolic transformation (AsinhTransform) was applied with parameters t = 10000, m = 4.5, a = 0.
    FlowKit60
    suggested: None
    To cluster the abundance of immune populations based on their dynamics over the pseudotime axis, the same kernels were used to fit a Mixture of Hierarchical Gaussian Processes (MOHGP) as implemented in the GPClust package67,68 using 8 as an initial guess of number of clusters.
    GPClust
    suggested: None
    Visualizations were produced using Gephi version 0.9.2 with the Force Atlas2 layout with parameters “LinLog mode”, “scaling factor” 8.0, and “gravity” 11.0.
    Gephi
    suggested: (Gephi, RRID:SCR_004293)
    Prediction of disease severity from immunotypes: A Random Forest Classifier was trained as implemented in scikit-learn framework64 (version 0.23.0) to distinguish between cases with “mild” and “severe” disease using 10-fold cross validation.
    scikit-learn
    suggested: (scikit-learn, RRID:SCR_002577)

    Results from OddPub: Thank you for sharing your code and data.

    Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:
    Although our study confirmed some findings and provides new data on the innate immune landscape of COVID-19 patients, we recognize limitations that could be overcome with larger sample sizes and matched control populations. Indeed, large population studies may discover new immune populations and find a relationship with disease progression or clinical factors. In this current pandemic, profiling a larger sample of the population and investigating multiple time points may help identify viral adaptation to the host (particularly when coupled to analysis of viral sequence) in patients with different outcomes. These programs may be achieved if an effective institutional organization, multicentric networking, and substantial financial support are available. The targeted nature of flow cytometry interrogates limited sets of immune populations and implies that only certain molecules can be effectively profiled. In our study we used mainly proportional data when comparing the abundance of immune populations between patient groups. While this may not necessarily imply absolute changes in cell numbers, we observed good overall agreement between changes in proportions and absolute counts when comparing severe and mild disease status (Supplementary Figure 6 and Supplementary Table 7). This highlights the importance of studies employing orthogonal modalities such as cytokine profiling51, single-cell RNA sequencing52–56, and their integration57,58. Nevertheless, even without orthogonal stu...

    Results from TrialIdentifier: No clinical trial numbers were referenced.

    Results from Barzooka: We did not find any issues relating to the usage of bar graphs.

    Results from JetFighter: We did not find any issues relating to colormaps.

    Results from rtransparent:
    • Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
    • Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
    • No protocol registration statement was detected.

    About SciScore

    SciScore is an automated tool that is designed to assist expert reviewers by finding and presenting formulaic information scattered throughout a paper in a standard, easy to digest format. SciScore checks for the presence and correctness of RRIDs (research resource identifiers), and for rigor criteria such as sex and investigator blinding. For details on the theoretical underpinning of rigor criteria and the tools shown here, including references cited, please follow this link.

  3. Version published to 10.1101/2020.09.08.20189092v1 on medRxiv