Topological data analysis identifies distinct biomarker phenotypes during the ‘inflammatory’ phase of COVID-19
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (ScreenIT)
Abstract
OBJECTIVES
The relationships between baseline clinical phenotypes and the cytokine milieu of the peak ‘inflammatory’ phase of coronavirus 2019 (COVID-19) are not yet well understood. We used Topological Data Analysis (TDA), a dimensionality reduction technique to identify patterns of inflammation associated with COVID-19 severity and clinical characteristics.
DESIGN
Exploratory analysis from a multi-center prospective cohort study.
SETTING
Eight military hospitals across the United States between April 2020 and January 2021.
PATIENTS
Adult (≥18 years of age) SARS-CoV-2 positive inpatient and outpatient participants were enrolled with plasma samples selected from the putative ‘inflammatory’ phase of COVID-19, defined as 15-28 days post symptom onset.
INTERVENTIONS
None.
MEASUREMENTS AND MAIN RESULTS
Concentrations of 12 inflammatory protein biomarkers were measured using a broad dynamic range immunoassay. TDA identified 3 distinct inflammatory protein expression clusters. Peak severity (outpatient, hospitalized, ICU admission or death), Charlson Comorbidity Index (CCI), and body mass index (BMI) were evaluated with logistic regression for associations with each cluster. The study population (n=129, 33.3% female, median 41.3 years of age) included 77 outpatient, 31 inpatient, 16 ICU-level, and 5 fatal cases. Three distinct clusters were found that differed by peak disease severity (p <0.001), age (p <0.001), BMI (p<0.001), and CCI (p=0.001).
CONCLUSIONS
Exploratory clustering methods can stratify heterogeneous patient populations and identify distinct inflammation patterns associated with comorbid disease, obesity, and severe illness due to COVID-19.
Article activity feed
-
SciScore for 10.1101/2021.12.25.21268206: (What is this?)
Please note, not all rigor criteria are appropriate for all manuscripts.
Table 1: Rigor
Ethics IRB: The protocol was approved by the Uniformed Services University Institutional Review Board (IDCRP-085)(13).
Consent: All patients provided written informed consent.Sex as a biological variable not detected. Randomization not detected. Blinding not detected. Power Analysis not detected. Table 2: Resources
Software and Algorithms Sentences Resources All statistical analyses were performed in Stata (version 15.0; StataCorp LLC, College Station, TX, USA) and R version 4.0.2 (21) StataCorpsuggested: (Stata, RRID:SCR_012763)Results from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).
Results from Limitation…SciScore for 10.1101/2021.12.25.21268206: (What is this?)
Please note, not all rigor criteria are appropriate for all manuscripts.
Table 1: Rigor
Ethics IRB: The protocol was approved by the Uniformed Services University Institutional Review Board (IDCRP-085)(13).
Consent: All patients provided written informed consent.Sex as a biological variable not detected. Randomization not detected. Blinding not detected. Power Analysis not detected. Table 2: Resources
Software and Algorithms Sentences Resources All statistical analyses were performed in Stata (version 15.0; StataCorp LLC, College Station, TX, USA) and R version 4.0.2 (21) StataCorpsuggested: (Stata, RRID:SCR_012763)Results from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).
Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:While this study, to our knowledge, is the first to use an advanced dimensionality reduction approach to understand relationships between biomarker patterns and clinical phenotypes during the inflammatory phase of COVID-19, there are limitations worth noting. Samples were collected from April 2020 to January 2021 and treatment practices and epidemiologic changes over time may have affected inflammation patterns. Hence, we incorporated a sensitivity analysis excluding those that received systemic steroids in Cluster 1 to aid in interpreting the findings. In addition, the sample size may limit our ability to identify uncommon biomarker patterns and external validation is needed of patterns identified. Additionally, regression was used to adjust for peak severity to identify biomarker and comorbid condition associations with TDA clusters distinct from severity trajectory differences. While this is a novel feature of this biomarker study, residual confounding related to peak severity remains possible. Despite limitations, results presented here are hypothesis generating and should be evaluated further in additional cohorts. This approach constitutes an early exploratory step in identifying host biomarker patterns that may be leveraged for personalized interventions, and offers new insights for COVID19 prognosis, therapy, and prevention with techniques that could be extended to understanding other severe infections. Using analytes identified from our international sepsis cohort re...
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.
Results from scite Reference Check: We found no unreliable references.
-
