Demonstrates CCL20 effectively distinguishes inflammatory subphenotypes in critically ill children, suggesting its prognostic significance.
Rationale Molecular subphenotyping using protein biomarkers has emerged as an effective tool for prognostic enrichment and risk stratification in critically ill patients. Such subphenotypes have largely been derived from a limited set of preestablished pro- and anti-inflammatory cytokines and chemokines. However, the broader proteomic context underlying subphenotypes has not been adequately tested. Here, we studied a large panel of inflammatory markers to more comprehensively determine discrimination of subphenotypes in the plasma proteome. Methods We analyzed the baseline expression of 249 inflammation related biomarkers from pediatric patients (n = 84) in the CAF-PINT study of heart and lung failure using a multiplex panel (NULISA, Alamar Biosciences). Age- and sex-matched samples were grouped according to subphenotypes (hypo- or hyper-inflammatory) previously derived through latent class analysis (LCA), and undersampling was applied to generate evenly split subsets (n = 42 per group). We compared a significant outlier marker, CCL20, with known subphenotype-defining markers (IL6, IL10) to assess discrimination of subphenotypes by each single marker. Receiver operating characteristic (ROC) curves were built using linear regression models of each of these markers and were compared against each other using DeLong tests. To validate these results, we assayed CCL20 levels using a rapid immunoassay in an age- and sex-matched subset of critically ill, pediatric patients (n = 116) in the PALI study of ARDS patients with previously identified LCA subphenotypes (n = 59 hyperinflamed, n = 57 hypoinflamed). Subphenotypes predicted by CCL20 alone were compared to true subphenotypes by enrichment for 90-day mortality. Results With CAF-PINT, we identified 137 differentially expressed markers (False Discovery Rate (FDR)<0.05). CCL20 was the most enriched among hyperinflammatory patients both in the differential analysis (FDR=8.2e-15) and penalized regression model (Odds Ratio (OR)=1.70). In CAF-PINT, CCL20 (AUC=0.952) had better discrimination of inflammatory subphenotypes as a single marker than IL6 (AUC=0.893, DeLong p = 0.04) and trended higher than IL10 (AUC=0.887, DeLong p = 0.09) [See Figure 1]. CCL20 was validated as a better single marker model in the PALI cohort (AUC=0.908) compared to IL6 (AUC=0.723, DeLong p = 0.001) and IL10 (AUC=0.735, DeLong p = 0.002). Mortality enrichment was higher among hyper inflammatory classes predicted by CCL20 compared to true hyper inflammatory classes (90-day mortality OR: 2.83 vs. 2.12). Conclusion CCL20, a chemoattractant for dendritic cells, B and T lymphocytes, especially IL-17 producing Th17 cells, results in better discrimination of molecular subphenotypes than IL6 and IL10 in two cohorts of critically ill children suggesting prognostic and mechanistic significance of this novel marker in the context of systemic inflammation and treatment heterogeneity.
No takes yet. Share an insight, caveat, or question.
Reddy et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: