Evaluating the prognostic value of admission IL-8 and sST2 as biomarkers of early myocardial injury in severe community-acquired pneumonia

Background Myocardial injury in severe community-acquired pneumonia (SCAP) is often under-recognized, and interpretation of high-sensitivity cardiac troponin I (hs-cTnI) may be complicated by renal dysfunction and critical illness. We evaluated whether admission interleukin-8 (IL-8) and soluble suppression of tumorigenicity 2 (sST2) could support early risk stratification of guideline-consistent myocardial injury in patients with SCAP.Methods In this prospective cohort of 146 patients with SCAP, guideline-consistent myocardial injury was defined as hs-cTnI elevation above the assay-specific 99th percentile upper reference limit. Admission IL-8, sST2, and their combination were assessed using receiver operating characteristic curve analysis and 10-fold cross-validation. Fixed and estimated glomerular filtration rate (eGFR)-stratified hs-cTnI thresholds were compared in an exploratory analysis. An exploratory subgroup analysis was performed among patients with CURB-65 ≤ 1.Results The median age was 71 years; 91 patients (62.3%) were male and 55 (37.7%) were female. Myocardial injury occurred in 88 patients (60.3%). The IL-8 + sST2 model showed good discrimination for myocardial injury, with an AUC of 0.825 (95% CI: 0.751–0.895) and an internally cross-validated AUC of 0.812. Admission IL-8 showed stronger individual discrimination than sST2. eGFR-stratified hs-cTnI thresholds did not reduce discordant hs-cTnI elevations in this cohort, likely because few patients had moderate-to-severe renal dysfunction. Among patients with CURB-65 ≤ 1 (n = 38), IL-8 > 43.53 pg/mL was associated with a higher prevalence of myocardial injury (76.2% vs. 29.4%, p = 0.004). An exploratory 0–2 biomarker score showed a graded increase in myocardial injury prevalence across score categories: 15.4%, 62.5%, and 78.6%.Conclusions Admission IL-8 was associated with guideline-consistent myocardial injury in patients with SCAP, whereas sST2 showed weaker individual discrimination. The combined model had numerically higher discrimination than IL-8 alone, but the incremental contribution of sST2 remains uncertain. These findings represent an initial evaluation and require external validation before clinical implementation.