Inflammatory cytokine signatures and factors associated with cardiogenic shock in acute coronary syndrome: Insights from an observational study

Abstract

Background: Cardiogenic shock (CS) complicating acute coronary syndrome (ACS) remains a leading cause of in-hospital cardiovascular mortality. Systemic inflammatory cytokines may contribute to CS pathophysiology but remain incompletely characterized. We aimed to delineate profiles of ACS patients with and without CS and evaluate their utility for CS identification.

Methods: In this single-centre observational study with a comparator sampling design, 150 ACS patients (August 2024-December 2025) were enrolled as ACS-CS (n = 60) and ACS-noCS (n = 90). A 13-analyte cytokine panel, selected a priori from an independent bioinformatics analysis of six GEO datasets, was quantified in serum by Luminex multiplex immunoassay within 24 hours of admission and before definitive therapy. Analyses comprised group comparison with Benjamini–Hochberg false discovery rate (FDR) correction, receiver operating characteristic (ROC) analysis, random-forest classification and multivariable logistic regression with Firth-penalised sensitivity analysis.

Results: ACS-CS patients were more frequently female (46.7% vs 22.2%; p = 0.002) and had higher diabetes prevalence (68.3% vs 37.8%; p < 0.001). Median SBP was lower (85 vs 127 mmHg; p < 0.001), LVEF reduced (27.5% vs 44.5%), BNP elevated (1009 vs 237.5 pg/mL; p < 0.001) and in-hospital mortality higher (35.0% vs 6.7%; p < 0.001). Eleven cytokines met both pre-specified criteria (FDR < 0.05 and log₂FC > 0.5). IL-6 was the strongest individual discriminator (log₂FC 2.62; AUC 0.870). A random-forest model incorporating the 13-cytokine panel achieved an internally cross-validated AUC of 0.958 (95% CI 0.92-0.98). IL-6, IL-10 and TNF-α remained independently associated with CS after adjustment for age and sex.

Conclusion: CS complicating ACS is accompanied by broad, coordinated systemic immune activation alongside hemodynamic failure. Multi-cytokine modelling showed high discrimination within this cohort, but all performance estimates are internal and potentially optimistic. These findings are hypothesis-generating and support prospective external validation of inflammatory profiling for risk stratification and as a rationale for immunomodulatory strategies.