Identifying biomarker-driven subphenotypes of cardiogenic shock: analysis of prospective cohorts and randomized controlled trials.

Summary

Unsupervised clustering of plasma biomarkers in two prospective cohorts identified four reproducible cardiogenic shock subphenotypes with distinct biology and mortality risk. Applying a simplified classifier to three completed RCTs improved risk discrimination beyond SCAI staging and suggested heterogeneity of treatment effect.

Key Findings

• Four biomarker-driven subphenotypes (adaptive, non-inflammatory, cardiopathic, inflammatory) were identified independently in two cohorts.
• Inflammatory and cardiopathic subphenotypes had the highest 28-day mortality; adding subphenotype membership improved Harrell’s C-index over SCAI stages.
• A simplified classifier assigned subphenotypes in three RCTs, enabling exploration of heterogeneity of treatment effect.

Clinical Implications

Subphenotype assignment may identify patients at highest risk (inflammatory/cardiopathic) and refine enrollment and therapy selection in trials (e.g., vasopressor/inotrope strategies, MCS timing), ultimately enabling targeted care pathways.

Limitations

• Retrospective biomarker availability and assay variability across cohorts and trials.
• Classifier simplification may lead to misclassification; no interventional testing of subphenotype-guided therapy.

Future Directions

Prospective, biomarker-integrated adaptive trials testing subphenotype-guided therapies; development of parsimonious clinical-biomarker panels and EHR integration for bedside classification.