The adrenomedullin system predicts cardiogenic shock in acute coronary syndromes

Background Cardiogenic shock (CS) is among the most dreadful complications of acute coronary syndromes (ACS). Endothelial dysfunction and vascular leakage are hallmarks of CS pathophysiology; however, biomarkers reflecting these early processes are lacking. The adrenomedullin (ADM) system—the inactive precursor glycine-extended ADM (ADM-Gly), the activating enzyme peptidylglycine α-amidating monooxygenase (PAM), and bioactive ADM (bio-ADM)—modulates vascular tone and permeability. Purpose To investigate associations of ADM system components with CS risk after ACS. Methods ADM-Gly, PAM, and bio-ADM were measured in 4,098 (SPUM-ACS, Switzerland) and 824 (FAST-MI, France) ACS patients without CS on admission. The primary endpoint was in-hospital CS. Associations were evaluated using multivariable-adjusted logistic regression models with sequential adjustment for demographic and clinical risk factors, NT-proBNP, and haemodynamic indices. Restricted cubic spline (RCS) analyses were applied to assess potential nonlinear dose–response relationships across biomarker concentrations. Incremental prognostic value beyond the ORBI risk score was quantified by comparing models with and without ADM biomarkers, assessing changes in discrimination (ΔAUC), reclassification (NRI), and model fit (ΔAIC), with external validation in FAST-MI. Results Higher ADM-Gly and bio-ADM, but not PAM, were associated with systemic inflammation and haemodynamic compromise (Figure 1). In RCS analyses, both biomarkers showed non-linear dose–response relationships with CS risk. Both ADM-Gly and bio-ADM independently predicted CS in SPUM-ACS (adjusted odds ratio (OR) per log2 increase, 1.44, 95% confidence interval (CI) 1.22–1.71, P<.001 and OR 1.37, 95% CI 1.10–1.70, P=.004) and FAST-MI (OR 1.82, 95% CI 1.34–2.48, P<.001 and OR 1.55, 95% CI 1.15–2.10, P=.004). Addition of these two biomarkers to the ORBI score improved discrimination (ΔAUC 0.02), reclassification (NRI 0.229), and model fit (ΔAIC −26.9), with consistent findings in the external validation cohort (Figure 2). PAM showed no independent prognostic value. Conclusions ADM-Gly and bio-ADM, but not PAM, independently predict in-hospital CS in patients with ACS and improve early risk stratification.