The adrenomedullin system and risk of cardiogenic shock or death complicating acute coronary syndromes

Abstract Introduction Cardiogenic shock (CS) is the leading cause of death due to acute coronary syndromes (ACS) and mortality remained nearly unchanged over the past two decades. Immediate revascularization of the infarct-related artery is the treatment strategy of choice to reduce CS-related mortality, but early identification of patients at high risk of CS and death is challenging. The adrenomedullin (ADM) system modulates vascular tone and endothelial barrier function and monitoring its activity may enhance early risk assessment of ACS patients prone to experience CS and ultimately death. Methods In 4293 ACS patients, prospectively recruited at 4 Swiss university hospitals, the 3 major ADM system components were assessed in plasma at the time of acute presentation by blinded study personnel: ADM-glycine (its inactive substrate), peptidylglycine α-amidating monooxygenase (PAM; the ADM-amidating enzyme), and the biologically active ADM-amide (bio-ADM). Patients were monitored for in-hospital development of CS and followed at 30 days and 1 year with external event adjudication. The relationship of each ADM system component with features of CS pathophysiology (such as proxies for the decrease in cardiac contractile mass, increase in cardiac stress or systemic inflammation) was estimated by means of partial correlation coefficients. Multivariable-adjusted regression models were fit to assess their independent association with CS development and death. Results Baseline ADM-glycine, PAM and bio-ADM levels were tightly linked to features of CS (Fig. 1), with each ADM system component associating with in-hospital CS (odds ratio, 95% confidence interval [CI] per doubling in each, 2.04, 1.79-2.33, P<.001; 1.91, 1.63-2.23, P<.001; 1.72, 1.11-2.66, P=.016). The association of each marker with CS remained robust after adjusting for established risk factors for in-hospital development of CS, including all hemodynamic-, patient-, and procedure-related variables informing the ORBI risk score (1.65, 1.35-2.00, P<.001; 1.50, 1.26-1.77, P<.001; and 1.87, 1.13-3.11, P=.015). Notably, integration of ADM-glycine, PAM, and bio-ADM into the ORBI risk prediction model conferred increased discriminative performance for the prediction of in-hospital CS (AUC, 0.76 vs. 0.81, P<.001), ranking the ADM system second among all variables informing the score (Fig. 2). While both ADM-glycine and bio-ADM independently associated with mortality risk at 30-days (multivariable-adjusted hazard ratio per doubling, 95% CI, 1.93, 1.54-2.42, P<.001; and 1.81, 1.34-2.44, P<.001) and 1-year (1.41, 1.18-1.68, P<.001; and 1.52, 1.24-1.87, P<.001), PAM showed no association with these outcomes (P>.05). Conclusions Monitoring key regulators of the ADM system improves early risk assessment of ACS patients at high risk of CS, with admission ADM-glycine and bio-ADM representing independent risk factors for 30-day and 1-year mortality post-ACS.