Article Text

Admission mid-regional pro-adrenomedullin is a more potent predictor of early adverse events than Grace clinical risk score in patients with non-ST elevation acute coronary syndromes
  1. OS Dhillon1,
  2. SQ Khan1,
  3. S Ponnambath2,
  4. PA Quinn1,
  5. IB Squire1,
  6. J Davies1,
  7. LL Ng1
  1. 1University of Leicester, Leicester, UK
  2. 2Leeds Teaching Hospital NHS Trust, Leeds, UK


Introduction Adrenomedullin is elevated in heart failure (HF) and post-acute coronary syndrome (ACS). Mid-regional pro-adrenomedullin (MRproADM) is more stable and has been shown to have prognostic value after ST elevation ACS.

Aims To assess the prognostic value of admission (A) and discharge (D) MRproADM levels in unselected patients with non-ST elevation ACS (NSTE-ACS) for major adverse cardiac events (MACE) including all-cause mortality, readmission with HF or recurrent myocardial infarction (re-MI) both as a primary composite endpoint and separate secondary endpoints and then compare it with the well-established Grace clinical risk score (calculated on discharge) to evaluate its clinical utility as a risk marker.

Method and Results Plasma MR-proADM was measured on admission and discharge in 745 (514 men, median age 70.0 ± 12.7 years) consecutive patients with NSTE-ACS. During follow-up (1–2837 days) 120 (16.1%) patients died, 65 (8.7%) were readmitted with HF and 77 (10.3%) had re-MI. A-MRproADM and D-MRproADM levels were entered into two separate multivariate Cox regression models adjusted for important clinical and biochemical factors to compare their predictive power. Receiver operator characteristic (ROC) analyses evaluating areas under the curve (AUC) were also performed, with the results shown in the table. Both were strong predictors of MACE along with age (hazard ratio 1.02, p = 0.019). A-MRproADM appears to be a more powerful predictor of death than D-MRproADM levels; however, D-MRproADM appears to predict readmission with HF more strongly, leading to a similar ability to predict MACE overall. For early mortality up to 30 days, Cox modelling yielded HR 116.88, p<0.001 for A-MRproADM and HR 27.89, p = 0.001 for D-MRproADM (see table 1). As A-MRproADM is a stronger predictor of death than D-MRproADM it was used as a primary comparator against Grace scoring. Both Grace and A-MRproADM were significant independent predictors of MACE (HR 6.86 and HR 1.01, respectively, both p<0.001). Comparison of the MACE rates according to tertiles of A-MRproADM and Grace scores were evaluated, with the results shown in the fig, which reveals statistically significant stratification of patients into low to high-risk groups (Mantel–Cox log rank 18.39 p<0.001 pooled comparison across model). The Grace score is designed to predict mortality from discharge to 6 months, therefore we performed Cox analyses with death censored at 1 and 6 months comparing it with A-MRproADM with results shown in table 2. At 1 month the ROC AUC was 0.90 for A-MRproADM and 0.74 for the Grace score, both p<0.001.

Conclusion Both admission and discharge MRproADM levels are strong prognostic markers of adverse events in NSTE-ACS and aid further risk stratification of patients when used in combination with the Grace score. A-MRproADM is a more potent predictor of early events than Grace scoring in this cohort and may represent an important clinical tool.

Abstract 077 Figure

MCE, major cardiac event.

Abstract 077 Table 1

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.