Predictive value of soluble urokinase-type plasminogen activator receptor for mortality in patients with suspected myocardial infarction

Nils A Sörensen,Julius Nikorowitsch,Johannes T Neumann,Nicole Rübsamen,Alina Goßling,Tau S Hartikainen,Stefan Blankenberg,Dirk Westermann,Tanja Zeller,Mahir Karakas

DOI: https://doi.org/10.1007/s00392-019-01475-1

Abstract:Background: Early risk stratification of patients with suspected acute myocardial infarction (AMI) constitutes an unmet need in current daily clinical practice. We aimed to evaluate the predictive value of soluble urokinase-type plasminogen activator receptor (suPAR) levels for 1-year mortality in patients with suspected AMI. Methods and results: suPAR levels were determined in 1314 patients presenting to the emergency department with suspected AMI. Patients were followed up for 12 months to assess all-cause mortality. Of 1314 patients included, 308 were diagnosed with AMI. Median suPAR levels did not differ between subjects with AMI compared to non-AMI (3.5 ng/ml vs. 3.2 ng/ml, p = 0.066). suPAR levels reliably predicted all-cause mortality after 1 year. Hazard ratio for 1-year mortality was 12.6 (p < 0.001) in the quartile with the highest suPAR levels compared to the first quartile. The prognostic value for 6-month mortality was comparable to an established risk prediction model, the Global Registry of Acute Coronary Events (GRACE) score, with an AUC of 0.79 (95% CI 0.72-0.86) for the GRACE score and 0.77 (95% CI 0.69-0.84) for suPAR. Addition of suPAR improved the GRACE score, as shown by integrated discrimination improvement statistics of 0.036 (p = 0.03) suggesting a further discrimination of events from non-events by the addition of suPAR. Conclusions: suPAR levels reliably predicted mortality in patients with suspected AMI. Study registration: http://www.clinicaltrials.gov (NCT02355457).