https://doi.org/10.1007/s00392-025-02737-x
1Herz- und Diabeteszentrum NRW Allgemeine und Interventionelle Kardiologie/Angiologie Bad Oeynhausen, Deutschland; 2Herz- und Diabeteszentrum NRW Klinik für Thorax- und Kardiovaskularchirurgie Bad Oeynhausen, Deutschland; 3Herz- und Diabeteszentrum NRW Klinik für Kardiologie Bad Oeynhausen, Deutschland
Objective
To assess the prognostic utility of peak postoperative high-sensitivity cardiac troponin i (hs-cTnI) among patients undergoing cardiac valvular (CV) or thoracic aortic (TA) surgery without concomitant coronary revascularization.
Methods
We conducted an analysis using data from a prospective, single-center registry of cardiac surgery patients at a tertiary care hospital in Germany. The study population included adult patients (≥18 years) who underwent elective and CV or TA surgery without concomitant coronary revascularization between between January 1st 2013, and May 30th 2019. Hs-cTnI levels were measured preoperatively (Baseline) and postoperatively at 4-hour intervals for the first 24 hours, followed by 8-hour intervals up to 48 hours after surgery. The association between peak postoperative hs-cTnI concentrations and clinical outcomes - all-cause long-term mortality and all-cause 30-day mortality - was assessed using cox proportional hazards models with regression splines to account for potential non-linear relationships. All models were adjusted for baseline risk using the European System for Cardiac Operative Risk Evaluation II Score (EuroSCORE II), which estimates perioperative mortality based on 18 clinical variables, including patient age and sex.
Results
Of the 4,232 patients included in the study (median age 69 years [interquartile rande (IQR): 59–76]; 38.4% female), all-cause 30-day mortality occurred in 61 patients (1.4%). during a median follow-up of 3.1 years [IQR: 1.9–5.0], 499 patients (11.8%) died from all causes. Multivariable analyses identified peak hs-cTnI thresholds of 282 times the upper limit of normal (ULN) for all-cause long-term mortality (hazard ratio (HR): 1.11, 95% CI: 1.00-1.24), and 194 times the ULN (HR: 1.25, 95% CI: 1.01-1.55) for all-cause 30-day mortality. Discriminatory capacity was moderate, with area under the curve (AUC) values of 0.56 for all-cause long-term and 0.55 for all-cause 30-day mortality. Sensitivity and specificity were 71% and 41% for long-term mortality, and 55% and 54% for 30-day mortality. Positive predictive values were consistently high at 90%, while negative predictive values remained at 16% and 14%.
Conclusion
Peak hs-cTnI levels are significantly associated with adverse outcomes in patients undergoing CV or TA surgery without concomitant coronary revascularization, with identified thresholds notably higher than current definitions for postoperative myocardial injury. These findings suggest that standard cutoffs may overestimate risk in this population, highlighting the need for refined hs-cTnI thresholds for more accurate perioperative risk assessment.
Figure 1: Unadjusted relationships between peak high-sensitivity cardiac troponin i (hs-cTnI) measurements 48 hours post-operatively as multiples of the upper limit of normal of 26 pg/ml (× ULN) and all-cause long-term mortality (Panel A) and all-cause 30-day mortality (Panel B).
Figure 2: Adjusted hazard ratio curves for all-cause long-term mortality (Panel A) and all-cause 30-day mortality (Panel B) as a function of peak high-sensitivity cardiac troponin i (hs-cTnI) measurements 48 hours post-operatively as multiples of the upper limit of normal of 26 pg/ml (× ULN) among patients who underwent isolated cardiac valvular or thoracic aortic surgery.