Prognostic value of high-sensitivity cardiac troponin I after isolated non-CABG cardiac surgery

To evaluate the prognostic value of peak postoperative high-sensitivity cardiac troponin I (hs-cTnI) in patients undergoing cardiac valvular (CV) or thoracic aortic (TA) surgery without concomitant coronary artery bypass grafting (CABG).

We analyzed data from a prospective, single-center registry of adult patients (≥ 18 years) who underwent elective, isolated CV or TA surgery at a tertiary hospital in Germany between January 2013 and May 2019. Hs-cTnI levels were measured preoperatively and postoperatively at regular intervals up to 48 h. The associations between peak hs-cTnI and long-term and 30-day all-cause mortality were assessed using Cox proportional hazards models with regression splines, adjusted for baseline risk via EuroSCORE II.

Among 4232 patients (median age 69 years [IQR 59-76]; 38.4% female), 30-day all-cause mortality occurred in 61 patients (1.4%). Over a median follow-up of 3.1 years [IQR 1.9-5.0], 499 patients (11.8%) died. Peak hs-cTnI thresholds of 282 × the upper limit of normal (ULN) and 194 × ULN were associated with increased risk of long-term (HR 1.11, 95% CI 1.00-1.24) and 30-day mortality (hazard ratio (HR) 1.25, 95% CI 1.01-1.55), respectively. However, predictive performance was limited with area under the receiver operating curve (AUC) values of 0.56 for long-term mortality and 0.55 for 30-day mortality. Sensitivity and specificity were 41% and 71% for long-term mortality, and 54% and 55% for 30-day all-cause mortality. Negative predictive values remained high (both at 90%), but negative predictive values were low at 16% and 14%, respectively. When applying the sex-specific reference levels (16 pg/ml for women, 34 pg/ml for men), thresholds of 223 × ULN and 194 × ULN were identified for long-term and 30-day mortality, with corresponding HRs of 1.08 and 1.62 and AUCs of 0.58 and 0.57, respectively.

Peak hs-cTnI levels are independently associated with adverse outcomes following CV or TA surgery without CABG. However, effective risk discrimination is poor, and the identified thresholds far exceed conventional definitions of myocardial infarction. These findings support the need for surgery-specific hs-cTnI thresholds to improve perioperative risk stratification.