Abstract: Objective To evaluate the accuracy of cardiac troponin (cTn) levels in the diagnosis of acute myocardial infarction (AMI) in patients with chronic kidney disease (CKD) and explore a potential strategy for improving the diagnostic accuracy. Methods We retrospectively analyzed the data from patients with high- risk chest pain admitted in Zhujiang Hospital from January, 2018 to December, 2020, including 126 patients with and 272 patients without CKD, and 122 patients diagnosed to have AMI and 276 patients without AMI. The baseline clinical data of the patients and blood test results within 12 h after admission were collected. Results In patients without AMI, cTnT level was significantly higher in those with co-morbid CKD than in those without CKD (P<0.001), and showed a moderate negative correlation with eGFR (rs=? 0.501, P<0.001), while cTnI level did not differ significantly between the two groups (P=0.72). In patients with CKD, the optimal cutoff level was 0.177 μg/L for cTnT and 0.415 ng/mL for cTnI for diagnosis of AMI, for which cTnI had a higher specificity than cTnT. The diagnostic model combining both cTnT and cTnI levels [P=eY/(1+eY), Y=6.928 (cTnT)-0.5 (cTnI)-1.491] had a higher AUC value than cTn level alone. Conclusion In CKD patients, the cutoff level of cTn is increased for diagnosing AMI, and cTnI has a higher diagnostic specificity than cTnT. The combination of cTnT and cTnI levels may further improve diagnostic efficacy for AMI.

Key words: acute myocardial infarction; chronic kidney disease; troponin T; troponin I