Background and Aims

Elevated cardiac troponin levels are a frequent finding in emergency department patients, often without a clear cause. Current high-sensitivity cardiac troponin T (cTnT) assays measure intact and fragmented cardiac troponin T (total cTnT) molecules, without distinguishing between them. This study investigated whether measuring only intact and minimally fragmented cTnT (long cTnT) provides additional value for myocardial infarction (MI) identification.

Methods

Consecutive emergency department patients with standard high-sensitivity cTnT levels (Roche Diagnostics) above the upper reference limit (≥14 ng/L) were recruited. Long cTnT levels were measured using a novel immunoassay. The additional diagnostic value of long cTnT in identifying patients with type 1 MI or any MI was assessed.

Results

A total of 1811 patients participated in the study, 1145 (63.2%) presenting with chest pain or dyspnoea. Overall, 205 (11.3%) had MI, including 148 classified as type 1 MI. Only .7% of patients in the lowest long cTnT tertile (<3.7 ng/L) had type 1 MI. The discriminative ability of long cTnT was superior to total cTnT in identifying patients with MI (area under curve [95% confidence intervals]) for any MI: .833 (.804–.863) vs .782 (.744–.819), and for type 1 MI: .839 (.807–.872) vs .777 (.735–.819), both (P < .001). Integrating the predictive data from long cTnT with total cTnT provided additional value in both reclassification and decision curve analyses, compared to total cTnT data alone.

Conclusions

The long cTnT assay demonstrated good diagnostic performance in identifying MI in patients with elevated total cTnT levels, with the potential to improve the accuracy of MI diagnosis.