Diagnostic Performance of Novel Troponin Algorithms for the Rule-Out of Non-ST-Elevation Acute Coronary Syndrome

Abstract

Background The European Society of Cardiology (ESC) rule-out algorithms use cutoffs optimized for exclusion of non-ST elevation myocardial infarction (NSTEMI). We investigated these and several novel algorithms for the rule-out of non-ST elevation acute coronary syndrome (NSTE-ACS) including less urgent coronary ischemia.

Method A total of 1504 unselected patients with suspected NSTE-ACS were included and divided into a derivation cohort (n = 988) and validation cohort (n = 516). The primary endpoint was the diagnostic performance to rule-out NSTEMI and unstable angina pectoris during index hospitalization. The secondary endpoint was combined MI, all-cause mortality (within 30 days) and urgent (24 h) revascularization. The ESC algorithms for high-sensitivity cardiac troponin T (hs-cTnT) and I (hs-cTnI) were compared to different novel low-baseline (limit of detection), low-delta (based on the assay analytical and biological variation), and 0–1-h and 0–3-h algorithms.

Results The prevalence of NSTE-ACS was 24.8%, 60.0% had noncardiac chest pain, and 15.2% other diseases. The 0–1/0–3-h algorithms had superior clinical sensitivity for the primary endpoint compared to the ESC algorithm (validation cohort); hs-cTnT: 95% vs 63%, and hs-cTnI: 87% vs 64%, respectively. Regarding the secondary endpoint, the algorithms had similar clinical sensitivity (100% vs 94%–96%) but lower clinical specificity (41%–19%) compared to the ESC algorithms (77%–74%). The rule-out rates decreased by a factor of 2–4.

Conclusion Low concentration/low-delta troponin algorithms improve the clinical sensitivity for a combined endpoint of NSTEMI and unstable angina pectoris, with the cost of a substantial reduction in total rule-out rate. There was no clear benefit compared to ESC for diagnosing high-risk events.