Background: Rapid diagnosis of acute coronary syndrome is a clinical and operational priority in busy emergency departments (ED). We examined the performance of an investigational troponin I (TnI) assay with 10-100-times greater sensitivity than current commercial assays.
Methods: Among patients with non-traumatic chest pain enrolled in the BWH-TIMI ED Chest Pain Study, we measured TnI (n=381) at baseline, 4-6 h, and 12-24 h with an investigational assay (S-TnI; Singulex, detection-limit 0.0002 µg/l, 99th percentile 0.009 µg/l) and a contemporary sensitive assay (TnI-Ultra; Siemens, detection-limit 0.006 µg/l, 99th percentile 0.04 µg/l). Final diagnosis was adjudicated using all diagnostic data and local hospital-based cardiac TnI (Siemens), blinded to investigational cardiac Tn.
Results: The adjudicated diagnosis was myocardial infarction (MI) in 96 patients, unstable angina in 41, and acute non-coronary cardiovascular conditions in 50 patients. Baseline S-TnI was highly sensitive for MI (97%, 95% CI 91-99%) with specificity 81% (95% CI 76-86%) and positive predictive value 63% (95% CI 55-71%). The negative predictive value with S-TnI was 99% (95% CI 96-100%). S-TnI had better diagnostic accuracy than the local assay (area under the curve 0.976 vs. 0.916, p=0.003). Among 20 patients with negative baseline TnI and diagnosis of MI, 19 had elevated baseline S-TnI. Compared to TnI-Ultra, S-TnI trended toward higher sensitivity (97 vs. 94%, p=NS) but did not differ significantly in negative predictive value (99 vs. 98%) or area under the curve (p=0.29).
Conclusion: Current and investigational Tn assays substantially increased clinical sensitivity and improved diagnostic accuracy for MI, despite a decline in specificity. A contemporary sensitive assay delivered similar overall accuracy to the investigational test, suggesting that we have reached a point of maximum diagnostic return with increasing analytical sensitivity.
Keywords: Biomarkers; chest pain; myocardial infarction; unstable angina.