Supplementary Material for: Clinical Evaluation of a New High-Sensitivity Cardiac Troponin I Assay for Diagnosis and Risk Assessment of Patients with Suspected Acute Myocardial Infarction
datasetposted on 18.01.2021, 11:35 by Shiozaki M., Inoue K., Suwa S., Lee C.-C., Chiang S.-J., Sato A., Fukuda K., Kubota N., Tamura H., Fujiwara Y., Yamaguchi K., Sato T., Sumiyoshi M., Daida H.
Introduction: Current assays based on the 0-hour/1-hour (0-/1-h) algorithm using high-sensitivity cardiac troponin (hs-cTn) are limited to only Abbott Architect hs-cTnI, Siemens Vista hs-cTnI, and Roche Elecsys hs-cTnT. Objective: This study aimed to evaluate this new hs-cTnI assay, LumipulsePresto hs Troponin I, for diagnosis of acute myocardial infarction (AMI) on admission and on 0-/1-h algorithm to stratify AMI patients precisely. Methods: This prospective cohort study included 442 patients with suspected non-ST-elevation myocardial infarction in three hospitals in Japan and Taiwan from June 2016 to January 2019. We enrolled patients presenting to the emergency department with symptoms suggestive of AMI and collected blood samples on admission and 1 hour later. Two independent cardiologists centrally adjudicated final diagnoses; all clinical information was reviewed twice: first, using serial hs-cTnT (Roche-Elecsys, primary analysis) and Lumipulse Presto Lumipulse Presto, second, using the Lumipulse Presto hs-cTnI measurements. At first, we compared diagnostic accuracy quantified using receiver operating characteristic (ROC) curves for AMI. Then, we evaluated major adverse cardiovascular events (cardiac death, AMI) in the rule-out group according to a 0-hour/1-hour algorithm at the 30-day follow-up. Results: Diagnostic accuracy at presentation by the ROC curve for AMI was very high and similar for the LumipulsePresto hs-cTnI and hs-cTnT,(area under the curve [AUC]: LumipulsePresto hs-cTnI, 0.89, 95% confidence interval [CI] 0.86–0.93; hs-cTnT, 0.89, 95% CI 0.85–0.93; p = 0.82). In early presenters, the LumipulsePresto hs-cTnI appeared to maintain the diagnostic performance of hs-cTn for patients with <3 h (AUC: LumipulsePresto hs-cTnI, 0.87, 95% CI 0.81–0.92; hs-cTnT, 0.86, 95% CI 0.80–0.92; p = 0.81). The algorithm using the LumipulsePresto hs-cTnI ruled out AMI in 200 patients with negative predictive value and sensitivity of 100% (95% CI 97.3%–100%) and 100% (95% CI 92.7%–100%), respectively, in the rule-out group. Conclusion: Diagnostic accuracy and clinical utility of the novel LumipulsePresto hs-cTnI assay are high and comparable with the established hs-cTn assays.