3303Classification of patients with acute chest pain by analytical benchmarks and subsequent management recommendations - A comparison of three highly-sensitivity troponin assays in the ROMICAT trials

Background Concordance of different highly-sensitive troponin (hsTn) assays in stratifying acute chest pain (ACP) patients according to analytical benchmarks and subsequent emergency department (ED) management recommendations are unknown. Methods We included patients enrolled in the ROMICAT (Rule Out Myocardial Infarction/Ischemia Using Computer-Assisted Tomography) I and II trials (n=624; 39.1% female; 52.8±10.0 years), who presented to the ED and were referred to further non-invasive diagnostic testing. In ROMICAT I, blood was obtained at 4 hours, and in ROMICAT II at ED presentation, and at 2 and 4 hours and was tested with three state-of-the-art hsTn assays (Roche Diagnostics, Elecsys 2010; Abbott ARCHITECT i2000SR; Siemens Diagnostics, HsVista). In a per sample analysis, we compared the concordance of assays for analytic benchmarks (below the level of detection (LOD)/LOD to 99th percentile/99th percentile to myocardial infarction (<LOD], ><LOD], >MI)/above MI). In a per patient analysis of serial hsTn testing in ROMICAT II, we determined concordance of management recommendations (rule out/observe/rule in) based on 2015ESC guidelines, results of coronary CT angiography and stress myocardial perfusion, and adjudicated endpoints of ACS. Results Overall, only 34.4% (353/1027) of samples were classified into the same benchmark category by all assays (table 1). In a per patient analysis, all assays agreed on the same of management recommendation in 25.3% (49/242) patients after the 1st hsTn, mostly driven by differences in discharge (6.6%, 21.1%, and 61.2%; respectively; all p<0.001). The concordance of management recommendations improved significantly to 67.4% (163/242; p<0.001) after the 2nd hsTn measurement but the final agreement for “observe” and “rule in” remained limited (13/90 and 5/18). Among patients in whom discharge was recommended, at least 18.8% (range: 18.8–21.0% across assays) had a positive imaging/stress test and at least 2.9% (range: 2.9–3.4%) had ACS. Table 1. Classification of measurements from three hsTn assays in 1027 samples of patients with ACP according to analytical benchmarks <LOD LOD – 99th %tile* 99th %tile* – AMI** AMI**< Roche Elecsys, n (%) 578 (56.3) 328 (31.9) 95 (9.3) 26 (2.5) Abbott ARCHITECT, n (%) 176 (17.1) 788 (76.7) 14 (1.7) 49 (4.8) Siemens Vista, n (%) 96 (9.4) 861 (83.9) 33 (3.2) 37 (3.6) *Assay specific 99th percentiles; **as per ESC 2015 guidelines. Conclusion Differences in hsTn assay concordance for analytical benchmarks and subsequent ED management recommendations in patients with ACP are substantial raising concerns about the lack of uniform management. Acknowledgement/Funding Fulbright Visiting Student Researcher Grant