Troponin T Percentiles from a Random Population Sample, Emergency Room Patients and Patients with Myocardial Infarction Ola Hammarsten, 1* Michael L.X. Fu, 2 Runa Sigurjonsdottir, 2 Max Petzold, 2 Lina Said, 1 Kerstin Landin-Wilhelmsen, 3 Bengt Widgren, 4 Mårten Larsson, 1 and Per Johanson 5 BACKGROUND: High-sensitivity cardiac troponin T (cTnT) assays detect small clinically important myo- cardial infarctions (MI) but also yield higher rates of false-positive results owing to increased concentrations sometimes present in patients without MI. Better un- derstanding is needed of factors influencing the 99th percentile of cTnT concentrations across populations and the frequency of changes in cTnT concentrations 20% often used in combination with increased cTnT concentrations for diagnosis of MI. METHODS: cTnT percentiles were determined by use of the Elecsys ® hscTnT immunoassay (Modular ® Analyt- ics E170) in a random population sample, in emer- gency room (ER) patients, and in patients with non– ST-elevation MI (NSTEMI). Changes in cTnT concentrations were determined in hospitalized pa- tients without MI. RESULTS: The 99th cTnT percentile in a random popu- lation sample (median age, 65 years) was 24 ng/L. In ER patients 65 years old without obvious conditions that increase cTnT, the 99th cTnT percentile was 12 ng/L with little age dependence, whereas in those 65 years old it was 82 ng/L and highly age dependent. In hospi- talized patients without MI the 97.5th percentile for change in the cTnT concentration was 51%– 67%. cTnT remained below the 99th percentile (12 ng/L) in 1% of patients with NSTEMI until 8.5 h after symptom onset and 6 h after ER arrival. CONCLUSIONS: Age 65 years was the dominant factor associated with increased cTnT in ER patients. This age association was more prominent in ER patients than in a random population sample. Changes in serial cTnT concentrations 20% were common in hospitalized patients without MI. © 2012 American Association for Clinical Chemistry High-sensitivity troponin assays allow reliable mea- surements of troponin concentrations below the 99th percentile (1) and help clinicians find small myocardial infarctions (MIs) 6 with improved diagnostic sensitiv- ity. These assays shorten the time to diagnosis by 3 h compared to the use of conventional troponin assays with blood sampling at admission and after 3 and 6 h (2, 3 ). On the other hand, high-sensitivity troponin as- says also increase the number of patients presenting with increases in troponin from other causes than MI (4, 5 ). Such increases are particularly common in el- derly patients (6), and lead to reductions in the diag- nostic specificity of troponin results. However, increases in troponin that can be found only with high- sensitivity assays have strong prognostic implications even in the absence of MI (7, 8 ). In addition, a recent report indicated that if patients with suspected coro- nary syndromes and troponin concentrations above the 99th percentile as measured with a sensitive tro- ponin assay are treated as having suspected MIs, mor- tality rates and recurrent MIs decrease by 50% (9). This report underscores the importance of finding and treating the smallest MIs and of implementing high- sensitivity troponin assays in clinical routines. To in- crease the diagnostic accuracy it is also critical to assess the changes in troponin concentrations over time. Al- though the magnitude of the change that is required for a diagnosis of MI remains unclear, relative changes that range from 20% to 117% (3, 4, 10 –12 ) or absolute 2-h changes of 7 ng/L (13 ) have been suggested. Against 1 Department of Clinical Chemistry and Transfusion Medicine, 2 Department of Medicine, and 3 Section for Endocrinology, Sahlgrenska University Hospital at Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden; 4 Head Department of Research, Development and Education, County of Halland, Sweden; 5 Department of Cardiology, Sahlgrenska University Hospital at Sahl- grenska Academy, University of Gothenburg, Gothenburg, Sweden. * Address correspondence to this author at: University of Gothenburg, Depart- ment of Clinical Chemistry and Transfusion Medicine, Bruna stråket 16, Sahlgrenska University Hospital, S-413 45 Gothenburg, Sweden. Fax +46- 31-82-84-58; e-mail ola.hammarsten@clinchem.gu.se. Received July 7, 2011; accepted January 4, 2012. Previously published online at DOI: 10.1373/clinchem.2011.171496 6 Nonstandard abbreviations: MI, myocardial infarction; cTnT, cardiac troponin T; ER, emergency room; NSTEMI, non–ST-elevation MI; MONICA, MONItoring of trends and determinants for CArdiovascular disease; METTS, Medical Emer- gency Triage and Treatment System. Clinical Chemistry 58:3 628–637 (2012) Proteomics and Protein Markers 628 Downloaded from https://academic.oup.com/clinchem/article/58/3/628/5620525 by guest on 11 June 2022