Effect of Repeat Measurements of High-Sensitivity Cardiac Troponin on the Same Sample Using the European Society of Cardiology 0-Hour/1-Hour or 2-Hour Algorithms for Early Rule-Out and Rule-In for Myocardial Infarction

To the Editor: There is debate on the clinical applicability of the European Society of Cardiology (ESC)1 0/1 h algorithm to rule-out and rule-in myocardial infarction (MI) using high-sensitivity cardiac troponin (hs-cTn) assays (1). External validations have not achieved the same diagnostic accuracy as studies referenced in the ESC guidelines (2). A critique applicable to all early rule-out/rule-in algorithms is whether the precision of hs-cTn assays is sufficient to achieve diagnostic accuracy at the values proposed (1). If not, approaches such as the 2 h algorithm might be more robust, although the latter also may employ changes that are small enough to challenge the analytical variation of assays (3, 4). Our objective was to evaluate the analytical variation of results in the same samples measured 3 times within 3.5 h and to determine the misclassification rate associated with the ESC 0/1 h algorithm and the 2 h algorithm due to short-term analytical variation. Briefly, 50 fresh centrifuged lithium heparin samples (stored at room temperature; not frozen) were measured for hs-cTnI (Abbott Diagnostics) as the first measurement (reported as a whole number, ng/L). The sample was reanalyzed again approximately 1.5 h later (second measurement) and a third measurement approximately 1.5 h after the second measurement (Table 1). The only selection criterion was that the lithium heparin sample was a full draw with sufficient sample volume to allow repeat testing without sampling error. Results were interpreted from the ESC recommended cutoff values for the Abbott hs-cTnI assay based on the following criteria: for the rule-out group all measurements <2 ng/L or <5 ng/L with differences between measurements <2 ng/L; for the rule-in group all measurements ≥52 ng/L or measurements with differences ≥6 ng/L from concentrations below 52 ng/L. The remaining cases fell into the observe group of the algorithm (2). For the 2 h algorithm the following criteria were used: rule-out group all measurements <6 ng/L with differences between measurements <2 ng/L; rule-in group all measurements ≥64 ng/L or measurements with differences ≥15 ng/L from concentrations <64 ng/L. The remaining cases fell into the observe group of the 2 h algorithm (3). Measurement of hs-cTnI in 50 heparin plasma samples at 3 different times (all within 3.5 h from first measurement).a Shaded rows represent misclassification by the ESC 0/1-h algorithm, bolded row by the 2 h algorithm. Measurement of hs-cTnI in 50 heparin plasma samples at 3 different times (all within 3.5 h from first measurement).a Shaded rows represent misclassification by the ESC 0/1-h algorithm, bolded row by the 2 h algorithm. Our experimental setup substitutes a single sample for the repeated samples that would be drawn from patients being evaluated by the ESC protocols, thereby assuring a stable clinical situation. With a perfect assay, repeated analyses of the same sample would yield identical cTn concentrations, and thus consistent sample categorizations. However, from the 50 patient samples, for the ESC 0/1 h algorithm, 7 samples yielded repeat measurements that would have reclassified patients into different groups (shaded rows in Table 1). Utilizing the first measurement for group assignment, 1 patient would be reclassified from rule-out to observe using <2 ng/L and 2 patients would be reclassified from observe to rule-out. At <5 ng/L, 1 patient would be reclassified from rule-out to observe. For rule-in, 1 patient had the first measurement ≥52 ng/L but on subsequent measurements it was less than this concentration, thus reclassifying the patient to the observe group. Finally, 2 other patients assigned to the observe group would be reclassified into the rule-in group as differences ≥6 ng/L were observed between repeated measurements. By comparison, for the 2 h algorithm only 1 sample (bolded in Table 1) yielded repeat measurements that reclassified a patient from the rule-out group to the observe group. These data demonstrate that repeat testing on the same sample yields changes in results that could lead to reclassification of more than 10% of patients using the ESC 0/1 h algorithm as compared to 2% using the 2 h algorithm. These differences occurred using the same analyzer and reagent kit, over a short time frame––all factors that mitigate sources of variation. Multiple analyzers performing measurements and different lots of reagents would have led to further variation and misclassification (4). These data support recent mathematical modeling experiments that suggest that minor shifts at low hs-cTnI concentrations will lead to misclassification of patients using early rule-in/rule-out algorithms (5). Importantly, there is always measurement uncertainty, even at the 99th percentile. The closer the cutoff is to the concentration range of the majority of healthy individuals, the more likely misclassification will occur. The analytical realities, which are intrinsic to hs-cTn assays when measuring low cTn concentrations below the 99th percentile are important for clinicians to understand. Our study represents an analytical approach to test different hs-cTn algorithms for early rule-out/rule-in. Notwithstanding the pure analytical nature of our study, the results do substantiate prior concerns about the hs-cTn cutoffs proposed by the ESC 0/1 h criteria and attest to the need for further research (1, 2, 5). European Society of Cardiology myocardial infarction high-sensitivity cardiac troponin.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.