Concern regarding missed non-ST-segment elevation myocardial infarctions when applying a single, pre-hospital cardiac troponin measurement

This commentary refers to ‘Rule out of non-ST-segment elevation acute coronary syndrome by a single, pre-hospital troponin measurement: a randomized trial’, by C. Camaro et al., https://doi.org/10.1093/eurheartj/ehad056 and the discussion piece ‘Less concern about missing NSTEMIs, more prospective randomized controlled trials warranted’, by C. Camaro et al., https://doi.org/10.1093/eurheartj/ehad646. We wish to congratulate Camaro et al.1 for their important contribution to the possible application of point-of-care (POC) cardiac troponin (cTn) assays in the pre-hospital setting.2 We are concerned that methodological limitations might have led to inappropriate conclusions. May we kindly ask the authors to provide additional information to help put their findings into better clinical perspective? First, when considering transferring the findings into clinical practice, it is of major importance to highlight the rather low sensitivity of the POC-assay used. This conventional cTn assay has a limit of detection (LOD) of 40 ng/L. This means that all concentrations <40 ng/L are reported as ‘normal’. Unfortunately, the LOD is higher than the 99th percentile upper reference limit recommended for clinical use (14 ng/L). Thereby, patients with concentrations between 14 and 40 ng/L, who in fact have elevated cardiac Troponin T concentrations possibly indicating myocardial infarction, are incorrectly reported as ‘normal’ by the POC assay. When applying 40 ng/L, as done in their study, as a triage decision in the APACE study2 (ClinicalTrials.gov number, NCT00470587), a multi-centre, international, prospective diagnostic study centrally adjudicating myocardial infarction, the resulting sensitivity would be 59.4% [95% confidence interval (56.8–62)]. Translating this number into clinical practice would mean that among 100 patients presenting with non-ST-segment elevation myocardial infarction (NSTEMI), 41 would be missed with the POC assay used. May we ask the authors whether they think that the clinical criteria used in their study in addition to the POC measurements are capable of compensating for the lack of sensitivity on the POC part? Second, when considering that the time of the blood analysis in the ambulance setting is even closer to the symptom onset, an additional drop in sensitivity must be expected due to the time-dependent nature in the release of cardiac troponin. May we ask the authors if they collected blood samples for later analysis using a high-sensitivity cardiac troponin (hs-cTn) assay to estimate the number of missed NSTEMIs that had hs-cTn concentrations below 40 ng/L at the time of blood draw in the ambulance? Third, the authors report 147 patients (13%) in which a HEAR(T)-score of >3 was the reason for exclusion. Thus, it has come to our attention that the reported proportion of 76% of low-risk-patients, in which the proposed rule-out pathway might be applicable, is much higher than previously reported in studies with similar inclusion criteria in the pre-hospital setting. For comparison, in previous studies from the UK, and The Netherlands, only 25.1%, 32%, and 26.5%, of patients had a HEART-score of ≤3.3–5 May we ask the authors to speculate why in their study the proportion of patients with a HEART-score ≤3 was three-fold higher compared with other studies? The above points are of great importance; if the calculation for cost effectiveness is based on an over-estimated prevalence of low-risk patients, the concluded amount of cost savings would possibly be misleading. J.G. has nothing to declare. P.L.-A. has received research grants from the Swiss Heart Foundation (FF20079 and FF21103) and speaker’s honoraria from Quidel, paid to the institution and outside the submitted work. C.M. reports grants and non-financial support from the Swiss National Science Foundation, the Swiss Heart Foundation, the University Hospital Basel, Basel University, Abbott, Astra Zeneca, Beckman Coulter, Boehringer Ingelheim, Idorsia, LSI-Medience, Novartis, Ortho Diagnostics, Quidel, Roche, Siemens, Singulex, SpinChip, and BRAHMS outside the submitted work, all paid to the institution.