2022 Volume 4 Issue 11 Pages 560-561
We appreciate the comments from Tada et al regarding our recent publication in Circulation Reports,1 which evaluated the diagnostic performance of the 0-hour/1-hour (0/1-h) algorithm with both high-sensitivity cardiac troponin (hs-cTn) T and I, as recommended in the 2015 European Society of Cardiology (ESC) guideline, and showed that the algorithm could effectively rule-in or rule-out patients with acute myocardial infarction (AMI).
First, Tada et al mentioned that hs-cTnI assays have different cut-off values and diagnostic performance, and that the diagnostic performance of the 0/1-h algorithms is assay specific, and they are concerned our meta-analysis of 6 studies using different hs-cTnI assays is probably incorrect and that the results of studies using different assays should be reported separately. In this regard, a meta-analysis according to the assay method may be warranted, but we have determined that there are not enough studies at this stage to allow for such a meta-analysis. We used the same meta-analysis approach as in previous research papers.2–4
Conversely, another opinion suggested that a systematic review and meta-analysis should be conducted integrating hs-cTnI and T data, but this was not done for the following reasons: the greater potential for heterogeneity in the meta-analysis; none of the previous published reports as such combined the 2 troponin assays; and the anticipation that clinicians would not be able to translate the integrated results into actual patient care, even if they could be presented.
In the first place, detection of an elevated hs-cTn value above the 99th percentile upper reference limit (URL) is defined as myocardial injury,5 and the 99th percentile should be determined in a healthy population. The 99th percentile URL (i.e., the cut-off value) is derived from hs-cTn measurements in an apparently healthy cohort and is dependent on participant selection, statistical analyses, and analytical and biological variability.6 It has been shown to be influenced by age, sex, and the presence of comorbidities;7 for example, it is higher with increasing age, higher in males than in females, and appears to vary by race.8 In statistical analyses, robust statistical methods are also considered inappropriate, and non-parametric methods are preferred.7 However, at least, the criteria to select healthy reference subjects and statistical analyses have not yet been standardized across assay methods, and those are issues for future investigation. In the future, a guidance will be provided to define the establishment of a uniform or standards for these hs-cTn assays, which would standardize the 99th percentile URL and increase the accuracy of the assay.
In our study, a systematic review and meta-analysis, albeit somewhat imprecise, we confirmed that the ESC 0/1-h algorithm with both hs-cTnI and hs-cTnT has valid diagnostic performance for the diagnosis and exclusion of AMI in patients with chest pain without ST-segment elevation.1 At this present time, we reiterate the need to understand the diagnostic performance of the assay methods employed at each institution and to use them clinically.
Second, Tada et al pointed out an error for the fifth-generation troponin T assay. We are grateful for this and are considering issuing an Erratum.
Funding was provided by the Japan Resuscitation Council and the Japanese Circulation Society of Emergency and Critical Care Committee.
T.M. is a member of Circulation Reports’ Editorial Team. The other authors declare no conflicts of interest with regard to this article.
