Editorial

Cardiac troponin testing in the acute care setting: Ordering, reporting, and high sensitivity assays—An update from the Canadian society of clinical chemists (CSCC)

Recently, Ortho Clinical Diagnostics have launched a high sensitivity cardiac troponin I assay adapted on Vitros3600 / 5600. We aimed at evaluating the analytical and clinical performances using the 0/3-hour algorithm, of the Ortho hs-cTnI assay on the Vitros 3600® (Ortho Clinical Diagnostics, Illkirch, France) analyzer with comparison to Roche hs-cTnT assay on the Cobas8000/e801® module (Roche diagnostics, Meylan, France) and the Abbott STAT hs-cTnI assay on the Alinity i® (Abbott, Rungis, France) analyzer.

Imprecision studies, linearity, limit of detection, and the interference to hemolysis were performed for Ortho hs-cTnI assay. The concordance study was based on results of troponin obtained from 160 patients with chest pain to the emergency department. Testing was performed simultaneously measuring the Roche hs-cTnT and the Ortho hs-cTnI, then on remaining sample the concentration of Abbott hs-cTnI (n = 150). We applied the 0/3h algorithm to rule out/rule in, in acute myocardial infarction.

Analytical performances were acceptable and in accordance with manufacturer’s data. For late presenters 100, 92.8 and 88.8% patients could be rule-out with Roche, Ortho and Abbott assay, respectively with one NSTEMI missed with both hs-TnI assays. Applying the hs-cTn cut-offs from 0/3-hour algorithm, 107 (66.8%) could be rule out with Roche hs-cTnT with no AMI missed (sensitivity 100% [95%CI : 90.5 to 100] and NPV 100%); 89 with Ortho hs-cTnI (sensitivity 97.3% [95%CI : 85.8–99.9] and NPV 98.8% [95%CI :92.7–99.8]); and 61 with Abbott hs-cTnI (sensitivity 97% [95%CI : 84.2–99.9] and NPV 98.4% [95%CI : 89.8–99.7]). For rule-in, 23.7% (n = 37) from the total population would be designated in this group, with a specificity of 86.9% (95%CI: 79.7–92.4) and PPV of 69.8% (95%CI: 59.4–78.5) vs specificity of 72.3% (95%CI:63.5–80.0) and PPV of 51.4% (95%CI: 44.1–58.2) with Roche hs-cTnT vs Ortho hs-cTnI respectively; and with Abbott, 33 patients had constitued this group with a specificity of 52.1% (95%CI:42.7–61.4) and PPV of 36.4% (95%CI: 32.0–41.0).

In conclusion, according to ESC guidelines, our results indicate that the Ortho hs-cTnI assay for the Vitros 3600 instrument is a method that may be used in the clinical practice using 0/3-hour algorithm.

Because the 99th percentile is of such importance in defining myocardial injury and myocardial infarction, it is important to know whether there are real age-related differences in troponin 99th percentiles.

We went to our database from the Canberra Heart Study where 1062 apparently healthy subjects were extensively screened for occult cardiac disease, and looking at persons aged <65 years and >65 years, for men and women separately, we compared a variety of cutpoints from the 99th percentile down to the 50th percentile.

With our rigorous criteria for defining cardiac health, we excluded 67.2% of males aged >65 years and 53.8% of women aged 65 years and older. Even with these rigorous exclusions we found that at every cutpoint examined between the 99th percentile and the 50th percentile, persons aged <65 years had lower troponin I concentrations that persons aged 65 years and older. Similarly, at every cutpoint examined, women had lower troponin I concentrations than did men.

For the 4 separate groups examined (men and women, age < 65 years and 65 years and older) after the exclusions of persons with subclinical cardiac disease, the distributions were not significantly different to a Gaussian distribution.

With the rigorous exclusions of persons with subclinical cardiac disease, and the fact that our populations have a Gaussian distribution, our data suggests that age-related hs-cTnI concentrations are real. This has important implications particularly when assessing older persons in the Emergency Department.

i) To assess the relationship between lipid markers and high-sensitivity C-reactive protein (hs-CRP), and high-sensitivity cardiac troponin I (hs-cTnI) in the reference population, and ii) to evaluate the impact of lipid markers and hs-CRP on the 99th percentile upper reference limit (URL) for hs-cTnI.

531 questionnaire-identified presumably healthy individuals were enrolled in a single-center, cross-sectional study. Surrogate biomarkers for diabetes, myocardial and renal dysfunction were used to refine the healthy cohort (n = 408). Lipid profile, total cholesterol:high-density lipoprotein cholesterol (HDL-C) ratio, non-HDL-C, apolipoprotein AI (apoAI), apolipoprotein B (apoB), apoB:apoAI ratio, lipoprotein(a), small dense low-density lipoprotein cholesterol (LDL-C) and hs-CRP were determined.

Individuals with detectable vs. non-detectable hs-cTnI concentrations more often showed elevated LDL-C (60% vs. 46%; p = 0.002), apoB (73% vs. 61%; p = 0.008), apoB:apoAI ratio (53% vs. 40%; p = 0.005) and lipoprotein(a) (15% vs. 7%; p = 0.015). The apoB:apoAI ratio and to a lesser extent other lipid markers, but not hs-CRP, were positively associated with hs-cTnI concentration in univariate and multivariate analyses. Exclusion of individuals with elevated apoB:apoAI ratio or apoB, but not hs-CRP, lowered the 99th percentile URL in the healthy cohort respectively by 12.9% (6.2 vs. 5.4 ng/L) and 14.5% (6.2 vs. 5.3 ng/L). The corresponding reduction for both lipid biomarkers in the presumably healthy population was 24.0% (7.5 vs. 5.7 ng/L).

Our study demonstrates that atherogenic lipid markers, particularly apoB:apoAI ratio or apoB, influence the 99th percentile URL for hs-cTnI.