How does the sensitivity, specificity, and precision cardiac troponin assays vary?

Updated: Nov 20, 2018
  • Author: Donald Schreiber, MD, CM; Chief Editor: Barry E Brenner, MD, PhD, FACEP  more...
  • Print


The sensitivity, specificity, and precision of the different commercially available troponin assays vary considerably. These differences are related to a lack of standardization, the use of different monoclonal antibodies, the presence of modified TnI and TnT in the serum, and variations in antibody cross-reactivity to the various detectable forms of TnI that result from its degradation.

Only one manufacturer produces the TnT assay, and its 99th percentile cutoffs and the 10% CV are well established. However, up to 20-fold variation has occurred in results obtained with the multitude of commercial TnI assays currently available, each with their own 99th percentile upper reference limits and 10% CV levels.

In the GUSTO IV study, a relatively insensitive point-of-care TnI assay was used to screen patients for study eligibility. In a subsequent study, the blood samples were reanalyzed using the 99th percentile cutoff of a far more sensitive central laboratory TnT assay. The more sensitive 99th percentile cutoff of this TnT assay identified an additional 96 (28%) of 337 patients with a positive TnT result but negative point-of-care TnI; these patients had higher rates of death or MI at 30 days. [15]

In a similar reanalysis of the TACTICS-TIMI 18 trial, 3 different TnI cutoffs were compared on 1821 patients to evaluate the 30-day risk of death or MI: the 99th percentile, 10% CV, and the World Health Organization (WHO) acute MI cutoffs. (The WHO cutoffs define acute MI using CK-MB and report troponin levels as either a higher “acute MI level” or a lower “intermediate level” that is correlated with “leak” or “minor myocardial injury.”)

Using the 10% CV cutoff identified, an additional 12% more cases were identified relative to the WHO acute MI cutoff. The 99th percentile cutoff identified an additional 10% of cases relative to the 10% CV cutoff, as well as a 22% increase in the number of cases over the WHO acute MI cutoff. Nevertheless, the odds ratios for the adverse cardiac event rates of death or MI at 30 days were similar for all 3 cutoffs, suggesting that the lower cutoffs detected more patients with cardiovascular risk without sacrificing specificity. [16, 17]

The National Academy of Clinical Biochemistry (NACB) working with the ACC/ESC guidelines has recommended adoption of the 99th percentile upper reference limit as the recommended cutoff for a positive troponin result. Ideally, the precision of the assay at this cutoff level should be measured by a CV that is less than 10%.

However, most TnI assays are imprecise at the 99th percentile reference limit. [18] Some have therefore recommended that the cutoff level be raised to the slightly higher 10% CV level instead of the 99th percentile reference limit to ensure adequate assay precision.

In addition, studies have shown that populations within the 99th percentile reference limit include patients with low troponin levels who nevertheless have an elevated cardiac risk, and that the true 99th percentile cutoff for a healthy patient population is actually a factor of 10-50 lower. Accordingly, these investigations suggest that higher sensitivity or ultrasensitive troponin assays are necessary. [17] The advantage of ultrasensitive troponins is based on the premise that lower cutoff levels achieve higher sensitivity that will allow earlier diagnosis, often within 90 minutes of presentation.

Did this answer your question?
Additional feedback? (Optional)
Thank you for your feedback!