The study shows that the 99th percentile for the upper reference limit (used to define myocardial injury) for high-sensitivity (hs)–troponin T in the new analysis matched those reported by manufacturers. However, the same threshold for hs–troponin I was lower than was manufacturer-reported levels when considering the whole population.
And for both hs–troponin T and hs–troponin I, there were significant differences in 99th percentile levels by age.
“Our data suggest that some cases of myocardial injury may be missed in the whole population by using current non–age specific thresholds of troponin I,” lead author, John McEvoy, MB, University of Galway (Ireland), said in an interview. “If the non–age specific threshold was lowered to that in our cohort, then we would pick up more people with myocardial injury.”
“However,” Dr. McEvoy added, “if age-specific thresholds were deployed, then our data suggest that thresholds used to diagnose myocardial injury would need to be higher in older adults, somewhat lower in middle-aged individuals and much lower in younger people.”
The study was published online in the Journal of the American College of Cardiology.
The authors explain that the 99th percentile upper–reference limit threshold is the common benchmark of abnormality for all troponin assays. Five high-sensitivity cardiac troponin assays have been cleared by the Food and Drug Administration for clinical use and allow for earlier diagnosis of MI.
However, there has been variability in the approach used to define the 99th percentile upper reference limits for these assays, with definitions of healthy reference populations differing and the various assays available are not standardized or harmonized. So troponin concentrations at 99th percentiles do not align across assays, and the generalizability of manufacturer-reported reference upper reference limits for hs-troponin assays to the U.S. adult population is unknown.
They note that though sex-specific 99th percentile upper reference limits for hs-troponin have been recommended since 2018, age-specific thresholds are not yet endorsed, and whether thresholds differ by race or ethnicity is also controversial.
They aimed to investigate these issues using stored serum samples from adults aged 18 or older who participated in the 1999-2004 National Health and Nutrition Examination Survey (NHANES).
Dr. McEvoy described the NHANES database as “the gold standard cohort for representation of the U.S. adult population,” noting that other studies conducted by the manufacturers of the troponin tests have often used convenience samples from patients attending hospital clinics and blood donors, which he said were not representative of the whole population.
For the study, the researchers estimated that the 99th percentile upper reference limit for four hs-troponin assays (one troponin T and three troponin I) in a strictly defined healthy reference subgroup of 2,746 individuals from the NHANES cohort.
Results showed that the NHANES 99th percentile upper reference limit for hs–troponin T (19 ng/L) matched the manufacturer-reported level (19 ng/L). But, the NHANES upper reference levels for three troponin I assays were lower than were levels stated by the manufacturers.
The NHANES levels were 13 ng/L for the Abbott hs–troponin I assay (manufacturer: 28 ng/L); 5 ng/L for the Ortho hs–troponin I assay (manufacturer: 11 ng/L); and 37 ng/L for the Siemens hs–troponin I assay (manufacturer: 46.5 ng/L).
Furthermore, the 99th percentile upper reference limits for all four hs-troponin assays were statistically significantly lower in healthy adults younger than 40 years, compared with healthy adults older than 60 years.
There were also significant differences in upper reference limits by sex, but none by race/ethnicity.
Dr. McEvoy explained that NHANES is a very well phenotyped database with information on individuals’ health, body mass index, and other biomarkers. “This allows us to define a completely healthy subgroup of people, which could explain why the 99th percentile threshold for hs–troponin I was lower than previously reported from other cohorts,” he added.
Though there may be concern that such a healthy subgroup would mean the sample is enriched with younger people, whereas the typical person having their troponin measured would be older, Dr. McEvoy pointed out that there were more than 400 people older than 60 years in the healthy group. “This is probably the biggest cohort of super healthy older U.S. adults ever sampled in this regard,” he commented.
Dr. McEvoy said that the overall results from the study suggested that different thresholds might need to be considered for troponin I. “This could lead to threshold levels used to diagnose myocardial injury being cut in the population as a whole.”
But, he said a more important message was the need for age-specific thresholds.
“We found that troponin levels track with age. Even in individuals who age in a very healthy way, their troponin levels are greater than in younger people. This is the first time this has been shown with such clear statistical significance,” Dr. McEvoy said. “We think this data provides a compelling case for the use of age-specific cut-offs.”
He explained that, if age-specific thresholds were used to diagnose myocardial injury, the cut point from the current data would be higher than it would be from current manufacturers’ recommendations in those older than 60 years, so fewer people in this age group would be labeled as having myocardial injury.
“Our results suggest that, at present, we are seeing more false positives in older people leading to more unnecessary tests.” Using age-specific cut off points will reduce the number of false positives in older people. Dr. McEvoy noted a similar change in the way D-Dimer blood tests have been used to diagnose pulmonary embolism in recent years.
Using age-specific cut-offs for hs-troponin would also reduce the number of false negatives in younger people, Dr. McEvoy added.