乌鸦传媒

Relying on the sex-specific 99th percentile upper reference limit of cardiac troponin may incorrectly suggest myocardial infarction (MI) in older adults with suspected heart attacks, though no diagnostic approach emerged as a better alternative, an analysis from High-STEACS found.

The guideline-recommended thresholds for increased high-sensitivity cardiac troponin I (hs-cTnI), indicating myocardial injury and potential infarction, performed at a similar sensitivity around 80% in people across age groups. However, specificity for MI fell with increasing age: 98.3% in people under 50 years, 95.5% in those 50-74, and 82.6% in those age 75 and older.

For that oldest group, age-adjusted 99th percentile thresholds improved specificity to 91.3% at the cost of a major loss in sensitivity (down from 81.6% to 55.9%) for an overall positive predictive value (PPV) of 59.3%, reported Atul Anand, MD, PhD, of the BHF Centre for Cardiovascular Science at University of Edinburgh, Scotland, and colleagues. Their manuscript was published in .

In short, "we do not support the adoption of age-adjusted thresholds for the diagnosis of myocardial infarction," the researchers wrote.

Other alternative hs-cTnI analyses they tried for people over 75 years old -- applying a universal rule-in threshold of 64 ng/L, and factoring in the person's cardiovascular comorbidities -- also failed to take MI diagnosis to an acceptable balance of sensitivity and specificity similar to that seen in younger patients.

"[N]o approach achieved parity in the diagnosis between older and younger patients with specificity and PPV reducing with advancing age regardless of the threshold adopted, and alternatives to the guideline-recommended approach resulted in a marked reduction in sensitivity in older persons," Anand and colleagues reported.

However, they reasoned that serial troponin testing may be a worthwhile approach, given the improvements to diagnostic performance when incorporating an absolute change in troponin concentration. This may also be practical because of the ease of early retesting within 1 hour and the easily identifiable clinical features of people actually needing immediate revascularization for MI, they said.

Study authors noted that the troponin 99th percentile had been derived from a cohort of healthy individuals and may be less helpful for diagnosing MI in older people; troponin concentrations higher than this threshold were common in older people, they observed, affecting nearly half of those older than 90.

They suggested that the diagnostic challenge can only be expected to grow with the aging population and is compounded by the higher frequency of atypical symptoms and nondiagnostic ECG findings of older patients with MI.

"Individualized diagnostic approaches and serial testing to determine absolute change in troponin concentration rather than adjustment of binary thresholds are needed to avoid disadvantaging older patients," they urged.

The main finding from High-STEACS had been that patients presenting with suspected acute coronary syndrome, and then ruled out based on different troponin assays, ultimately fared no differently based on cardiovascular outcomes within a year.

Back in 2018, when those results were reported, the investigators already suggested that the diagnostic threshold for MI being hs-cTnI above the 99th percentile may not be appropriate.

The multicenter trial had been conducted across 10 hospitals in Scotland. For this secondary analysis, Anand and colleagues included 46,435 consecutive adults who underwent cardiac troponin testing and excluded those with ST-segment elevation MI.

Anand and colleagues acknowledged that the limitations of the study include the predominantly white population and the results not being applicable to cardiac troponin T or a different assay.

Comment