乌鸦传媒

Even low levels of high-sensitivity cardiac troponin T (hs-cTnT) were associated with worse survival long term after an emergency visit for chest pain without myocardial infarction (MI) or other conditions known to impact troponin levels, a Swedish report suggested.

Compared with the 62% of patients who had hs-cTnT below the detection level of 5 ng/L, those with even slightly higher levels were at over follow-up, with adjusted hazard ratios as follows:

• 2.00 for the 21% with 5-9 ng/L (2.1 versus 0.5 deaths per 100 person-years)
• 2.92 for the 8.6% at 10-14 ng/L (5.1 versus 0.5 deaths per 100 person-years)
• 4.07 for the 5.7% at 15-29 ng/L (12 versus 0.5 deaths per 100 person-years)
• 6.77 for the 1.5% at 30-49 ng/L (23 versus 0.5 deaths per 100 person-years)
• 9.68 for the 0.7% at 50 and above (33 versus 0.5 deaths per 100 person-years)

"Our data indicate that any detectable hs-cTnT level is associated with an increased risk of death and cardiovascular events and may merit further attention," according to Martin Holzmann, MD, PhD, of Karolinska University Hospital in Stockholm, and colleagues in the October 31 issue Journal of the American College of Cardiology.

The study included all 19,460 patients over 25 years of age who had chest pain and hs-cTnT analyzed in the emergency department of Holzmann's institution in 2011-2014 but no acute conditions. Physicians took at least two hs-cTnT readings to determine if even slight hs-cTnT elevations were persistent, a scenario that is not discussed at length in current guidelines.

"In the absence of clinical guidelines, we strongly believe that persistently elevated hs-cTnT levels may be reason in itself to investigate patients for exclusion of previously undiagnosed heart disease. In addition, future research should focus on how to mitigate risk for patients with no detectable heart disease that may explain persistently elevated hs-cTnT levels," Holzmann's group suggested.

Marc Bonaca, MD, MPH, of Brigham and Women's Hospital in Boston, agreed in an an : "The data presented by [the authors] should remind clinicians not to be falsely reassured when acute coronary syndrome is 'ruled out' by a lack of dynamic changes or a troponin level that is measurable but below the 99th percentile threshold. Even in the presence of apparent clinical stability and no apparent acute condition, these patients remain at heightened intermediate- to long-term risk."

"For patients whose test results reveal no acute pathology, we must transition to considering [hs-cTnT] as a potent continuous marker of subclinical pathology," Bonaca added.

"Although there is no defined diagnostic algorithm in this setting, evaluation for underlying structural heart disease with echocardiography may be reasonable. In patients with [hs-cTnT] elevations, particularly patients with risk factors for coronary disease, testing for ischemia coupled with an assessment of left ventricular [LV] function may be warranted," he suggested. "In patients with abnormal echocardiography findings or concerning indicators for inflammatory or infiltrative disease, cardiac MRI may reveal underlying pathology."

Higher troponin levels were tied to especially high rates of cardiovascular mortality, with adjusted hazard ratios compared with undetectable levels ranging from 3.59-fold higher adjusted risk for the 5-9 ng/L group to 27-fold elevated risk for those with 50 ng/L and above compared with undetectable levels.

A hs-cTnT level of 10 ng/L and above was associated with doubled-to-tripled odds of MIs over follow-up. Heart failure rates were several-fold higher with hs-cTnT 5 ng/L and higher.

In all, 6.9% died over a mean follow-up of 3.3 years. Cancer and cardiovascular disease were each responsible for one-third of deaths.

With increasing troponin levels, groups rose stepwise in age, comorbidity burden, and the proportion of men -- baseline differences leading the investigators to acknowledge the possibility of residual confounding in their retrospective study.

Additionally, it is unclear how many patients experienced type 1 versus type 2 MI, and the authors lacked data on LV hypertrophy or LV ejection fraction.