Published in the June 2011 issue of Today’s Hospitalist
WHILE MISSED DIAGNOSES tops the list of reasons why physicians are sued, lowering biomarker thresholds to diagnose myocardial infarction continues to be controversial. Moving the needle down on troponin concentrations, detractors say, will result in too many false positives, exposing patients to needless testing and perhaps invasive procedures.
But that concern has been debunked in a recent trial from Scottish researchers. In a study that appeared in the March 23/30 Journal of the American Medical Association, authors validated the use of a more sensitive troponin assay and then dropped the threshold for diagnosing acute coronary syndrome from 0.20 ng/mL to 0.05 ng/mL.
According to Nicholas Mills, MD, PhD, the study’s lead author who’s with the University/British Heart Foundation Centre for Cardiovascular Science at the University of Edinburgh, that decision led to a 29% increase in the number of MI diagnoses.
But the lower threshold produced virtually no false positives. It did, however, cut in half the risk of death or recurring MI for patients with troponins in the 0.05 ng/mL-0.19 ng/mL range. Those patients’ high risk became apparent, Dr. Mills says, as he and his colleagues crunched the numbers collected during the trial’s validation phase.
Only troponin levels above the common diagnostic threshold of 0.20 ng/mL were reported to treating clinicians during that phase. But the researchers realized that 39% of patients within the 0.05 ng/mL to 0.19 ng/mL range had either died or had an MI within a year. That was compared to only 7% of those whose troponins were less than 0.05 ng/mL “and only 24% of those whose troponin concentrations were 0.20 ng/mL or higher.
Why did patients with slight troponin rises fare worse than those with much higher elevations?
“Clinicians were falsely reassured that these patients had apparently normal troponin, so they were less likely to refer them to cardiologists,” Dr. Mills says. “Patients were less likely to undergo coronary angiography, be prescribed clopidogrel or dual antiplatelet therapy, or be treated with beta-blockers, ACE inhibitors, or statins.”
Dr. Mills spoke to Today’s Hospitalist about the study results and about the challenges of implementing more sensitive assays.
Were you surprised by the results?
We were very surprised. We expected patients with small troponin rises below the diagnostic threshold to be at slightly higher risk than those with truly normal concentrations. But we did not anticipate that their risk would be higher than patients treated for more substantial transmural myocardial infarcts.
These patients had worse outcomes, but that additional risk was not initially apparent during the validation period because the investigators and clinicians were unaware of troponin concentrations below the accepted diagnostic threshold. By the time the clinical outcomes from patients admitted during the validation period were available, we’d already implemented the more sensitive assay and dropped the diagnostic threshold.
If you didn’t yet have that mortality data when you switched to the more sensitive assay, were clinicians reluctant to use it?
They were for a number of reasons. First, a more sensitive assay will definitely increase the number of referrals and the pressure on existing cardiac services, including the coronary catheter lab, the echo suite and the cardiac rehabilitation program. So that’s one fear.
The second is the genuine concern that patients will be falsely diagnosed with myocardial infarction who actually come in with an alternative acute medical illness, be it septicemia, pulmonary embolus or dysrhythimia, and will receive unnecessary and potentially harmful treatment.
In the first month or two after we dropped the diagnostic threshold, some clinicians still dismissed small troponin rises in patients with chest pain. But then, physicians’ personal experience with these patients led them to appreciate that a small troponin rise in a patient with acute coronary syndrome is just as significant as a more substantial increase.
These patients all have the same pathology, plaque rupture and coronary thrombosis. Whether they have a micro-infarct as a result of embolization from a proximal plaque rupture or a transmural infarct due to vessel occlusion does not matter. They’ve presented with the same underlying problem, and their risk of recurrent myocardial infarction is similar. You disregard small troponin rises in patients who present with chest pain at your peril.
You didn’t find many false positives?
We identified only one patient with a small troponin rise falsely diagnosed as acute coronary syndrome who subsequently represented with a pulmonary embolus. That was one out of 2,000 patients, which I would consider an acceptable number.
You mentioned that clinicians had concerns about resources when MI diagnoses increase 30%. Was that a problem in your hospitals?
That was something that our service was able to absorb. Remember, our study covered all consecutive patients with suspected cardiac chest pain, including very elderly patients who may not require invasive assessments and treatment. So not all these patients will have significantly increased treatment costs. But there’s no doubt that lowering the diagnostic threshold will increase the number of myocardial infarctions diagnosed. If you extrapolate our findings across the U.K., implementing a sensitive troponin assay would potentially result in an additional 50,000 MIs per annum.
And in time, perhaps the rate of revascularization of these patients will also increase. Initially, when you change your clinical practice, there’s a lag phase where clinicians don’t alter their approach as rapidly as you might think. Now, two years on, cardiologists and acute physicians will refer most patients with small troponin rises for invasive investigation of coronary disease.
Before the diagnostic threshold was lowered, were most patients with slight troponin elevations diagnosed with unstable angina?
Surprisingly, only about 30% were diagnosed with unstable angina. The rest were a composite of diagnoses that ranged from gastroesophageal reflux to musculoskeletal chest pain or even “chest pain of uncertain significance” in some cases. Now, we suspect that all those patients actually had acute coronary syndrome, and I think unstable angina will be diagnosed less frequently in the high-sensitive troponin era.
What do guidelines say the diagnostic threshold for troponin concentrations should be?
When the Global Task Force published the universal definition of myocardial infarction in 2007, it did not make a statement about sensitive assays, but it recommended troponin as the biomarker of choice. It also recommended that clinicians, where possible, use the 99th percentile, which is the normal reference population, as the diagnostic threshold.
But the majority of commercially available troponin assays can’t measure troponin at that very low threshold with sufficient diagnostic precision, so most hospitals use a considerably higher threshold.
Do you look forward to even more sensitive assays?
Absolutely. The one we currently use can measure troponin close to the 99th percentile but does not quite achieve that target. The next generation of assays will, I’m sure, allow reporting of even lower concentrations and could potentially improve clinical outcomes for acute coronary syndrome even more.
Phyllis Maguire is Executive Editor of Today’s Hospitalist.
