By Joel A. Miller, MD, FAAEM, FACEP
Recently, we reviewed a landmark study published in the New England Journal of Medicine (NEJM Missed diagnosis of ACS in the ED_2000) that formed much of the foundation and justification of the AHA guidelines that currently guide our practice in caring for patients with chest pain. In looking deeply at the study we found the conclusions and our subsequent actions may not have been justified based on their data and discussion.
The study concluded that we missed 2% of MIs in the ED, which is higher than what most of us would consider an acceptable miss rate. They looked at about 11,000 patients seen in the ED for “possible ACS.” 8% had an MI, and we missed 2% of those (or 19 patients). They conclude that we missed 2% of MIs, which is technically correct, but more importantly we only missed 19/10,689 patients who presented with possible ACS, or 0.18% of the total population. We also discussed how we have broadened the category of “possible ACS” since the publication of this study. In this study (where we only missed 0.18% of AMIs), the patients enrolled with “possible ACS” were found to have a cardiac etiology of their chest pain 45% of the time. This is likely an order of magnitude higher than our current population of “possible ACS” patients, and therefore we might speculate that our current miss rate is lower than that reported in the study. We also discussed how the diagnosis of MI and UA have evolved over the last 20 years. Troponins are far more sensitive for ACS than CK-MB, and the routine application of delta troponin testing further increases the sensitivity. It is likely that this has further decreased our miss rate for MI.
Unstable angina was originally conceived as a historical diagnosis based on worsening angina, new onset severe angina, or angina at rest. We have always said that any new chest pain could be possible ACS, but based on our reading of the UA literature, it’s unlikely that non-anginal chest pain is unstable angina. We are able to rule out AMI with enzymes, and may be able to rule out UA by history, EKG, and enzymes, thus closing the loop on ACS without necessitating a mandatory stress test.
In discussing the evolution of unstable angina, we briefly reviewed 3 articles published by Braunwald in Circulation. In 1989 he proposed a standard definition for unstable angina, during the era when MI was diagnosed with CK-MB. After the publication of that article, troponins began to gain wider acceptance, and in 2000 he published an article further subdividing UA into troponin positive and troponin negative unstable angina. He noted that troponin positive unstable angina had a very high 30 day event rate, but that troponin negative unstable angina had a very low 30 day event rate (arguably below our test threshold). Following his second article, we started defining anyone with a positive troponin as having an MI, which left only the troponin negative patients in the unstable angina group. By moving all the troponin positive unstable angina patients into the category of MI, we lower the adverse event rate in both the MI and the UA groups (the Will Rogers phenomenon – google it). So over the last decade or so as troponins have gotten better and we have begun routinely using delta troponins to maximize the test’s characteristics, we have moved more and more patients from UA to MI, leaving progressively healthier patients in the UA group. Finally, in 2013, Braunwald published another article in which he suggests that we may soon be able to retire the term unstable angina. He notes that in studies of high sensitivity troponin, patients with a negative HS troponin do not have death or MI within 30 days, thus a patient with a negative HS troponin cannot have an unstable lesion.
It will be interesting to watch this unfold over the next few years. I would encourage you all to ask questions and stay engaged in discussions on this topic. When in doubt however, fall back on the recommendations of the AHA guidelines:
High risk (+enzymes, ischemic EKG): admit to a cath capable facility
Intermediate risk (advanced age, multiple risk factors, angina by history): admit for further evaluation
Low risk: When in doubt, rule out with enzymes (single set after 8 hours, or 2 hour delta troponin and CK-MB) and stress within 72 hours
If you are feeling a little more adventurous, you can apply any of the numerous studies to define a very low risk population who may not need further testing. (ie, HEART score, North American CP Rule, ASPECT, ADAPT, or any of the Hollander studies). For example, a patient under 40, not using cocaine, with no risk factors, a non-ischemic EKG, and no history of cardiac disease has a <1% chance of a MACE in 30 days and likely does not need referral for stress testing. Adding a negative troponin brings the 30 day event rate to 0.14%.
Below I will attach a few key quotes from the Braunwald trifecta:
Braunwald 1989: Circulation-1989-Braunwald-410-4
“This classification of UA is based on two premises: 1) the patient’s symptoms are actually caused by myocardial ischemia, 2)in pt’s with prolonged ischemia, the dx of AMI is excluded by EKG and enzymes”
Braunwald 2000: ACC_AHA guidelines for UA and NSTEMI 2000
“Pathological studies in pts with UA who died suddenly have demonstrated that the fatal event is often preceded by repetitive embolization of thrombi from an unstable atheroma. This results in focal myocardial necroses that are not large enough to be detected by CK or CK-MB measurements. The detection of this so called minor myocardial injury in UA may therefore reflect the presence of an unstable plaque containing platelet-rich thrombus in the proximal coronary artery and can be detected by measurement of serum cardiac troponin as surrogate markers for thrombus formation.”
Braunwald 2013: Circulation-2013-Braunwald-2452-7
“It appears that we have now come full circle in our definition of symptomatic ischemic heart disease. Before the 1930s, 2 manifestations, stable angina and AMI, were recognized. Patients in the gray zone between stable angina and AMI that we now call UA were described 75 years ago and at first appeared to be quite rare. Over the next half-century, they were recognized with increasing frequency, and by 25 years ago, about one half of all patients with NSTE-ACS were considered to have UA. However, use of ever more sensitive biomarkers of myocardial necrosis, especially cTn, has steadily chipped away at the fraction of patients with NSTE-ACS without MI who therefore are still considered to have UA. Indeed, in its 2008-2009 report, the World Health Organization revision of the definition of MI stated: “Many patients who in the past would have been diagnosed as having unstable angina will now be diagnosed as having had an MI.” In the next few years, there will likely be much wider use of higher-sensitivity assays for cTn and acceptance of the universal definition of MI. As a consequence, UA is likely to be further marginalized, its definition will become highly dependent on the particular assay for cTn used, and the term will become ever more ambiguous and cause confusion because it will mean different things to different people. Indeed, it is not clear that ACS events can occur without some increase in circulating cTn when measured by a high-sensitivity assay.
As a consequence, we are likely to return to the situation that existed before attention was directed toward patients we now consider to have UA, and patients with ischemic heart disease will again be divided into the original 2 rather than 3 major groups. One group will be patients with angina pectoris, whose angina may be of widely varying severity and classified by the Canadian Cardiovascular Society system. The second group will comprise patients with AMI as defined by the third universal definition, which includes the type of MI (type I, type II, etc), the ECG changes (ie, STEMI and NSTEMI), and the extent of myocardial damage that is related to the magnitude of cTn release.