Friday, September 24, 2010

STEMI versus NSTEMI---meaningless distinction??

Brendan Phibbs, ever the towering intellect and still in possession of a steel trap mind at over 90 years of age, continues to contribute to the medical literature. His latest paper, in the July 2010 issue of Annals of Noninvasive Electrocardiography, is quite remarkable. It is titled Differential Classification of Acute Myocardial Infarction into ST- and Non-ST Segment Elevation Is Not Valid or Rational. Phibbs is being true to form. Never one for weasel words, he makes strong declarative statements. I don't have access to the full text of the paper, but let's look at the abstract:

Background: The distinction between ST elevation and nonST elevation infarcts is widely accepted and is employed as a guide to management...

Finding: The pathology of the two subsets is identical as are the morbidity, mortality and clinical course. Non-ST elevation infarcts are likely to be subsequent, to occur in older patients and to involve the circumflex artery: this subset therefore includes a high-risk group. ST deviation in any part of the electric field of the heart will predictably be accompanied by reciprocal deviation if the entire field of the heart is mapped. Further, ST deviation of infarction is often transient, resolving in minutes so that infarcts will be predictably misclassified. ST deviation per se is therefore not a rational basis for classification of infarcts. In fact, invasive therapy is indicated in both subsets with identical results.

Conclusion: The distinction between ST elevation and non-ST elevation infarcts is baseless. The high risk subgroup included in the non-ST elevation infarct set should not be denied the benefit of early invasive therapy.

So here's what he's saying: In his review of the world's literature, clinical and pathological correlations do not support a meaningful electrocardiographic distinction between STEMI and NSTEMI. At first that seems counterintuitive but as I think about it I can think of several explanations for why the distinction would be artificial. He's not arguing that there's no such thing as a STEMI or a NSTEMI (after all, some MI patients have ST elevations and some don't), but that the distinction is not clinically or pathologically meaningful. There might be a number of reasons.

First, many circumflex occlusions (epicardial coronary artery occlusion was long believed to be the pathologic correlate of STEMI) are associated with no or subtle ST elevation. Second, the presence or absence of ST elevation often depends on time to presentation with ST segments elevated one minute and isoelectric the next in the same patient, or vice versa, reflecting the unstable coronary plaque in a dynamic state of thrombosis and lysis. Third, there are several known “STEMI equivalents” which represent acute coronary occlusion without ST elevation, such as “true posterior” infarctions and hyperacute T wave infarctions seen in the very early stages of coronary occlusion. Finally, certain conditions which displace the ST segments such as bundle branch block and left ventricular strain may play “tug-o-war on the stylus” and mask ST elevation.

So do ST segment elevations have meaning at all? Of course. They are the most powerful initial diagnostic finding in acute MI. Often they help localize the culprit lesion and predict outcome. But if the distinction is meaningless how does one know whether to send the patient to the cath lab? In Phibbs' line of reasoning, absent contraindications, all patients with acute MI should go promptly to the cath lab. But in the real world it's usually only the ones who declare their diagnosis at the door who are eligible---and who are those patients? The ones with ST segment elevation! So is this paper of any practical importance? Maybe, in that it informs us that we should not be so rigid in our categorization. Furthermore, some patients without ST segment elevation but with a strong clinical picture or early biomarker elevation should be sent to the lab (remember what Grunt Doc did).

I think the distinction between STEMI and NSTEMI will persist for a while, just out of convenience. After all, once Phibbs' view gains traction it will open up a can of worms. The guidelines, which now lump NSTEMI with unstable angina, will have to be reorganized.

But I have a feeling it will eventually gain traction, because I think Phibbs is right. We wrestled with Phibbs in the area of electrocardiographic classification of MI once before and he won. In 1999 he published a paper demonstrating that the distinction between Q wave and non Q wave MI was meaningless. That view was not immediately accepted either, and Phibbs published several more papers on the subject over the next several years, until the concept of Q wave versus non Q wave infarction became obsolete. R.I.P.


Tom B said...

The distinction is not meaningless. The problem is that many "high-risk NSTEMI" patients are actually "STEMI patients". The admission comes from the statement that "NSTEMI is likely to be in the distribution of the circumflex artery." That's absurd. It's more correct to say that STEMI is more likely to be missed in the distribution of the circumfex artery. I've seen many acute posterior STEMIs misclassified as NSTEMI. That's the problem and it could be easily rectified if ED physicians were more skilled at reading 12-lead ECGs or employed modified chest leads V7-V9.


R. W. Donnell said...

I appreciate your points. I think any areas of disagreement we have may be semantic. In other words it depends on what you mean by STEMI. So does STEMI mean a certain pathophysiological condition (occlusion of a major epicardial coronary artery) or is it merely descriptive (ST elevation or no ST elevation)?

So let's take your example of posterior MI. You and I know it's a STEMI equivalent, because it's a major coronary occlusion and should be treated like a STEMI. However, if used in a purely descriptive sense it's not a STEMI because there's no ST elevation in the indicative leads!

Another example. I can recall being called to the ER to admit a patient for "unstable angina/NSTEMI." I looked at the electrocardiogram and noted the major change was ST depression in AVL. Changes in the inferior leads were subtle (mainly the large amplitude, broad based T waves)but there was no actual ST elevation. I knew this was a "STEMI equivalent" and said "that's for all practical purposes a STEMI; this patient needs her RCA opened up." The patient went to the lab and that's what she had. So that was a patient who pathophysiologically had a STEMI but by strict descriptive criteria did not.

Your statement about leads V7-V9 is well taken. If additional leads were utilized STEMI vs NSTEMI would undoubtedly take on more meaning. But your statement about V7-V9 also reinforces the point that to assign pathophysiologic meaning based on the mere presence or absence of ST segment elevation on a 12 lead electrocardiogram is fallacious.

A new classification, type 1 vs type 2 MI, (see this article which separates MIs based on whether they are due to coronary occlusion or demand ischemia, is more meaningful though not yet in common usage.

I could only access the abstract of the Phibbs paper. It claimed that, at least in the published literature, ST elevation had no correlation with pathophysiology or outcome. I am willing to take his word for it, as I know him to be a scientifically rigorous and honest investigator, although it would be interesting to read the body of the paper.

I look forward to browsing your blogs, BTW.

Tom B said...

Okay, I see your point.

Yes, I think a "STEMI" is an acute and occlusive thrombotic lesion in an epicardial coronary artery, and the "1 mm of ST-elevation in 2 or more anatomically contiguous leads" criterion is problematic at best.

I myself meet this criteria (2 mm of ST-elevation in lead V2 and 1.5 mm in lead V3) and I certainly don't need to be whisked away to the cardiac cath lab!

Then there are the situations you rightly pointed out. Acute isolated posterior STEMI, hyperacute T-waves without ST-elevation, or < 1 mm elevation in a lead with a very small QRS complex (rule of proportionality).

Finally, there is the problem of identifying acute STEMI in the presence of a baseline abnormality (e.g., paced rhythm, LBBB, LVH, and so on).

So I do think it's a problem of nomenclature. But I also think it's an error to classify these patients as having had a NSTEMI. It implies an imperfect understanding of the issue and a less-than-ideal grasp of ECG interpretation.

Great blog! Found you with a Twitter search.


Tom B said...

P.S. Thanks for the link! I will read it with interest.


burned-out medic said...

i like the part about how best to categorize AMI. he is right that we tend to rigidly separate pts into different groups; STE being the most accepted method.

however, even current AHA guidelines allow for sending pts to cath labs without STE, based on other factors, particularly how crappy a pt looks.

in our 12-lead adventures around here, it does seem to me that many EDs have forgotten about these considerations and have rigidly stuck to a cath-only-if-STE-present mindset. that seems to me is what dr. phibbs is most troubled by.