Beware confusing the diagnosis of posterior STEMI by using posterior leads…

Or at least understand their limitations!!

This patient had chest pain and this ECG:
There is no significant ST elevation. But there is ischemic ST depression in V2-V6.  It is maximal in V3 and V4, which represents posterior STEMI until proven otherwise.

(This can be due to subendocardial ischemia, but less likely.  Most subendocardial ischemia is diffuse, with the ST depression vector pointing towards the apex of the heart (II, V5) and thus has maximal ST depression out in V5 and V6.  

The ST elevation vector is clearly away from V3 and V4, towards the posterior wall.  If you put posterior leads on, you MUST get posterior ST elevation.  HOWEVER, it may be of very small magnitude and thus is may be a false negative.

Posterior leads were applied:
V3 is misplaced here, so ignore it.  Notice there is some ST elevation in V7-V9, but it is minimal.  It does not even meet the "criteria" of 0.5 mm in posterior leads.  V2 has the same ST depression in had on the first ECG, confirming that the artery is still closed.

So this gives a false negative.  Posterior leads lead you astray.  The voltage is so small that you might be dissuaded from your diagnosis of posterior MI.

This was indeed a circuflex occlusion with a posterior wall motion abnormality.

So when should you use posterior leads?

There is some literature showing that some posterior MI show up only on anterior precordial leads as ST depression, and some show only on posterior leads.

I think some of this is due to timing.  Arteries open and close and if you don't record the ECGs simultaneous, then you might be recording two different conditions of ischemia.

It may be useful to record posterior leads in a patient in whom you are convinced has an MI but has no significant abnormalities on the standard 12-lead. 

Subtle LAD Occlusion with Pseudonormalization of Wellens’ Waves.

Here is another patient I saw a few weeks ago  He is in his 30's and presented with many hours of chest pain.  Here is his initial ECG:
The computerized QTc is 361.  There is ST elevation in V2-V4.  Is it normal ST elevation?  Normal ST elevation should have good R-waves and V2 and V3 have Q-waves.  So this is not normal ST elevation.
STEMI criteria: 2 consecutive leads with ST elevation, defined as 1 mm in V1 and V4-V6, and, for a 35 year old, 2.5 mm in V2 and V3 (at the J-point relative to the PQ junction).   This ECG has at most 1 mm in V2, 1.5-2.0 in V3, and 1.5 in V4, so it does not meet STEMI criteria.  LAD occlusion frequently does NOT present as STEMI on the ECG.

With the presence of Q-waves, one should assume this ST elevation is due to LAD occlusion, not to normal variant.  Do not even use the formula.

Let's go back in time

It turns out he had been in the hospital less than a month prior with a NonSTEMI with Wellens' syndrome.  Here is his ED ECG from that visit, after resolution of chest pain:
Classic Wellens'.  Patient is at high risk of closure of his LAD.

The physicians wanted to do an angiogram, but in spite of pleading with the patient, the patient would not agree to undergo any further testing because he felt fine.  They discharged him on aspirin and clopidogrel.

This was his pre-discharge ECG after having a peak troponin I of 0.78 ng/mL (99% = 0.030 ng/mL)
Notice that much of the T-wave inversion has normalized, but there is residual biphasic T-wave inversion in V4 and V5.

A pre-discharge echo showed LVH, no wall motion abnormality, and an EF of 60%.

Back to the 2nd ED presentation

And then he presented to me with this ECG, which I post again here so that it is in sequence:
I did not see this ECG immediately.  The computer of course read it as early repolarization.  The resident saw it and was worried and ordered a repeat ECG.      

--Had the resident compared with the previous ECG, the loss of R-wave would have made the diagnosis immediately.
--Had the resident seen the pseudonormalization of T-waves (now upright in V4 and V5, all but diagnostic of re-occlusion), the diagnosis would have been made sooner.
--Serial ECGs should not wait an hour

This is ECG #2, 54 minutes after the first:
Obvious anterior STEMI

Angiogram showed a proximal subtotal thrombotic occlusion which was stented.

Here is the post cath ECG:
Marked loss of R-waves with persistent ST elevation.  These may resolve over time but are highly correlated with development of LV aneurysm.

Echo 5 days later showed anterolateral, septal, and apical wall motion abnormalities and an EF of 40%.  Peak troponin I was 176 ng/mL (very high).

Learning Points:

1. As I've endlessly repeated here, LAD occlusion may be very subtle.  Any delay in diagnosis can result in significantly worse outcome.

2.  Wellens' syndrome is the result of reperfusion of a briefly occluded LAD.  That LAD is at high risk for re-occlusion

3. Re-occlusion results in pseudonormalization (see link for other cases) of the inverted Wellens' waves.  The apparently normal, upright, T-waves are in fact NOT normal and are the result of re-occlusion.

It is Far too Early for a Requiem for Unstable Angina

Dr. Braunwald declared that high sensitivity troponins will make Unstable Angina a thing of the past.  That very well may be true, but in the United States we are not yet in the era of high sensitivity troponins.  And we do not know what their effect will be.

Here are many more cases of Unstable Angina.

Unstable Angina

A male in his 50's with no past history presented with new intermittent burning left chest pain, lasting 10 minutes at a time, radiating to the left arm, for 24 hours.  He had chest pain just prior to arrival in the ED, but it resolved prior to physician evaluation.  He did use cocaine a few days prior.  He had blood pressures in the 170/100 range.  Here is his initial ECG:
There is terminal T-wave inversion in V2-V5, highly suggestive of "Wellens' syndrome."
Patient had pain, is pain free, has intact R-waves, and terminal T-wave inversion
(Wellens' Pattern A)

There was no further pain, but a second ECG was recorded 1 hour later:
There is some evolution of T-wave inversion (V3 has a deeper negative deflection), increasing suspicion for Wellens' syndrome.

The ED was concerned about "Wellens' pattern," and gave aspirin, clopidogrel, and heparin.  He was not given anything for his blood pressure.  He was admitted to the hospital. The consultants were very worried as well, and discussed angiogram vs. stress test in the morning.

Comment: when a patient has cardiac ischemia and is hypertensive, the blood pressure should be controlled.  There are many choices. One is metoprolol.  There is an admonition not to use a beta blocker in patients who have cocaine in their system because if may produce "unopposd alpha" effects by blocking the vasodilating effects of beta-2. However 1) that admonition is based on faulty data from combined beta-1 and beta-2 blockade (namely propranolol -- metolprolol has almost no beta-2 blocking effects) and 2) the patient has not had cocaine for days.

The patient ruled out by 4 serial contemporary troponins over 9 hours.

A cardiac echo was normal, showing no wall motion abnormality and EF of 65%.

No more ECGs were done (was there further evolution of T-wave inversion?  Was there resolution?)  In my experience, if troponins are all negative, the T-waves rarely evolve to Pattern B (deep symmetric inversion).  Instead, they often resolve.  Such resolution is NOT pseudonormalization (which is due to re-occlusion of the vessel), but rather due to resolution of ischemia.

The assessment in the light of a normal echo and normal troponins was that this was noncardiac chest pain or cocaine chest pain with nonspecific T-wave inversions.

There had been discussion of a stress test but the patient was discharged.  It seems he had been instructed to go to his primary MD and get a stress test.

Case progression.

Two days after discharge, the patient presented to another hospital ED with the same complaints of intermittent chest pain (3 episodes).  He had not seen his MD nor had a stress test yet.  His last episode of chest pain was just prior to ED arrival and was resolved upon arrival.  Here is the initial ECG:
Classic Evolution of Wellens' Waves

The physician had no access to previous ECGs.

The first troponin was negative.  The physician consulted the on-call cardiologist who thought it was very unlikely that this would be Wellens.'  He did not think it possible that the patient could have 4 days of intermittent chest pain with normal troponins throughout and without wall motion abnormality, and with a normal ejection fraction.

He thought the most likely cause was related to the patient's recent cocaine use and vasospasm. He did agree with the plan for stress test but felt like it could be done as an outpatient.

The patient was told to return for increased or longer lasting pain.

He did return the next day after another episode of pain, and was pain-free on arrival.

Here is the ECG:
More Wellens' Evolution, but with some loss of R-wave amplitude

This time the initial troponin I returned at 1.5 ng/mL.  The peak troponin is not available.

He had an angiogram which showed a 95% thrombotic LAD stenosis with preprocedure TIMI-3 flow.  It was stented.  There was also an 80% proximal LAD stenosis.

Here is the angiogram:

Here is the post cath ECG:
There is loss of R-wave in V2 and V3.  The T-waves have evolved as expected from Pattern A (biphasic) to deep and symmetric (Pattern B). 

A post cath echo showed anterior and apical WMA with EF of 55%.  Will this be permanent?  It is uncertain if there was irreversible loss of significant myocardium.  The loss of R-waves suggests that this is the case.

Learning Points:

1. Unstable Angina still exists.  Contemporary troponins are CERTAINLY not good enough to rule it out.

2.  Regional wall motion recovers rapidly after reperfusion of the infarct artery.  After chest pain is resolved, a normal echocardiogram should not reassure you!

3.  Pay attention to specific ECG findings!

4. Chest pain in a patient who uses cocaine is just as likely to be due to ACS as it is in a patient who does not use cocaine.  See this article and editorial.  

5. Chest pain that occurs more than one day after cocaine use should not be attributed to cocaine.

I was reading a stack of ECGs yesterday, and saw this one, with no clinical information….

I was reading a stack of ECGs yesterday, and saw this one, with no clinical information:
What do you think? (computerized QTc is 429 ms)
See followup below.

My thought was: "This is an acute LAD occlusion, and I am afraid it would have been missed."

Why did I think that this apparently benign looking ECG represents LAD occlusion?  It is an ECG that does not meet any millimeter "criteria" for STEMI.

I am sorry if I repeat myself too often, but there are many LAD occlusions which are missed at initial presentation because it is thought to be normal variant ST elevation.  That is why I studied it and produced the formula for differentiating benign from ischemic ST elevation in "anterior" leads.  Many are not diagnosed until the patient "rules in" by biomarkers, goes to cath later, and has the artery opened only after the damage is done.  And these get a final diagnosis of "NonSTEMI" because they do not meet "criteria" for STEMI and yet they are due to complete occlusion of the artery. You don't get a message from anyone that you "missed a STEMI."  And yet the patient lost a lot of myocardium unnecessarily.

See these references:
Wang T et al.  Am Heart J 2009;157(4):716-23.
From AM, et al. Am J Cardiol 2010;106(8):1081-5.

Pride YB et al.  JACC: Cardiovasc Interventions 2010; 3(8):806-11.

Almost all of these occlusions can be diagnosed, but only with a high suspicion and with careful investigation, serial ECGs, and high quality echocardiography.

On this ECG there is ST elevation in lead V2, and it must be explained.  Is it benign?  If it were benign, then there should be excellent R-wave progression.  There is not.  In fact, the R-wave gets smaller from V2 to V3.  Furthermore, the T-waves are much more symmetric than the T-waves of early repolarization.  If you plug these numbers into the excel applet on the sidebar:
--ST elevation at 60 ms after the J-point in lead V3
--computerized QTc
--R-wave amplitude in V4

The resulting value is 25.0.  Any value greater than 23.4 is LAD occlusion until proven otherwise.  Less than 4% of LAD occlusion have a value < 22.0.

So I went to the chart:

The patient had acute onset of chest discomfort, called 911, and had this prehospital ECG approximately 15-20 minutes prior:
Besides a PVC, there is no evidence of ischemia.  But the fact that it is so different from the one above tells us that the one above one has evolved injury (ischemic ST elevation).

It turns out that the clinical information made this case much easier than the ECG: the patient looked distressed, was diaphoretic, vomiting, and ill appearing.

The clinicians looked at this first ECG and the prehospital one and did not think much of it.  So they recorded another one shortly thereafter:
The ST elevation has evolved into a definite STEMI.

The cath lab was activated.

There was a 100% occlusion of the mid-LAD.

Here is the post cath ECG:
This is a normal amount of ST elevation, and normal T-waves in V2-V4.  There is some T-wave inversion in V5 and V6.

The management was so fast that the peak troponin I was only 9.2 ng/mL.

Subsequent echo showed:

Left ventricular hypertrophy concentric .
Regional wall motion abnormality-distal inferior wall .
Regional wall motion abnormality-apex.

I suspect there will be nearly full recovery of wall motion because the post cath ECG is nearly normal.

Here is another identical case that did not turn out so well.

Learning Point:

ST elevation must be adequately explained, even if it appears benign.  This case was relatively easy because of the clinical presentation.

Many others are not so easy and require subtle ECG diagnosis.

People often think that I read these with a retrospectoscope.  I do not.  I found this in a "stack" (reading through a list on the computer) of undifferentiated ECGs.  And I do this often.  There are ways to make the diagnosis from the ECG. They can be learned.  And the LAD occlusion, early repol formula can greatly help.

Bilateral Arm and Mid Back Pain, BP 250/140, in a 50-something

A 50 something presented with 10/10 bilateral arm and pain between the shoulder blades that started at work.  It felt like a "tearing," and was associated with diaphoresis.

BP was 250/140.  He appeared distressed.

Here is the initial ECG:
This is diagnostic of inferoposterior MI, with acute coronary occlusion.  
There are hyperacute T-waves in inferior leads, ST depression and T-wave inversion in aVL, and ST depression with T-wave inversion in V2 and V3.  There is also ST depression in V4-V6.

Of course the initial fear was aortic dissection, with the classic pain and severely elevated BP.  And it is not uncommon for a dissection to dissect down the coronary arteries, especially the RCA, so one might find an ECG like this in aortic dissection.  (Though only a tiny minority of RCA STEMI are caused by aortic dissection!)

The patient was given labetolol without much effect, then esmolol and nitroprusside until the BP was under control.

He was sent to the CT scanner where no dissection was found.

At this point the ECG findings must be attributed to acute coronary syndrome.  A repeat ECG is indicated to look for evolution or resolution. 

Before another ECG was obtained, the pain resolved.  This was recorded at 90 minutes:
Normalization was attributed to normalization of the severely elevated BP.

The first troponin was slightly elevated.  The ischemia was thought to be due to severe hypertension (a type 2 MI).  The cath lab was not activated.

Only less than 5% of type 2 MI present with subepicardial ischemia (ST elevation).  Demand ischemia from hypertension would be unlikely to manifest such focal STEMI on the ECG.


While waiting for a bed, he developed chest pain again, and this ECG was recorded at 120 minutes:
Now there is obvious inferoposterior STEMI

The cath lab was activated.

While waiting, the pain subsided again 10 minutes later (recorded at 130 minutes):
ST segments and hyperacute T-waves have normalized.

10 minutes later, the pain recurred again (time 140 minutes):
Obvious STEMI again.

Cath showed a 100% thrombotic occlusion of the RPAV off the RCA, stented.

Reperfusion was so fast that there was no wall motion abnormality the next day.