Wednesday, September 28, 2011

Is it STEMI or Non STEMI? What you call it has consequences.

This 53 yo male with no cardiac history developed chest pain 12 hours prior to admission.  It was coming and going, waxing and waning, all night.  At 8 AM, it increased to 9/10 and he called 911 and had this prehospital ECG recorded at 0858:
There is sinus rhythm with ST elevation in V1-V3, Q-waves in III and aVF, and ST depression in I, aVL, V5 and V6.  It is obviously ACS, and, to me, clearly a STEMI.  The computerized QTc is 429 and another ECG recorded shortly thereafter looked similar, with a QTc of 450.  
The equation value was about 30, which clearly reflects LAD STEMI.  In fact, this one is so obvious (due to the ST depression) that I would not have included it in my study comparing LAD occlusion to early repolarization.  The computer (which often misses MI) read it as Acute MI.  The cath lab was activated prehospital.

The patient received nitroglycerin and his pain greatly diminished (possibly even resolved). He arrived in the ED and had this ECG at 0947:
The ST elevation is somewhat less.  There is now ST elevation in lead III, with reciprocal ST depression in aVL (which was also present on the prehospital, showing how aVL is more sensitive for inferior STEMI than are inferior leads).  The QTc is now only 388 ms, and the equation value has dropped to 25.4 (STEMI less obvious, but still above 23.4).
The initial troponin was 2.4 ng/ml.  The patient went to the cath lab with a door to balloon time of 50 minutes, where he had a 4 cm long 95% narrowing of the LAD that was nitroglycerin responsive, and thought to be spasm.  It was a type III (wraparound) LAD to the inferior wall, explaining the inferior ST elevation.  There is no mention of thrombus.  It was stented.  There was also a chronically occluded circumflex and RCA which both filled via collaterals from the LAD (he was hanging on by the 5% of the LAD that was open).

Of course the ECG cannot tell you the cause of the ischemia (thrombus, spasm, or even demand ischemia), so the fact that this was spasm does not affect the ECG diagnosis, and there would be no way to know this without doing the cath.

Troponin I peaked at 82 ng/ml.  Echo showed anterior and apical, and inferior, WMA and EF of 35%.

The ECG the after PCI had the same appearance, and the next day there was slightly less STE and evolving T-wave inversions, seen here:
Persistent ST elevation in V1 and V2, with T-wave inversion in V3 - V6.


A cardiologists called this a NonSTEMI because he saw the ST elevation as normal.  Normal ST elevation, as I have shown in my study yet to be published, has an equation value that is less than 23.4 in almost all cases.  But not all.  Could this have been a false positive?

No. See below.

However, the troponin peaked at 82 (large MI consistent with STEMI), the convalescent echo showed a persistent anterior wall motion abnomality and a persistently poor EF of 44%.

So there was significant Myocardial wall loss, consistent with STEMI.

Finally, here is his ECG months later:
Most ST elevation is gone, equation value is now 22.1 (less than 23.4), and the R-wave in V4 has recovered some. 
I think this pretty clearly shows that the ST elevation was not baseline ST elevation, but rather a result of the acute coronary syndrome and therefore it was indeed a STEMI.

On review, the cardiologist sees that this is a STEMI.  Partly by just looking again and also because he had not seen the prehospital ECG.  

Learning points:
1. ACS should be called STEMI if there is ST elevation that is a result of the ACS.
2. When there is ST elevation that is the result of ACS, there is occlusion or near occlusion of the artery.
3. The diagnosis of STEMI vs. NonSTEMI should not be based on the degree of ST elevation; rather, it should be based on whether there is ST elevation due to occlusion or near occlusion. 1 mm, or 2 mm, are arbitrary measurements that do not have a physiologic basis.
4. Many STEMIs are erroneously called NonSTEMIs.
5. If you call it a NonSTEMI, you may delay cath lab activation that should be immediate.
6. Had this patient not gone immediately to the cath lab, he would not have shown up in missed STEMI statistics.  This is one reason why, in the literature, STEMI has such a low miss rate.  If true STEMIs are called NonSTEMIs, then of course the STEMIs are not missed so often.
7. If the door to balloon time had been 300 minutes, this would not have been reflected in the hospital statistics.  How many hospitals are accurately classifying their MI cases as STEMI?  When the door to balloon time is being tabulated, how often are cases like this called NonSTEMI just to buff the statistics?

Sunday, September 25, 2011

ST elevation of early repolarization may vary with the rate

This 49 yo black male presented with sudden substernal non radiating pleuritic chest pain on the day prior to presentation.  Here is the presenting ECG:
There is ST elevation in anterior leads that is classic for early repolarization: there is excellent R-wave progression, QT is not long, T-waves are assymmetric (slower upstroke than downstroke), and well-formed J-waves.  The equation value [1.196 x (ST elevation at 60 ms after the J-point)] + [QTc x 0.059] - [R amplitude in V4) x 0.326)] is low [= (2.0 x 1.196) + (418 x 0.059) - (25 x 0.326)], which is equal to 18.9.  A value less than 23.4 is unlikely to be MI.  Interestingly, the presence of J-waves did not add value to the prediction rule equation. 
Bedside  ultrasound showed no wall motion abnormality.  A troponin drawn the next AM was negative.  A repeat EKG showed increased ST elevation:
Equation value is: (4.5 x 1.196) + (0.059 x 400) - (0.326 x 28)  = 19.9 (early repol)

Repeat echo again showed no wall motion abnormality.

Why is the ST elevation greater in the second ECG?  One must remember that the ST elevation of early repolarization diminishes with increased sympathetic tone, such as during exercise.  When the heart rate is faster, as in the first ECG, the ST elevation is likely to be less pronounced than when the heart rate is slower.

Kambara found in his longitudinal study of patients with early repolarization that, in 26% the ST elevation disappeared on follow up ECG and in 74% the degree of ST elevation varied on followup ECGs.


I do not have proof in this case, but I'm pretty certain that the difference in ST elevation is due to the difference in heart rate.

Learning point: The ST elevation of early repol is not constant.  In particular, it may be diminished with exercise, sympathetic tone, and heart rate, and may be increased when the heart rate is slower.

Thanks to Steve Dunlop for this case.

Here is a good review article on early repolarization:

Tuesday, September 20, 2011

Cardiac arrest, LBBB with STEMI on the ECG, but no Acute Coronary Syndrome!



This 80 year old with a history of CABG had a cardiac arrest.  He was resuscitated after fairly prolonged down time, but regained consciousness, though he was confused.  He did not state he had chest pain, but, then again, he couldn't remember anything.  Here is the prehospital ECG at 1935:

Sinus rhythm with left bundle branch block (LBBB).  There is concordant ST elevation in all inferior leads.  Remember that, in LBBB, lead II should have discordant ST depression, and the fact that it does not indicates relative ST elevation.  In Sgarbossa's study, just 1 mm concordant STE in just 1 lead was 92% specific for MI and earned the ECG 5 points.  There is reciprocal ST depression (excessively discordant) in I and aVL, consistent with RCA occlusion.  There is proportionally excessively discordant ST elevation in V1, consistent with RV involvement.  There is concordant ST depression in V2 and V3, greater than 1 mm.  Just 1 mm concordant ST depression in just 1 lead of V1-V3 was 96% specific for MI and would earn the ECG 3 points.  So this ECG gets 8 points PLUS has excessive discordance in V1 PLUS has the finding in many leads, not just one.  3 points gets you an MI by Sgarbossa.   There are also marked Q-waves in lead III.
We did a bedside cardiac ultrasound.  Here it is:
This is a parasternal short axis view (a transverse cut through the heart; you see the left ventricle).  You can see that the part of the heart closest to the transducer (top) is the anterior wall and is moving and contracting well.  The part farthest (bottom) is the posterior wall and is not contracting.  This is a posterior wall motion abnormality.  This is consistent with MI, though one cannot tell if it is new or old.  The structure at the bottom that is moving is the mitral valve, with anterior and posterior leaflets.

This is as clear a STEMI as you can get.  Now, it is true that shortly after a non-ACS cardiac arrest, there can be transient diffuse ST depression, but not ST elevation in a coronary distribution, and there should not be a wall motion abnormality.

This was the initial ED ECG at 1951:
All the same findings are there, but they have diminished, consistent with reperfusion

There was some delay in cath team arrival since it was the middle of the night, so this right sided ECG was recorded at 2010:
The ST segments have normalized in the limb leads and in V1R (V2) and V2R (V1).  The Q-wave in lead III persists.
The angiogram showed a chronically occluded RCA and an occluded SVG to the RCA.  There was faint filling of the distal branches of the RCA by collateralss from the left coronary system.

As it turned out, the patient had an old inferoposterior MI that was scarred; this scar initiated primary V Fib arrest, which in turn resulted in temporary hypoperfusion of the inferior wall because of its very tenuous blood supply and resulting ST elevation on the ECG.

So this is classic inferoposterior STEMI on the ECG but is NOT acute coronary syndrome!

This could not have been known without the angiogram.  The ECG and ultrasound could not have been differentiated from acute plaque rupture with occlusion of the RCA.

Monday, September 19, 2011

Weakness due to Bradycardia due to Sick Sinus with Sinoatrial block, with Ashmann's phenomenon

This post is from last December, and shows sinoatrial block.  A reader was puzzled by why one p-wave conducted and another did not.

I had not thought about it at the time, but K. Wang noticed that the preceding R-R intervals were different, and this led to Ashmann's phenomenon.

Thursday, September 15, 2011

Respiratory Failure, Heart Failure, and Narrow Complex Tachycardia

A 60 yo female presented by ambulance in resp distress, requiring noninvasive ventilatory support. She had pulmonary edema and was near respiratory failure.

Here is the prehospital ECG (will not comment on the interpretation until the end of the post, so you can ponder it for yourself):
Computer read: "Sinus tachycardia" at a rate of 143, and "***Acute MI***". Obviously, there is no clear STEMI.

She was continued on respiratory support, treated for COPD and CHF with nebs, steroids, lasix and nitro. Here is her first ED ECG:
The computer reads "sinus tach".

An ED bedside echo was performed from the subcostal view:


Here is a legend for what you are seeing on the echocardiogram: LA = Left atrium, LV = left ventricle, RA = right atrium, RV = right ventricle, MV = mitral valve. The LA is greatly enlarged. The LV is very small. The mitral valve is very echogenic and highly suggests stenosis, which we confirmed with doppler.

Chart review confirmed h/o porcine mitral valve replacement with subsequent development of prosthetic mitral valve stenosis (that is to say, it recurred in the new valve). This conforms with our ED echo.

The heart rate continued at 143, and by this time there was much less artifact. This strip was printed out:
Now there is clear atrial flutter.
A bit of history could be obtained at this point, and she said she had had rather sudden SOB about 15 hours prior, and had had some pink and frothy sputum.

A better 12-lead was obtained:
Now, again, the atrial flutter is obvious, and there are no signs of ischemia
So, we have a patient who is in respiratory distress, due to mitral stenosis and complicated by atrial flutter, which diminishes LV filling that is already compromised by mitral stenosis. ACLS would say to do electrical cardioversion for a patient with atrial flutter and rapid ventricular response who is suffering respiratory failure, but patients with mitral stenosis are at very high risk of thromboembolism and stroke (old literature). Cardioversion would increase this risk. Therefore, we decided on slowing the ventricular rate with diltiazem. Here is the subsequent ECG:
There is now atrial flutter with 4:1 block.
With more time to fill, the LV was able to pump better. The patient improved gradually, and refused a valve replacement. She returned a few days later in distress and will now get a new valve.

Look again at the first ED ECG:


Knowing that it is atrial flutter, you can now see (if you didn't before) the atrial spikes in V1 (2:1) that might have been interpreted to be artifact.

Learning points:
1) When the heart rate does not change, but stays rapid and constant, it is probably not sinus tachycardia and then you should...
2) Look for atrial flutter waves
3) ACLS is guidelines only. Sometimes the patient does better with less aggressive care (and, of course, sometimes with more). One must always think it through.
4) Bedside echo can be very useful

Monday, September 12, 2011

Subacute MI masked by a wide complex

This is a post I put up about a year ago and had forgotten about until I noticed that it was getting a little traffic recently (how does this happen?). 

I find it very interesting and worth another look.

Click here to see these ECGs that represent a subacute MI masked by a wide complex.

Forgive me if you've seen it before and find it too repetitive to post again.


Friday, September 9, 2011

Chest pain, tachycardia, diffuse ST elevation. What is the diagnosis?

First, I want you to see the ECG without a lot of clinical information: this 45 year old male presented with left chest pain:

















More clinical information:
One week prior, he was stabbed in the left chest and a chest tube was placed.  His pulse was 120 and blood pressure 90/50.

ECG analysis:
There is sinus tachycardia. There is ST elevation in inferior, lateral, and anterior leads. The ST elevation in II is greater than III. There is no reciprocal ST depression in aVL. The ST elevation in anterior leads is marked, and scary. There is perhaps excessive PR depression (see esp. leads II, V5). This ECG is classic for pericarditis, but could conceivably be due to antero-infero-lateral STEMI due to proximal (before D1 to the lateral wall) occlusion of a type III (wraparound, to inferior wall) LAD.

Findings favoring pericarditis over diffuse MI:
1) no ST depression in aVL. We studied 160 inferior STEMI; only 1 had absence of any ST depression in aVL.  We also studied 39 consecutive cases of pericarditis who presented to the ED with chest pain and ST elevation meeting "reperfusion criteria."  None had any ST depression anywhere (except aVR, in which all had it).
2) The computerized QTc is 371 ms.  This makes anterior MI very unlikely.  In my study of 355 consecutive LAD occlusions, only 2 had a QTc less than 372 ms.
3) High anterior R-wave amplitude.  The mean R-wave amplitude in V2-V4 is at least 18 mm.  In my study of 355 consecutive LAD occlusions, zero had a mean R-wave amplitude this high.
4) STE in lead II greater than lead III.  This is not very accurate, and is found in inferior MI due to circumflex occlusion; i.e., inferolateral MI may have this.
5) The equation value (incorporating ST elevation, R-wave amplitude, and QTc) is 22.32 (less than 23.4 is unlikely to be LAD occlusion when early repol is the alternate diagnosis, but this has not been tested against pericarditis)

I was shown this ECG with no clinical info the next day, and due to the above considerations, after a glance at the ECG I said, "It's not an MI."

Verifying the diagnosis of pericarditis by ED bedside echo
If this is STEMI, it is a massive antero-infero-lateral STEMI and the patient would probably present in cardiogenic shock.  A bedside echo should show very poor LV function, and the patient should probably have pulmonary edema (which he did not).  A bedside echo was done: there was no pericardial effusion and, though the LV was not what the clinicians thought should be appropriately hyperdynamic, there was only "possibly decreased LV function."

Further evaluation revealed a fever of 101.  Fluids were given and the heart rate decreased, but an ECG 30 minutes later was unchanged.  The cath lab was activated.  Ultimately, he was not taken to the cath lab and troponins had a very tiny rise consistent with a very small degree of myocarditis in addition to pericarditis.  He grew out MSSA from his blood and had an infected hematoma in his chest.  The next day, he developed a pericardial friction rub, confirming the diagnosis of myo-pericarditis.  No more ECGs were recorded.

Here are more cases in which pericarditis was on the differential diagnosis.



Tuesday, September 6, 2011

Missed Acute MI, with coronary occlusion, evidence only by T-wave inversion in V2 and evolving ST depression in V3

A 39 yo otherwise healthy man with no risk factors was walking at the mall when he developed chest pressure.  He presented to the ED after 30 minutes, now also feeling weak.  He was diaphoretic.  Here was his initial ECG:

There is sinus rhythm.  There is abnormal T-wave inversion in V2 (in morphology, consistent with "persistent juvenile pattern" because the R-wave is not greater than the S-wave, but this would be very unusual in a 39 yo male). There is minimal, nondiagnostic ST elevation in inferior leads without any reciprocal ST depression in aVL.  There are thin and normal inferior Q-waves. Thus, there are some suspicious abnormalities, but no definite signs of ischemia. 

Because of persistent symptoms, another ECG was recorded 30 minutes later:

There is only one new finding on this ECG which suggests ischemia.  It is very subtle but real.  Look at lead V3, where there is now some ST depression.  The previous ECG has a small amount of appropriate ST elevation in V3.  Any ST depression in a young male is abnormal, especially if changed from previous.  This is especially worrisome when combined with the abnormal T-wave in V2.

Let's look at both V3's, magnified:

The later ECG (bottom panel) shows minimal ST depression in V3.  The top shows minimal ST elevation (normal).  The difference is significant and highly suggests posterior ischemia.

This abnormality in V3 was apparently not seen by the treating MD, who is a nationally recognized expert in STEMI care (showing how difficult these diagnoses can be).

The initial troponin was negative. The patient was admitted to telemetry.  At 4 AM, his second troponin returned at 1.8 ng/mL.  Another ECG was recorded:

T-waves in V2 and V3 are now upright and larger, evolving.  Are these posterior reperfusion T-waves?


He went for cath at 6 AM and had on occluded OM-2 that was opened and stented.  Troponin I peaked at 99 ng/mL (large MI)!

So this is a NonSTEMI, right?  Technically, yes, because there is not 1 mm of STE in 2 consecutive leads.  But the definition misses the point.  It is a coronary occlusion with a substantial myocardial territory at risk, that showed only very subtle ST changes.

Should you activate the cath lab for this?

Not from the ECG alone.  However, if you notice the ST depression, you then realize that this is ischemic chest pain, not esophageal spasm.  Once you know that the chest pain is ischemic in origin, and you cannot control it medically, then you must go urgently to the cath lab.

The patient should be treated with NTG, Aspirin (and clopidogrel, if your institution allows), metoprolol, antithrombotics, and GP IIb IIIa inhibitors.  If the pain persists, and the ST depression persists, then talk to your interventionalist immediately.

Here is the followup ECG:
The ST abnormalities have resolved.  There are new inferior Q-waves diagnostic of inferior MI.  The R-wave is increased in V2, consistent with MI (analog of a Q-wave).  There is no apparent resolution of the minimal and non-diagnostic inferior ST elevation.

I don't have all the data on this case, and do not know if there is an inferior wall motion abnormality, or if this OM-2 supplied the inferior wall.  It  probably did, as evidenced by the Q-waves; but it is very interesting that during the acute phase, there were no diagnostic ST changes in inferior leads, and the minimal ST elevation that was present did not evolve.

Many MIs are electrocardiographically "silent," especially when in the circumflex territory.  I do wonder whether, in the studies that show this phenomenon, if an ECG expert evaluated the ECG for the subtle signs of ischemia.  I suspect that many or most that are thought to be "silent" are really just "subtle."

Here are more electrocardiographically subtle MI.

Thursday, September 1, 2011

What is the rhythm? Answer at the bottom.

An elderly woman presented with abdominal pain.  Here is her ECG:
























This is easily mistaken for atrial flutter because the waves have a fluttering, wavy, sawtooth appearance.  However, 3 things make atrial flutter impossible, 2 are related:
1) the rate of the waves is > 400
2) the QRS does not appear at the same part of the atrial wave cycle each time (which it must do in flutter)
3) the ventricular rate is irregularly irregular.  In atrial flutter, the ventricular rate can be regular if there is a constant ratio (2:1 block, 3:1 block etc).  It can also be regularly irregular, in which every QRS comes at a multiple of the atrial wave rate (if atrial wave rate is 300, it happens every 200 ms, and every R-R interval must be a multiple of 200ms)

2) and 3) are related: since 2) is true, 3) must also be true.

The woman had atrial fibrillation.  It was misdiagnosed as flutter, and this may have contributed to a delayed diagnosis of mesenteric embolism.