Monday, February 29, 2016

Nausea and Vomiting, Again. This ECG is loaded with information.

A middle-aged diabetic dialysis patient presented with 24 hours of nausea and vomiting associated with ~6 pound weight loss.  Patient stated his dry weight is around 85 kg.  The emesis is nonbloody and nonbilious.  He did have one episode of diarrhea.  He denied fevers and chills, abdominal pain, chest pain, or SOB.  Has been able to intermittently tolerate PO intake.  He is worried that he may be dehydrated.  He did not finish his dialysis run today.
Someone thought this was slow VT with a "northwest" axis (towards aVR, or -135 degrees), but there are definite P-waves (with a long PR interval).  
P-waves are best seen (as always) in II and V1, biphasic in V1 (as always).

The QRS duration is 150 ms, with a large R-wave in V1 and a wide negative QRS in lateral leads.  
Thus there is RBBB

There appears to also be right ventricular hypertrophy, as the R' wave in V1 is greater than 15 mm.

Besides the above, what is your diagnosis?

This ECG was texted to me with no clinical information, and I texted back:

"RBBB with RVH and inferior-posterior-lateral subacute MI.  Duration of symptoms is greater than 12 hours.  First troponin I will be high, maybe 60 ng/mL (60,000 ng/L)" (See short discussion below of the relation of troponin to infarct size)

Why did I think there was any acute MI, and why subacute?   (Why did I think the symptoms were prolonged)?

Answer: There are well-formed Q-waves in inferior and lateral leads.  In RBBB, V5 and V6 usually have a narrow R-wave followed by a wide S-wave (see the old ECG below, Figure 2).  But in the above ECG, V5 and V6 only have a Q-wave.  Inferior leads have only a minimal residual r-wave, preceded by large Q-waves.

The R-waves in V1-V3 are very large, suggesting posterior completed infarction vs. RV hypertrophy.  Posterior MI leads to loss of posterior depolarization amplitude and thus to increase in anterior depolarization amplitude.  These would be Q-waves if recorded from posterior leads, but since they are recorded anterior, they appear as enlarged R-waves.  In figure 3, I have inverted the image vertically to simulate recording leads from the opposite polarity (see Figure 3).

Here is the patient's previous ECG (Figure 2):
Previous ECG.  Normal RBBB, no evidence of ischemia.  R-waves of of normal height.  There are no significant inferior or lateral Q-waves.  There is no ST elevation in any lead.

Figure 3 (Figure 1 flipped vertically)
This is the presentation ECG flipped vertically.  Doing so is analogous to recording all leads from the opposite polarity: for instance, it is similar to recording V1-V3 from posterior leads.  One difference is that actual posterior leads must be recorded with much intervening lung, which dampens the voltage (amplitude).  By inverting the image, the amplitudes recorded from the precordial leads are maintained.
One can now see that these large R-waves are Q-waves and the ST depression, as seen from the posterior wall, is really ST elevation!

Patient course

The first troponin returned at 200 ng/mL!!

The cath lab was activated.  The culprit was a 100% occluded circumflex.  It was opened and stented.  the Troponin peaked at 500 ng/mL!  One would think that such a high troponin would be correlated with very poor ejection fraction, but as I explain below, peak troponin is not a reliable indicator of infarct size.

There was also a calcified 95% RCA and calcified 60% LAD.

The echocardiographic ejection fraction (EF) was 50%, with a posterior-lateral- and inferior wall motion abnormality.

Several days later this ECG was recorded:
There is persistent ST deviation, but much decreased from before.  4mm of ST depression in V1 and V2 has diminished to 1-2 mm, consistent with reperfusion.
The persistent ST deviation, however, suggests possible development of a future ventricular aneursym.  However, it is too early to make any conclusions.  Much ST deviation may persist for weeks before entirely resolving.

This ECG was recorded 6 days after first

There is some persistent ST deviation.  This may be permanent and may be associated with echocardiographic dyskinesis (aneurysm).

LV aneurysm is common in completed, full thickness (transmural) MI, which is what we have here.  It is uncommon in the age of reperfusion therapy,  as most STEMI get treated reasonably early, before transmural infarct.

Patients with completed, transmural infarct are also at risk for post-infarction regional pericarditis and myocardial rupture.  LV aneurysm puts them at risk for a mural thrombus, which puts them at risk for embolism, especially embolic stroke.

Peak Troponin and Infarct Size:

60 ng/mL would be a very high first troponin.  Most STEMI peak at over 10 ng/mL; most NonSTEMI at less than 10 ng/mL.  A very large anterior MI may have peak troponins in the 80-300 range, and I have seen it up to 500 in a 24 hour old, non-reperfused left main occlusion (it would have been much higher still if measured after reperfusion).  However, peak troponin levels are very variable and not a reliable measure of infarct size because they are dependent on too many variables:
1) the amount of myocardium at risk
2) the duration of ischemia
3) the density of ischemia (are there collaterals?)
4) the persistence of occlusion (opening and closing spontaneously?), and
5) reperfusion (which releases large amount of troponin).

Friday, February 26, 2016

"I have food poisoning"

This patient in her 40's with type 1 DM presented with 16 hours of vomiting (x 10) and diarrhea (liquid/loose x 10).  She came to triage complaining of "food poisoning."  There was some diffuse abdominal pain, but no chest pain or shortness of breath.  She was tachycardic to 120 and had no Kussmaul respirations.  Her abdomen was diffusely tender but without guarding.

The intern presented the patient and a plan to check a chemistry panel, and to give antiemetics and fluids.

I said: "you better order an EKG; you never know when a vomiting patient, especially a diabetic, is having an MI."

Here is her ECG:
There is anterolateral ST elevation, with well-formed QS-waves in V1-V3.  This is diagnostic of a subacute anterolateral STEMI.
She did have a mild ketoacidosis as well.

At angiography, there was a 100% thrombotic occlusion of the mid (not proximal) LAD and had evidence of prolonged occlusion (subacute MI).  The initial troponin I was 88 ng/mL (88,000.00 ng/L), confirming prolonged occlusion and large MI.  The highest troponin, after artery opening, was 222 ng/mL.  Echo showed EF of 12% and multiple wall motion abnormalities.  Angiogram showed severe downstream microocclusion.

Next day ECG:
Persistent ST Elevation and Long QT.
Such persistent ST elevation with QS-waves is associated with mechanical complictions (e.g., myocardial rupture) and with development of LV aneurysm.

At 48 hours:
ST Elevation remains persistent

Learning Points:

1. ECGs are cheap and noninvasive.  As long as you know how to keep from over-interpreting ECGs in patients with a low pretest probability, it can't hurt to order one in nearly any patient who has acute chest or abdominal symptoms.

2. A clearly diagnostic ECG is diagnostic even when there are no clearly ischemic symptoms.  It may even be diagnostic when the patient has no symptoms!

Here is a case of a patient whose only symptom was hand numbness: 

Spontaneous Reperfusion and Re-occlusion - My Bad Thinking Contributes to a Death

Friday, February 19, 2016

Why is there ST depression in aVL in this case of Pericarditis?

This ECG is from an article in the New England Journal on pericarditis.  

Acute Pericarditis

Martin M. LeWinter, M.D.
N Engl J Med 2014; 371:2410-2416December 18, 2014DOI: 10.1056/NEJMcp1404070

A reader had read this post on ST depression in lead aVL in pericarditis vs. inferior STEMI

He sent this ECG recorded from a patient with pericarditis and asked why there is ST depression in lead aVL:
There is sinus rhythm with diffuse ST elevation (especially in inferolateral leads, which is typical of pericarditis).  There is diffuse PR depression and Spodick's sign.
There is a tiny amount of reciprocal ST depression in lead aVL.

My answer to the reader is below:

As there are always exceptions to every rule, one must look at every ECG individually.
There is minimal ST depression in aVL, perhaps 0.5 mm.
Put that into the context of the whole ECG: there is very significant ST elevation in inferior leads.  This is not subtle. When there is inferior MI with large ST elevation, there is nearly always marked reciprocal ST depression in aVL.

Proportional Analysis

Always interpret every ECG with proportional analysis.  Everything should be proportional to everything else for any given diagnosis.    In this case, you see very large inferior ST elevation, but only minimal (if any) reciprocal ST depression.  The proportions do not make sense.  This would be exceptionally unusual for inferior MI.

ST axis analysis

The ST elevation is higher in lead II (3mm) than lead III (2mm).  This is also a pretty good ECG sign for pericarditis, as it implies the ST axis is closer to lead II than III, and less than 90 degrees.  When there is ST depression in aVL, the ST axis must be to the right of lead II (greater than 90 degrees).  Thus, the ST axis is between leads II and aVF (60-90 degrees).  However, the fact that there is such minimal ST depression in lead aVL tells us that the ST axis is only BARELY greater than 60 degrees, perhaps 65 degrees.  Thus, it is still very close to the normal axis for pericarditis (up to 60 degrees) and it is unusually leftward for inferior STEMI.

But, as I always say, you diagnose pericarditis at your peril, so always approach this with caution!

Steve Smith

Sunday, February 14, 2016

Is it early repol, pericarditis, or anterior STEMI?

This was sent to me by a reader:

This patient was a 26y/o pulled over by Police when he complained of chest pain that radiated to left arm. EMS was called. The initial 12 lead is here:

Smith comment:
There is sinus tachycardia.  There is a high voltage QRS, particularly high voltage S-waves in V2 and V3.  There is marked ST elevation in V3-V5, but there are also very marked J-waves in those leads and the T-waves are not large relative to the QRS.
This is all highly suspicious for early repolarization, especially in a 26 year old.
One can use the LAD occlusion vs. early repolarization formula: I have magnified V1-V6 for better measurement:
STE at 60 ms after the J-point in lead V3 (STE60V3) = 4 mm, QTc = 386 ms, and R-wave amplitude in V4 (RAV4) = 16 mm, we get 22.342

A value of 22.342 strongly supports early repolarization (about 95% sensitive for LAD occlusion in all comers, but has a better negative predictive value in a 26 year old).

With this, I would do serial ECGs and an emergent echo to look for a wall motion abnormality.  I would not activate the cath lab, as LAD occlusion is very unlikely.  

Of course it is better to be safe than to miss an LAD occlusion, but a an echocardiogram done with contrast and interpreted by an expert will not miss an LAD occlusion and would spare the cath team a night time activation.

Crews became concerned about ACS and administered ASA, nitro, and serial ECG's showed no change in ECG.

The patient was taken to the cath lab upon arrival at hospital.

When I first saw this ECG I thought the elevation looked concerning but the QT was not that long and there is huge QRS in anterior and lateral leads. QRS in lead V4 is at least 15 according to my calculations give or take. I saw no reciprocal depression.

Pt went to the cath lab where his coronaries were found to be clean. Troponin's were not elevated. Cardio echo showed no wall motion abnormality and good function.  The patient was observed overnight and discharged the next day with diagnosis of pericarditis.


I doubt this was pericarditis.  Pericarditis has inferolateral ST elevation primarily.  This is classic early repolarization.

Friday, February 12, 2016

Pericarditis? Or STEMI? The difference can be fatal.

This was sent by a reader who wishes to remain anonymous.  Details are at a bare minimum, but it remains instructive.

An elderly patient had a good reason for demand ischemia and there was no suspicion of ACS.  This ECG was recorded:
There is sinus tachycardia.  Complexes 3, 6, 12, & 15 appear to have some pre-excitation, with secondary repolarization changes.
In leads without pre-excitation, there is
diffuse ST depression consistent with demand ischemia

Clinical Course

The pathology leading to the ischemia was treated and resolved.  Later, what may have been simply a routine ECG was recorded.  If the patient had symptoms at the time of this 2nd ECG, they were not documented.  It also is unclear if anyone involved with the patient saw the ECG at the time it was recorded -- no clinician commented on it in the chart.

Here it is:
The cardiologist who read the ECG on the system (who probably did not know the patient or the reason for the recording) read this as "Diffuse Nonspecific ST Elevation, consider Pericarditis"
What do you think?

There is subtle ST elevation in inferior and lateral leads, and subtle reciprocal ST depression in lead aVL and also subtle ST depression in lead V2.  This is highly suspicious for acute STEMI, and even meets STEMI criteria in leads V4-V6.  It is very unlikely to be pericarditis.


Not long after the ECG was recorded, the patient arrested.  She was unable to be resuscitated.  The autopsy showed an acute RCA thrombosis with 90% occlusion.

Learning Points:
1. You diagnose pericarditis at your (or your patient's) peril
2. Pericarditis does not have reciprocal ST depression, not in either lead aVL or in V2.  This ECG is very specific for STEMI.
3. Even when patients are asymptomatic, or apparently so, they can have life threatening MI.  If the ECG has specific signs of MI, as this one does, then do not be fooled by the absence of symptoms.
4. It is very easy to overlook ECG findings if you are interpreting a routine ECG, without knowledge of the patient's condition (this includes over-reading of ECGs from a list, as was done in this case).

Monday, February 8, 2016

Subacute STEMI. Should the patient go for emergent PCI, or can he wait until the next day?

This was sent by a very good medical student, who had a very good question.  He wishes to remain anonymous.

A 56 y/o male presented with Chest Pain radiating to the left jaw, starting at 8pm the previous night (15 hours prior), which was 10/10 at that time.  He decided to wait it out at home, then presented at around 11 AM, pain now reported at 2/10.

Here is the initial ECG (see the patient's previous ECG below for comparison):
There is sinus rhythm and new inferior QS-waves with less than 1 mm of inferior ST elevation, and reciprocal ST depression in aVL, and T-wave inversion.  
Such T-wave inversion is common not only in reperfusion, but in persistent and prolonged occlusion after formation of Q-waves, especially QS-waves.

Here is the patient's previous ECG:
This old ECG confirms that the first ECG shows a new MI

So this patient has a subacute STEMI.   

The initial troponin T returned at 0.47 ng/mL (quite high for Troponin T) and rose from there to 0.81, then 1.96 (typical of a large STEMI).  Whether the troponin continues to rise or not says nothing about ongoing ischemia: it takes time for complete troponin rise and fall even after infarction is completed.  Only the ECG and pain can tell you prospectively whether ischemia is ongoing.

The patient was admitted with "NonSTEMI" and did not undergo emergent angiogram and PCI.  He went the next day.

This is a nearly completed STEMI (a very advanced subacute STEMI). To call it a NonSTEMI is misleading.

Is it too late for emergency cath lab activation?  Should he go now, or with less urgency (tomorrow)?

My answer is this: 

If there is:
1) persistent ST elevation (as there is here) or 
2) persistent pain (as there is here)

Then the patient should go emergently.

But I know of no data to support this.

There are 2 excellent articles addressing whether a patient with completed MI should undergo PCI at all (vs. medical therapy alone), but none addressing this situation.

Schomig et al. published this in JAMA in 2005:  
Data presented shows benefit of PCI (vs. medical therapy alone) for patients who present between 12-48 hours after STEMI if there was persistent STE, or simply new Q-waveseven in the absence of pain.  But they did not assess the urgency of PCI.

Hochman et al. published the Occluded Arteries Trial (in the New England Journal).  
In this article, they assessed arteries occluded for 3 days or more, and found that PCI resulted in worse outcomes than medical therapy.

The question this student posed was slightly different: beyond 12 hours, is emergent PCI better than delayed PCI (both groups getting PCI).

ACC/AHA 2013 STEMI guidelines say this:

Primary PCI is reasonable in patients with STEMI if there is clinical and/or ECG evidence of ongoing ischemia between 12 and 24 hours after symptom onset (94,95). (Level of Evidence: B).  This means if there is chest pain, persistent ST elevation, or especially upright T-waves.

But the ACC/AHA guidelines do NOT address the timing (emergent vs. urgent) and they reference two studies, one of which is the Schomig article above, and the other of which also does not address emergent vs. urgent PCI.  

94. Schömig A, Mehilli J, Antoniucci D, et al. Mechanical reperfusion in patients with acute myocardial infarction presenting more than 12 hours from symptom onset: a randomized controlled trial. JAMA. 2005;293: 2865–72.

95. Gierlotka M, Gasior M, Wilczek K, et al. Reperfusion by primarypercutaneous coronary intervention in patients with ST-segment elevationmyocardial infarction within 12 to 24 hours of the onset ofsymptoms (from a prospective national observational study [PL-ACS]). Am J Cardiol. 2011;107:501–8.

Learning Points
1. I think if there is ongoing pain or ST Elevation, it is wise to go emergently to the cath lab to save any remaining viable myocardium.  Randomized data on this is lacking.  Probably, most researchers would be reluctant to test such management.  Similarly, no one has ever conducted a randomized trial of emergent vs. delayed PCI for NonSTEMI with refractory (continued) symptoms.
2. PCI at some point within the first 12-48 hours of STEMI is definitely indicated.

Dr. Goldsmith of our Cardiology Department added his opinion in an email:

This is a gray zone. Apropos of your general thinking that we should act to make a clinical difference and not just on the basis of guidelines (!) the issue of timing here relates to how much salvage is likely. Points in favor of going now: the relatively low trop on presentation (suggests possible opening/closing of an artery) with pain. If the trop had been 20 I probably would not have gone.....pain and ST elev are problematic guides because both can persist for many hours after there is no chance of successful reperfusion (MI pain in the old days required morphine for 24 hours, usually, as I remember from my residency!). So these are tough, and as you say there is no firm guidance. The low trop with a rise suggest a new event or recurrent event to me, rather than a done deal, so I probably would have argued for earlier study knowing it might or might not help but would not likely hurt. Perfect Pathway B case, by the way!

Saturday, February 6, 2016

What is going on here? There is more than immediately meets the eye.

This is a repost, but it was only posted as a video presentation on hqmeded:

A patient presented with chest pain:
What do you see? (Sorry for the obscured V6.  It doesn't show anything)

There is an obvious inferior STEMI.  The cath lab was activated.  I saw the ECG later and called the interventionalist to hear the outcome.  He said there was a mid-RCA occlusion.  I said, "That is strange, because the ECG has ST elevation in V1 and indicates a right ventricular MI."

He went back to look at the angiogram.  He then saw that the culprit was in the proximal RCA, proximal to the RV marginal branch, and that there had, indeed, been an RV MI.  The mid-RCA occlusion was an embolus from the proximal RCA.  Fortunately, the patient was still on the table and he could go back and place the stent at the correct lesion in the ulcerated plaque in the proximal RCA.

Learning Points:

1.  In inferior STEMI, ST elevation in V1 is a very specific finding for right ventricular MI.  We have studied this and are near completion of a manuscript on it.

2. In inferior STEMI, it pays to record a right sided ECG.

Friday, February 5, 2016

Apparent Pseudo-STEMI patterns are not necessarily Pseudo. Beware.

This was contributed by an ECG enthusiast who wishes to remain anonymous.

LVH is a well-described “mimic” of STEMI. However, a diagnosis of LVH does not exclude an acute coronary occlusion, and the clinical context, including symptoms and old ECGs, must be taken into account.

A 50 year-old woman came to the ED with recent-onset chest pain.:
She had a history of hypertension, as well as concentric LVH on a very recent echo. Furthermore, she had markedly elevated systolic BP > 200 mm Hg.

Her initial ECG:
There is STE in lead III, < 1 mm, as well as STD with inverted T waves in leads I and aVL. This pattern of STE and STD could represent an early or subtle inferior coronary occlusion. However, the R wave in aVL is ≥ 11 mm, basically diagnostic for LVH.  

Smith comment: the voltage only barely meets LVH criteria, but the morphology of the T-wave inversion in aVL has the "hockey stick" shape often seen in LVH.  Furthermore, the inferior T-waves are not as large as one would suspect with a coronary occlusion (see example at bottom).  I would agree with the contributor that it does not look like the typical ST-T of subtle inferior MI.  But, remember, coronary occlusion can show almost nothing on the ECG.  If it can show nothing, then very subtle abnormalities could be something.

The ECG raised concerns for inferior MI, but it was decided that the STE and STD were due to repolarization changes of LVH. Nonetheless, a repeat ECG was obtained 1 hour later:
Not significantly evolved from the prior ECG.

These ECGs were compared with an ECG from 1 week prior:
Although there is T wave flattening in leads I and aVL, there is no suggestion of LVH on this ECG. In fact, the ST segments in I and aVL are slightly elevated, and those in the inferior leads II and aVF subtly depressed.

Smith comment: The change from one month ago is diagnostic of ischemia.  

I very much doubt that LVH on the ECG, with ST-T repolarization abnormalities, can develop over one month.  Serial ECGs are very helpful, but frequently show no or minimal evolution over one hour.  One need only see this recent case I posted for another example of absence of evolution.

Contributor continued:
Given the previously documented LVH on echo, as well as the lack of evolution over 1 hour, the new ECG changes were attributed to a combination of acute hypertension, LVH, and different patient and limb-lead positioning. The patient received aspirin, morphine, and nitroglycerin. The first troponin was below the 99% cut off, and she was admitted to telemetry for serial troponins.

Smith comment: I think it would have been appropriate to dial up the Nitro until the BP is normal, then repeat the ECG.  If the ST segment abnormalities remain, then either angiography or emergency formal echo is indicated.  If they resolve, then perhaps the findings are due to the severe hypertension.  That is exactly what happened in this really interesting case.

Contributer continued
Wait, why did she have an ECG from a week ago?
She had been admitted earlier in the month for an NSTEMI without ECG changes. On angiography, a drug-eluting stent was used to open a total RCA occlusion. (The ECG above was post-PCI.)

With this knowledge that she had received PCI to the coronary region of concern on the ECG (the RCA typically supplies the inferior wall) only weeks ago, the plan was made to manage her ongoing pain medically, and she was admitted to a telemetry bed.

ECGs were repeated at 5 hours post-arrival: 
Although the R wave in aVL is still diagnostic for LVH, the STD/TWI in leads I and aVL have resolved, as has the STE in III and aVF. This proves that they were due to ACS, not LVH.

And 6 hours post-arrival:
The R wave in aVL has regressed to her baseline height, while T wave inversions in III and aVF suggest spontaneous reperfusion of the inferior wall. 

The 8-hour troponin returned at 6.7 ng/mL. (Smith comment: we don't know the peak, and it is not a reliable indicator of infarct size as it depends on reperfusion, which makes it rise high and quickly, infarct size, collateral circulation, and assay.  However, this certainly is consistent with an acute coronary occlusion, although her ECGs did not meet criteria for STEMI).  She went to angiogram the next day and the RCA was indeed (persistently) occluded.  The RCA territory, however, was now well supplied by left to right collaterals, so that the damage was not extensive.  No PCI was done except for on two unrelated diagonals (off the LAD).

Why did she have an acute occlusion after getting stented?
Unfortunately, after her PCI earlier in the month, she had been discharged home without a Plavix prescription. In combination with aspirin, this medication is essential for preventing in-stent thrombosis. Her echo, which during her earlier admission had shown only concentric LVH, now had an akinetic basal inferior segment. She was presumed to have had an early in-stent thrombosis of the RCA DES due to insufficient platelet inhibition.

Brief review: LVH and inferior STEMI
LVH is well appreciated as producing STE in leads V1-V3, mimicking anterior STEMI. However, no proven criteria permit a clear distinction between acute anterior ischemia and LVH.

This situation is even less clear for inferior STE when LVH criteria are met. A recent review of LVH and STEMI highlights that “there is confusion whether the exclusion of STE thresholds in patients with LVH is limited to leads V2-V3 or applies to all leads.”

On the one hand...

On the one hand, LVH frequently manifests ST changes that mimic inferior STEMI. 
For example:
Example 2 from Life in the Fast Lane:

Case 4 from this post on Dr. Smith's ECG blog:

Case 6 from this post on Dr. Smith's ECG Blog:

On the other hand...
On the other hand, these clear examples of LVH easily met precordial-lead criteria for LVH, as well as limb-lead criteria. Our patient, however, only met limb-lead criteria, and just barely.

Furthermore, it is plausible that the change in axis was because the limb-lead electrodes had been placed differently (e.g. on chest instead of arms) than the prior ECG, or the patient positioned differently (e.g. seated versus supine). However, such changes would not produce the ST changes we see, especially over so short a period of time.

Dr Smith has a great case that also demonstrates the peril of attributing new inferior ST changes to LVH. A patient with chest pain had this ECG:
It meets criteria for LVH based on aVL, but the ST-T are far out of proportion to the QRS amplitude.  This was diagnosed as MI by the emergency physician, but the interventionalist would not take the patient to the cath lab because he was convinced it was due to LVH.
Furthermore, there was a previous ECG for comparison:
Again, the previous ECG makes the diagnosis definite

Recommended Resources