Friday, May 31, 2013

Woman in her 60's with 1 hour of chest pain

A woman in her 60's with h/o DM and HTN complains of 1 hour of sudden onset chest pain.  Here is her prehospital ECG:

Sinus rhythm.  Computerized QTc is 422.  What do you think?

There is ST segment elevation (STE) at the J-point, 1.0 mm in lead V2 and 1.5 mm in lead V3, relative to the PR segment (PQ junction).  This does not meet the most commonly used "criteria" for diagnosis of anterior ST elevation MI (see below).  So is this normal variant ST elevation?  Or is it abnormal?  Is it early repolarization?  Could it be anterior STEMI?

Remember that normal variant ST elevation is less common as age increases.

The standard "criteria" are (1): At least 2 Consecutive Leads With ST elevation of:
V1, V4-V6: 1 mm  
V2, V3:   for men over 40 yo: 2 mm
                for men under 40 yo: 2.5 mm
                for women, any age: 1.5 mm 

"The threshold value for abnormal J-point elevation in V2 and V3 recommended in that part is 0.2 mV for men 40 years of age and older and 0.25 mV for men less than 40 years of age. The recommended threshold value for adult women in V2 and V3 is 0.15 mV. The threshold recommended for abnormal J-point elevation for men and women in all other standard leads is 0.1 mV. These threshold values appear to be an appropriate compromise for practical clinical use in the evaluation of ST elevation."

We found these criteria to be very insensitive and nonspecific in differentiating subtle anterior STEMI from early repol.  However, if we use the formula for differentiating early repolarization from subtle LAD occlusion (see sidebar of this blog for excel applet), and use these numbers: 
1. STE at 60 ms after the J-point in lead V3 = 2.0 mm, 
2. QTc = 422 ms, and 
3. R-wave amplitude in lead V4 = 4mm, 
 --we get a value of  25.99 (greater than 23.4 correlates highly with LAD occlusion).  It is the relatively long QT interval and the low R-wave amplitude which make this likely to be STEMI.

Case continued:

The medics (and computer) interpreted the ECG as normal and brought the patient to the ED without alerting me.  By protocol, an ED ECG was recorded and this ECG was brought to me:
Now there is slightly more ST elevation, the QTc is slightly longer at 427ms, and lateral precordial T-waves are larger.
This is the first ECG that I saw and it is clearly diagnostic of LAD occlusion, with anterolateral STEMI.  I immediately activated the cath lab.  The formula value for this is now 27.2 (STE60V3=2.5, QTc=427, RAV4=3)

While waiting for the cath lab, we did a bedside echo and the anterior wall motion abnormality was very subtle and would not have helped with the diagnosis.

She went to the cath lab and had an ostial occlusion of a type III (wraparound) LAD to the inferior wall.  There was a large ramus intermedius (artery between the LAD and circ which functions like a large first diagnonal) which supplied the lateral wall.

Subsequent formal echo with Definity contrast showed an akinetic distal septum anterior and apex as well as akinetic distal inferior wall, so there was a dense MI.  The peak troponin was 257 ng/mL, so this was a large anterior MI.

The inferior wall involvement does not show on the ECG and I find this difficult to explain.

Here is the ECG 2 days later:
Interestingly, the T-wave inversion is consistent with only anterolateral (not antero-infero-lateral) MI.  And T inversions in I and aVL are present in spite of absence of clear evidence of injury on the previous ECGs.


1. Wagner GS, Macfarlane P, Wellens H, et al. AHA/ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram: part VI: acute ischemia/infarction: a scientific statement from the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; the American College of Cardiology Foundation; and the Heart Rhythm Society. Endorsed by the International Society for Computerized  Electrocardiology. J Am Coll Cardiol. 2009;53:1003-1011.

Tuesday, May 28, 2013

Brugada pattern induced by tricyclic antidepressant

A 31 year old male presented obtunded after a suspected intentional overdose with amitryptiline.  He was intubated, and this ECG was recorded:
Time zero: QRS duration = 162 ms, QRS axis = 147.  There is a wide QRS and large R-wave in lead aVR, typical of amitryptiline toxicity.  But the ECG also looks like classic type I Brugada pattern, in V1 and V2 but especially in V2: there is an rSR' with downsloping ST elevation.

After immediate intravenous bicarbonate, this ECG was recorded 14 minutes later:
Time 14 minutes: after 100 ml of 8.4% bicarbonate: QRS = 146 ms.  The QRS has narrowed slightly.  The Brugada pattern is still present

At 2 hours, after more bicarbonate, this was recorded:
Time = 2 hours, after more bicarbonate: QRS = 117 ms.  There is still a hint of Brugada pattern in V1.

Day 4 this was recorded:
Again, there is a hint of Brugada in V1, and even moreso in V2, with rSR' and downsloping ST elevation.

Day 5 this was recorded:
Still a hint of Brugada in V1

One month later:
When the patient awoke, he and his family reported a family history of close relatives with cardiac arrest of uncertain etiology.  The patient himself had never had syncope or dysrhythmias.  Without Brugada pattern on the baseline ECG, there is no Brugada syndrome.  However, there is an uncertain history of sudden death as well as inducible Brugada pattern.

There is no clear management solution for such patients.  The electrophysiologist was of the strong opinion that an IVCD was not indicated without further information.  The management of inducible type I Brugada pattern is not at all clear.

The electrophysiologist felt that if the pattern was also inducible with procainamide, then genetic testing would be indicated.  If not, then avoidance of the TCA would be enough.  He proceeded also with an implantable loop recorder to detect subclinical dysrhythmias, and this was negative.

Drug-Induced Brugada Syndrome

Suffice it to say that this is a very complex topic and I know little more about it than can be read in this 2009 article from Europace, the most recent article I could find on the topic.  It is free full text.

Here are a very few salient points from the article:

Brugada syndrome is due to a genetic defect in the gene (SCN5A) that codes for tha alpha unit of the sodium channel.  There appear to be many kinds of defects with variable manifestations and penetrance, and variable susceptibility to drugs, including TCAs (or also to other states such as fever).

TCAs even at therapeutic doses have been reported to unmask latent Brugada, and such reported cases have polymorphism of the gene. 

Thursday, May 23, 2013

24 yo woman with chest pain: Is this STEMI? Pericarditis?

A 24 yo with no past medical history and no risk factors except for tobacco smoking presented with chest pain.  The pain started in her left chest at 6:20 AM and radiated up through her arms and into her back. She admitted to drinking heavily the previous night and returned home about 1:30 or 2 AM.  She had multiple episodes of vomiting overnight. She's never had chest pain like this before.  BP was 140/100 and pulse of 93.  There was no rub on exam.

Here is her prehospital ECG at 0720:

Sinus rhythm.  Inferior and lateral ST elevation with reciprocal ST depression in aVL.  There is no significant PR segment depression

She was given 2 sublingual NTG and her pain improved from 10/10 to 7/10, and continued to improve.  Here is her first ED ECG at 0748:
Not much change except for less STE in lateral leads.

Chest X-ray confirmed absence of Boerhaave's syndrome.

Some might suspect pericarditis in a young person with diffuse ST elevation.  However, you diagnose pericarditis at your peril!  I believe pericarditis is over diagnosed, even in the literature, and that many cases assumed to be pericarditis in the past would not be proved to be Acute MI.  This is conjecture based on many cases that I have seen, not based on peer-reviewed evidence.

Furthermore, in our study of benign inferior ST elevation vs. inferior STEMI, ST depression in aVL was nearly perfect in diagnosing MI.

I activated the cath lab immediately. A bedside echo performed by a world expert, Dr. Asinger, showed no effusion, and he could not discern a wall motion abnormality.  The patients pain was almost gone by this time.

While waiting for the team, we recorded another ECG at 0826:

Again, not much change.

And a right sided ECG at 0827:
This appears to show right sided ST elevation, though perhaps not a full mm.  This strongly suggests an RV infarct. 

At cath, the culprit was a proximal LAD lesion (open, with TIMI-3 fllow)!  It had embolized to the distal LAD, which was a "type III" or "wraparound" LAD supplying the inferor wall.  So this was an antero-infero-lateral MI.  The proximal lesion was stented and the distal was treated with antiplatelet and antithrombotic therapy.

The next day she has reperfusion T-waves in the anterior leads, as well as inferior and  lateral leads:

Looks like Wellens' syndrome in anterior leads, because it is analogous.

Formal Echo later showed moderate hypokinesis of the septum and dense hypokinesis of the apex and inferior wall.  Peak troponin I was 24 ng/ml.

We always look for explanations when young people have MI.  "They must have done cocaine."  "They must have high cholesterol."  But sometimes coronary atherosclerosis is accelerated, or there is one focal lesion, without good reason.  Her lipids were: LDL 139, VLDL 10, HDL 70, total cholesterol 219.


1) Myocardial Infarction does occur in young women!  They need not have a lot of risk factors or take cocaine.  In young women with STEMI, the diagnosis is more often missed or delayed than in any other demographic group, probably because of the bias that leads us to say, as my chief of EM always expresses it: "Nah, couldn't be."  MI is much more common in young people, including women, than previously recognized.  Many MI in very young women may be due to spontaneous coronary dissection, but this is not the most common etiology.  If the ECG is diagnostic, as in this case, believe it!  Do not try to convince yourself that it "couldn't be."

There was an important article published in fall of 2012 showing that young women with STEMI have longer Door to Balloon times and higher mortality.  There are a couple smaller studies with the same findings here and here.

2) You diagnose pericarditis at your peril

3) Absence of Wall motion abnormality may be misleading, especially if the pain and/or ECG abnormalities have resolved.  In this case, the difference between the informal and formal echo was the use of Definity contrast.

Friday, May 17, 2013

PVC or Aberrant Conduction? (Another Guest Post from Dr. Wang)!

Question: This tracing shows (choose one from the list below)

               a) atrial bigeminy with aberrant conduction

               b) ventricular bigeminy

 Annotated ECG:

Answer:   a) atrial bigeminy with aberrant conduction.

Discussion: Every other QRS is wide, different looking from the sinus beats and, at a quick glance, the tracing seems to show ventricular bigeminy. However, careful observation reveals a prematurely occuring P wave (↓) in front of these wide QRSs and this is an examle of atrial bigeminy with aberrant conduction. Why does aberrant conduction ever occur? It is simply because the two bundle branches (BBs) have different length of refractory periods. If an impulse occurs when one BB is still refractory while the other has recovered from the refractory period, the impulse will conduct thru the recovered BB, bypasssing the refractory BB: aberrant conduction results. Of course, if an impulse occurs at a time when both BBs have recovered from the refractory period, it will conduct normally, while if both BBs are refractory, it won't conduct at all (non-conducted PAC). This was well diagrammed and explained in the Medscape ECG of the week posted on 3/7/2012.

Monday, May 13, 2013

Series of Prehospital ECGs Showing Reperfusion

Medics recorded this ECG at time 0 in a patient with chest pain:

There are inferior hyperacute T-waves, with reciprocally inverted hyperacute T-wave in aVL, and a biphasic (down-up) T-wave in V2 and V3, as well as ST depression in I and V4-V6.

"Down-up" biphasic T-waves should always be thought to be reciprocal of "up-down" (Wellens' type) reperfusion T-waves.  (Alternatively, if you see a down-up T-wave, it may actually be a down T-wave followed by an upright U-wave, if it happens very late.)  Thus, this is probably inferior-posterior STEMI, possibly with some degree of reperfusion.

The chest pain began to resolve after NTG (t = 2 minutes):

The inferior T-waves are slightly smaller, and there is less negativity to the T-waves in V2 and V3

Another was recorded at t = 4 minutes:
Now the inferior T-waves are not at all hyperacute and the precordial T-waves are mostly upright.  Clearly, this is a reperfusing inferoposterior MI.  There are down-up T-waves in V4-V6.

Here is t = 13 minutes, with posterior leads V7-V9:
Now there are inverted (reperfusion) T-waves in the inferior leads, with a reciprocally upright T-wave in aVL. The T-waves in V2 and V3 are upright.  There are mostly inverted T-waves in V7-V9, pretty much reciprocal  to (opposite of) V2 and V3.

Here is t = 14 minutes (simultaneous to previous):
T-waves are upright in V4-V6, reciprocal to the inverted T-waves in V7-V9, but not exactly because these are not 180 degrees opposite each other.    V2 and V3 are another example of posterior reperfusion T-waves.
Angiogram showed an open culprit artery supplying the inferior and posterior walls

Thursday, May 9, 2013

Tuesday, May 7, 2013

Atypical Chest Pain: Suspicious ECG, and a Left Main ACS is found in a 30-something

A male in his 30s presented with chest pain, cough, and sore throat.

He had been seen in clinic the day prior for cough x 1 month and sore throat.  A strep test was positive and he was treated with penicillin, and also with acetaminophen and ibuprofen.

On the day of ED (ER) presentation, he presented with 5 hours of intermittent sudden onset left side chest pressure unrelieved by ibuprofen, with associated vomiting and SOB.  There was some association with moving and palpation, but also some improvement with NTG.  An ECG was obtained immediately:
Sinus rhythm.  There is some inferior and lateral ST elevation that is very subtle, and is associated with a very small amount of ST depression in lead aVL (which we have found to always be associated with inferior MI).  There are large T-waves in II, aVF, and V4-V6.  There is also a tiny amount of terminal T-wave inversion in lead III (see comparison below).  There is a deformed T-wave in V2 (see discussion and graphic below)

 I was highly suspicious of MI, so an ACS workup was initiated and we recorded a bedside echo (unavailable) which showed no inferior wall motion abnormality (or anywhere else).  He was given aspirin and clopidogrel 600 mg.  Because of the absence of wall motion abnormality, the patient's age, and the atypical pain, we decided to serially evaluate him and not activate the cath lab.  A second ECG was recorded at 37 minutes.
This is identical, except the terminal T-wave inversion in III is gone.
Here is a magnification of lead III, for comparison:

Around this time, the initial troponin returned at 2.0 ng/ml.  Because he had persistent chest pain and persistent ECG abnormalities, we activated the cath lab.  He was taken and found to have open coronaries.  The angiographer was concerned about inadequate opacification of the distal left main and proximal LAD, and did intravascular ultrasound.  He confirmed a plaque with fissure in the distal left main coronary artery.  There was no indication for stenting, but aggressive and prolonged antithrombotic and antiplatelet therapy (to "cool down" the fissured plaque) was instituted for this very high risk lesion.

Here is the ECG after the cath:
The STE and reciprocal ST depression are gone, and T-waves are smaller.  The previously deformed T-wave in V2 is normal now.

Some have asked what I mean by deformed T-wave in V2.  Here are V2 and V3 from the first 3 ECGs side-by-side:
Notice in the 1st and 2nd recordings, the initial part of the T-wave is flattened, and the upright part is very small.  This is possibly a result of anterior wall ischemia, or (what I initially thought and the interpretation I favor) a result of subtle posterior STEMI which is pulling the ST segment down (ST segment starting to elevate if recorded from posterior leads).  I don't know that any of this interpretation is true, but is based on observing scores of ECGs with known posterior transmural ischemia and reperfusion.

Troponin I peaked at 8 ng/ml.  A formal echo showed no wall motion abnormality.  An ECG at 24 hours was recorded:

This confirms inferior and lateral infarction, with increased ST elevation and evolution of T-waves (reperfusion T-waves).

At 48 hours:

Further evolution

The troponins trended down from a peak of 8 ng/ml.

The ECG suggests inferolateral MI.  The T-wave in V2 suggests either posterior or anterior ischemia.  The angiogram is consistent with anterior (LAD) and posterolateral MI (left main supplies the LAD and circumflex), and of inferior MI if the the circumflex is dominant (the cath report did not specify  this).

In any case, what would be considered by most to be a "nonspecific" ECG was highly suspicious and prompted rapid evaluation and management in a young man who might otherwise be thought "low risk" with young age and atypical chest pain.  The finding of reciprocal ST depression in lead aVL confirmed the pathologic nature of the inferior ST elevation.

Friday, May 3, 2013

Left Bundle Branch Block with Less Than 1 mm of Concordant ST Elevation (in the Setting of Hypertrophic Cardiomyopathy)

A male in his 60's with a known history of hypertrophic cardiomyopathy (HOCM) presented for chest pain.  He thought  it was "acid reflux," and it started a few hours prior to his presentation to a non-PCI capable hospital.  He denied any associated symptoms.  His pain was not relieved with NTG.  He had an minimally elevated troponin at the referral institution, (assay unknown).  He was diaphoretic and clutching his chest.  Here is the first ECG:

There is sinus rhythm with LBBB and high voltage, and all leads have appropriate and proportional discordance EXCEPT for lead V4, which has less than 1 mm of (inappropriate) concordant ST elevation.  Lead V5 also has significantly less discordance than expected.

Definitions in Left Bundle Branch Block:

Appropriate Discordance: The ST segment (and usually the T-wave) should be in the opposite direction of the majority (not the end) of the QRS in all leads.  It should be proportionate, with the ST/S ratio being < 0.25 or ST/R ratio < 0.30.

Concordance: When the ST segment and/or T-wave is in the same direction as the majority of the QRS. This is abnormal for LBBB and, if 1 mm, a specific sign of STEMI-equivalent.

Modified Sgarbossa Rule:
Any one of:
1.     1 mm of concordant ST elevation in at least one lead.  Usually leads I, aVL, V5 and V6.  May also involve inferior leads.
2.     1 mm of concordant ST depression in at least one of leads V1-V3.
3.     Excessively discordant ST elevation, as measured by ST elevation (at the J-point, relative to the PR segment) to S-wave ratio of > 0.25

What should the ECG look like if there is both HOCM and LBBB?  I don't think anyone knows.

The patient was transferred to a PCI capable hospital where this ECG was recorded:

Now the concordant ST elevation is in lead V5, but since V4 now has a predominantly negative QRS, the difference is a result of lead placement differences between the two ECGs.  V6 has an inappropriately isoelectric ST segment.

An ECG from 5 years previous was available:

This shows inappropriate concordance of less than 1 mm in V4, and inappropriate isoelectric ST segments in V5, so all this configuration is not new for this patient who has known HOCM.

Strictly speaking, none of these meet the first Sgarbossa criterion of 1mm of concordant ST elevation.  In our study of patients with ischemic symptoms and LBBB, of 33 occlusions and 129 without occlusion, we had initially hypothesized that ANY concordant ST elevation (not just 1 mm) would be specific for coronary occlusion.  We did not publish the data (yet), but less than 1 mm of concordant STE occurred too frequently among the control group to be used as a specific criterion for STEMI-equivalent.

Out of an abundance of caution, the cath lab was activated and the coronaries were clean.

The troponin at the referral institution was normal.

The echocardiogram showed:

Normal estimated left ventricular ejection fraction, about 60%.
Left ventricular hypertrophy concentric, marked.
Dynamic left ventricular outflow/subvalvular obstruction (due to HOCM)
Asynchronous interventricular septal motion due to left bundle branch block.

The outcome

The patient developed abdominal pain in addition to chest pain, had a tender abdomen, and was diagnosed with cholecystitis!!!


1. Perhaps HOCM in the setting of LBBB can have concordant ST elevation
2. Concordant ST elevation of less than 1 mm has a specificity less than if it is greater than 1 mm
3. Troponins at or near the 99% cutpoint may give discordant results, especially with different assays.

Wednesday, May 1, 2013

2 hours of lecture from Sunday April 28 ECG Workshop - the Best of Smith at Best of Hennepin

This is most of the best stuff I have on subtle findings of coronary occlusion on the ECG.

It starts with a 25 minute powerpoint lecture on the diagnosis of STEMI-equivalent in  Left Bundle Branch Block.

Then it goes on to a second Powerpoint that starts with Hyperacute T-waves, Goes on to Subtle LAD occlusion and  the formula in depth, then to subtle inferior, lateral, and posterior MI, then to T-wave inversion.

What it does not cover: I did not have time to go into the 5 primary patterns of ischemic ST depression, but that can be found at this post.  I also do not talk about right bundle branch block or pseudoinfarction patterns such as LVH and LV aneurysm.

Here are the lectures. You won't see this anywhere else.  If you value the ECG as a tool to identify patients who need reperfusion therapy immediately, this is essential for you.

I will also put this link on the sidebar.

Another interpretation of the last rhythm strip, by Ken Grauer. In any case, no AV Block!

One very sophisticated reader, Ken Grauer, came up with another interpretation of the last rhythm strip.

Here is the strip:

Here is K. Wang's ladder diagram and interpretation:

Answer: b) NSR and AV junctional acceleration with AV dissociation and occasional capture beats.
Discussion. As diagrammed, P waves from the sinus impulse occur regularly at a rate of 65/m (P1,2,3...etc). The blips pointed by the arrows are not r' of rSr', but sinus P waves, judging from the timing.  If it were an r', then all QRSs would have it.   The primary problem in this patient is AV junctional acceleration. P1 is conducted to R1. P2 was going to be conducted but accelerated junctional beat (R2) occurred sooner than that. P2 and R2 are dissociated because they occur close together during the physiologic refractory periods of each other. R3 is the accelerated junctional beat which failed to conduct to the atria and next sinus P wave is uninterrupted and occurs on time and conducts to the ventricle (R4). And the cycle repeats. 

Thus, there is AV dissociation without any AV block.

The primary problem is junctional acceleration. What is the clinical significance of junctional acceleration? One has to consider digitalis intoxication, myocardial ischemia or infarction, or excess amount of catecholamines circulating which means any stressful condition. Again, a given ECG tracing can be dissected into primary disorder and secondary manifestations and ask what is causing the primary disorder, so that the patient can be treated appropriately, promoting quality patient care.

Here is Ken Grauer's ladder diagram and interpretation:


What bothered me about this being an accelerated junctional rhythm (interrupted each 3rd beat by sinus capture) - was that the R-R interval for the 1st junctional beat in each sequence (= the R-R between beats #3-4; 6-7; 9-10; and 12-13) is slightly different (shorter) than the R-R interval of the subsequent beat (= the R-R between beats #4-5; 7-8; 10-11; and 13-14) - whereas if this was AV dissociation by usurpation with junctional acceleration I would have expected them to be the same ....

Why couldn't the mechanism be the one I drew - whereby there are PJCs (beats #4,7, 10, 13) that conduct retrograde enough to slow down forward conduction of the next sinus impulse. Admittedly - that next sinus impulse is slowed down to a greater extent than usually occurs with such concealed conduction (perhaps due to switching to some alternate slow conduction pathway ... ).

Just wondering if Dr. Wang might consider my alternate mechanism plausible. In any case - GREAT TRACING - and what IS agreed is that there is AV dissociation from some junctional intervention and no evidence of any AV block. 

Here is K. Wang's Respsonse:

An interesting comment and a legitimate argument by observing slight difference in the R-R intervals of what I claim to be junctional beats, an astute observation. Yes, junctional rhythm is very regular rhythm, and I would have been happier if they occurred with exactly the same interval. The slight difference in this case may be due to some degree of penetration of the impulse from P2,4 and 6 into the junction (some degree of concealed conduction). In my view, the argument against your diagram is that that long P2-R3 interval will become too short (P3-R4) too quickly. Also, a premature junctional beat would have occurred more prematurely. His bundle recording would have settled the issue, which this patient did not have. It is an important issue because an accelerated junctional rhythm and junctional trigeminy (which is what you are advocating) have different clinical implications (see the discussion of the case). It is also interesting that we two are applying concealed conduction at different places of the same tracing.

Thank you for showing interest in our program and generating interesting discussion.

K. Wang.

My impression is that Ken's interpretation is very viable, but that the differences in R-R interval are tiny and that K.'s interpretation is more likely.  

As Ken says, the important thing is not which of these is correct, but that in both interpretations, there is no AV block, only AV Dissociation.