Wednesday, December 5, 2018

Two cases of ST Elevation with Terminal T-wave Inversion - do either, neither, or both need reperfusion?

Written by Pendell Meyers with edits by Steve Smith

I was sent these 2 ECGs with no clinical information other than chest pain:

Do either or both of these ECGs show ischemic changes? If so, what should you do and why?

Let's take them one at a time.

What would your response be?

I responded: "Awesome classic benign T wave inversion! That's the patient's baseline normal variant. ... But if it were a good story with exertional syncope or something you'd have to treat it like it could be HOCM, etc. Tell me more."

There is sinus rhythm with very large voltage and associated repolarization abnormalities. In V3-V6 there are classic and dramatic findings of BTWI including small or no S-waves, large R-waves, pronounced J-waves, ST-elevation, and steep dramatic T-wave inversions, all with a short QT interval.

Later, I was told this history:

The patient was a 17 year old African American male with history of asthma who presented with chest tightness and shortness of breath for 6 hours. He improved almost completely after a duoneb (albuterol and ipratropium). However the clinicians were surprised by his unusual ECG findings. 

He had no prior ECG on file. He had no family history of cardiac issues or sudden death. Labs were normal, including troponin x2. He is able to play basketball routinely without any issues.

In the ED he had a CT angio which was negative for aortic dissection. He was admitted for observation and cardiology consult. He had a normal CT coronary angiogram, as well as a normal cardiac echo. Serial troponins were negative.

Smith Comment: Although this ECG pattern is not ischemic, it may sometimes be seen in individuals with hypertrophic cardiomyopathy. For this reason it may be prudent to obtain an echocardiogram in these patients before they are allowed to engage in vigorous athletic activity. [1,2]

Let's return to the other case:

There is sinus rhythm with QS-waves in V1-V3. There is ST elevation in V1-V3 as well, with convex appearance and some terminal T-wave inversion. 

I advised the team that the patient has had a large anterior MI, but I cannot tell how long ago it occurred. It could be 6-12 hours after completion or reperfusion of anterior MI, or it could be days to years after the event. I advised them that I do not see any signs of reocclusion on this ECG, meaning that I do not think this ECG at this time represents acute LAD occlusion that would benefit from emergent reperfusion.

Let's use the LV aneurysm rule on this ECG to see if it supports our theory:

Here is the rule, which should be used if there is suspicion of aneurysm (there are Q-waves in V1-V4, especially QS waves): "If there is one lead of V1-V4 with a T/QRS ratio greater than 0.36, then it is acute MI. If less than 0.36, it is either subacute (over 6 hours) or old." [3,4]

V1 = 2/10 = 0.20
V2 = 2/12 = 0.17
V3 = 1.5/7 = 0.21

So it agrees with our visual subjective interpretation that this is either subacute or old.

I later found out that the patient had history of an LAD occlusion MI two months ago, with a stent placed.

He was taken for emergent cath due to his history and the ECG findings including ST elevation (although ST elevation is a terrible way to distinguish ischemia from non-ischemia). 

His cath showed perfectly patent stents.

Three serial troponins were undetectable.

His serial ECGs did not change.

He was discharged home.

Learning Points:

There are many causes of ST elevation, terminal T-wave inversions, and both simultaneously. Experience with the cases on this blog can teach you how to differentiate them.

Wellens syndrome (or reperfusion in general) is an important cause of terminal T-wave inversions. See these cases below for examples. However, Wellens' syndrome includes resolution of chest pain and preservation of R-waves, in addition to T-wave inversion. Thus, this second case would not be an example of Wellens' syndrome.

Classic Evolution of Wellens' T-waves over 26 hours

Also see below for more cases of LV aneurysm morphology:

Tachycardia and ST Elevation.


T/QRS ratio to differentiate anterior STEMI from anterior LV aneurysm:
1. Papadakis M, Carre F, Kervio G, et al. The prevalence, distribution, and clinical outcomes of electrocardiographic repolarization patterns in male athletes of African/Afro-Caribbean origin. Eur Heart J. 2011;32(18):2304-2313. doi:10.1093/eurheartj/ehr140

2. Wells S, Rowin EJ, Bhatt V, Maron MS, Maron BJ. Association Between Race and Clinical Profile of Patients Referred for Hypertrophic Cardiomyopathy. Circulation. 2018;137(18):1973-1975.

Comment by KEN GRAUER, MD (12/5/2018):
It’s always great practice to be given tracings and “Put to the Test” as to whether you think acute changes are, or are not present. I came to the identical conclusion as did Drs. Smith & Meyers about the 2 ECGs in this case — albeit via a slightly different intuitive pathway. For clarity — I repeat the tracings in Figure-1.
Figure-1: Two ECGs we are asked to assess, but without benefit of any history (See text).
ECG #1:
  • Knowing some History is KEY. As per Dr. Meyers — If the history of this patient was worrisome, then the onus of “proof” still rests upon us to prove there are no acute ECG changes, rather than the other way around. REMEMBER — Patients with early repolarization on their baseline ECG may develop superimposed changes of acute STEMI, that can at times be difficult to discern because of pronounced baseline benign ST-T wave repolarization changes.
  • That said, in addition to the benign-appearing findings described in detail by Drs. Meyers & Smith — I thought these findings in ECG #1 were probably not acute because: iThe Q waves in the inferolateral leads are very small (these look like normal septal q waves in a patient like this with an inferior frontal plane axis); iiThere is no reciprocal ST depression; iiiExcept for lead V3 — the leads with the deepest T inversion manifest the tallest R waves; andivLead V3 is perfectly consistent with being a transition lead” (ie, V3 looks precisely as I’d expect for a lead located in between the upright T wave and predominantly negative QRS seen in V2 — and — the huge R wave with deep T inversion in V4).
  • The ECG appearance of patients with hypertrophic cardiomyopathy (HCMis highly variable. Among potential ECG findings in such patients are LBBB, RBBB, IVCD, large Q waves in multiple leads (especially in lateral and/or inferior leads); prominent septal forces (ie, relatively tall R waves in anterior leads) — and/or marked increase in QRS amplitude in various leads with notable ST-T wave abnormalities (as are seen here in ECG #1). And sometimes, the ECG of patients with hypertrophic cardiomyopathy is surprisingly unremarkable. TAKE-HOME Point: The prominent QRS voltage and marked ST-T wave abnormalities in ECG #1 should clearly prompt consideration of possible HCM. An Echo is indicated if there is any concern.
ECG #2:
  • Knowing some History is KEY. As per Dr. Meyers — the correct clinical interpretation of ECG #2 is, “Anterior MI of uncertain age, possibly recent” — with need for clinical correlation.
  • Reasons we KNOW there has been prior anterior MI in ECG #2 include the following — iThere is a fragmented initial Q wave deflection in lead V3 (and when you see this initial fragmented qrS deflection — it is highly suggestive of true prior infarction); iiThere is loss of r wave (a tiny r looks to be present in V1 — but it is wiped out by the q wave in V3)iii) The q wave in V4 shouldn’t be there (because this q in V4 is larger than the tiny q in V5 — and it looks like there is no q in V6); and, ivST coving and T inversion is most prominent in the anterior leads that manifest abnormal initial (q wave) deflections.
  • Reasons for suspecting that even if ECG changes in ECG #2 were to be recent — they probably do not reflect acute coronary occlusion, include the following — iIt usually takes a little time to develop the fragmentation and the amount of lost anterior forces as is seen in leads V1, V2 and V3 of ECG #2; and, iiThe amount of ST elevation in the anterior leads is modest compared to significant loss of anterior forces and the extent of T wave inversion. NOTE: Exceptions always exist, and it is possible to develop the ECG picture seen in ECG #2 over a fairly short period of time … — BUT  “the look” of this tracing is such that I thought the changes in ECG #2 were probably not acute.

Friday, November 30, 2018

The 4 Physiologic Etiologies of Shock, and the 3 Etiologies of Cardiogenic Shock

A 60-something presented with hypotension, bradycardia, chest pain and back pain.

She had a h/o aortic aneurysm, aortic insufficiency, peripheral vascular disease, and hypertension.  She had a mechanical aortic valve.  She was on anti-hypertensives including atenolol, and on coumadin, with an INR of 2.3. 

She was ill appearing.  BP was 70/49, pulse 60.

A bedside echo showed good ejection fraction and normal right ventricle and no pericardial fluid.

Here is the initial ECG:
What do you think?

This ECG actually looks like a left main occlusion (which rarely presents to the ED alive):  ST Elevation in aVR, but also in V1, and what appears to be "coving" of the ST segment in aVL, which suggests ST Elevation in that lead as well.

There is bradycardia, which is ominous in such a sick patient.  This may be due to atenolol, but could simply be a sign of severe illness.

Whereas ST elevation in aVR is usually reciprocal to the the ST depression of subendocardial ischemia, with a negative ST vector towards leads II and V5, and may be accompanied by STE in V1 (which is in the same direction as aVR), when there is also STE in aVL it implies a more directly superior ST axis (supported by ST depression in all of II, III, aVF).

When the ST axis is directly superior, there may actually be transmural ischemia of the "base" of the heart (the base is actually the top of the heart, which really does not have a wall -- the ventricles have openings to the atria at the base, and so no complete wall).  However, the anterior, lateral, posterior, and septal walls all have a superior portion that may result in ST elevation in a superior direction if all walls have subepicardial ischemia, as in left main occlusion.

It is important to remember that an ECG like this represents possible left main occlusion only if it is a result of ACS!! 

Knotts et al. found that such ECG findings only represented left main ACS in 14% of such ECGs: 
Only 23% of patients with the aVR STE pattern had any LM disease (fewer if defined as ≥ 50% stenosis). Only 28% of patients had ACS of any vessel, and, of those patients, the LM was the culprit in just 49% (14% of all cases).  It was a baseline finding in 62% of patients, usually due to LVH. 

Reference: Knotts RJ, Wilson JM, Kim E, Huang HD, Birnbaum Y. Diffuse ST depression with ST elevation in aVR: Is this pattern specific for global ischemia due to left main coronary artery disease? J Electrocardiol 2013;46:240-8.    

Case continued:

The chest X-ray was consistent with pulmonary edema, and the ultrasound showed a "plump" inferior vena cava.  (This data is all consistent with high right and left sided filling pressures, which is the hallmark of cardiogenic shock -- not obstructive, distributive, or hypovolemic)

Aortic dissection was considered (with involvement of the coronary arteries and aortic outflow obstruction), but a CT scan was negative for dissection.

The cath lab was activated.

It is a very long story, but I will make it brief:

The angiogram could not be done because of inability to access the arterial system (unexplained).

The interventionalist listened to the heart and heard both a systolic flow murmur across the valve and a diastolic murmur.

A formal echo showed aortic insufficiency, then TEE showed a mass on the aortic valve, and surgical exploration showed aortic valve thrombosis.  

This was not ACS.  This was not Left Main occlusion.


Shock has 4 basic etiologies:

1. Obstructive (pulm embolism, tension pneumothorax, tamponade, atrial myxoma, etc.)
These were all ruled out
2. Distributive (sepsis, neurogenic, beriberi, etc.).  The patient showed no signs of high output
3. Hypovolemic (dehydration, hemorrhage, third spacing)  The IVC was plump

4. Cardiogenic -- the likely culprit

Cardiogenic shock has 3 basic physiologic etiologies:

1. Poor pump function (diastolic or systolic)  The bedside echo showed good pump function, though that does not completely rule out diastolic dysfunction.
2. Dysrhythmia (too fast or too slow to adequately fill and pump)  There was only some mild bradycardia
3. Valvular.  The pump works but the blood goes the wrong way.  This is what we are left with: Valvular etiology is frequently forgotten.  Any valve dysfunction can result in shock, especially aortic, mitral, or tricuspid (I have never seen a case of acute pulmonic valve dysfunction and cursory search found nothing)

Aortic valve thrombosis should be considered in any patient who has a prosthetic aortic valve, especially a mechanical one.  It is a catastrophic complication, which is why we are so careful to maintain a high therapeutic INR on these patients.

One could say that this case is a hybrid of cardiogenic and obstructive shock, but since it is all valvular, let's call it cardiogenic.

Suffice it to say that one should suspect valvular thrombosis in patients with artificial valves who have:

1. Dyspnea
2. New or worsening heart failure
3. Hypotension or shock

One should listen for muffled heart sounds or a new murmur, but this is not sensitive.

One should be suspicious especially if the INR is not at the goal of:

2.5  - for bi-leaflet or current-generation single tilting disk mechanical aortic valves is 2.5 if they have no risk factors for thromboembolism 

3.0  - if they have a ball-cage mechanical valve (because of the associated higher risk of thrombosis) or risk factors for thromboembolism such as atrial fibrillation, left ventricular systolic dysfunction, prior thromboembolism or hypercoagulable state.

3.0  - (range, 2.5 to 3.5) for patients with mechanical mitral valves 
3.5 to 4.0 -  for patients with mechanical tricuspid valves. 
Transesophageal echo may be necessary for diagnosis.

Always think about doing Doppler on the valves during POCUS.  It is not that difficult to see regurgitation.  Had that been seen, the diagnosis would have been made much earlier.
Complications of aortic thrombosis include, of course, severe obstruction of aortic outflow (with shock), aortic insufficiency and regurgitation, and embolism to any artery in the tree, resulting in stroke, myocardial infarction, or other catastrophic infarction.
The combination of outflow obstruction (decreased cardiac output) and regurgitation (low diastolic pressure) results in severe coronary ischemia.  The coronaries are perfused during diastole when the myocardium is not contracting.
There is no reason to write more about this complication, as there are many good articles on the topic.  See this one:

Comment by KEN GRAUER, MD (11/30/2018):
Interesting case with an excellent discussion and timely reminder by Dr. Smith of the principal etiologies of shock and cardiogenic shock. I limit my comments to the initial ECG, which I’ve reproduced for clarity in Figure-1.
Figure-1: Initial ECG on a woman in her 60s, presenting with shock (See text). 
My interpretation of the ECG in Figure-1 was as follows:
  • Baseline artifact. This is most marked in leads I and III, suggesting the cause might be the right upper extremity.
  • Sinus bradycardia at a rate just over 50/minute. All intervals (PR, QRS, QT) are normal. Slight leftward axis, though not quite leftward enough to qualify for LAHB (since the QRS complex is still predominantly positive in lead II).
  • Probable LVH (See below).
  • Diffuse ST-T wave depression, with especially deep T wave inversion in leads V4-thru-6 — and biphasic T waves in leads III, aVF and V3. Significant ST elevation in leads aVR and V1, with fatter-than-they-should-be T wave peaks in these leads, and a pattern for the ST-T waves in leads aVR and V1 that looks “reciprocal” to the ST-T wave depression in many of the other leads.
IMPRESSION: Sinus bradycardia with probable LVH + ST-T wave changes suggestive of LV “strain” + diffuse subendocardial ischemia. Urge clinical correlation.
  • Why probable” LVHThe ECG is an imperfect tool for assessment of LVH. At best — sensitivity of the ECG for picking up LVH is no more than 55%. Echo is far superior to the ECG for assessment of chamber enlargement. That said, when the clinical situation is “right” — and, voltage criteria for LVH are met — and, ST-T wave repolarization changes of LV “strain” are present — then specificity of the ECG for detecting LVH may be more than 90-95%! Admittedly, voltage criteria for LVH are not met in Figure-1 (ie, the R wave in lead V6 falls shy of the 18-20 mm needed in this lead). But, given the high prevalence clinical situation (ie, this patient has known hypertension plus aortic insufficiency) + nearly enough R wave amplitude in lead V6 (16mm) + ST-T wave changes consistent with LV “strain” in all lateral leads (ie, leads I, aVL, V4-6) — true chamber LVH is likely (CLICK HERE  for more on the ECG diagnosis of LVH).
  • It would be extremely interesting to see a baseline ECG on this patient, taken at a time when she is not in acute cardiogenic shock! My GUESS as to what such a “baseline” ECG on this patient would show is: iST-T wave changes of LV “strain” in the same 5 lateral leads with similar shape for the ST-T waves, but with less J-point depression than the 2-4mm seen in leads V4-thru-V6 in Figure-1; iimuch less (if any) ST elevation in leads aVR and V1; iiipossibly an upright T wave in lead V2, with a shape reciprocal to the ST-T wave depression seen in the lateral chest leads in Figure-1 (ie, LV “strain” often results in mirror-image upright ST-T waves in anterior leads, compared to the ST-T depression seen in lateral chest leads); ivmuch less (if any) ST depression in leads II, III, aVF and V3 (with loss of biphasic T waves); and, v) possibly greater R wave amplitude in lateral chest leads (sometimes ischemia reduces R wave amplitude ... ).
  • Without the luxury of seeing this patient’s “baseline ECG” — we are reduced to educated guessing about which ECG findings in Figure-1 are “new” vs “old” vs “new on-top-of old”. My hunch is that what we are seeing in Figure-1, is a combination of longstanding LVH with LV “strain” superimposed diffuse subendocardial ischemia in this patient with acute cardiogenic shock from complications of her valvular heart disease. Cardiac catheterization would have told us if significant underlying coronary disease was also contributing to the subendocardial ischemia.
Our THANKS to Dr. Smith for presentation of this interesting case.

Wednesday, November 28, 2018

Saddleback ST Elevation. Is it STEMI? Is it type II Brugada?

A 50-something presented with epigastric and chest pain.

Here is his ECG:
What do you think?  QTc 388 ms.

Computer interpretation:

***ACUTE MI***

There is a saddleback, which is rarely due to MI.  V2 has the morphology of type II Brugada, as there is a relatively large beta angle, described here:

1. Draw a horizontal line from top of r' wave (black line 1)
2. Draw a horizontal line 5 mm below this (green line 2)
3. Extend the downsloping r'-ST segment (black line 3) until it intersects the green line
4. Measure the base.  

If greater than 3.5 mm, then meets criteria (this is equivalent to a 35 degree beta angle)

However, whenever you see an rSR', especially with a saddleback, think of lead placement.

Then look at the P-wave in V2.  Is it fully upright?  If not, then there is probable high placement of lead V2.

I went back to look and, indeed, V1 and V2 were placed too high.

I put them in the correct position and we recorded another ECG:
Now the P-wave in V2 is upright.  The lead is placed correctly.
Looks like typical normal variant ST elevation (otherwise known as early repolarization)
QTc 384

You can use the LAD-Early Repol Formula to differentiate this from LAD occlusion:

ST elevation at 60 ms after the J-point, relative to PQ junction (STE60V3), = 2.5mm
QTc by computer = 384
R-wave amplitude in V4 (RAV4) = 19mm
Total QRS amplitude in V2 (QRSV2) = 17.5mm

Any of these calculators work:
Use the iPhone app: SubtleSTEMI
Use the Android app: ECG SMITH
Use the link at the top of the blog.

4-variable formula: 15.1 (A value less than 18.2 favors early repol.  This value is very low.)

The patient was diagnosed with reflux

Learning Points:

1. Saddleback ST Elevation is almost never STEMI
2. Saddleback STE may be type II Brugada syndrome
3. A Type II mimic may result from leads V1 and V2 placed too high
4.  An inverted P-wave in lead V2 implies lead misplacement too high

Saddleback in STEMI:

Here are the only 2 ECGs with V2 "saddleback" that I have ever seen which really represented an LAD Occlusion:

Anatomy of a Missed LAD Occlusion (classified as a NonSTEMI)

A Very Subtle LAD Occlusion....T-wave in V1??

Here are other cases of saddleback STE:

Is this Saddleback a STEMI??

Comment by KEN GRAUER, MD (11/28/2018):
Great case for illustrating a number of important concepts — albeit it raises a series of questions without definite answer. For clarity — I put both tracings together in Figure-1. For the purpose of fostering discussion — I’ll play “Devil’s Advocate” with my comments below:
Figure-1: TOP ( = ECG #1) — Initial ECG in this case. BOTTOM ( = ECG #2) — Repeat ECG after verifying correct lead placement. The lead V1 and V2 electrodes had been placed too high on the chest in ECG #1 (See text). 
ECG #1: ( = the Initial Tracing):
The initial concern at the time ECG #1 was obtained was whether the ST elevation in leads V1 and V2 was indicative of acute OMI. As it turned out — leads V1 and V2 were placed too high on the chest, so that ECG #1 was not a valid tracing for such assessment. That said — I think it worthwhile to specify reasons why even if lead placement had been correct, acute OMI would still be unlikely on the basis of this initial ECG:
  • The shape of the ST elevation in leads V1 and V2 in ECG #1 is concave up, which is usually a benign morphology. As per Dr. Smith — the “saddleback” configuration (seen best here in lead V2) is rarely due to acute MI.
  • Apart from leads V1 and V2 — there really is no abnormal ST segment deviation. The slight ST elevation seen in lead V3 does not exceed the 1-2 mm of concave-up ST elevation commonly seen in leads V2 and/or V3 with normal repolarization variants. No other lead shows any ST elevation. There is no reciprocal ST depression. So, even if all leads had been correctly placed in ECG #1 — I would not have suspected acute OMI.
Assessing for a Brugada-2 ECG Pattern:
saddleback” (ie, Brugada-2ECG pattern is assessed for by applying the criteria in the Figure explained by Dr. Smith above. I illustrate application of these measurements in the magnified insert of the middle complex in lead V2.
  • In ECG #1 — the base of the triangle (horizontal PINK line) formed by drawing lines from the peak of the r’ in lead V2 looks to be extremely close to 3.5 mm, which is right at the upper limit qualifying for a Brugada-2 pattern.
  • Criteria for a Brugada-2 ECG pattern disappeared once the V1 and V2 electrodes were correctly positioned (See ECG #2 in Figure-1). While disappearance of the saddleback pattern in lead V2 after repositioning of the V1, V2 electrodes is comforting — it should be appreciated that in some patients with Brugada patterns, the ECG abnormality may only be seen with higher-than-usual electrode lead positioning (ie, in the 2nd or 3rd interspaces). This is because the abnormal electrical activity leading to a Brugada ECG pattern arises from a very limited zone located in the RVOT — and in some patients this electrical zone requires higher lead placement for detection. This raises the difficult-to-answer clinical question of when to do additional ECGs on a patient with placement of leads V1,V2 in the 2nd and/or 3rd interspaces?
  • When associated with appropriate clinical features (ie, personal history of cardiac arrest, polymorphic VT, non-vagal syncope, positive family history of sudden death at an early age, etc.) — then spontaneous occurrence of a Brugada-1 ECG pattern is diagnostic of Brugada Syndrome. On the other hand — isolated occurrence of a Brugada-2 pattern is non-diagnostic, although when associated with other features, it may raise clinical suspicion and warrant pharmacologic challenge and/or referral to EP cardiology.
For More on Brugada ECG Patterns:
  • Dr. Smith’s March 26, 2015 post has great illustrations reviewing assessment for a Brugada-2 ECG pattern.
  • My 29-minute ECG Video on the topic reviews the basics of recognizing Brugada-1 and Brugada-2 patterns, with detailed discussion of Brugada Syndrome.
Regarding Malposition of Leads V1 & V2:
Dr. Smith’s suspicion of lead malposition in this case was invaluable for accurate diagnosis. He then confirmed his suspicion by checking the patient himself to see where the V1, V2 electrodes had been placed — and followed this up by repeat ECG that revealed the patient’s “true” ECG.
  • We’ve previously reviewed clues to quickly Recognize V1, V2 Misplacement (Click HERE).
  • I was admittedly less suspicious about lead V1,V2 misplacement in this case — because: iThe negative component of the P wave in lead V2 is relatively modest; iiThe P wave and QRST morphology in leads V1 and V2 looks quite different from the all-negative P wave and QRST complex appearance in lead aVR; andiiiThe distinct terminal S waves in both leads I and V6 suggested to me that the r’ in leads V1 and V2 might simply be explained by the presence of incomplete RBBB. That said, Dr. Smith’s suspicion was correct — and the saddleback pattern disappeared on the repeat ECG.
ECG #2Some Unusual Findings:
As discussed by Dr. Smith — this patient’s “true ECG” ( = ECG #2) no longer manifests a saddleback pattern, and does not show any acute changes. Nevertheless — there are some additional ECG findings worthy of mention.
  • There is a surprising amount of fragmentation of the QRS complex in ECG #2. For example, the QRS complex shows a 4-component (rSr’S’ ) QRS deflection in lead III. Multi-fragmented complexes are also evident in leads aVL and aVF. S waves are present in 11 of the 12 leads on this tracing. Clinically — the significance of these unusual findings is uncertain — though “fragmentation” often indicates “scar”, and might reflect underlying cardiac disease. Whether to explore this possibility further would depend on clinical correlation.

Recommended Resources