Monday, April 17, 2023

Compare these two ECGs. Do either, neither, or both show anything important?

One case sent by Dr. Sean Rees MD, written by Pendell Meyers, other case by Sam Ghali and Steve Smith

Take a look at these two ECGs below from two patients in the ED, first without any clinical context. Full case details and outcomes are below.

Case 1:

Case 2: 

Case 1:

What do you think?

This was sent to Dr. Smith by SamGhali (@EM_RESUS) with zero other info.  

Smith's response was: "OMI Mimic."

Later, this info was supplied by Sam:

This ECG was recorded in a 23-year-old African American man with a history of psychiatric illness, acute alcohol/drug intoxication, brought in by police officers status post being tazed. He had no symptoms of ACS. This pattern was recognized as an OMI mimic and no further workup was pursued.  The remainder of his Emergency Department stay was uneventful.

The Queen of Hearts correctly says:

Smith: Why is this ECG which manifests so much ST Elevation NOT a STEMI (even if it were a 60 year old with chest pain)?   
--There is even a convex ST segment in V3!! (when there is ST Elevation, convexity in any one of leads V2-V6 is fairly specific sign of LAD occlusion).  
The reasons are: 
1) there is too much voltage in the QRS (deep S-wave in V3 especially).  
2) The T(volume)/QRS ratio is not large enough for the T-waves to be considered hyperacute.  
3) The shape of the T-wave is just "not right".  This is something that is hard to teach, but with hundreds of such cases, we have taught the artificial intelligence algorithm to recognize this.
4) There is well formed J-point notching.

Case 2:

What do you think? In addition to sinus tachycardia, the only abnormalities listed by the computer were "low voltage, precordial leads" and "anteroseptal infarct, old...Q greater than 40mS, V1-V2"

Meyers interpretation: I was sent this ECG with no clinical information whatsoever, and I responded: "Easily diagnostic of acute LAD occlusion." There are hyperacute T waves in V2-V5, I, II, aVL. The STD also present in V2 completes the description of de Winter T waves (hyperacute T wave with depressed ST takeoff). 

Here is the clinical informaton on ECG 2:

A man in his 50s presented to the Emergency Department with acute chest pain that started within the past few hours. He also stated he "just didn't feel well" starting about 14 hours prior to presentation, but the chest pain was only noticed about 2 hours prior to presentation. He was mildly tachycardic (105-110 bpm) and hypertensive (157/92 mm Hg) on arrival.

His triage at 0127 is the ECG above.

Based on approximately 18,000 training ECGs with expert interpretation and outcomes (trained by Smith and Meyers), PM Cardio's Queen of Hearts AI instantly makes this diagnosis on any smartphone, even just by taking a screenshot of a piece of paper in suboptimal conditions:

The ECG was interpreted by the clinicians as "No STEMI." 

His HEAR score (before troponin resulted) was documented at 3, with documentation stating "low suspicion for ACS."

Repeat 0157 with ongoing chest pain:

Basically the same features diagnostic of LAD occlusion. QOH: "OMI High confidence". Physician interpretation: "No STEMI."

Initial troponin I returned elevated at 350 ng/L (not sure exactly which high sensitivity troponin I assay this is, but many of them have a male upper limit of normal around 20-35 ng/L). A troponin this high in a patient with no known chronic troponin elevation, and active acute ACS symptoms, has a very high likelihood of type 1 ACS regardless of the ECG.

Repeat ECG at 0221:
Basically unchanged. QOH: "OMI Mid confidence." Physician: "No STEMI."

With ongoing chest pain and positive troponin, the physician consulted with the cardiologist at the nearby PCI center, who replied that there was no indication for STEMI activation, and no indication for thrombolytics. The patient was transferred to the PCI center non-emergently over the next few hours. 

Upon arrival to the PCI center, the repeat troponin returned at 13,962 ng/L. Chest pain is documented as ongoing.

Repeat ECG at 1624 (shortly before cath):
QS waves now present in V2-V3, with slight STE, showing the pattern of left ventricular aneurysm morphology. No significant signal of terminal T wave inversion to suggest any reperfusion; this MI is in the process of completing. With this ECG alone, and no clinical context, it would be difficult for me to say if this OMI is subacute or old. QOH: "Not OMI, High confidence." (probably because QOH has not had enough training yet to differentiate subacute OMI vs. old LVA morphology, which is very difficult).

Cardiologist interpretation: "Technically does not meet STEMI criteria but concerning for ischemia."

Cath around 1730:

Left main: normal
LAD: prox LAD 30% stenosis immediately followed by a "99-100% thrombus filled occlusion" involving the proximal and mid LAD, as well as a large diagonal branch, with TIMI 1 flow pre-procedure. 
LCX: minimal luminal irregularities 
RCA: minimal luminal irregularities

Aspiration thrombectomy and PCI performed with resultant TIMI 3 flow in LAD and D1.

Next trop 85,528 ng/L (consistent with a massive myocardial infarction), none further measured.

1824 post cath:
Mot much different than the ECG above, but QOH: "OMI, Low confidence."

Echo: EF 40-45%, anterior and apical WMA. 

The patient was discharged 3 days later, and long term follow up is not available. But you could imagine the long term prognosis difference between this patient and one who received reperfusion therapy at the time of OMI diagnosis upon arrival to the initial ED.

Final Diagnosis: "NSTEMI"

Smith:  think about this: the patient had an occluded LAD and lost his entire anterior wall, and yet received a diagnosis of "NSTEMI".  

NSTEMI is worthless garbage diagnosis!  

The diagnosis should be either OMI or Non-OMI.  In this case it is OMI.

Learning Points:

OMI is not solely an ECG diagnosis, but expert ECG interpretation can greatly help to improve recognition of STEMI(-) OMI and conversely to recognize false positive STEMI criteria.

The day seems near when AI ECG interpretation will be a reliable and objective solution for any provider to rapidly differentiate subtle OMI from OMI mimics. 

Ongoing ischemia despite medical management has always been an indication for emergent cath, which is frequently ignored as it was in this case.

According to the 2022 ACC Expert Consensus Decision Pathway on the Evaluation and Disposition of Acute Chest Pain in the Emergency Department (reference below), hyperacute T waves and De Winter sign are both now considered "STEMI equivalents" and warrant the same management as STEMI. 


MY Comment, by KEN GRAUER, MD (4/17/2023):


Today's post by Dr. Meyers (with contributions from Drs. Rees, Ghali and Smith) — serves as a lesson in the recognition of acute OMI — and in distinction from one of the most commonly encountered "OMI mimics".

Like Dr. Smith — I initially viewed the 2 ECGs from Case #1 and Case #2 without the benefit of any history. The Take-Home Point from these 2 Cases lies with how to immediately recognize that Case #1 is an "OMI mimic" — whereas Case #2 shows what should be easily recognizable ECG findings of deWinter T waves, that in a patient with new chest pain should mandate prompt cath.

  • For clarity in Figure-1 — I've reproduced these 2 tracings from today. I focus my comments on a few additional points to the above excellent discussion by Dr. Meyers.

Figure-1: Comparison of the 2 ECGs in Case #1 and Case #2.

CASE #1:
We have presented many examples of LVH that masquerades as an OMI "mimic" (See My Comment in the April 2, 2023 post — and the April 14, 2023 postto name just the most recent examples). Although there are "many variations on this theme" — the basic premise is that although the ST-T wave changes of LV "strain" most often manifest in one or more of the lateral leads — some patients manifest a "mirror-image" of strain in the anterior leads.
  • The first "tip-off" that the ECG in question is unlikely to represent acute OMI — is that one or more of the anterior S waves are disproportionately deep. Whereas in most cases of ECG LVH — the diagnosis is made by tall lateral R waves — the "mirror-image" of this increased lateral lead R wave amplitude is unexpectedly deep S waves in one or more of the anterior leads. For example, in Case #1 — the S wave in lead V3 is 33 mm deep!
  • Doesn't the RED insert of the mirror-image of lead V3 — show precisely what you might expect to see for the ST-T wave of LV "strain" in a lateral chest lead?
  • The other clue that the ST elevation seen in multiple other leads on the Case #1 ECG is unlikely to be due to acute OMI — is the prominent J-point notching in leads II,III,aVF; and in leads V5,V6. In association with upsloping (ie, "smiley"-configuration) ST segments in these leads — this appearance is very characteristic of a repolarization variant (and very unlikely to indicate acute OMI).

  • PEARL #1: I suspect that the ECG in Case #1 represents both benign repolarization changes (J-point notching in the leads with upward sloping ST elevation) — and — probable LVH (ie, extreme increased voltage that surpasses that expected even for a younger adult + LV "strain", as seen in the RED insert in lead V3 in the lead with extreme S wave amplitude). It turns out that the patient whose ECG is shown in Case #1 was a young black male with known alcohol abuse (ie, predisposition to alcoholic cardiomyopathy) — and IF this younger black male had significant longstanding hypertension (common in this demographic) — we would then have an explanation for his abnormal ECG.

  • PEARL #2: Perhaps you were initially concerned not only by the coved ST elevation we see in lead V3 of the Case #1 ECG — but also by the ST elevation and straightening of the ST segment take-off in lead V4? If so — I strongly suspect this ST elevation and straightening in lead V4 simply reflects a "transition" lead between the coved ST elevation in lead V3 (curved RED line in this lead) — and the upward-sloping ST elevation in lead V5 (curved BLUE line in this lead).

  • BOTTOM LINE: Once familiar with how common it is for LVH and repolarization variants to serve as "OMI mimics" — the combination of findings seen in this Case #1 ECG will immediately suggest that this tracing is an "OMI mimic". This impression will then be solidified as soon as you learn that the patient in question is a young black male who is predisposed to both repolarization variants and LVH, and who is without cardiac-sounding chest pain.

CASE #2:
I am more disturbed by learning of the events associated with the ECG in Case #2. As noted by Dr. Meyers in his above discussion — Dr. Meyers instantly recognized the Case #2 ECG as completely characteristic of deWinter T waves — which in a patient with new chest pain, has to be immediately recognized as indicative of acute LAD OMI until proven otherwise.
  • There are certain patterns in ECG interpretation that emergency providers must be able to instantly recognize. Among these patterns are deWinter T waves. In the same way that Brugada-1 ECG patterns — Wellens' Syndrome — acute STEMIs — and Ventricular Tachycardia — must be immediately recognized by emergency providers (to name just a few of these "immediate recognize" patterns) — so it is in a patient with new chest pain who presents with an ECG looking like that seen in Case #2. Yet multiple providers (including at least 1 cardiologist) failed to recognize the deWinter T wave pattern in this acutely ill patient with new, severe and persistent chest pain.

We'll continue to present variations on the "theme" of deWinter T waves. For review — I've copied below my summary of this entity from My Comment in the February 10, 2023 post of Dr. Smith's ECG Blog. Features of deWinter T waves seen in the Case #2 ECG include the following:
  • Disproportionately tall T waves in multiple precordial leads (most marked in leads V2 and V3 — but also seen in leads V4 and V5). These T waves manifest "hyperacute" characteristics of being not only taller-than-expected (given relative amplitude of the QRS in the respective lead) — but "fatter"-at-their-peak and wider-at-their-base than one would expect. For example — the T wave in lead V3 is more than twice the height of the tiny R wave in this lead.

  • J-point ST depression at the onset of the ST segment in leads with tall deWinter T waves (ie, most marked in leads V2 and V3 — but also suggested in the lateral chest leads).

  • PEARL #3: Easy to overlook in this initial ECG in Case #2 — is not only the QS complex in leads V1,V2 — but the fact that the S wave in lead V2 is fragmented (slurred) in the initial part of its downslope, strongly suggesting (much more than a simple QS in this lead would) — that there has already been infarction. This is further supported by the finding of a small-but-present Q wave in both leads V3 and V4! While these Q waves are indeed tiny — the fact that they follow QS complexes in leads V1,V2 in association with deWinter T waves and, in the absence of septal q waves in more lateral chest leads V5,V6 — is indication that infarction has already occurred.

  • PEARL #4: ST-T waves are also abnormal in the limb leads. Although the QRS complex in both leads I and aVL is tiny — the T wave in these leads is clearly hyperacute (ie, disproportionately taller and "fatter" than it should be given the tiny size of the QRS complex). Subtle ST segment flattening with slight ST depression is also seen in the inferior leads. In the context of deWinter T waves in the chest leads — these more subtle limb lead ST-T wave abnormalities confirm acuity of this tracing.

  • BOTTOM LINE: As per Dr. Meyers — the 2022 ACC Expert Consensus Decision Pathway for acute CP Management (on Page 1933 in the J Am Coll Cardiol 2022 manuscript) — now specifically includes deWinter T waves — as potential indication of "proximal LAD occlusion that warrants immediate angiography". Emergency providers need to instinctively recognize this ECG pattern in a patient with new chest pain!


KEY POINTS regarding DWinter T Waves:

In 2008 — Robert J. de Winter and colleagues (Drs. Verouden, Wellens, and Wilde) submitted a Letter to the Editor to the New England Journal of Medicine (N Engl J Med 359:2071-2073, 2008) — in which they described a “new ECG pattern” without ST elevation that signifies acute occlusion of the proximal LAD (Left Anterior Descending) coronary artery.

  • The authors recognized this pattern in 30 of 1532 (~2%) patients with acute anterior MI. Cardiac cath confirmed LAD occlusion in all cases — with ~50% of patients having a "wraparound" LAD. Left mainstem occlusion was not present.
  • This was the authors’ original description of the new ECG pattern: “Instead of the signature ST-segment elevation — the ST segment showed 1-3 mm of upsloping ST depression at the J point in leads V1-to-V6 — that continued into tall, positive symmetrical T waves”.
  • The QRS complex was usually not widened (or no more than minimally widened).
  • Most patients also manifested 1-2 mm of ST elevation in lead aVR.

 For illustrative purposes — I’ve adapted Figure-1 from the original manuscript by deWinter et al, published in this 2008 NEJM citation.
Figure-2: The deWinter T Wave Pattern, as first described by Robbert J. de Winter et al in N Engl J Med 359:2071-2073, 2008. ECGs for the 8 patients shown here were obtained between 26 and 141 minutes after the onset of symptoms. (See text).

In their original 2008 manuscript — deWinter et al went on to describe the following additional features:

  • “Although tall, symmetrical T waves have been recognized as a transient early feature that changes into overt ST elevation in the precordial leads — in this group of patients, this new pattern was static, persisting from time of the 1st ECG until the pre-cath ECG.”
  • Hyperkalemia was not a contributing factor to this ECG pattern (ie, Serum K+ levels on admission were normal for these patients).   


NOTE: Technically speaking — the deWinter T wave pattern as described in 2008 by deWinter et al differs from the finding of simple "hyperacute" anterior T waves — because ECG findings with a strict deWinter T wave pattern persist for an hour or more until the “culprit” LAD vessel has been reperfused. 

  • As I note above (and as illustrated in the example ECGs taken from the deWinter manuscript that are shown in Figure-2) — there should be involvement in all 6 chest leads with the strict deWinter pattern, with most leads showing several mm of upsloping J-point ST depression and giant T waves.

MY OBSERVATIONS regarding De Winter T Waves: Over the past decade — I have observed literally hundreds of cases in numerous international ECG-internet Forums of deWinter-like T waves in patients with new cardiac symptoms.

  • Many (most) of these cases do not fit strict definition of “deWinter T waves” — in that fewer than all 6 chest leads may be involved — J-point ST depression is often minimal (if present at all) in many of the chest leads — and, the number of leads that manifest giant T waves is limited.
  • ECG changes in many of the cases that I have observed are not “static” until reperfusion (as had been initially reported in 2008 by deWinter et al.). Neveretheless, cath follow-up has routinely confirmed LAD occlusion in almost all cases.

 on the Phenomenon of deWinter T Waves: 

I believe there is a spectrum of ECG findings, that in the setting of new-onset cardiac symptoms is predictive of acute LAD occlusion as the cause. I suspect that to a large extent — what is seen on ECG depends greatly on when during the process the ECG is obtained.
  • While many of these patients do not manifest “true" deWinter T waves” (because their ECG pattern does not remain static until reperfusion by coronary angioplasty) — for the practical purpose of promptly recognizing acute OMI — I do not feel that it matters ( = my opinion) whether a “true” deWinter T wave pattern is present — vs the presence of simple “hyperacute” T waves (that are deWinter-like)

No comments:

Post a Comment

DEAR READER: I have loved receiving your comments, but I am no longer able to moderate them. Since the vast majority are SPAM, I need to moderate them all. Therefore, comments will rarely be published any more. So Sorry.

Recommended Resources