Thursday, November 14, 2019

A 50-something with left shoulder pain and diffuse ST elevation

A 50-something presented with left shoulder pain.

He had an ECG recorded:
What do you think?

There is a huge amount of ST Elevation, but to my eye it was typical of normal variant.  One might say there is terminal QRS distortion, but, although there is indeed absence of S-waves in BOTH V2 and V3, there is in fact very clear notching at the end of the QRS in both V2 and V3.

Fortunately, there was an ECG from about a year prior:

And here is one from about 10 years prior:

You can see here that the computer says "suggests pericarditis" but that I changed it to early repolarization.

Just because there is diffuse ST elevation does NOT mean it is pericarditis.
The most common etiology of diffuse ST elevation is, by far, early repolarization.

The patient had musculoskeletal shoulder pain.

MY Comment by KEN GRAUER, MD (11/14/2019):
A subject well worth periodic review — is the concept of Terminal QRS Distortion (T-QRS-D). Prior to working with Dr. Stephen Smith — this concept was unknown to me. When present — T-QRS-D may provide invaluable assistance for distinguishing between early repolarization vs acute OMI (ie, When true T-QRS-D is present in a patient with new symptoms — it is virtually diagnostic of acute OMI). To review:
  • T-QRS-— is defined as the absence of both a J-wave and an S-wave in either lead V2 or lead V3 (and according to Drs. Smith and Meyers — probably also in lead V4). Although simple to define — this finding may be subtle! I fully acknowledge that it has taken me a while to become comfortable and confident in its recognition.
A picture is worth 1,000 words. I’ve taken the lead V3 examples in Figure-1 from previous cases posted on Dr. Smith’s ECG Blog:
  • TOP in Figure-1 — Despite marked ST elevation in this lead V3 — this is not T-QRS-D, because there is well-defined J-point notching (BLUE arrow). This patient had a repolarization variant as the reason for ST elevation.
  •  BOTTOM in Figure-1 — This is T-QRS-D, because in this V3 lead there is no J-point notching — and, there is no S wave (RED arrow showing that the last QRS deflection never descends below the baseline).
Figure-1: Comparison between ST elevation in lead V3 due to a repolarization variant (TOPfrom 4/27/2019) — vs acute OMI (BOTTOMfrom 9/20/2015), which manifests T-QRS-D (See text).

Regarding the current case — I have put together the first 2 ECGs that were shown above (Figure-2).
  • Isn’t it tempting to say there is T-QRS-D in the initial ECG that was done in the ED ( = ECG #1 in Figure-2)? After all, there is no S wave in lead V3 ...
Figure-2: The first 2 ECGs shown in this case (See text).

COMMENT on ECG #1: The patient in this case was a 50-something man, who presented with left shoulder pain. As per Dr. Smith — there is marked ST elevation in lead V3 — with a lesser amount of ST elevation in inferior and lateral chest leads.
  • The reason the ST-T wave appearance in lead V3 of ECG #1 does not qualify as T-QRS-D — is that despite lack of an S wave in this lead, there is J-point notching (or at least J-point slurring) that is characteristic of repolarization variants.
  • Other ECG features in ECG #1 in favor of a repolarization variant instead of acute OMI include: i) A relatively short QTc interval and tall R waves in the mid-chest leads; ii) Lack of reciprocal ST depression; iii) A similar “look” to the peaked T waves that we see in at least 9 of the 12 leads in ECG #1 (compared to a more localized ST-T wave picture that is typical with acute infarction); and, iv) J-point notching or slurring that is typical for repolarization variants in no less than 7 of the 12 leads in ECG #1 (BLUE arrows in Figure-1).
BOTTOM Line: While the composite of the above features makes it more likely that ECG #1 does not reflect acute OMI — there nevertheless is a significant amount of ST elevation in multiple leads in this 50-something man who presented with new symptoms.
  • Therefore — more information was needed to attain greater certainty (ie, stat Echo looking for wall motion abnormality; additional ECGs on this patient; serial troponins).
  •  In this case — finding a prior ECG on this patient from a year earlier was revealing (ECG #2 in Figure-2). Neither the lack of S wave in lead V3, nor J-point notching or slurring were new findings (RED arrows in ECG #2). This confirmed the impression that the ST-T wave appearance in ECG #1 reflected a longterm repolarization variant in this patient.
  • P.S.: Given that there is a very small-but-present percentage of lethal cardiac arrhythmic events in otherwise healthy young adut individuals who have repolarization variants — I no longer use the term, "BER" (Benign Early Repolarization) when I see this type of ECG finding (Zakka & Refaat-ACC, 2016). Instead, I simply call it "repolarization variant" — because this ECG finding is usually, but not 100% "benign". If a patient with the ECG findings in ECG #1 presented with new, cardiac-sounding syncope — full evaluation would be in order.

Our THANKS to Dr. Smith for presenting this case!
  • For additional examples illustrating distinction between T-QRS-D vs repolarization variants — Review of the October 6,2015 post may prove insightful.


  1. Interesting case...

    (My comments are based on the assumption that the patient did not appear in a lot of distress.)

    I'm wondering what the indication was for the ECG to be done? Was the patient also having other signs or symptoms of an OMI? Was the left shoulder exam so equivocal that it was felt prudent to look further?

    Did the exam of the left shoulder not reveal any tenderness or pain with active ROM? The likelihood of having an acute OMI AND acute onset of musculoskeletal pain of the left shoulder at exactly the same time is profoundly low.

    As far as the ECG itself is concerned - I agree with you 100%. It's a textbook case of early repolarization (or now, normal variant) - one of those diagnoses (along with benign T wave inversion) that has to be studied again and again until one has the self-confidence to make the diagnosis.

    Also there is another very interesting finding on the tracing: there is a P-mitrale in the inferior leads, yet the P-terminal force in V1 is quite normal. This man has an interatrial delay within Bachmann's bundle and is at risk for future atrial fibrillation if the delay progresses.

    1. THANKS (as always!) for your comment Jerry. My biggest “plea” regarding an ED history of “pain” (be it chest, arm, neck, etc) — is brief indication of when symptoms began, and IF symptoms were truly concerning for possible acute coronary disease. Unfortunately, with cases that are passed on — this essential information is often lacking (as it is here). Despite the appearance of a repolarization variant — IF the history was that of new-onset left shoulder pain that was felt to be potentially indicative of acute coronary disease — my approach to this patient would be different (ie, Was this ECG truly needed? — or was the clinician team “sorry” after ordering it and seeing what it showed …?). Otherwise — I also saw that peculiar P wave notching in the inferior leads (albeit without associated significant negative component to the P in V1) — but I chose not to comment on this. Sensitivity and specificity of the ECG for atrial enlargement is wanting — and this might be another example of an ECG finding (if this was an otherwise healthy 50-year old) that I would have preferred not knowing about. THANKS again for your input! — :)

  2. Great case!, I have a question regarding terminal QRS distortion and the subtleSTEMI formula. I understand that with terminal QRS distortion on V2 AND V3, the formula should not be used, is this correct? or the correct thing would be that with terminal QRS distortion on V2 OR V3 to not use de formula?. The app SubtleSTEMI literally says: "Is terminal QRS distortion (absence of both A-wave and J-wave) in V2 AND V3?" I am confused. Hope to get help from you, and btw thank for creating such an incredible ECG blog (sorry for the english, it is not my mother tongue)

  3. THANKS for your comment Felipe. If I understand your question correctly — my understanding of Dr. Smith’s formula is that IF there is clear evidence elsewhere on the ECG of acute OMI — then the formula should not be used because the formula wasn’t studied under these circumstances — and because you ALREADY have your answer (ie, if there is definite T-QRS-D = Terminal QRS Distortion) — then you KNOW this patient with symptoms has acute OMI. It’s possible you sent in your question in before I wrote My Comment — Please take a look above at my Figure-1, which hopefully clarifies your understanding of how to recognize T-QRS-D. I’ll pass on your Question to Drs. Smith & Meyers to see if they would add anything to my answer. Gracias otra vez por tu pregunta! (Thanks again for your question) — :)


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