Wednesday, June 29, 2022

When Normal is only Pseudo-Normal, it can deceive the caregivers.

An Aussie friend down under sent me this case. 

A 70-something male presented with sharp stabbing central CP.  Here is the initial ECG with 7/10 CP:


The pain was resolving, and another ECG was recorded with 3/10 CP:

Normal, but compared to the first, the T-waves in V2 and V3 are not as tall.  Hmm.......

The pain completely resolved:

What do you think?

Now there is an all-but-diagnostic terminal T-wave inversion in V2.  This is very subtle Wellens'.

The patient then developed 8/10 stabbing CP again:

The T-wave in V2 is again upright, indicating re-occlusion of the LAD.  There is also less T-wave inversion in aVL, so it is likely a proximal LAD.  There are subtle ST-T abnormalities in III, V5, and V6 as well.  
This is "Pseudonormalization" because it looks normal but it is not!!

The patient again became pain free:

Now there are unequivocal Wellens' pattern A waves (terminal T-wave inversion, biphasic up-down) in V2 and V3.  The T-wave is again deeper in aVL

Here is the high sensitivity troponin I profile (99% URL = 20 ng/L):

0852    8

1215    10

1730     8

2030     12

0535     12

The patient had chest pain again at 24 hours after presentation:

On its own, this ECG appears almost completely normal (for those with sharp eyes, there is abnormal ischemic ST depression in inferior and lateral precordial leads.  However, compared with all previous ECG, the T-waves are now upright, and there is much more ST Elevation than on any previous ECG.
Thus: Pseudonormalization AGAIN!

There is again Reocclusion.  I don't believe this was recognized as Re-occlusion because another troponin was measured, which returned at 30 ng/L (now elevated).  

Then this was recorded at 26 hours, without any intervention:

The ST elevation is resolved.  The T-wave is now normal.  There are Wellen's inverted waves in aVL.

Fortunately, the LAD spontaneously reperfused.  Otherwise, there would have been a lot of myocardial loss.

Here is the angiogram:

The Proximal LAD has an ulcerated plaque with 90% stenosis and there is TIMI II-III flow.  
There was successful PCI.

Here is the entire progression again.  Be sure to click on the image to enlarge it!!

Learning Points

1. Wellens' is reperfusion of OMI

2. The artery is has "hot" thrombus which can propagate and/or lyse, over and over.

3. This is really a Transient STEMI (better: Transient OMI, as it never meets STEMI criteria!)

4. Wellens' implies some (sometime tiny) amount of myocardial loss (infarction) and therefore some change in troponin.  In this case, it did rise by 4 ng/L, which is a significant rise (i.e., not due to natural variation or laboratory imprecision), but it did NOT rise above the 99th percentile, so this was unstable angina, until the later re-occlusion at which point the troponin DID rise above the 99th percentile and it was a type I MI.

5.  The re-occlusion was missed because the ECG looks nearly normal during LAD Occlusion.  The best clue that it was occluded was to look carefully for the VERY SPECIFIC changes (some changes in ECGs are not specific, but this morphology is very specific!)

6.  Optimally, patients with Wellens' would go immediately to angiography, because that hot thrombus is a transient STEMI.  Transient STEMI frequently need to go to the cath lab in the middle of the night if they don't go immediately

See this post:

Timing of revascularization in patients with transient STEMI: a randomized clinical trial

7. If you don't take them right away, they should get continuous 12-lead ST segment monitoring.

See these posts:

Why we need continuous 12-lead ST segment monitoring in Wellens' syndrome

Comment by KEN GRAUER, MD (6/29/2022):
Illustrative case presented by Dr. Smith that emphasizes once again the following important concepts:
  • That part of the definition of Wellens' Syndrome requires a history of the patient having had chest pain — that has now resolved!
  • That there is a spectrum of ECG changes that may be seen with Wellens' Syndrome — including extremely subtle findings (as evidenced by the terminal T wave inversion in lead V2 of the 3rd ECG shown above). This could have been all-too-easy to overlook had the clinician not integrated the history with new development of this terminal T wave inversion at the time that the patient's chest pain had totally resolved!

  • Subtle cases like the one presented today "tell a story" — that may only become clear through sequential clinical correlation of serial tracings — in which the timing of ECGs — the presence and relative severity of patient symptoms — of serial troponin values (and of bedside Echo if done) — are all integrated.
  • Understanding all parts of the story explains how concepts such as "pseudo-normalization" of the ECG — and reocclusion of the "culprit" vessel — can easily be overlooked.

We have emphasized the importance of serial ECG comparison on numerous occasions in Dr. Smith's ECG Blog. For reference — I'll just cite 3 posts in which I focused on the "art" of serial ECG comparison in My Comment ( = the November 15, 2018 post — the June 28, 2018 post, in which Dr. Smith worked through another Wellens' Syndrome evolution  — and the August 20, 2020 post).
  • For clarity — I wanted to focus on just 1 of the 7 tracings shown today — which is the 4th ECG that was done (Figure-1). As discussed above by Dr. Smith — this 4th tracing was obtained in association with the recurrence of chest pain (described by the patient as "8/10 and stabbing"). Without careful serial ECG comparison with the 3 previous tracings — the clinical significance of the ECG findings on this 4th ECG could easily be overlooked!
  • The fact that the patient's chest pain symptoms returned (with even greater severity than they had been) — in association with an element of "pseudo-normalization" (ie, the "tell-tale" terminal T wave inversion in lead V2 had been replaced by an upright T wave) — was diagnostic in this context of proximal LAD reocclusion.

The other point to emphasize about the ECG in Figure-1 — is that it is not a normal tracing.
  • Although there is no more than the most minimal ST depression — there is definite ST segment flattening — with an abrupt angulation at the junction of the flattened ST segments with the T wave that follows (RED lines in leads I,II; V4,V5,V6 in Figure-1). These are not normal ST-T waves.
  • What this ST segment flattening and angulation at the T wave junction means — is another matter. It depends. I've seen similar ST-T waves countless times in the thousands of primary care ECGs I interpreted in years past. By itself — it is purely a nonspecific finding, which can at times represent a subtle sign of underlying coronary disease — but at other times may be devoid of clinical significance.
  • That said — in the context of today's serial tracings — this ST segment flattening and abrupt angulation was clearly increased in ECG #4 compared to the 3 prior ECGs in today's case. Together with loss of the terminal T wave inversion in lead V2 of ECG #3 — this made for that many more leads in ECG #4 indicative of "dynamic" ST-T wave changes despite seeming "pseudo-normalization".

Figure-1: This is the 4th ECG shown in today's case — obtained shortly after the patient's chest pain returned (described as "8/10" and "stabbing" ).

Saturday, June 25, 2022

Acute chest pain and a difficult ECG

Written by Pendell Meyers

A man in his late 40s with several CAD risk factors presented with acute chest pain. He had several episodes lasting several hours over the past few days, with a more intense episode starting 1 hour prior to arrival. He described the pain as substernal squeezing and nonradiating, and associated with diaphoresis. He had some recent cocaine use 3 days ago. 

Here is his triage ECG:

I sent this ECG with no clinical context to several colleagues, with their responses below:

Smith: "I'm going to say LVH, but it's a tough one." (Meaning: I think that's not OMI but rather a mimic caused by LVH)
McLaren: "Looks like LVH, are V2-3 switched? What's currently labelled V3 looks hyperacute T, but other leads ok."
Grauer: "Assuming the history is acute chest pain...clear LVH...lead V2 is malpositioned (the abrupt tall R makes no sense) - but I'm concerned about the amount of ST elevation and the straight ST segment take-off in V3 - with ST coving and T inversion in V4 - so if history is concerning, this could be OMI in progress."

Smith: Notice that the T-wave inversion in V4-V6 is classic Benign T-wave Inversion MorphologySee countless examples here.

His baseline ECG from 1 year ago was available:

Triage ECG interpretation:
Sinus rhythm with LVH that appears slightly more severe than baseline ECG from 1 year ago. Next, for ischemia interpretation, there are many questions that must be considered:

Does the STE and potentially plus size T waves in V1-3 mean anterior OMI?
Do the terminal T wave inversions in V4-6 indicate reperfusion of OMI? Wellens?
Is the slight STD and T wave inversions in aVL worrisome for subtle inferior inferior OMI with tiny STE and budding hyperacute T waves in inferior leads?

Or are all these findings due to LVH?

LVH causes some of the most difficult OMI and STEMI mimics out there. In my opinion based on many ECGs I've seen in the past, all the questions above are answered by LVH alone, except potentially the concern for possible hyperacute T wave in V3. In other words, the terminal T wave inversion in V4-6 is a Wellens mimic caused by LVH, the TWI in I and aVL are appropriately discordant from the LVH. As for V3, I am slightly reassured when I look to its next-door neighbor, lead V4, and see no corroborating signs of hyperacute T waves. So I would be concerned and keep an eye on this area in subsequent repeat ECGs, but I would not yet consider it diagnostic for OMI.

Case continued:

Initial high sensitivity troponin I = 40 ng/L (99% upper reference limit for men is 20 ng/L for this assay)

He was given aspirin and nitroglycerin shortly after arrival, and approximately 20 minutes later, his pain completely resolved.

The case and the ECG were discussed with cardiology who basically said "there's no STEMI, so let us know what the troponin is."

After pain resolution, another ECG was recorded:

The patient was admitted for high risk chest pain and further evaluation.

Later that day:

Next day

Again, the benign T-wave inversion morphology shows itself in V4-V6

Repeat troponins:

39 ng/L

40 ng/L

40 ng/L

39 ng/L


Moderate concentric LV hypertrophy

EF 61%

No segmental wall motion abnormalities

No pericardial effusion

The medicine team consulted cardiology, who advised that no further workup was indicated at this time. They referred to a coronary CT angio performed about 1 year ago which showed several coronary arteries with nonobstructive CAD. Somehow this was used as part of the reasoning not to perform further workup.

No definitive cause of chest pain was found. The discharge note described possible musculoskeletal, esophageal, and cocaine-related causes as possibilities.

Learning Points:

Abnormal depolarization usually results in abnormal repolarization, and learning the expected repolarization patterns is critical for improving diagnostic performance in ECG interpretation.

LVH produces some of the most difficult STEMI and OMI mimics out there.

OMI is not just an ECG diagnosis.

We must focus on learning false positive ECG patterns in parallel with learning the false negative STEMI(-) OMI patterns.

Comment by KEN GRAUER, MD (6/25/2022):
Today's case by Dr. Meyers provides yet another example that illustrates the challenge of distinguishing between acute OMI vs LVH. I am not aware of any "magic formula" that instantly makes this distinction — and therein lies the difficulty!
  • Most of the time (as was true for today's case) — a single ECG will not suffice for telling you if the cath lab needs to be immediately activated. Instead — comparison with prior and/or serial tracings + troponins (and optimally stat Echo) — will be needed for optimal decision-making.

Dr. Meyers' discussion above makes the excellent point that LVH may account for all of the ECG findings in today's initial ECG tracing. While I completely agree with this assessment — I wanted to offer another perspective on how I would have proceeded in today's case.
  • For clarity — I've reproduced the initial ECG in Figure-1.

Figure-1: The initial ECG in today's case.

The Initial ECG:
The rhythm in ECG #1 is sinus at ~85/minute. Intervals (PR, QRS, QTc) are all normal. The frontal plane axis is slightly leftward (about -10 degrees) — but this is not enough to qualify for LAHB.
  • Small and narrow q waves are seen in leads I, aVL; V4 and V5 — but strangely enough, not in lead V6. That said — I still suspect these are normal septal q waves.
  • Criteria for LVH are easily satisfied in multiple leads (ie, very tall R in aVL ~14 mm; very tall R in V5 22 mm + S in V1 >35 mm). For "My Take" on a user-friendly approach to the ECG diagnosis of LVH — See My Comments in the June 20, 2020 and April 27, 2019 posts in Dr. Smith's ECG Blog.

  • ST-T wave changes in high-lateral leads I and aVL look completely typical for LV "strain" — because there is no more than minimal J-point depression with a slow downslope to the ST segment, followed by a more rapid upslope. The amount of ST-T wave depression in these leads is perfectly proportional to what I'd expect given the R wave amplitude in these same leads.
  • In contrast — the ST-T wave changes in lateral chest leads V5 and V6 look different. That is, the ST segment is flat — and the T waves are symmetrically inverted, consistent with either LV "strain" and/or ischemia. That said, the important point given this patient's presentation to the ED with new chest pain — is that nothing I've described thus far suggests acute OMI or the need for immediate cath.

As Dr. Meyers noted above — the one lead I was most concerned about was lead V3. This lead clearly shows significant J-point ST elevation, with straightening of the ST segment "takeoff" and a wide T wave base — which taken together could represent a hyperacute T wave in this patient with new-onset chest pain. 
  • Given modest depth of the S wave in lead V3 — the reason for my concern was that I thought the amount of J-point ST elevation in this lead V3 was more than I'd expect for the anterior ST elevation sometimes seen with LVH.
  • I suspected the ST-T wave appearance in lead V4 represented a transition between the positive ST-T wave of lead V3 — and the T wave inversion of lead V5. That said — the ST segment in lead V4 was clearly coved, so I thought it impossible to say that this was not acute.

My problem was with lead V2. The nearly isoelectric QRS complex in this lead made no physiologic sense given predominantly negative QRS complexes before it (in lead V1) — and after it (in lead V3).
  • To Summarize: The patient in today's case presented with new-onset chest pain. While I was not at all convinced that ECG #1 represented an ongoing acute process — I felt there was no denying that: i) Lead V3 could represent a hyperacute ST-T wave; ii) If lead V3 did represent a hyperacute ST-T wave — then the ST segment coving in lead V4 could be the tail end of this process; iii) While not necessarily abnormal taken by itself — the ST-T wave in lead V1 was clearly taller than is usually seen in this lead; andiv) I had no doubt that lead V2 was malpositioned
  • Given i), ii) and iii) in the above bullet Summary — I felt the "true" appearance of lead V2 was essential for accurate interpretation of this initial ECG. Therefore — I would have verified lead placement — and immediately repeated the initial ECG on this patient.

  • Of note — R wave progression was normal on the baseline ECG done 1 year earlier (ie, the 2nd ECG shown above in this case).
  • After pain resolution in the ED — the ECG was repeated. This is the 3rd ECG shown above in today's case — and it once again strongly suggests malposition of lead V2 (once again rendering this lead useless in our interpretation). Since this repeat ECG was recorded in the ED not long after the initial tracing — I suspect that the same technician (who was responsible for malposition of lead V2 in the initial tracing) — was probably also the one to record this repeat ECG after chest pain resolved.
  • Later that day — and again on the next day — a 4th and 5th ECG was obtained. Both of these last 2 tracings showed normal R wave progression. Since these tracings were done presumably hours (or moreafter the first 2 ECGs — presumably a different technician recorded these tracings (which is why I suspect that R wave progression is now normal).

  • In Conclusion: Whereas I did not feel we could rule out an OMI from the initial tracing shown in Figure-1 — none of the subsequent tracings in today's case were suggestive of an acute event. I suspect that IF the initial ECG would have been immediately repeated with accurate chest lead electrode placement — that there may not have been any need for concern about a possible acute event from the initial ECG.

  • Learning Point: When clinical decision-making hangs in the balance and you strongly suspect an error in lead placement — it is best to immediately repeat the ECG — after correct lead placement is verified. Sometimes, you may need to verify correct placement yourself. I know of no other way to ensure accurate interpretation of the ECG in question.

Thursday, June 23, 2022

Acute Chest pain with LBBB. What is going on?

This history and ECG were texted to me from a far away ED on a Friday night:

"Very elderly make with history of coronary disease on an angiogram 1 year prior presents with 2 hours of chest pain, sternal, pressure, and mild diaphoresis."

What do you think?

There is sinus tachycardia with LBBB.  There is concordant ST depression (STD) in lead V2, and excessively proportionally discordant STD in all of leads V3-V6.  Thus, there is one lead (V2) that meets the Sgarbossa criteria and the Smith modified criteria, and 4 other leads (V3-V6) that meet the alternative Smith modified criteria (proportionally excessively discordant ST depression at an ST/R ratio >0.30). 

Additionally, there is ST Elevation in aVR.  The ECG is low resolution and so it is difficult to determine the exact measurements, but I believe it measures at 1.5 mm.  The S-wave is 3.0 mm, and so this lead ALSO meets the Smith Modified Sgargossa criteria.  So the criteria are met in 6 leads, when only one is necessary!! In both of our studies of the Modified Sgarbossa criteria, not only was the presence of the criteria in just one lead much better than the original Sgarbossa criteria, but the proportionally excessively discordant findings were present in far more leads than in the Original criteria, and so it was more visually obvious and stunning.

In the case, the maximal ST depression is in lead V4, both in an absolute millimeter measurement and as a proportion to the preceding R-wave amplitude.  Whether normal conduction or LBBB, STD maximal in V1-V4 is very specific for posterior OMI, whereas V5-V6 is typical of non-occlusive but obstructive ischemia.     

In either case, cath lab activation is indicated!!

The physician who sent this tried to transfer the patient to an interventional center, but initially failed.

He wrote: "I'm going to try interventional cardiologist at XXXXXX as the cardiologist at the first hospital was not impressed."

So he called the interventional cardiologist at the other receiving center, and this was the response: "Interventional cardiologist at YYYYYY said that is is not a real STEMI."

I told him: "Send the patient!!  Do not take "no" for an answer!!"

And so he did.   And forced their hand.

Subsequent events

When the patient arrived, the troponin (not high sensitivity) was 29.4 ng/mL (very high, typical of large STEMI).  On arrival, the symptoms were improved, but we do not have another ECG to see if it, too, is improved.

At angiogram, they found this:

Here is the critical info:

The Ramus Intermedius had a 99% stenosis with TIMI-2 flow and they did NOT intervene because the "Ramus is unlikely to cause the ST changes and chest symptoms!"  

But these ECG findings are EXACTLY what you would expect from severe ACS of the Ramus.  (The Ramus Intermedius is a large coronary artery not present in all individuals that originates at the angle between the circumflex and the LAD; it supplies roughly the same territory as the first diagonal (D1) off the LAD.)

Furthermore, whenever there is a lesion with flow less than TIMI-3 (TIMI 0-2), it is an acute lesion; if it is resulting in ischemia and can be stented, my understanding is that it generally should be stented.

I don't understand this.

Left Main and 3-vessel disease in LBBB

In our validation study of the Smith Modified Sgarbossa criteria, we identified 4 patients who did not meet our definition of ACO yet did have notable catheterization findings and outcomes warranting immediate diagnosis and intervention.  These 4 patients were found to have AMI and new 3-vessel or left main coronary artery disease with either 1) an acute but non-occlusive culprit lesion, or 2) very high troponin. Peak 24-hour troponin T levels were 0.25, 0.75, 1.04, and 6.61 ng/mL. Two of these 4 required urgent or emergent coronary artery bypass graft. We categorized these patients as having acute 3-vessel/left main disease myocardial infarction(3V/LMD). Three of these 4 patients had discordant ST elevation (STE/S ratio > 20% in lead aVR), but only 1 of these 3 was positive by the 25% modified criteria. Two of the four 3 V/LMD patients also met the ST depression criteria (ST Depression to R-wave > 30%) due to widespread ST depression in various other leads, including III, aVF, V5, and V6.  None of the 4 met the weighted or unweighted original Sgarbossa criteria. If new 3V/LMD(without a specific culprit lesion identified) had been classified as equivalent to a culprit lesion in our definition of ACO, the sensitivity of each rule would be reduced by 1% to 3%, and specificity would remain unchanged, but more patients with potential need for urgent revascularization would have been identified.

Learning Points:

1. Formerly, it was thought that myocardial infarction (MI) could not be diagnosed in LBBB.  "MI" formerly meant "Old MI" as represented by Q waves.  "Acute transmural ischemia" is different from "MI".  In fact, infarction is the end result of completed transmural ischemia.

2. The STEMI criteria used in normal conduction (absence of BBB) are NOT very sensitive for Acute Transmural Ischemia (OMI) because they do not use proportionality or any other features of acute ischemia other than a fixed millimeter amount of ST Elevation.  

In fact, the Smith Modified Sgarbossa criteria in LBBB are far more sensitive for OMI than are the STEMI criteria in normal conduction!!!  This is because they use proportionality and because, contrary to conventional wisdom, LBBB does NOT hide transmural ischemia if you use proportions.

Comment by KEN GRAUER, MD (6/23/2022):
In the past, it was thought that one could not diagnose an acute STEMI (ST Elevation Myocardial Infarction) in the presence of LBBB. This notion has been completely refuted — and we have posted numerous illustrative cases demonstrating this in Dr. Smith's ECG Blog. Although it may be more difficult to diagnose acute ischemia or infarction in a patient with chest pain who presents in complete LBBB — in a surprising number of such patients, there will be at least strong suggestion on the initial ECG of acute STEMI despite the presence of underlying LBBB.
  • When acute ECG changes in a patient with LBBB are subtle — diagnostic aids such as Smith-modified-Sgarbossa criteria may be helpful. At other times (such as for the ECG in Figure-1) — the diagnosis of an acute cardiac event is obvious, even without the need to invoke modified Smith-Sgarbossa criteria.

Figure-1: The initial ECG in today's case.

MThoughts on the ECG in Figure-1:
The rhythm is sinus at ~90-95/minute. The PR interval is at the upper limit of normal — but the QRS is obviously wide, and consistent with complete LBBB. As per Dr. Smith's discussion above — Smith-modified-Sgarbossa Criteria are easily satisfied in no less than 6 leads — when only 1 lead is necessary!
  • In addition to use of Smith-modified-Sgarbossa criteria — I routinely favor a qualitative approach. Realizing that conduction defects (LBBB, RBBB, IVCD) and ventricular pacing may each alter expected ST-T wave appearance — I look for those leads that clearly show ST-wave findings that should not bthere

  • Among the flagrant ST-T wave changes in Figure-1 that should be immediately apparent — the ST segment coving and huge amount of J-point ST depression (that exceeds 5 mm in lead V4!) — in this elderly patient with new-onset chest pain — should prompt the need for immediate cath without having to delay for additional data.

Additional Observations:
Although I completely agree that today's tracing is best classified as "LBBB" — it is worthwhile pointing out that QRS morphology is not completely typical for this conduction defect.
  • For typical LBBB — there should be predominant negativity (ie, no more than a tiny initial r wave) in right-sided lead V1 — as well as in other anterior leads. The fact that a significant R wave is already present in lead V3 is not expected with a "typical" LBBB.
  • Looking closer at QRS morphology in lead V3 — there is alternation for the 4 beats seen in this lead between a qRS pattern (for the 1st and 3rd beats in this lead) — with a "fat" initial R-slur-R'-S for the 2nd and 4th beats in this lead. This is bizarre. It doesn't seem to be artifact. If "true" QRS morphology in lead V3 is the qRS pattern — then this initial q wave is abnormal and suggestive of infarction. If instead the R-slur-R'-S pattern is the "true" morphology — then this is marked fragmentation, consistent with "scar" from underlying heart disease.

  • In left-sided leads with typical LBBB morphology — there should be a monophasic (all upright) R wave. We see this in leads I and aVL — but there is a fairly deep S wave (of 5 mm) in lead V6. While true that many patients with LBBB and LVH only develop a monophasic R wave in more laterally placed leads (such as lead V7 or V8) — awareness that QRS morphology is indeed somewhat atypical for LBBB is relevant for helping us to interpret the significance of ST-T wave changes. There is simply no way that the ST segment coving in leads V3-thru-V6 could be "normal" (not to mention the huge inappropriate amount of ST depression).

  • BOTTOM Line: As per Dr. Smith — this elderly man with new-onset chest pain should have been accepted for immediate cath on the basis of his initial ECG.

Monday, June 20, 2022

LVH and Anterior ST Elevation: is it OMI and would you activate the cath lab?

Written by Jesse McLaren (@ECGCases), with comments by Smith and Grauer


A 50 year old presented with chest pain radiating to the shoulder. They had a history of an LAD stent 10 years ago and alcohol use disorder, with repeated visits for chest pain and two code STEMIs two years ago that found no occlusive disease.


What do you think? 

There’s normal sinus rhythm, normal intervals, normal axis, and normal R wave progression. There’s LVH with repolarization abnormalities, including discordant ST depression and T wave inversion inferolaterally and discordant ST elevation and tall T wave in V2.


Are there any primary ischemic changes? How can we identify OMI in the presence of LVH?


1. prior ECGs


Below are the first and last ECGs from that the visit resulting in a code STEMI, which found a patent LAD stent but no occlusive disease (as per the discharge summary, further angiogram details not available):

This shows the benefits and limitations of prior ECG. On the one hand, we can see prior secondary repolarization abnormalities. But we can also see that they fluctuate over time, and can vary with differences in lead placement.


2. proportionality


The secondary ST/T changes from LVH can make it difficult to assess primary ischemic abnormalities, and the STEMI paradigm doesn’t even try—defining STEMI as ST elevation in the absence of LVH. But the OMI paradigm can draw from the principle of proportionality. As Drs. Aslanger, Meyers and Smith explain in Recognizing electrocardiographically subtle occlusion myocardial infarction and differentiating it from mimics: ten steps to or away from the cath lab: If there is high voltage indicating left ventricular hypertrophy (LVH), some STE may be seen in leads with deep S waves (usually V1-V3) and may mimic STEMI. If the amplitude of STE is more than one-sixth of the amplitude of S-wave in one of these leads, it is highly suspicious for OMI.”


This patient has a baseline ST/T ratio in V2 of >15% but it is greater on the new ECG. Both the second ECG from the last visit and new ECG have 3mm of ST elevation in V3, but the S wave has diminished from 14mm to 11mm, raising the ST/T ratio from 21% to 27%.

Very important post for understanding anterior ST Elevation in the presence of LVH: 

LVH with anterior ST Elevation. When is it anterior STEMI?


3. patient


With the patient's repeated visits with chest pain, prior stable angiograms and repolarization changes on ECG, it would have been easy to dismiss. But the cardiologist was concerned about the patient’s description of the pain, which was worse than prior episodes and in a patient with prior coronary artery disease, so they activated the cath lab. The LAD stent was patent but there was a 90% ulcerated plaque in the right posterolateral branch, which was stented. There was also 30% proximal RCA and 50% proximal circumflex lesions.


Because of a cath lab activation and culprit lesion requiring a stent, the patient was diagnosed as “STEMI”. But they never had a rise in troponin, and the initial and post-cath trop were the same: 14ng/L prior (normal <26 in males and <16 in females) and 13 after. So was this unstable angina? 


Smith: I don't think you can even make this conclusion.  It might or might not be unstable angina, probably not, and with fluctuating ECG.


Here is the post-cath ECG:



There’s the same ST elevation in V2 and similar S wave, but now the T wave seems even larger. Is this a posterior reperfusion T wave, given the location of the culprit? But there should be at least a tiny increase in troponin with reperfusion injury.


A few days later the patient presented again with chest pain, and EMS brought in again as code STEMI based on anterior ST elevation and inferior ST depression. This time the patient declined cath and serial trops remained at 13. Echo findings: "a normal size of the LV with concentric remodeling and overall normal left systolic ejection fraction. Mild anterior septum segment, mid septum segment, and mid inferior segment wall motion abnormalities were noted. RV was normal in size and function, and there was no relevant valve pathology." The patient was diagnosed with non-cardiac chest pain. Below are the initial and discharge ECGs:



Again we see some of the variations in repolarization abnormalities, exaggerated by lead placement (R/S ratio is different V2-4 and the V1-2 are placed too high on the second—with fully negative P in V1 and biphasic P in V2).


Take home

1.     Problem: LVH produces secondary repolarization abnormalities which can make it difficult to identify primary ischemic changes, and STEMI criteria don’t help

2.     Prior ECGs can show baseline repolarization changes, but these can fluctuate over time and with variations in lead placement and patient position

3.     Proportionality: disproportionate ST elevation can help identify OMI—with ST/S>15% concerning—but there can still be false positives

4.     Patient: patients with high-pretest probability and ischemic symptoms deserve investigation even if prior angiograms have been unremarkable or ECGs are nonspecific

5.     Paradigm: code STEMIs leading to stents does not mean the patient had a “STEMI”. ACS should not be classified by whether the presenting ECG has ST elevation, but by whether the patient had Occlusion MI (OMI) or non-Occlusive MI (NOMI)-- or unstable angina.


Smith: How would we conclude in retrospect that this was, or was not, OMI?  

    1. Any ECG which manifests ischemia should have evolution.  This fluctuating LVH probably would not fulfill that criterion.  

    2. There should be at least some troponin rise and/or fall, even if they do not reach the 99th percentile.  Furthermore, it would be very rare for this to be OMI without at least one troponin going over the 99th percentile.  Our research definition of OMI requires either a) diminished flow in the artery (TIMI 0-2) or b) culprit plus a very high troponin. There can be OMI that does not meet these criteria (but that does manifest on the ECG and has evolution) due to a) occlusion that is too brief, or b) inability to identify a culprit, but there must be SOME troponin rise and/or fall, even if it is unstable angina with all of them below the 99th percentile.  The definition of acute MI includes a rise and/or fall of troponin with at least one value over the 99th percentile for that assay. 

    Therefore, this patient had Non-ischemic ST Elevation and a chronic, stable stenosis that was intervened upon.  There was some non-ischemic fluctuation in the ECG. 

This patient should carry around a copy of their ECG!

Comment by KEN GRAUER, MD (6/20/2022):
Superb discussion by Dr. McLaren about the difficult topic of how to assess and manage the patient who has a history of coronary disease — but who also has a history of repeated ED visits for chest pain, which have led to cardiac catheterization that previously failed to show occlusive coronary disease. Additional features of today's case that complicate assessment include:
  • The history of "alcohol use disorder". While this most definitely does not rule out the possibility of new acute cardiac disease — in my experience, it always made it more difficult for me to assess such patients.
  • A series of prior ECGs on the patient that showed fluctuation of ST-T wave abnormalities. That said — differences in lead placement made assessment of these fluctuations more difficult.
  • Some unusual findings on the initial ECG (including LVH) — that left me uncertain about how to proceed.

I thought the initial ECG in today's case was worthy of further comment. For clarity — I have reproduced this initial tracing in Figure-1.
  • ECG #1 shows a normal sinus rhythm at a rate just over 60/minute. The PR interval is normal, and the QRS is narrow — but the QTc is prolonged (I estimate ~480 msec.). The frontal plane axis is normal at +20 degrees.

As noted above by Dr. McLaren — there is LVH on this initial ECG. That said — there are some atypical features about this ECG diagnosis:
  • I have previously reviewed "My Take" on a user-friendly approach to the ECG diagnosis of LVH (See My Comments in the June 20, 2020 and April 27, 2019 posts in Dr. Smith's ECG Blog). Voltage criteria are not quite met in the chest leads in Figure-1. There are met in lead aVL (R ≥12 mm in aVL) — and the very tall R wave in lead I is probably also qualifying.
  • As I've emphasized in previous posts — the fact that more there are more than 50 criteria in the literature for the ECG diagnosis of LVH means that none of these criteria are optimal. It also means that the ECG is far from a perfect assessment tool for LVH. For practical purposes — when QRS amplitude appears increased and ST-T wave changes consistent with LV "strain" are present — the probability is high that "true" LVH is present if the patient is an adult of a certain age who has a history of heart disease. I therefore thought the diagnosis of LVH to be secure.

I found the most striking ECG finding in Figure-1 — to be the 2+ mm of J-point ST elevation in lead V2, that occurs in association with a disproportionately tall T wave with wide base. That said — I was not convinced that the upward concavity ("smiley"-configuration) ST segment with angled J-point in ECG #1 was acute. While the overall shape of this ST-T wave in lead V2 looked most consistent with the type of ST elevation seen in anterior leads with LVH (June 20, 2020) — the S wave in lead V2 was not nearly as deep as I would expect if the ST-T wave in lead V2 was simply the result of LVH.
  • There was subtle ST elevation in the high-lateral leads I and aVL of ECG #1 — with what appeared to be some mirror-image opposite ST depression in the inferior leads. Could this be acute? (especially since the ST-T wave depression of LV "strain" was nowhere to be seen in these high-lateral leads?).
  • There was moderately deep and symmetric T wave inversion in lateral chest leads V5,V6 (with terminal negativity of the T wave in lead V4). Was this consistent with ischemia? with LV "strain"? — or — with a combination of the two?

  • My Impression of ECG #1: I was not at all certain from assessment of this initial ECG as to how I would answer Dr. McLaren question about whether to activate the cath lab. My "hunch" — was that acute OMI was less likely — but I felt I'd need more information (ie, the degree of concern from the patient's history and symptoms, comparison with other tracings, troponins, etc.) to make my decision.
  • Review of the 2 prior tracings from the patient's last code STEMI visit did not resolve my uncertainty. While ST-T wave changes seemed a bit more marked on ECG #1 than on the prior tracings — I did not think they provided a definitive answer.

Figure-1: The initial ECG in today's case.

Bottom LINE in Today's Case:
As emphasized by Dr. McLaren — Patients with high pre-test probability and ischemic symptoms sometimes deserve investigation, even if prior caths have been unremarkable and/or ECGs are nonspecific. This was the case for today's patient.
  • Unfortunately — assessment of this patient in the future may remain problematic if he again returns for chest pain, but without definitive ECG changes or troponin elevation.

  • P.S.: Very important point highlighted by Dr. Smith's last comment = Just because there is a high-grade stenosis (ie, a 90% narrowing in today's case) — does not necessarily mean that this is a "culprit" lesion responsible for symptoms.

Recommended Resources