Tuesday, August 4, 2020

A man in his 40s with chest pain reproducible with palpation

Written by Pendell Meyers, submitted by George Konstantinou

A man in his early 40s with history of smoking and hypertension presented to the ED with substernal and right sided chest pain of 8 hours duration. The pain had first started after a stressful event and had waxed and waned several times over the 8 hours. The pain was reproducible with palpation on the right side of the chest.

Here is his initial ECG:

Notice the leads configuration (this ECG comes to us from Greece).

There is sinus rhythm with very small STE in V2-V3. The T waves in V2-V5 are very concerning for hyperacute T waves with increased area under the curve. Comparison with a prior ECG (unavailable in this case) would almost certainly show dramatically increased area under the T wave. There is the smallest hint of STD in V5-V6 and III. There is very poor R wave progression, with minimal R waves throughout the precordium.

Here, Ken Grauer has calculated the 4-variable formula results for the presentation ECG:

For the four-variable formula, 18.2 was the derived cutpoint with the highest accuracy at 92%, sensitivity 88.8%, specificity 94.7%. A value > 18.2 supports LAD occlusion, while a value < 18.2 supports normal ST elevation. The closer the score is to 18.2 (e.g. >17.7 or <18.7) the more likely it is to represent a false negative or false positive. 19.62 is very concerning for true LAD occlusion.

Simplified alternative formulaThere is also a simplified formula developed by Dr. Emre Aslanger, based on the original formula and published in the American Journal of Cardiology in 2018; 122(8):1303-1309.  The benefit of the new formula is that it is independent of any QT Correction, which would depend on the 4 different QT correction formulas. Also, this formula uses the QT in millimeters, not in milliseconds!

Here it is: (RAV4 + QRSV2) - [(QT in mm) + STE60V3); a value < 12 is indicative of LAD occlusion.

The area under the curve for this formula was 0.963, with sens, spec, and acc of 87%, 92%, and 90%.

For this case: (3 + 16) - 10 + 0.5) = 19 - 10.5 = 8.5 which is less than 12.

Case Continued The attending cardiologist found the ECG "not convincing for STEMI" and no emergent angiogram was performed. The patient was given ticagrelor, aspirin, and heparin after the high senstivity troponon I returned elevated at 857 ng/L (URL <28ng/L).

Note: Even without the ECG, if a patient has new chest pain, then such a high troponin is diagnostic of type 1 MI.  If the pain is ongoing, then cath lab activation is necessary regardless of the ECG.

The patient was admitted to the coronary ICU. Pain persistent throughout the ED course and for the next few hours in the coronary ICU, during which time serial ECGs were recorded:


Both repeat ECGs do not show any clear progression to STEMI criteria, mostly unchanged from the presentation ECG.

Due to the persisting pain, angiography was finally done at about 3 hours after admission to the coronary ICU. There was complete occlusion of the mid-LAD, with PCI performed successfully.

Pre-intervention, total mid LAD occlusion.

Pre-intervention, annotated.

Post-intervention, showing the opened LAD.

ECG the next day:

There are new Q waves in V1-V3, with very poor R wave progression in V4-V6.

Peak troponin and clinical course are not available.

Learning Points:

Those who care to learn expert ECG interpretation can use these lessons in practice to help identify OMI when it would otherwise be missed.

Use the 4-variable LAD OMI vs. normal variant STE formula to help identify subtle LAD occlusions such as this one.

This patient could have been reperfused approximately 3 hours sooner, which is likely to reduce infarct size and improve long term outcome.

In a patient with MI diagnosed by troponin whose chest pain is persistent, emergent cath lab activation is indicated by both European and American guidelines (ACC/AHA, ESC).


4-Variable Derivation:
Driver BE, Khalil A, Henry T, Kazmi F, Adil A, Smith SW. A new 4-variable formula to differentiate normal variant ST segment elevation in V2-V4 (early repolarization) from subtle left anterior descending coronary occlusion – adding QRS amplitude of V2 improves the model. J Electrocardiol. 2017;50(5):561–569. 

4-Variable Validation:
Bozbeyoğlu E, Aslanger E, Yıldırımtürk Ö, et al. A tale of two formulas: differentiation of subtle anterior MI from benign ST segment elevation. Ann Noninvasive Electrocardiol. 2018;23(6):e12568.

MY Comment by KEN GRAUER, MD (8/4/2020):
Today’s case (presented by Dr. Meyers) follows the “theme” of several of our recent posts — namely resetting the paradigm away from a “STEMI” vs “non-STEMI” approach (which misses ~25-30% of acute coronary occlusions) — with our goal toward increased acceptance of the more effective OMI-NOMI paradigm.
  • We showed one example of a missed OMI in our recent August 1, 2020 post. Today’s case illustrates another ...

I thought the ECG findings in today’s case were subtle. Because of limited tracing-to-tracing variation on the serial ECGs — I decided to focus My Comment on the initial ECG seen in the ED, which for clarity, I’ve labeled and show again in Figure-1.
  • The history in today’s case was somewhat equivocal. Chest pain was a new symptom for this 40-something year old male smoker. This chest pain had been intermittently present over a number of hours prior to presenting to the ED, albeit with some atypical features.

Figure-1: The initial ECG in this case (See text).

MY THOUGHTS on ECG #1: The important clinical question in this case is despite not satisfying criteria for a “STEMI” — Should the patient whose ECG is shown in Figure-1 undergo cardiac catheterization?
  • The rhythm in ECG #1 is sinus at ~70-75/minute. The PR, QRS and QTc intervals are all normal. There is LAD (Left Axis Deviation) — though not quite enough to qualify as LAHB (because there is not clear predominant negativity in lead II). There is no chamber enlargement.

Regarding Q-R-S-T Changes: I noted the following descriptive findings — all of which are subtle!
  • There are small and narrow septal Q waves in leads I and aVL.
  • Transition is delayed in the chest leads. There probably is more net positive area within the R wave in lead V5 than within the S wave in this lead — but barely. Predominant positivity of the QRS complex never develops in the chest leads (ie, there is poor R wave progression”).
  • There is straightening of the takeoff of the ST segment in lead II (angled RED line in this lead). This clearly looks abnormal — probably because of the disproportionately large (fat-at-its-peak and wide-at-its-base) T wave in this lead.
  • There is also straightening of the ST segment in lead aVF. That said — the ST segment in aVF did not strike me as much as being clearly abnormal, probably because of the much more “modest” accompanying T wave in this lead.
  • There appears to be a small amount of upsloping ST elevation in high lateral leads I and aVL. I didn’t interpret this as necessarily abnormal — given that repolarization variants sometimes show similar ST elevation of this shape.
  • The T wave is inverted in lead III — but there is no ST segment depression, and given the predominantly negative QRS complex in this lead — the T wave inversion in lead III is not necessarily abnormal.
  • T waves in leads V2-thru-V6 look to be disproportionately tall. This is subtle — probably because of the modest R wave amplitude across all chest leads. But regardless of the R wave amplitude — I would not expect T wave amplitude to surpass R wave amplitude, as it does in ECG #1 in leads V4, V5 and V6. Support that these lateral chest lead T waves are abnormal (and in this setting, hyperacute!) is forthcoming from: i) ST segment straightening in these leads (angled RED lines); andii) Recognition that T waves dwarf the S waves in these leads.

BOTTOM Line Regarding ECG #1: There are a number of subtle findings in ECG #1None of them in isolation would be enough to compel me to recommend prompt cath lab activation. However — taken together + this patient’s history ( = new chest pain that morning that was still persistent at the time ECG #1 was obtained) + the positive high-sensitivity troponin — should suggest an ongoing acute coronary syndrome until proven otherwise.
  • Putting It All Together — I’d interpret the T waves in leads II and V2-thru-V6 as hyperacute in the setting of an ongoing acute coronary syndrome.
  • Taken together — the lack of inferior lead reciprocal ST depression with unimpressive ST elevation in aVL lack of hyperacute T waves in lead V1 are all consistent with a mid-LAD lesion (as was found on cath).

Our appreciation to Dr. George Konstantinou (of Greece) for submitting this case.

Saturday, August 1, 2020

Computer and transferring physician say "normal." What do you think?

This was sent by Guillaume Debaty, from the Grenoble Region of France.  He has done a lot of great work on cardiac arrest, including as co-author of our study on esmolol in refractory cardiac arrest, and much more with Keith Lurie.

See his Google Scholar profile here.

A 40-something woman was in a remote alpine location when she complained of crushing chest pain.

The physicians said the ECG was normal, as did the computer.

Guillaume asked for it to be sent.

What do you think?

Guillaume thought that the minimal ST elevation in inferior leads was accompanied by abnormal ST depression and T-wave inversion in aVL.  Note that the voltage of the STD and TWI are VERY small -- but so is the voltage of the QRS.   All ST depression and T-wave inversion is proportional, and must be evaluated in the context of a large or small QRS voltage.

T-wave inversion in aVL and in V2 are both "soft" signs of inferior OMI.

Here the STD in aVL is significant.  He saw it.

They activated the cath lab and found a 100% obtuse marginal occlusion.  It could have been a false positive ECG, but if it had been, no harm done.  It was not.

Of course we do not know for certain that the inferior findings represent ischemia.  It could be that this is the baseline normal ECG and that the marginal occlusion was electrocardiographically silent.  The only way to tell would be to see evolution on subsequent ECGs: if no evolution, then these findings have nothing to do with the occlusion. I will try to get those ECGs!

Learning Points

1. The ECG in OMI may be normal.
2. The ECG in OMI may appear normal at first glance, but actually be highly suspicious or diagnostic
3. Any ST depression in aVL with ANY STE in inferior leads is highly suspicious for OMI.
4. The computer makes this mistake very frequently.  See here over 20 cases of the computer calling the ECG completely normal when in fact it is diagnostic of OMI (some of the many cases here are of other disorders than OMI).

MY Comment by KEN GRAUER, MD (8/1/2020):
For illustrative purposes — the timing of this case is optimal, coming as it does just 1 day after Dr. Smith posted his latest contribution to the challenge of resetting the paradigm away from the problematic “STEMI” vs “non-STEMI” approach — and towards acceptance of the much more practical and effective OMI-NOMI paradigm.
  • Today’s case could have been one of the estimated 25-30% of acute coronary occlusions that are regularly missed by insistence on millimeter-defined STEMI criteria before considering intervention.

For clarity — I’ve labeled and reproduced the initial ECG in today case (Figure-1). It should be emphasized that ECG findings in Figure-1 are subtle — and none of these findings in isolation would compel me to activate the cath lab.
  • As emphasized in My Comment to the July 31, 2020 post — in the setting of new cardiac symptoms — One looks for any of a number of other ECG findings that should raise concern for an acute ongoing event despite lack of millimeter-criteria for STEMI. Several of these are present in this case.

NOTE: The computer did not “make a mistake” in this case. The mistake was made by the clinicians who accepted at face value a computer interpretation of “normal” for an ECG with subtle but definitely-present ECG findings.
  • KEY Point: We should expect that the computerized interpretation will not be sensitive for detecting subtle acute changes. Therefore — Do not look at the computerized interpretation until after you have made your clinical decision. Even then, if you are concerned by the ECG and the clinical history — Do not be swayed by a negative computer report!
  • P.S. — Despite the best of programming — the computer is still not able to “touch” and talk to the patient in front of them. And, in today’s case — the fact that this previously healthy 40-something woman presented with new-onset “crushing” chest pain was of critical importance for optimal clinical decision-making — because this type of history needs to heighten your acuity for seeking out subtle ECG findings that we should not expect the computerized report to pick up.

Figure-1: The initial ECG in today’s case (See text).

MY THOUGHTS on ECG #1: The rhythm is sinus at ~60/minute. All intervals and the axis are normal. There is no chamber enlargement. Regarding Q-R-S-T Changes — I noted the following:
  • There are very small and narrow Q waves in inferolateral leads. These are clinically insignificant.
  • R wave progression is normal — with transition occurring between leads V3-to-V4.
  • There is subtle flattening of the ST segment in a number of leads (horizontal RED lines). Normally, there should be a gradual upsloping of the ST segment in most leads, that then goes on to blend almost imperceptibly into an upright T wave. This is not what we see in leads V2 and V3, and by extension in lead I.
  • It’s a bit more challenging to assess the ST segment in leads V4 and V5 — because ST-T wave morphology is different for the 2 beats that we see. There should be no doubt that the ST segment in leads V4 and V5 for complex -A- is flattened — but not so much for complex -B- (and unfortunately we don’t know which of these 2 complexes show the true ST-T wave morphology in these leads).
  • As noted by Dr. Smith — the T wave in lead aVL is inverted. Although the amplitude of this inverted T wave is tiny — so is the QRS! While true that in isolation — the T wave may normally be inverted in lead aVL when the QRS is predominantly negative (as it is in ECG #1) — the base of this tiny T wave inversion looks disproportionately wide. Conclusion — I did not know for certain if the T wave inversion in lead aVL of ECG #1 was abnormal — but it could be!
  • Also as noted by Dr. Smith — the T wave in lead V2 is inverted. While the T wave may normally be inverted in lead V1 in adults — most of the time, the T wave should not be inverted in lead V2. Adding to my suspicion that this was indeed an abnormal finding is the distinct ST segment straightening in lead V2, that continues on in lead V3.
  • Finally — there are the subtle findings in the inferior leads. I was not as concerned by the slight-but-real upsloping ST elevation in each of these leads. But I wondered (after more than a few looks) if those peaked T waves in leads III and aVF might not be a little bit taller-than-expected given R wave amplitude in these leads ...

BOTTOM Line Regarding ECG #1: If this tracing had appeared in my “daily stack” during the 30 years that I routinely read all out-patient tracings for our 35 medical providers — I would not have thought twice about saying, “Changes do not appear to be acute”. But that is not the clinical scenario that occurred in this case.
  • All of the above findings are subtle! But taken together in the context of a patient presenting with new-onset “crushing” chest pain — I was concerned about a few potentially hyperacute T waves + nonspecific ST segment flattening in multiple leads + potential reciprocal changes in leads aVL and V2 — which taken together, was fortunately enough for Dr. Guillaume Debaty to wisely activate the cath lab.

Recommended Resources