Monday, December 25, 2017

See what happens when one fails to diagnose STEMI in LBBB and Paced Rhythm

Written by Meyers, edits by Smith:

A female in her 60s presented complaining of 2-3 days of fatigue and shortness of breath. She called EMS when her symptoms acutely worsened while she was shopping. EMS arrived and recorded a heart rate of 27.

On arrival to the ED she was noted to be in complete heart block. She was given atropine with transient increase in HR to 80s.

This ECG was recorded after response to atropine:
What do you think?


Sinus rhythm with 2nd degree type 1 (Wenckebach) AV block, best visualized in lead I.

Note: Wenckeback AV block occurs at the AV node (as opposed to the His bundle or below). Complete heart block that responds to atropine is also high (AV node, not His bundle or below). The AV node blood supply comes from a branch of the RCA.

On the other hand, 2nd degree type II and third degree (complete) block are likely to be due to LAD ischemia (in third degree block, especially if the escape is wide), and are unlikely to respond to atropine.

So the Wenckebach rhythm alone, and its response to atropine, can provide important information as to the etiology of the block.

How about the QRST?

LBBB is present with concordant ST elevation in leads III and aVF, with excessively discordant reciprocal ST depression in aVL. There is also concordant ST depression seen in the second QRS complex in V2. These findings meet both the original and modified Sgarbossa criteria, and are diagnostic of inferior and posterior wall acute coronary occlusion.

As a reminder or review for new readers, the modified Sgarbossa criteria are designed to identify acute coronary occlusion in the setting of LBBB, and are considered positive if any of the following three criteria are met in any lead when measuring ST elevation (at the J-point, relative to the PR interval):

1) Concordant ST elevation of 1mm or greater in any lead
2) Concordant ST depression of 1mm or greater in leads V1-V3
3) Excessively discordant ST elevation in any lead, such that the STE is 25% or greater of the preceding S-wave amplitude (20% was also still quite specific and even more sensitive)

There is a second rule which is less sensitive but very specific:
Any single lead with proportionally excessively discordant ST elevation or ST depression that is at least 30% of the previous S- or R-wave

See this post for further details on the modified Sgarbossa criteria.

The AV block is almost certainly attributable to the fact that the RCA almost always supplies the AV node.  Her prior recorded heart rate of 27 before atropine suggests she was in complete heart block at that time (no tracings available).

These findings were apparently not noticed in the setting of LBBB. A transvenous pacemaker was placed as the heart rate continued to drop.

What do you think?

Now there is a paced rhythm, as expected.  Do ST segments meet the modified Sgarbossa criteria?  In other words, are the ST segments either 1. concordant or 2. proportionally excessively discordant?  (See the third component of the first rule, or the second rule).

1. The STD ratio in lead I is indeed excessive by the second criteria (one lead with STE or STD greater than 30% of the preceding S- or R-wave).

2. Moreover, the first complex in lead III appears to 4.5 mm of ST Elevation with a 19 mm S-wave, for a ratio of 23.7% (almost 25%). The second complex appears to have a lower ratio (3mm/20mm = 15%).

3. The first complex in aVF seems to have 3mm STE / 14 mm S wave = 21%.

So the third criterion (excessively proportionally discordant STE at 25% of preceding S-wave) is almost met. Remember that, in the validation study, the rule using the third criterion at a 25% ratio was 99% specific and 84% sensitive, but a criterion of 20% was still 80% specific (and 94% sensitive) and thus is almost always due to occlusion.

So although the ischemia is much less visible in the paced rhythm, it is still diagnostic.

Look at lead V2: it is the transition lead in this case (where the QRS transitions from mostly negative to mostly positive), and the positive component of the QRS is slightly larger than the negative component. Thus, one expects ST depression in V2, and in order to indicate ischemia, it has to be proportionally excessive ST depression. Since the R-wave is 8 mm, and excessive STD is 30%, the ST depression would have to be at least 2.4 (2.5) mm to be ischemic. I don't believe it is.

This ECG by itself is be diagnostic of inferior and posterior acute coronary occlusion.

However, in this case, you get the benefit of seeing the ischemic findings both with and without pacing. For the vast majority of our chronic pacemaker patients, we don't get to see this, but they still deserve immediate diagnosis and treatment.

The initial troponin I was elevated at 0.44 ng/mL. After temporarily improving the heart rate with transvenous pacemaker, she was noted to be hemodynamically stable and to have "decreased shortness of breath but persistent fatigue." Coronary ischemia apparently was still not considered the most likely etiology at this time. Electrophysiology consult was placed, and together the physicians felt that the situation was more likely explained by acute renal failure and hyperkalemia (despite the fact that the highest potassium level measured was 5.9 mEq/L). Her second troponin I drawn 3 hours later was 0.38 ng/mL, which was interpreted as reassuring. She was admitted to the CCU and consented for a permanent pacemaker planned for the following morning.

Another ECG was recorded that night, with no mention of what symptoms prompted it:

This shows a huge amount of inferior ST elevation (ST/S ratio 1.0 or greater!) with reciprocal depression in aVL and I. This confirms ongoing transmural infarction of the inferior wall. The ST segment in V2 is inappropriately isoelectric (it should be elevated), suggesting ongoing posterior infarction.

Repeat troponin I around this time was rising at 0.68 ng/mL. Another troponin 4 hours later was 3.04 ng/mL. Unfortunately there is no available documentation about the thought process throughout the evening and night concerning the worsening ECGs and rising troponins. It seems that they scheduled her for cath the next morning.

Her last recorded troponin at 1:30am returned at 181.23 ng/mL!

At 7:45 am the next morning she went into asystole. She received CPR, atropine, epinephrine and was then noted to be in ventricular fibrillation. She was shocked once with return of spontaneous circulation, however her blood pressure continued to trend downward over the next 10 minutes. She was placed on dopamine and epinephrine drips, as well as receiving further hyperkalemia therapy (hyperkalemia had resolved on labs at this point). Another VF arrest was corrected with defibrillation. The patient remained critically hypotensive on multiple drips.

Finally the patient was taken to the cath lab where she was noted to have a completely occluded proximal RCA. Thrombectomy was performed with excellent angiographic result, however the patient continued to deteriorate despite balloon pump placement and continued pressors.

She expired on the cath lab table.

You may think this case is rare, that you and your colleagues would never miss ECG changes like the ones in this case, when the clinical scenario is so obvious. Yet I find these cases far too often. I cannot comment on the clinical decision making about NOT doing angiography when such patients have obvious clinical and ECG indications. What I can help with is helping clinicians to know what ECG findings to look for in the setting of LBBB and/or paced rhythms.

In one online quiz in Medscape, 50% of physicians believed the STEMI cannot be diagnosed in the presence of a paced rhythm.

The modified Sgarbossa criteria have been validated as the most accurate criteria currently available for identifying acute coronary occlusion in the setting of LBBB (but like all other criteria, are not perfect). We believe (based on cases like this and many others) that they are also applicable in paced rhythm, and are currently in the process of studying this. Preliminary results are positive, and the full multicenter results are on the way.


  1. First of all sad outcome.
    I would like to add the following point (a bit academic, given that the first ECG was diagnostic of an ongoing STEMI): I suspect that RV involvement was very likely: ECG#1 shows ST depression in V2 and in the presence of posterior involvement we should expect STD in V1 also, here instead we see clear ST elevation. Moreover RV infarction may well be one of the etiologic factors causing hemodynamic instability.
    Thanks for presenting this very instructive case!

    1. Mario, Great comment, and we should have mentioned it, but it is not as definite as one might think.

      There is no ischemic STE in V1; instead it is the normal amount of expected ST elevation for LBBB. V2 is much more sensitive to posterior MI than V1, and V1 often has no relative ST depression even when V2 manifests STD of posterior MI.

      We studied this very topic (manuscript under consideration). We found that in inferior MI patients without STE in V1 (as in this case, no excessive STE) who had ST depression in V2 (n=82, 16 with RVMI and 68 without), the NPV was 72% and PPV was 25%. So your suggestion is correct 72% of the time, but not always.

      Thanks for the great comment!


  2. Very helpful and relevant post.. Thanks a lot

  3. Really helpful and relevant case discussion.. Thanks a lot

  4. Hi i am a french med student (6th year), thank you very much for your great post !
    However I have 2 questions that may appear very "silly" so i apologize in advance:
    - I have been teached that we could not interpretate ST segment if there is a bundle branch block, so that LBBB + ischemic symptoms equals urgent angiography.
    In this case, i am surprised by the absence of ischemic pain and would have never guess STEMI as first etiology. Do STEMI without the typical academic pain happen often ?

    - Although both first and second troponin were elevated, the second came lower than the first one. How should we interpret this decrease ?

    Thks again !

    1. Tam,
      Great questions. One third of MI patient present without chest pain. However, one is presented with a patient with symptoms, and if the symptom is chest pain, vs. weakness, the probability of MI is far lower for weakness. So the ECG must be much more convincing when there is no chest pain.
      We abandoned the idea that a new LBBB is an indication for cath. Read my paper: The discussion tells you all about this.

      where are you in France? It is my favorite country!! J'aime la France, et j'ai beaucoup des amis partout!

      Steve Smith

    2. I am from Tours, université Francois Rabelais.

      Thks for the reply !

  5. Manuel Vindas. Costa Rica.March 1, 2018 at 6:28 AM

    Hi Steve; Greatings from Costa Rica. We are looking forward your visit. Thanks for this magnificent case.

    1. Manuel,
      thanks. do I have a visit planned that I am unaware of?


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