Monday, March 8, 2010

New Left Bundle Branch Block is a poor indicator of coronary occlusion

For a couple other very interesting posts on LBBB and STEMI, click here, and here and here for the most recent post.

Case 1:

A 55 yo male with a history of hypertension but no MI or CHF presented with 5 days of progressive dyspnea without chest pain. Here is his ECG:

There is sinus tachycardia with left bundle branch block (LBBB). A previous ECG from one year prior was normal without LBBB. According to the ACC and AHA, new LBBB in the presence of ischemic symptoms is an indication for reperfusion therapy. Cardiologists widely realize that, in reality, this is not a specific indication. Physicians do not follow this indication because, in their experience, most patients with new LBBB do not have coronary occlusion unless they also have some specific indicators of occlusion, known as the Sgarbossa criteria, or the Sgarbossa criteria as modified by Smith (data to be published). These criteria are dependent on the fact that, in LBBB, the ST segment and T wave (repolarization) are in the opposite direction from the major deflection of the QRS.

Sgarbossa criteria are 1) concordant ST elevation on at least 1 mm in at least 1 lead (5 points) 2) concordant ST depression in V1-V3 (leads that always have a negative QRS in LBBB)(3 points) and 3) 5 mm of discordant ST elevation in leads with a negative QRS (excessive discordance) (2 points because it was "only" 89% specific). I have modified only the last criterion by making it proportional to the QRS. Using angiographic data, I have adjusted the excessive discordance criteria number 3 so that the discordant ST elevation is only significant if it is out of proportion to the QRS. Specifically, I have found that if the ST segment as measured at the J point and relative to the PR segment is > 0.20 the depth of the preceding S-wave, in just one lead of V1-V4, that it is likely to be coronary occlusion.

This new LBBB shown above in case 1 has no concordance (all ST segments and T waves are in the opposite direction from the QRS; this is the normal configuration for LBBB). However, according to the Sgarbossa criteria, there is excessive discordance as defined by ST elevation of 5 mm in lead V2 (although this did not satisfy Sgarbossa's point system criterion, which gives only 2 of the necessary 3 points for excessive discordance). By the new criterion, the discordant ST elevation is not excessive (5/35 = 0.14 which is less than 0.20). Thus, by the new criterion this is a true negative, but by Sgarbossa's criterion, it would get 2 points and be highly suspicious for acute STEMI. The reason that Sgarbossa's criterion was not specific enough is that many patients have high QRS voltage and resulting high discordant ST elevation, at baseline.

The BNP was 1200; an echo revealed global LV dysfunction with EF of 20% and no evidence of wall motion abnormality. The troponins were negative. The use of a 5 mm absolute criterion would have resulted in a false positive cath lab activation.

Case 2:

This 89 yo m presented after syncope during exertion, with no chest pain. He had a loud systolic murmur at the left sternal border that radiated to the carotids, consistent with severe aortic stenosis.
Here is the initial ECG:

The first troponin returned positive at 0.20 ng/ml (LoD .04 ng/ml, 99% reference less than 0.1 ng/ml) and a second ECG was recorded. At no time did the patient have chest pain.

This is also new LBBB, this time without any specific criteria positive. The patient did have release of troponin but never had chest pain. The clinicians discussed activating the cath lab based on new LBBB, but this patient had neither CP nor dyspnea and thus is not a candidate even under ACC AHA guidelines. His only symptom was syncope, which was explained by aortic stenosis.

Furthermore, the troponin elevation can be explained exertion in the presence of aortic stenosis, due to demand ischemia (or "type II" MI, not caused by acute coronary syndrome and thus not requiring intensive antiplatelet or antithrombotic therapy).

A great majority of patients with ischemic symptoms and new LBBB do not have acute coronary occlusion and do not need immediate reperfusion therapy. Some do have "NSTEMI," but the vast majority who have coronary occlusion will meet the modified Sgarbossa criteria.

How this should affect global guidelines such as ACC AHA is unclear. But what is clear is that clinicians find new LBBB akin to crying wolf, and no longer follow the guidelines.


  1. I think most people get the wrong impression from the common mantra that a LBBB with ischemic symptoms warrants a trip to the cath lab. Its not that ischemia or MI causes a LBBB, but rather that a LBBB makes the ability of detecting underlying ischemia more difficult. That is what makes the modified criteria mentioned above very helpful.

  2. 1) LBBB may indeed be caused by coronary occlusion, especially large anterior STEMI that affects the conducting system. This may be reversed by reperfusion therapy. That is why NEW LBBB is more of an "indication" than old LBBB.
    2) It is a misunderstanding that LBBB significantly obscures the EKG diagnosis of coronary occlusion. This conception is a leftover from the pre-imaging days when old MI (diagnosed with Q waves) was diagnosed by EKG, and LBBB does indeed obscure Q-waves. In the reperfusion era, we are looking for ST segment changes, not Q-waves. However, the notion that MI cannot be diagnosed in the presence of LBBB carried over to Acute MI. Additionally, the diagnosis of MI by biomarkers is confused with the diagnosis of complete coronary occlusion. In normal conduction, the diagnosis of MI as defined by biomarker elevation is very insensitive. There is a lot of evidence (that is not widely known) that the ECG in the presence of LBBB is nearly as sensitive for the diagnosis of occlusion as it is in normal conduction. See my chapter with references in Brady and Truwit: Critical Decisions in Emergency and Acute Care Electrocardiography.

  3. Dr. Smith -

    For some reason I thought your modified criterion was ST-elevation that is > 0.25 the QRS complex!

    I've been misquoting you. I'll make the necessary corrections on my blog.


  4. Tom,
    You're not misquoting me.
    I've updated the data (n was 13 and is now 34) and found that 0.20 is more sensitive without losing specificity.

  5. Dr.Smith,
    Do you have any post explaining how to diagnose MI in presence of LBBB?


  6. Here is a post that describes it:

    Tom Bouthillet has done a great job of describing my ratio rule here:

  7. Dear Dr. Smith!

    Case 2:
    How do we know that the syncope was the result of the aortic stenosis and not an arrhythmia or altered ventricular function due to a transient silent ischaemia (transient occlusion, unstable "angina")?
    I mean OK, a severe aortic stenosis can cause symptoms like this and slightly elevated troponin levels, but I really don't want a patient like this that dies next day because of coronary occlusion...Is it a real danger, what do you think?
    (Sorry, if my question is stupid or the answer is trivial...)

    Thank for your answer!

    med. student at Semmelweis University, Hungary

  8. Marton,
    No, it's a good question. The absence of any of the modified Sgarbossa criteria (see the link to the abstract of my recent paper -- it is on the upper right of the blog) makes coronary occlusion very unlikely. Add that to the absence of chest pain or dyspnea, and it is more unlikely still. OK?
    Steve Smith

  9. Dear Dr Smith,

    In case 2s 1st trace is left anterior hemi/fasicular block present, with qR lead I, rS lead III and left axis deviation? Does this help explain the sudden change to complete LBBB? Are the conduction abnormalities likely due to LVH 2nd to the AS? Lastly does the presence of 1st deg AV block with LBBB constitute trifasicular block? (RBBB with 1st deg and R or L axis deviation often used for PPM indication in my area).

    Your blog = my bible


    1. Charlie,
      good observations!
      1. Yes, you are right, there is LAFB.
      2. Of course now all he needs to do is block his posterior fascicle and it is LBBB. Notice the rate is faster (75 vs. 60) when there is LBBB. There is probably a rate-related LPFB.
      3. It constitutes incomplete trifascicular block. Full trifascicular block is 3rd degree heart block.

      thanks for your insights!

      Steve Smith