Monday, June 21, 2021

Neck and Jaw Pain in a patient with a Pacemaker. Sgarbossa Negative. But How about the Modified Sgarbossa Criteria?

I was at home on a late Saturday evening when this first ED ECG was texted to me:

Atrial and Right Ventricular Paced Rhythm
(most pacing is RV pacing --- there is increasing use of biventricular pacing)
What do you think?  What did I say?





"It looks like Occlusion Myocardial Infarction (OMI).  If the clinical presentation is consistent with acute MI, Activate the Cath Lab."


I added this to my text response: "The EKG meets the Smith modified Sgarbossa criteria, so I think there is no choice but to take a look at his coronary arteries, but for some reason I do not feel convinced in my own mind as I look at it." 


I think I was not totally convinced because all the ST segments have very normal upward concavity.  But we have shown that, in LBBB, upward convexity is only present in 50% of leads which meet the Modified Sgarbossa Criteria and are True Positives.


Full Analysis: Leads II, and V4-V6 all have discordant ST Elevation with an ST/S ratio greater than 25%.  In our PERFECT study by Dodd KW et al. (Paced ECG Requiring Fast Emergent Coronary Therapy), which I am told will be published online today in the Annals of Emergency Medicine, just ONE LEAD with such a ratio was highly specific for OMI.   

See post on the the PERFECT study paper here coming soon.  

  

As for ST Depression in V2: It is NOT concordant, as the Original Sgarbossa Criteria requires.  It is Discordant, but it is excessively discordant (the modified criteria in LBBB established that a single lead with discordant ST/S (if QRS mostly negative) or ST/R (if QRS mostly positive) ratio greater than 30% is nearly 100% specific for OMI -- this has not been validated in the group with a paced rhythm)


Moreover, although there is no definition of hyperacute T-waves for EITHER paced rhythm, LBBB, or normal conduction, these T-waves do appear to be hyperacute.


Important: there is 1) no concordant ST elevation, 2) no concordant ST depression in V1-V3 (or anywhere), and 3) no discordant ST elevation meeting the 5 mm Sgarbossa criteria.  Thus it is NEGATIVE by the Original Sgarbossa Criteria.  In the PERFECT study, 17 diagnoses were made EXCLUSIVELY by the 25% rule; only 2 could be made exclusively by the 5 mm rule.  That is why the Modified Criteria are so much more sensitive.


Also: in the study, the Modified Sgarbossa criteria worked as well on biventricular pacing as on right ventricular pacing.


So this is an inferior-posterior-lateral OMI until proven otherwise, usually due to a very large RCA with lateral branches, but could be a dominant circumflex.


Here is the clinical context:

Presenting complaint: Weakness, Dental Pain

A 50-something male presented with neck and jaw pain for one hour.  

Past History: CAD (coronary artery disease), HLD (hyperlipidemia), Hypertension, Mobitz II with pacemaker placed in 2014 (Medtronic MRI compatible), STEMI (ST elevation myocardial infarction), RCA dissection --- no intervention, Tobacco dependence, Type 2 diabetes mellitus.

He had been seen one day prior at another hospital with the same complaint.

He stated that, when he had the RCA dissection, his symptoms had been similar, but he also suffered from cervical radiculopathy and was followed by a neurosurgeon; he complained that those symptoms had not been adquately addressed.  

On the visit one day prior, an ECG had been done and was negative for OMI/ischemia.  The patient had left AMA.  No troponins were drawn at that visit.


The cath lab was activated.


Later, the ECG from one day prior was obtained:

Sinus rhythm with RBBB, without pacing


45 minutes later, while waiting for the cath team, this was recorded:


There is diagnostic evolution of ST Segments




Angiography:  100% Thrombotic Occlusion of a Dominant Circumflex in the AV groove.


First Abbott Architect high sensitivity troponin returned after the patient left for the cath lab = 70 ng/L.  Upper reference limit = 34 ng/L.


After PCI, with right ventricular pacing:

As you can see, all ratios are in the normal range now.
Highest ratio is in V3 at 2/14 = 14%
There is no longer excessively discordant STD in V2

Notice how the T-waves appear almost as hyperacute as before.  
In V3 and V4, they are nearly 20 mm, almost as tall as before
This suggests that T-wave size is less important in paced rhythm that is the ST/S ratio.
This needs further study in our database.



Post PCI ECG with some conduction of atrial paced beats (without ventricular pacing):

This shows resolution of the ischemia, with inverted (reperfusion) T-waves in inferior and lateral leads.  

There is another interesting finding, too!




There are P-waves which conduct, and RBBB, but there are also pacer spikes just a few milliseconds after the onset of the QRS.  This means that the impulse did not reach the ventricular sensing before the activation of the ventricle, and thus could not inhibit firing of the ventricular lead.  So there is fusion of intrinsic RBBB and a paced beat.

But all ischemia is gone.


Learning Points:

1. Use the Smith Modified Sgarbossa Criteria for diagnosing OMI in ventricular paced rhythm.

2. Stay tuned for the article in Annals of EM online this week.


Thursday, June 17, 2021

Acute Chest pain. All P-waves are not conducting. Is it OMI?

A 50-something male complained of acute onset chest discomfort about 30 minutes PTA while at rest with radiation described as numbness to the back of his neck and both arms.  He denied history of CAD, but he reported that he has history of smoking, hyperlipidemia, and pre-diabetes. He reports some shortness of breath and anxiety.

Here is his first ED ECG, ECG 1:

What is the rhythm?  Are there any signs of OMI?
There is an annotated version below, and Ken Grauer does a laddergram of this one at the bottom of the post.
Peter Hammarlund noticed that there is lead reversal (LA/LL)!!










Notice that there is high degree AV block.  

This is the essential feature.  Acute chest pain with high degree AV block is Occlusion MI (OMI) until proven otherwise.  One much less plausible scenario is that there is supply ischemia -- that is, some other cause of AV block with bradycardia, bradycardia causing hypotension, hypotension causing decreased coronary perfusion which causes chest pain).

In other words, you don't need to have the exact rhythm diagnosis in order to know that there is a need for angiogram +/- coronary intervention.

But here I give more analysis of the rhythm:

Below, the red arrows point out all the P-waves, which are regular.  However, none of them clearly conduct.  Every 2nd P-waves appears to conduct, but if you look closely, the PR interval gradually shortens.  The QRSs are nearly regular.  It may be that some of the P-waves that appear to conduct are actually conducting, in which case it is NOT 3rd degree (complete) AV block.   There is also some variability of QRS morphology, suggesting that there is some degree of fusion (automatic infra-nodal escape fused with conducted beat).

Ken Grauer is the master of rhythms, and he gives a very detailed analysis of this ECG below, with a laddergram.  Even he is not certain of the exact rhythm.

Moreover, the first QRS is of a different morphology that then others, which suggests that it is NOT an escape beat or that the others are fusion beats.

Annotated version of ECG 2


But whether is 2nd degree Mobitz II or 3rd degree does not really matter.

There is acute chest pain and high degree AV block, so this is acute OMI until proven otherwise.

More on this ECG: There is RBBB, but the R'-wave is limited to V1.  Therefore, the ST depression in V2 and V3 cannot be blamed on the RBBB (RBBB only has expected ST depression in V2 and V3 when it is "discordant" to a positive R'-wave)

The cath lab was activated.

While waiting for the cath team, this was recorded at 37 minutes:
2nd Degree Mobitz II
Subtle STE in III with reciprocal STD in aVL.  
These ST-T changes, by themselves, are also diagnostic of inferior OMI.


Here the rhythm is analyzed:

Red arrows point out the somewhat difficult to discern P-waves.
The red line shows the relation between the P-waves in leads V1 and II 
(II is across the bottom, and simultaneous with other leads above.)


Last ECG before cath lab:

What do you see?









There is now STE in V1, and an enlarged T-wave in V2. So this is also a right ventricular MI.

Angiogram: Culprit Lesion (s): Thrombotic occlusion of the proximal RCA.  Proximal confirms that the RV is involved 

(Proximal means proximal to the RV marginal branch, which supplies the RV.  Because most people have some collateral circulation to the RV from the LAD, most proximal occlusions do NOT result in STE in V1 or in right sided leads.  And not all which do have such STE have hemodynamic significance.  This one did not.)


Large MI.  
See how the troponin rose briskly after artery opening!


Formal contrast echo

The estimated left ventricular ejection fraction is 48%.

Regional wall motion abnormality-inferior, and inferolateral.

Right ventricle not optimally visualized, probably normal size/function.


Learning Points:

1. Chest pain and high grade AV block [does not include 2nd degree type I (Wenckebach)] is due to OMI until proven otherwise.

2. As above: Proximal RCA occlusion means proximal to the RV marginal branch, which supplies the RV.  Because most people have some collateral circulation to the RV from the LAD, most proximal occlusions do NOT result in STE in V1 or in right sided leads.  And not even all of them with such STE result in hemodynamic significance.  

=====================================

Ken Grauer's analysis of ECG 1 

(By the way, this is the only ECG I sent him, so he did not have then benefit of the others -- I don't think they would have helped) 

— Definitely NOT complete AV block because as you note the R-R intervals are NOT all the same. Instead, beats #2 and 3 both clearly occur earlier-than-expected — therefore this is to my eye definite evidence of at least some conduction.

— QRS morphology is definitely within the conduction system, given the picture-perfect rSR’ in lead V1 with terminal S waves in lateral leads I, V6. I don’t know the history … but if there was recent chest pain — then the ST segment in lead V1 is abnormal (should be slightly depressed) and those T waves in leads V2,V3 look taller-than-they-should be with RBBB conduction — so could represent possible hyperacute T waves … (T waves in leads III and aVF for beats #2, 3,4 also look abnormal, even accounting for fusion … ).

— Unfortunately, we don’t see what happened before beat #1 …. so I’m just guessing on this beat — but it is notable that we KNOW the P wave before beat #1 is NOT conducting (1st RED arrow in the long lead II showing an ultra-short PR interval) — and QRS morphology shows LAHB morphology not seen for the 2nd beat in leads II and III. I interpreted this different morphology in leads I, II, III (in view of the P wave too short to be conducting at all) as indicative of a fascicular escape beat from the left posterior hemifascicle.

— The P waves before beats #2 and 3 are definitely conducting (2nd and 4th RED arrows in the long lead II). I believe these beats are manifesting Wenckebach periodicity. Marriott always emphasized that in addition to progressively increasing PR interval until  a beat is dropped — AV Wenckebach manifests “PR/RP reciprocity”. By this, he meant that the shorter the RP’ interval — the longer the next PR interval — and vice versa. So as a result of beat #1 presumably being a left posterior hemifascicular escape beat — the 2nd P wave ( = 2nd RED arrow) has a shorter RP’ interval (measured from beat #1) — than we see for the RP’ interval between beat #2 and the 4th P wave — which is why the PR interval preceding conducted beat #2 is LONGER than the PR interval preceding beat #3.

— The R-R interval between ALL OTHER BEATS in this tracing is the same ( = 1100 msec) — therefore, I propose these are all escape beats arising from the left posterior hemifascicle — and that the reason for slightly varying QRS morphology between these beats, is that they are ALL manifesting slightly different degrees of fusion.

— I wasn’t quite sure how to represent a left posterior hemifascicular escape focus on a laddergram (never did that before … ) — because it is BELOW the AV nodal tier — yet not arising from ventricular myocardium … so after several trials, I decided on showing the RED circle arising from the bottom of the ventricular tier — as this is the easiest way to show variable degrees of fusion.

— I have NO IDEA of what a “pure” sinus-conducted QRS would look like — but I think beat #3 comes closest to pure sinus conduction — which is why you see the shortest distance traveling upward from the RED circle of beat #3.

— Although subtle — Note that the distance between the QRS and the closest P wave IS changing for each beat on this tracing. Therefore, I believe we have 2nd-degree AV Block with 2:1 AV conduction — showing potention for AV Wenckebach (that PR/RP reciprocity), therefore making this Mobitz I despite the wide QRS — with an accelerated left posterior hemifascicular escape rhythm and variable degrees of fusion — and, as mentioned earlier — the possibility of hyperacute T waves IF the history at all suggests a possible recent event.

— If you make a blog post on this, feel free to use this laddergram.

Again — I am NOT certain of the above … but that’s my best guess.

Take care — :) Ken




ADDENDUM (Added on 6/18/2021 by Ken Grauer): Peter Hammarlund questioned whether this 1st ECG might represent LA/LL Reversal. As per Dr. Smith — I did not know the history, and my focus was purely on the fascinating cardiac rhythm ( = My OVERSIGHT! ).


Yes — this DOES look like LA/LL reversal (as per Peter). I’ve reproduced the Learning Points from the November 19, 2020 post in Dr. Smith’s ECG Blog — in which I presented another case of LA/LL reversal (See Figure-1 below). This 1st ECG in today’s case is consistent with LA/LL reversal because:

  • Those dissociated P waves in lead I are larger than the P waves in lead II.
  • The QRS is surprisingly negative in lead III.
  • In the 2 follow-up ECGs in today’s case (these ECGs are shown above in Dr. Smith's discussion) — the QRS is positive in lead III, and the P wave returns to being larger in lead II than in lead I (as is expected with normal sinus rhythm).


NOTE: For additional details regarding the changes to expect when there is LA/LL reversal — Please check out my discussion at the above link in the Nov. 19, 2020 post.


Figure-1: Learning Points from the November 19, 2020 post in Dr. Smith’s ECG Blog — taken from another case of LA/LL lead reversal.




Friday, June 11, 2021

LBBB: Using the (Smith) Modified Sgarbossa Criteria would have saved this man's life

Case submitted and written by Dr. Jesse McLaren (@ECGcases), of Emergency Medicine Cases

Reviewed by Pendell Meyers and Steve Smith

An 85yo with a history of hypertension developed chest pain and collapsed, and had bystander CPR. The paramedics found the patient with ROSC and a GCS 7, and an ECG showing LBBB with possible lateral ST elevation. The patient was brought to the ED as a possible Code STEMI and was seen directly by cardiology. On arrival, GCS was 13 and the patient complained of ongoing chest pain. Vitals were HR 58 BP 167/70 R20 sat 96%. Below is the first ED ECG, labeled LBBB by the machine. Are there any indications to activate the cath lab?












There is sinus rhythm, LBBB and disproportionately discordant STE in V4 (ST/S = 2/6 = 0.33) that meets the Modified Sgarbossa Criteria. There is also a hint of concordant STE in V5 but it might be less than 1mm. There is also a hint of concordant STE in III.


In III, there is slightly less than 1 mm STE in the context of an isoelectric QRS. In LBBB, when the QRS is isoelectric (neither positive nor negative), the ST segment should also be isoelectric. 1 mm of STE here is highly suggestive of OMI.

Lead aVL has at least 1 mm of reciprocal STD in the context of a QRS which is under 4 mm. Thus, there is nearly 30% discordant STD, which is very specific for OMI (see both the derivation and the validation of Smith Modified Sgarbossa criteria - see validation results below).  


According to the validation of the Modified Sgarbossa Criteria, the traditional weighted Sgarbossa criteria only has a negative likelihood ratio of 0.51, which is insufficient to rule out Occlusion MI in a patient with a high pre-test probability. The Modified Sgarbossa Criteria has a better negative likelihood ratio (though still insufficient in a high risk patient), and a very high positive likelihood ratio.


Accordingly, in the algorithm by Cai et al for patients with LBBB and ischemic symptoms, the first indication for PCI is clinical: patients with hemodynamic instability or acute heart failure. Similarly, STEMI guidelines call for urgent angiography for refractory ischemia or electrical/hemodynamic instability, regardless of ECG findings. If the patient is stable, the next consideration is ECG criteria: Sgarbossa criteria of 3 or more (concordant STE or concordant STD in anterior leads), or Modified Sgarbossa Criteria (disproportionate STE). If none of these are present then proceed with serial ECGs, troponins, and bedside echo.




In this case, cardiology noted “old LBBB, negative Sgarbossa”, so they kept the patient in the ED for repeat ECGs and troponin levels. But the lack of traditional Sgarbossa criteria is not reassuring enough for such high pretest probability (elderly patient with chest pain, out of hospital cardiac arrest and LBBB), and the Modified Sgarbossa Criteria confirms Occlusion MI in this case.

 

The patient still had chest pain and the ECG was repeated one hour later, labeled LBBB by the machine. Are there any indications for cath lab activation?




Disproportionately discordant STE remains in V4 and has now progressed to V3 (ST/S = 3/11 = 0.27), while V5 is back to isoelectric. So there is now high pre-test probability + refractory ischemia + Modified Sgarbossa + dynamic ECG changes. But it didn’t meet traditional Sgarbossa criteria so the cath lab was not activated.

 

Thirty minutes later the first Troponin I came back elevated at 650 ng/L (normal <26), and bedside ultrasound found anteroseptal akinesia. The patient still had chest pain and a third ECG was performed. Any indications for cath lab activation?

An initial level of 650 ng/L represents a too rapid rise to be attributable only to Type II MI from brief cardiac arrest.  This ALONE is very strong evidence of acute coronary occlusion.




There is still disproportionate STE in V3-4, and there is also inferior concordant STE in III/aVF (or if you think the QRS in III is isoelectric, not technically concordant but still diagnostic), with reciprocal STD in aVL. So now there is high-pretest probability + refractory ischemia (with biochemical and echo confirmation) + Modified Sgarbossa + original Sgarbossa + dynamic changes. (Using Barcelona criteria V4 had discordant change of 1mm in a lead of 6mm or less, and the reciprocal STD in aVL had excessive discordant STD, but neither of these would have added anything to the modified Sgarbossa criteria in this case).

 

The cath lab was activated three hours after arrival in the ED, and found a 100% mid LAD occlusion and 90-95% RCA occlusion. Which was the culprit lesion?

 

Occlusion MI with both anterior and inferior injury can be from 1) occlusion of the LAD which wraps around the apex to supply the inferior wall (our colleague Emre Aslanger has shown us that the LAD does not have to literally be a type III "wraparound" LAD to supply the anterior and inferior leads' territories), 2) occlusion of a large RCA which supplies the inferior and lateral walls, or 3) occlusion in one of these arteries that interrupts collateral circulation to the other. 

Because only the inferior leads met Sgarbossa criteria on the third ECG, the RCA was felt to be the culprit lesion and the LAD was assumed to be chronic. So the RCA was stented. This is faulty reasoning because the anterior leads meet the better disproportional discordant STE criteria of the Modified criteria.

But the first two ECGs met Modified Sgarbossa Criteria in leads V3-4, and bedside echo found anteroseptal wall motion abnormalities corresponding with the LAD occlusion. The patient continued to have chest pain after the RCA was reperfused, so the LAD was then stented. But by this time the patient went into cardiogenic shock and passed away. As the discharge note acknowledged in hindsight, the inferior ECG changes resulted from LAD occlusion that compromised collateral circulation to a previously diseased RCA.

Meyers note: As Dr. McLaren points out, skilled ECG interpretation helps us see prospectively that the LAD was the clear culprit in this case. RCA occlusion typically should not produce OMI in V3, for example, unless there is chronic LAD occlusion with RCA-to-LAD collaterals. It is clear to me that the LAD is at least one culprit. The anterior leads are more diagnostic in this case than the inferior leads. It seems to me that the angiographers' lack of knowledge of the modified Sgarbossa criteria may have been a critical missing piece of information for this patient's care.


Learning points

1.      Refractory ischemia or electrical/hemodynamic instability are indications for the PCI. VF arrest is of course "electrical instability"!

2.      Relying on traditional Sgarbossa criteria can lead to reperfusion delays

3.      Modified Sgarbossa Criteria can help identify Occlusion MI and the culprit vessel, complemented with bedside echo

4.      Anterior and inferior STE can result from wraparound LAD occlusion, RCA occlusion, or occlusion of one compromising collateral circulation to the other


Sunday, June 6, 2021

A man in his 50s with chest pain and shortness of breath

Submitted by Ali Khan MD, written by Pendell Meyers


A man in his early 50s presented with exertional chest pain and dyspnea. He had family history of early CAD. Otherwise, no clear risk factors. Vitals were within normal limits. 

No prior ECG was available. Here is his triage ECG:

What do you think?


















This is yet another subtle inferior (and likely also posterior) OMI. There is a small and narrow QRS complex with reasonable axis and R wave progression, therefore the QRS cannot explain any abnormalities of the ST segment and/or T waves. The T waves in II, III, and aVF are subtly too large for their QRS, and the most important finding is that, in aVL, there is reciprocal STD and proportionally-large-volume T wave inversion (a "reciprocal negative hyperacute T wave"). There is also slight STD in I. The picture is not of perfect quality, but I believe there could be a hint of STD in V3-V4. Sinus brady at 60 bpm is another piece of evidence pointing to an RCA OMI, which supplies the SA and AV nodes, causing bradycardia and heart blocks during occlusion. 


Dr. Khan immediately understood this ECG and called his cardiologist asking for emergent cath. 

He also sent this picture to me in real time with no information at all, and I responded: "It is diagnostic of inferior OMI to me. But it will be a hard sell to most cardiologists." This was not meant to be critical of cardiology, but it does reflect my truthful expectation based on prior experience and cases that we are sent from all over the world. Most people, EM or cardiology or other specialty, cannot see this subtle STEMI(-) OMI.


Here is a zoomed-in picture of the inferior leads showing that it truly does not meet STEMI criteria (although humans have never been able to agree on measuring STE anyway):





Dr. Khan and the interventionalist on call agreed that this patient needed emergent cath. 

A complete acute occlusion (TIMI 0) of the RPL (right posterolateral artery), requiring 3 stents to restore TIMI 3 flow.

The high sensitivity troponin I returned at 1468 ng/L. No further troponins were measured.

Echocardiogram showed an inferior wall motion abnormality.


Here is his ECG after the intervention:


There are up-down biphasic T-waves in III and aVF.
There are enlarged upright T-waves in V2 and V3
These are all findings which reflect reperfusion of the inferior and posterior walls. 




The patient did well.



Learning Points:

Hyperacute T waves are not well defined, but involve an increased size/bulk/area of the T wave compared to the QRS complex, and compared to prior T waves if available. They of course cause reciprocal findings on the opposite wall.

aVL is a reliable reciprocal partner for lead III. When there is OMI in one of these two areas, the other lead will reliably show reciprocal findings. In this case, I need lead aVL (showing reciprocal negative hyperacute T waves) to be sure that the inferior T waves are hyperacute.

Check out these other cases of subtle inferior OMI where aVL is very helpful:










See this great post for some examples of high lateral OMI and compared to some false positives:





Recommended Resources