Friday, July 27, 2012

Paced Rhythm with Dynamic ST segments and Concordant T-waves - High Risk ACS

A 77 yo woman with h/o pacer for unknown reasons (no medical records available) and no h/o coronary disease presented with 24 hours of constant chest pain.  She decided to call 911 because she just couldn't walk any more.  Paramedics found her with normal vital signs, chest pain 8/10, and the following ECG:

There are p-waves followed by a paced QRS.  It is slightly unusual that there is an RBBB pattern.  Most pacers are in the apex of the RV and thus have all negative QRS in V1-V6.  Here the RBBB pattern suggests that the pacing wire is on the left (see below).  More importanly, although the ST segments are discordant in V4-V6, the T-waves are concordant.  This suggests ischemia, usually Non-STEMI in this context.  That the ST segments are discordantly elevated at the same time the T-waves are inverted suggests injury in these leads, but is not diagnostic.
When there is RBBB pattern in pacing, there are two distinct possibilities: 1) the lead has perforated into the LV or 2) there is a lead going through the coronary sinus to the LV.  In this case, fortunately, x-ray confirmed the latter.

I was worried that she had suffered a very large MI because her pain had lasted so long.  It did decrease to 3/10 after sublingual NTG.  She arrived in the ED and had the following ECG at 1030:
There are again p-waves with a paced QRS (one p-wave comes early, PAC), and the ST elevation in V4-V6 is largely resolved, and the T-wave inversion deeper. Along with the resolution of chest pain, this is nearly diagnostic of ACS.  Later, the pacer was interrogated and found to be 100% atrial synchronous biventricular pacing.

The troponin I returned at 12 ng/ml.  I was relieved that it was not much higher.  I gave her ASA, clopidogrel, and heparin and she went immediately to the cath lab and had an active ostial LAD lesion with a long hazy calcified segment that could not be stented.  She went for 2-vessel CABG.

Here is an ECG from 1530:
There is some continued dynamism of the T-waves

Echo showed an inferoposterior and a probable anteroapical wall motion abnormality.

She did well.


Here is an addition to this post in response to a question/comment:  "I am sorry to say that I do not find any literature about that topic. Therefore I kindly ask, why you comment on the dynamic ST-segment changes in this patient as a typical sign of acute MI. Could you comment on this or give additional literature."

There is some previous literature on this, but not a lot.  I have summarized the most recent article below.  There is also one from the 90’s, referenced  here.1

There is little reason to suspect that paced rhythms act any differently than left bundle branch block or, if the activation is from the left, like RBBB.  In my experience of multiple cases (see here for a couple), and of cases of PVC’s (see here and here), concordant ST segments and excessively discordant ST segments (> 25% of the preceding S-wave or R-wave, as measured at the J-point) are very good for injury.  Even better are dynamic ST segments, as in this case.  There is no way for pacing alone to result in dynamic ST segments; one must assume dynamic ischemia until proven otherwise.

This is the more recent article: Maloy et al.(Maloy 2011)2 assessed the Sgarbossa criteria1 in 57 cases of paced ECG in biomarker-diagnosed AMI (not angiographic occlusions), and compared them to 99 troponin-negative paced controls who were otherwise of comparable age and sex.  There were no cases of concordant ST elevation.  Concordant ST depression in one of leads V1-V3 had sensitivity of 19% (95% CI 11–31%) and specificity 81% (95% CI 72–87%).  For ST-segment elevation >5mm and discordant with QRS complex, the sensitivity was 10% (95% CI 5–21%) and specificity 99% (95% CI 93–99%).  The authors do not publish the numbers of patients who met each criterion.  Only one patient (a control) met both criteria, so it appears that the sensitivity of the overall rule was 29%.  Given that only about 30% of troponin-diagnosed AMI are STEMI-equivalents (occlusion or near occlusion, as opposed to Non STEMI), this is surprisingly high sensitivity for a group that had no concordant ST elevation.

1.            Sgarbossa EB, Pinski SL, Gates KB, Wagner GS. Early electrocardiographic diagnosis of acute myocardial infarction in the presence of ventricular paced rhythm.  GUSTO-I investigators. Am J Cardiol 1996;77(5):423-4.
2.            Maloy KR, Bhat R, Davis J, Reed K, Morrissey R. Sgarbossa Criteria are Highly Specific for Acute Myocardial Infarction with Pacemakers. The western journal of emergency medicine 2011;11(4):354-7.

Sunday, July 22, 2012

A Child with Blunt Trauma

A 6 yo girl had significant trauma from  an MVC, with head injury (initial GCS 10, but no intracranial bleeding) and mild orthopedic injuries.  She was intubated.  A FAST exam was normal (no pericardial or peritoneal fluid).  As part of her workup, this ECG was recorded:

What do you think?

Interpretation:  There is sinus tachycardia, with right bundle branch block (RBBB).  This would be an unusual baseline finding in a child, and therefore myocardial contusion should be highly suspected.  There appears to be lead misplacement, V2 to V3, as the R' wave disappears in V2, then reappears in V3.  V1 and V2 (labelled V3) have the typical downsloping ST depression of RBBB, with a typical negative T-wave.  However, V3 (labelled V2) has ST segment elevation, which is not normal for RBBB.  Notice lead aVL also has ST elevation.  Normal RBBB has no ST elevation in any lead, though sometimes there is an exception when there is abscnce of R'-wave, as in V2 here.

Course: A CT of the head, neck, chest, abdomen and pelvis showed no other unanticipated injuries and she was admitted to the ICU.

Blunt cardiac injury my result in:
1) Acute myocardial rupture with tamponade
2) Valve rupture (tricuspid, aortic, mitral)
3) Coronary thrombosis or dissection (and thus Acute MI) from direct coronary blunt injury
4) Dysrhythmias of all kinds.
5) Myocardial contusion (edema and hemorrhage in the myocardium) which may result in dysrhythmias, blocks (especially RBBB as here), and poor cardiac contractility, including wall motion abnormalities.  Localized bleeding from contusion could cause hemopericardium even without rupture.

The literature on myocardial contusion is confusing, as, unless there are hemodynamic or dysrhythmic events at presentation, it is difficult to predict who will have complications except for those who have echocardiographic abnormalities.  There is no reference standard for diagnosis in patients without such echocardiographic abnormalities, and there are no definite clinical findings (with the probable exception of ST elevation or depression on the ECG) which consistently predict adverse events in succeeding hours or days.  Patients may have positive biomarkers (troponin) from the related stress of multisystem trauma, and low levels of troponin probably do not predict adverse events any better than the patients clinical condition.  Hemodynamic instability in trauma is usually due to bleeding, but if ultrasound shows poor contractility, then this may be due to cardiac contusion.   More commonly, however, ECG and/or echo wall motion abnormalities are due to pre-existing disease, especially coronary disease complicating trauma.

The finding of RBBB on this ECG is highly suggestive, if not diagnostic, of myocardial contusion.  The much more subtle finding of the ST elevation, not noticed by any physician caring for her, is further suggestive of significant contusion.  There is some old literature which I read in the 1980's (but cannot find now), that suggested that ST abnormalities due to myocardial contusion have the highest risk of any ECG finding for adverse outcomes and a higher risk than positive cardiac biomarkers (at that time, the biomarker was CK or CK-MB).  

Inpatient stay: She quickly awoke and was observed for a couple days, treated for minor orthopedic injuries, and recovered well from the head injury.  No further ECG, troponin, or echocardiogram was done because she was asymptomatic, and had a normal rhythm and rate.  She was discharged to home feeling just fine.


Three weeks later, shortly after having been physically active (bouncing on a trampoline), she was found unresponsive.  She was pulseless, with a narrow complex tachycardia on the monitor.  In the ED, ultrasound showed hemopericardium with tamponade.  An ED thoracotomy was done and the pericardium drained and internal compressions continued.  A cardiothoracic surgeon was called and together they found a 2 mm hole in the LV high in the circumflex distribution (corresponds to the STE in aVL) with a large amount of surrounding thinned, necrotic muscle.  This was too macerated to be repaired and she could not be resuscitated.

Needless to say, all involved were emotionally devastated.

Could the myocardial damage have been diagnosed by echo?  If it had been, could this outcome have been anticipated?  I am aware of no similar case reports in the literature (of delayed myocardial rupture after cardiac contusion), but I have not done an exhaustive review.  If a wall motion abnormality had been found, what could have been done to prevent the rupture?  If anticipated, would a prescription of bedrest have prevented it?  Was there traumatic circumflex artery occlusion that contributed?  Could an angiogram have found this?  And, then, so what?  Could subsequent myocardial rupture have been predicted or suspected?   

Before this case, I doubt any such outcome could have been anticipated, even though myocardial contusion (due to RBBB and STE) is evident.

After this case, were I to see a similar one, I would:

1) do a formal echocardiogram on anyone with new significant ECG abnormalities.
2) consider an angiogram if there is ST elevation (as here) and a wall motion abnormality
3) measure troponins, as a very high peak troponin would confirm large territory of contusion
3) limit physical activity if there is concern for a large area of contusion.

I do think this ECG gave a solid clue to the pathology.  ST elevation is an alarming finding of myocardial contusion. ST elevation is difficult to notice in the context of RBBB and one must specifically look for it to notice it.  In this case, it appears that the over-reading cardiologist also did not see it.

Finally, this may just be an incredibly rare event that none of us will ever see again. 

Monday, July 16, 2012

Electrical Alternans? From Archives of Internal Medicine.

Here is a nice case from Archives of Internal Medicine.  Surprisingly, one can read it without a subscription.  For the ECG, page down and click on the "View Large" link.



Tuesday, July 3, 2012

pdf full text of article on early repol vs. anterior STEMI

This article is now in print in July annals.

Smith SW et al.  Electrocardiographic Differentiation of Early Repolarization From Subtle Anterior ST-Segment Elevation Myocardial Infarction.  Annals of Emergency Medicine 60(1):45-56, July 2012.

Here is the full text of the article in pdf.

Monday, July 2, 2012

LVH and New ST-segment elevation: False Positive or True Positive STEMI?

A 72 year old man fell from standing and had head trauma.  He was initially disoriented, but cleared.  BP was 80/50 prehospital.  On arrival, it was 90/70.  There was a scalp laceration.  Head CT was negative.  Patient stated he had had experienced some chest pain prior, but had no active chest pain.  He has a past history of hypertension, chronic renal insufficiency, prior anterior MI with proximal LAD stent, and cardiomyopathy with EF of 35%.

Here is his ED ECG:

There is sinus rhythm at rate of about 85 bpm.  There is profound LVH, with a V2 S-wave of 55 mm.  There is typical ST elevation of LVH ("secondary repolarization abnormalities) in V2 and V3.  There are Q-waves of his previous MI in V4 and V5, with ST elevation in V4 (discordant) and also V5 (concordant).

V4 and V5 make this ECG worrisome for STEMI, but by no means diagnostic.  Whenever there is this much voltage, it can greatly distort repolarization and result in a false positive cath lab activation.  The fact that the patient has no ongoing chest pain is very important.

A previous ECG from 11 days prior was obtained:

Here there is bradycardia at a rate of 42, with a PVC.  The LVH is present, but voltage is not as profound.  The ST segments are not nearly as abnormal here as they are on the presentation ECG.

Now there is much more worry for STEMI.  However, the difference in heart rate and voltage, and the hypotension and/or syncope, can entirely account for the change in ST segments.

He went to the cath lab and had no change in his coronary arteries from the previous angiogram.  Troponin I peaked at 0.50 ng/ml (baseline values were about 0.15 ng/ml from chronic renal failure).  Echocardiogram showed no wall motion abnormality, with normal biventricular function, LVEF of greater than 55%, and severe concentric LVH.

Here is his ECG then next AM.
No great change from presentation.
His presentation was explained by an excess of antihypertensive medications.

Now, you might think reading this as negative for STEMI can only be done by the retrospectoscope.  But this case was sent to me as an unknown (by an undergraduate student who will start med school in the fall, and who is a whiz with EKGs!).   This was my response (again, without knowing anything about the outcome):

"It looks like ischemia superimposed on LVH, but not necessarily due to ACS.  Could be demand ischemia from a number of causes.  I would not be surprised with either a positive or negative troponin, but I would be surprised if the trop was very high (> 10 for troponin I).  What was the troponin?  Was there  ultimately a cath?  What was the outcome?  It’s a good one for posting if we get all the details."

I only add this in order to show that a careful informed reading of even the most scary ECGs is possible.

What would I do in this case?

1. I would have obtained an emergent echocardiogram to look for new wall motion abnormality.
2. Had the patient had active chest pain, or not had the change in voltage or heart rate, I would have activated the cath lab.


Recommended Resources