Monday, October 28, 2013

Cardiac arrest: even after the angiogram, the diagnosis is not always clear

A woman in her 40's who was healthy, except for hypertension, was at work when she suddenly complained of neck and shoulder pain and then collapsed.  It was witnessed, and CPR was performed by trained individuals.  She was found to be in ventricular fibrillation and was defibrillated 8 times without a single, even transient, conversion out of fibrillation.

Fine ventricular fibrillation
She received 2 mg epinephrine, 150 mg amiodarone and underwent chest compressions with the LUCAS device.

She arrived in the ED 37 minutes after 911 was called, with continuing CPR.  She was immediately intubated during continued compressions, then underwent a 9th defibrillation, which resulted in an organized rhythm at 42 minutes after initial arrest. The following 12-lead ECG was recorded at 11 minutes after ROSC.  The patient had a combined respiratory and metabolic acidosis (as we commonly find in those with prolonged arrest), and a K of 4.1, at the time of the ECG.  Mg was 1.6.
The rhythm is nearly regular, but there are no P-waves (it is too regular to be atrial fibrillation).  One might be tempted to think this complex is very wide, that the entire complex, as seen in lead II, is QRS.  However, the QRS is barely wide, if at all.  The remainder of what appears to be QRS at first glance is due to ST segment deviation.  (See image with lines below).  The tall R-wave in lead V1, with S-waves in V5 and V6 is consistent with RBBB or incomplete RBBB, depending on the exact QRS duration.  Once the end of the QRS is correctly identified, then the presence of ST deviation becomes apparent: there is ST elevation in aVR and V1, and diffuse ST depression (in I, II, III, aVF and V3-V6). 
Here is the annotated ECG from above, with lines drawn.
The end of the QRS is best seen in leads V4 to V6.  I have drawn a line through the end (actually, my software missed it so that the real end is a fraction of a box to the right). I extend this line down to the lead II rhythm strip at the bottom, so that I know on lead II where the end of the QRS is.  Then I can draw 4 other lines to find the end of the QRS in all leads.  Here is another excellent example of this.

So, in this patient with incessant ventricular fibrillation that has been converted, there is now a supraventricular rhythm with RBBB.  There is ST elevation in aVR, but also in V1, and diffuse ST depression.  STE limited to aVR is due to diffuse subendocardial ischemia, but what of STE in both aVR and V1? 

[Many would say that such ST elevation in aVR is diagnostic of left main occlusion.  I repeat that ST elevation in aVR is not diagnostic of left main occlusion.  Here is an article I wrote: Updates on the ECG in ACS.  The last section is a detailed discussion of the research on aVR in both STEMI and NonSTEMI.  If you want to understand aVR, read this.]

The additional ST Elevation in V1 is not usually seen with diffuse subendocardial ischemia, and suggests that something else, like STEMI from LAD occlusion, could be present.  In left main occlusion, by blocking flow to both the anterior wall (LAD) and posterior wall (circ), the ST depression of posterior ischemia could theoretically diminish the ST elevation of anterior ischemia and leave only V1 with significant ST elevation (Nikus, et al. see below).  I have never seen this, but it is possible.  Alternatively, it is a variant of diffuse subendocardial ischemia, with STE in V1 reciprocal to ST depression in inferior and lateral leads.

The ST elevation in V1, with RBBB pattern and inverted T-wave also suggests Brugada syndrome, but does not fit with the remainder of the ECG.  

Cardiac arrest can cause diffuse subendocardial ischemia, usually transient (it often resolves as time goes by after ROSC).  The patient was resuscitated for 50 minutes, and then this ECG was recorded:
Sinus rhythm.  ST segment abnormalities are almost all resolved.
Was this:
1) ACS with ischemia and spontaneous reperfusion?  Or
2) The ischemia of an ECG post-arrest?
Only an angiogram can tell for certain.  Or can it?

The patient was taken for an angiogram and had an 80% LAD lesion, but it could not be definitely determined whether this was an acute thrombotic lesion or a chronic stable lesion.  It was stented.

Here is the post cath ECG
T-wave inversions consistent with anterior MI, but not diagnostic.

The troponin I peaked at 8.1.  An echocardiogram on day 3 showed no wall motion abnormality (but of course, these can resolved with reperfusion, and the more time it has to resolve from "stunning", the more likely it is to be resolved).

Also, anterior MI could result from 1) ACS, but also from 2) severe ischemia due to combination of a hemodynamically significant LAD stenosis + severe hypotension during cardiac arrest.  In spite of all the evidence, we are still left in the dark!

Here is a 24-hour ECG:
Developing T-wave inversions, still consistent with anterior MI, but not revealing of its mechanism

Here is a 48-hour ECG:
Now there is a very long QTc, at 571 ms.  Does the patient have a long QT disorder, or is this post-ischemia?


It could not be determined definitively whether this was primary ACS with arrest, or primary ventricular fibrillation with subsequent demand ischemia.

Therefore, to be safe, an internal defibrillator was placed.

The patient was discharged neurologically intact.


1.  Smith SW. Updates on the Electrocardiogram in Acute Coronary Syndromes. Current Emergency and Hospital Medicine Reports (2013) 1:43–52. 

2. Nikus KC, Eskola MJ. Electrocardiogram patterns in acute left main coronary artery occlusion. J Electrocardiol. 2008;41(6):626–9.


  1. The link to your paper results in an error. I can't tell if the link is bad or if it's a problem at Springer.

  2. Thanks for another fascinating case. All the way through I was expecting you to reveal another medical problem that had caused the arrest. One of those chicken and egg things I guess. What could have happened that made the 9th defibrillation successful, other than reperfusion of LAD? Would be interesting to interrogate the ICD in a year or so for arrhythmias. Must've been some high quality CPR for no neuro defecit with that much down time, 1st responders should be proud. Not sure if she would have qualified for the ICD in the UK, although would clearly be a special case for consideration.

    1. Thanks, Charles! From the info I have, I would have thought STEMI also, but I did not see the angiogram and am not an expert at interpreting that, so I can't be certain and must defer to the uncertainty of the cardiologists.

  3. Thanks for your interesting case. Following up with the serial EKG, could we conlude an anterior STEMI with the progression of T waves from positive to deep inverted? And the progression of cardiac enzymes also? Sometimes, Could we end up with an autolyzed thrombus/Prinzmetal induced STEMI diagnosis?

    1. Yes! There is no doubt that there could have been a thrombus which lysed or Prinzmetal's Angina. The ECG is consistent with such a mechanism. But I must defer to the ones who interpreted the angiogram that it is sufficiently likely to NOT be ACS that they had to place an IVCD.
      Steve Smith

  4. Very interesting case, thanks for posting. It doesn't count for much, but just out of curiosity, do you know if the patient received therapeutic hypothermia?

    Also, the final ECG is labelled as occurring 24 hrs later when I believe you mean 48.

    1. Vince,
      Yes, she did undergo hypothermia. Thanks for alerting me to my misprint which I will fix!


Recommended Resources