Tuesday, October 3, 2017

Chest pain and a "normal" ECG

This is another case written by Pendell Meyers (who is helping to edit the blog and has many great recent posts)


A 45 year old man was driving to work when he experienced acute onset sharp left sided chest pain with paresthesias of the left arm. He received aspirin en route via EMS, and no EMS ECGs are available.

Here is his ECG on presentation to the ED, approximately 45 minutes after onset of pain, but with pain improving since onset:
What is your interpretation?

I sent this ECG with no clinical information to Dr. Smith. He replied:

"If acute chest discomfort, then this is likely an acute proximal LAD occlusion
STE in V2 with a Q-wave
Subtle ST depression in III, aVF. down-up in aVF
Subtle STE in aVL"

Baseline ST elevation is common in the anterior leads, but to be called normal STE it must have a normal QRS. The QRS in this case is NOT normal. There is very poor R-wave progression, with no R-wave in V2, and tiny R-waves in V3 and V4. The Q-wave in V2 is followed by STE and large (though not quite hyperacute) T-waves. Therefore this STE cannot be called normal.

We would then look for reciprocal changes in the inferior leads to further confirm the hypothesis that this is abnormal STE in the anterior leads. There we find a relatively normal QRS with abnormal ST segments and T-waves which do not agree with their QRS complexes (which supports them being acute changes). There is minimal STD in III and aVF. There is a fully inverted T-wave in III, and "down-up" morphology in aVF.

This is very subtle, and only long-time readers and experts will see this as a easily-recognizable pattern. Luckily there are some equations which approach the skill of experts for this particular question, which only require a few simple measurements. There is a Q wave in lead V2 which technically prevents the use of these formulas, but for the sake of interest here is what happens if the formulas are applied anyway:

Both formulas predict LAD occlusion, even with the most conservative (small) values for STE 60 V3 used.

The initial ECG was interpreted as "normal" by the computer and "no acute ischemic changes" by the ED physician. The initial troponin was negative (drawn approximately 1 hour after onset of symptoms).

Side note: contemporary troponin drawn 1 hour after acute LAD occlusion should usually be negative, unless the event has been going on longer than the patients symptoms. The assay at my institution, for example, is frequently negative until 4-6 hours after acute coronary occlusion.

The patient had been chest pain free since shortly after arrival, and was sent to the observation unit for serial ECGs, troponins, and CTCA. One hour after arrival, while the patient was awaiting CTCA, he started to have returning chest pain.

A repeat ECG was recorded with pain 2/10:
Not much change. Maybe even improvement in the reciprocal changes in the inferior leads.

This was brought to the ED physician who interpreted this as unchanged from initial ECG.

10 minutes later the patient became diaphoretic with worsening pain. Another repeat ECG was performed:
No caption needed!

Shortly after this, the patient had a VF arrest. He was shocked out of VF twice, and just before the second shock this ECG was captured:
Although this is often erroneously called "Torsades (de Pointes)," it is really ventricular fibrillation.   It should have been shocked at least 10 seconds ago. 

After the second defibrillation the patient had an organized rhythm:
Bradycardic escape/agonal rhythm, with large ST deviations. Beats 4, 6, and 7 are narrow, as the rhythm is trying to resume from above the ventricles.

This rhythm reportedly produced no palpable pulse, and CPR was continued. 30 seconds later, however, the patient began spontaneously moving and CPR was discontinued.

A repeat ECG was done:
Obvious anterolateral wall STEMI. (Proximal LAD occlusion with dramatic ST Elevation)

The patient was intubated and taken to the cath lab. On arrival his BP was 70s/40s, so an intra-aortic ballon pump was placed. They found a 100% acute thrombotic proximal LAD lesion. Here are the cath images before and after intervention:

Serial ECGs after PCI show improvement in ST segments, but also some ongoing ST elevation implying active injury after cath (this could mean downstream small vessel occlusion):
There are also hyperacute T-waves.
Smith likes to say: "Hyperacute T-waves occur 'on the way up,' and 'on the way down.'"
This means that they occur shortly after onset of occlusion, but also may be the last remaining sign of ischemia after ST elevation resolves (after reperfusion).

T-waves less hyperacute, though ST elevation remains

Echo showed EF ~40%, with severe hypokinesis of the mid anterior septum and apex. Troponin T was recorded at 1.41 (very high), and was still climbing but no further measurements were available.

He recovered and was discharged 5 days later.

2 days after discharge, he awoke at 3am with sudden severe chest pain. He presented immediately to the ED and had this ECG:
There is large anterolateral STE with hyperacute T-waves, and reciprocal STD in III. This is diagnostic of re-occlusion. This ECG cannot be present 7 days after a single persistent acute occlusion.

The patient was taken back to the cath lab, where 100% proximal in-stent rethrombosis was found and treated.

Repeat ECG after PCI:
LAD occlusion resolving.

Take Home Points:

1) Computers are not currently sufficient to detect important ECG abnormalities
2) Dangerous LAD occlusion can be very subtle
3) "Normal ST elevation" in the anterior leads requires a normal QRS and R-wave progression
4) Use the LAD occlusion formulas
5) Don't confuse VF with torsades


  1. Great case! I think you mean "in-stent rethrombosis" as the second presentation EKG is a new STEMI within days, not a gradual narrowing of restenosis over months.

    Dan Lee

  2. Great case... There are Already ST dep TWIs in inferior leads and formed Q waves in septal leads; can we still use the formula?

    1. Good point! The TWI in the inferior leads wouldn't prevent use of the formula, but according to the derivation criteria, Q wave in V2 would. I have added that caveat above. So that makes this more easily recognizable. I would add that some people won't identify V2 as having a pathologic Q wave, so it's nice to see that the formula still may be helpful in some scenarios in which it wasn't exactly meant for.

  3. Excellent case! One thing I noticed immediately between the first and second ECGs was a marked attenuation of the S wave in V2. It doesn’t get much attention, but I have found this to be a sign of evolving anterior STEMI.



    1. I see what you're describing, but I simply haven't yet noticed (and probably haven't been looking for) attenuation of the S wave as a marker. If you have evidence or cases showing this please send them our way, thanks!

  4. "Echo showed EF ~40%, with severe hyperkinesis of the mid anterior septum and apex."
    You might mean HYPOkinesis.
    Excellent post!

    1. yes of course, thanks for the correction!

  5. Nice case. I also agree that many physicians and paramedics way overcall TDP instead of coarse VF. Would be interesting to publish a deep dive into this.


Recommended Resources