A man with chest pain off and on for two days, and "No STEMI" at triage.

 Written by Kaley El-Arab MD, edits by Pendell Meyers and Stephen Smith

A 61-year-old male with hypertension and hyperlipidemia presented to the emergency department for chest tightness radiating to the back of his neck that has been intermittent for the past day or two. Here is his triage ECG which was obtained at 20:34 during active pain.

What do you think?

This ECG was read as “No STEMI” with no prior available for comparison. It is true this ECG does not meet STEMI criteria (there is 1.0 mm STE in III, and possibly 0.5 mm in aVF), but there is clear evidence of OMI findings on this ECG. Leads II, III, and aVF have hyperacute T waves with reciprocal T wave inversions and down-sloping STD in I and aVL; along with ST depressions in the precordial leads that is maximal in V2-V4, this is diagnostic for inferior and posterior OMI.

Meyers note: The slightly complicating feature of this ECG is that the QRS likely represents LVH, with high positive voltage in I and aVL, and relatively high negative voltage in inferior leads. Some may have thought that these ST segment shifts were attributable to the LVH, but these shifts are out of proportion to the trained eye. As far as I am aware, we do not yet have great studies of the quantified "appropriate discordance" of LVH. We have tried to study it before, only to find that we cannot obtain enough cases of OMI with simultaneous high voltage in the leads with OMI findings, to study it. Armstrong et al. studied this with insufficient methodology (in our opinion) and advocated for a 25% ratio, but in our experience this is clearly far too high of a percentage. 

I sent this ECG to the Queen of Hearts (PMcardio OMI), and here is the verdict:

You can subscribe for news and early access (via participating in our studies) to the Queen of Hearts here: https://bit.ly/omi-queen-form

Because his ECG is read as "No STEMI", he was placed in an overcrowded waiting room without a monitor, which is basically normal at all times nowadays. It is unclear if he received aspirin at triage.


The patient’s chest pain spontaneously resolved before he was evaluated and has a repeat ECG obtained at 22:12 obtained shown below.

 

This ECG is more difficult. The ST depressions in the precordial leads are more subtle and the inferior T waves are less pronounced. 

Meyers note: To me, this ECG alone would be nonspecific for OMI, but there are ST depressions that are relatively equal from V2-V6, etc. I think it looks ischemic, but I would not have been sure of OMI from this ECG alone. In context, of course, it is clear that the patient is reperfusing, as pain has dissipated and the diagnostic findings of OMI have become more nonspecific.

Smith: Although the first ECG is indeed diagnostic of OMI, the change in this 2nd ECG, associated with resolution of chest pain, makes the first ECG even more diagnostic of an OMI that then later reperfused spontaneously.

The Queen of Hearts agrees:

Around this time his initial high sensitivity troponin I resulted at 231 ng/L.


Soon afterward, the patient’s symptoms return along with lightheadedness, bradycardia, and hypotension. The following is the patient’s third ECG which was obtained at 22:37.

 

The precordial leads again show STD from V2-V6, and this time perhaps it is again more maximal in V3-4 suggesting posterior OMI again. However the inferior leads show evidence of reperfusion especially in lead III which shows terminal T wave inversion that is highly specific for reperfusion of inferior OMI. The patient has also developed sinus bradycardia, which may result from right coronary artery ischemia to the SA node. Overall, with return of pain and possible active posterior OMI, obviously we must consider this reocclusion until proven otherwise.

The Queen of Hearts sees it of course:

Still none of these three ECGs meet STEMI criteria. The patient’s second troponin results at 273 ng/L and third troponin results at 320 ng/L. 

The patient is started on epinephrine infusion for cardiogenic shock and cardiology took the patient to the cath lab.


During angiogram in the cath lab, the patient suffered two episodes of ventricular fibrillation for which he was successfully defibrillated. 

Angiogram showed a culprit lesion of 100% stenosis to the right coronary artery and 100% stenosis of the right posterior descending artery, both with TIMI 0 flow. Two stents were placed with resultant TIMI 3 flow.


Echocardiogram the following day showed a left ventricular ejection fraction of 52% (+/- 5%) with hypokinesis of the basal-mid inferior and inferoseptal myocardium. Troponins peaked at approximately 20,000 ng/L. He improved clinically and was discharged home two days later.


What do you think the final diagnosis was? "NSTEMI" of course.

Do you think we discussed this patient's 2-3 hour delay to reperfusion in our quarterly "STEMI meeting"? Of course not. There is no place for discussion of delays for deadly NSTEMIs that can be found on the initial ECG. Instead we discussed 5 minute delays for the STEMI(+) OMI patients.

Learning Points:

LVH can make it more difficult to see OMI on the ECG, but it can still be done with training.

ST depression maximal in V1-V4, without a QRS abnormality clearly causing it, in the setting of ACS symptoms, is very concerning for posterior MI until proven otherwise.

The Queen of Hearts seems really really good at finding subtle OMI. Stay tuned for upcoming studies showing this. Can you imagine if she had been used at triage in this case?

Of course, don't forget that ACS with ongoing ischemia despite medical management, and ACS with electrical instability are indications for emergent reperfusion even in the absence of any ECG findings.

The STEMI vs. NSTEMI paradigm doesn't care if you have imminently deadly acute coronary occlusion, unless you have STEMI criteria, ongoing ischemia after hours of delay, or until you start to have shock or arrest from it. Even if it could have been identified with better ECG training or any other modality (clinical suspicion, echo, emergent CT coronary angiogram, etc). Just another NSTEMI.

Smith comment: Here is another case in which LVH in the limb leads (as opposed to precordial leads) led to the interventionalist cancelling the cath lab activation:

See full case here: ST changes due to limb lead LVH?

I sent this ECG to the Queen of Hearts (PMcardio OMI), and here is the verdict: