Tuesday, September 8, 2015

Low HEART score. Acute LAD occlusion. Detected only by analysis of subtle ECG.

A completely healthy 39 year old woman with no cardiac risk factors had sudden onset of bilateral trapezius pain that radiated around to her throat.  It resolved after about 5 minutes, but then recurred and was sustained for over an hour.  She called 911.

EMS recorded these ECGs:

Time 0:
In V2-V4, there is ST elevation that does not meet STEMI "criteria," of 1.5 mm at the J-point, relative to the PQ junction.  But there are also unusually Large T-waves
Time = 13 min
T-wave in V2 is now taller and fatter, the ST segment is more straight.
T-wave in V3 is no taller, but it is fatter due to a straighter ST segment
This is highly suspicious for early LAD occlusion
Time = 24 min
No significant change
Time = 25 min

These prehospital ECGs were lost and not seen.

The patient arrived in the ED.

The pain completely resolved after nitroglycerine, and just before this ECG was recorded in the ED (she had been pain free for moments only when this was recorded):
Computer read: Normal ECG.
However, T-waves are still unusually large; the computer almost never sees this.
The T-wave in V2 is smaller, but so is the QRS; so the proportion is not different.
QTc is 444 ms.
STE 60 V3 = 1.5 mm, R-wave amplitude V4 = 15 mm
Formula value is 23.1, which is close to being an LAD occlusion value.  It is below the cutoff of 23.4, but above my safe value of 22.0

This patient has a nondiagnostic ECG by most rules.  However, with attention to subtleties, it is very worrisome.

If you use something like the HEART score:
1. H  History: She has atypical pain (trapezius) (score = 0-1, depending on the physician)
2. E  EKG: a negative ECG (score = 0 or 1, most would say 0)
3. A  Age: = 0
4. R  Risk factors = 0
5. T:  Troponin = 0 [first troponin (contemporary, not high sensitivity) was less than the level of detection).
Total HEART score = 0-2.  Risk of 30-day adverse events is less than 1.7%.   Some might send her home.

But maybe she has an acute LAD occlusion that will get even worse.

The providers did a bedside echo and even used speckle tracking to look for strain:

I think maybe there is an anterior wall motion abnormality, but this is very difficult.  They read it as normal.

Here are a couple shots with strain, or "speckle tracking" on ED Echo:

To, me these look like anterior wall motion abnormality, but I showed them to one of our ultrasound fellows who is very interested in this.

She said:

This is a tough one. I see what you mean, initially when I looked at the image, I also thought there was an anterior wall motion abnormality.  But then on closer inspection, I suspect that maybe the anterior wall is just not being tracked well. In systole, you can see the anterior wall come down and outside of the area that is being tracked (more so than the other tracked walls). Even though the strain values are a little off in the graph (so is the posterior wall) it is still a value range (about -18) that would be considered non-ischemic by the cardiology literature, I believe.  I have been wrong before though! So it is possible that I am misinterpreting the clip. If it were me, I would get values at the level of the mitral valve, papillary muscles, and apex (all in PSS axis). Also, narrowing the area being tracked helps the walls get recognized much better.

As I wrote, the first troponin was below the Level of Detection.

She remained pain free, and was admitted without further serial ECGs.  

When in doubt, one should always get serial ECGs.  Bedside echo is not enough.

At time = 240 minutes (4 hours), the second troponin returned at 1.15 ng/mL.  That prompted recording of this ECG:
Back to normal for this patient.  This demonstrates that all ST elevation of the previous ECGs was ischemic, not normal.  She was having a transient STEMI, briefly.

It is very lucky that she spontaneously reperfused her LAD.  It did not progress to full STEMI with loss of the anterior wall, as in this case.

Also, persistence of a pain free state does not guarantee an open artery.  See this case.

A formal contrast echo was done at this point:
Normal estimated left ventricular ejection fraction, 65%.
Regional wall motion abnormality-distal septum and apex.

She was treated medically for NonSTEMI, pending next day cath, which showed  ulcerated plaque and a 60% thrombotic stenosis in the LAD distal to the first diagonal.  It was stented.

Learning Points:
1. Always get serial ECGs when there is any doubt about what is going on.
2. Always find and look at prehospital ECGs.  They give extremely valuable information.
3. Hyperacute T-waves remain for some time after reperfusion of an artery.  I always say that "you get hyperacute T-waves both 'on the way up' (before ST segment elevation) and 'on the way down' (as ST elevation is resolving).
4. Wall motion abnormalities are very hard to see, even with advanced Speckle Tracking technology.  They require a great contrast exam and expert interpretation.
5.  This case does not demonstrate it, but a wall motion abnormality may disappear after spontaneous reperfusion (see this case).
6. Patients with transient occlusion may manifest only transient STEMI on ECG.  Subseqent troponins may be all negative and subsequent formal echo may be normal.  See this case


  1. I have noticed also new upright T wave at 13 min, 24 min EKGs , which didn't present in 1st tracing (Time =0) , does this finding support LAD occlusion? if we did not lost those tracing

    1. I would not call this a new upright T-wave, as they are all terminally inverted. There is perhaps a bit of increased size of the positive part of the wave.

  2. great post doc Smith, love serial ecgs,it's meticulous observation, just as you would observe a concussed pt, or an asthmatic being given nebs, if the ecg is playing up often the heart is too, cheers

  3. What about the Q waves?

    Please understand that I'm a medic who tries hard to interpret this site. :)

    1. I don't see any pathologic Q-waves. A normal one of <30 ms in V4, but no others

  4. I think this is a case where too much knowledge can be a bad thing. We don't do bedside ECHOs for WMA in our hospital so we can't be distracted by it. I'd look at the first ECG and go 'Normal potassium, huh? Jeez, those look like hyperacute Ts! We'd better be pretty suspicious here!' I'm surprised she didn't have serial ECGs.

    Sometimes, broader decision rules and less attempt to stratify in a clever way can be better. It all comes down to specificity and sensitivity.

    1. The bedside echo has a good positive predictive value, but not negative. That is, good specificity if you see a WMA, but poor sensitivity.

  5. Hello Dr. Smith,

    regarding V3 in the prehospital ECGs... from time 0 to time 13 there was a significant loss of R wave... and then an addition minimal loss from time 13 to 24. This also makes the bulkier T wave seem even more hyperacute relative to the much smaller R wave. This would seem to be a significant finding, no?

    Thank you,

    1. Dave,
      I would say yes, except that it is isolated to lead V3 only, which makes me think this aspect is artifactual. See what I mean?
      Thanks, very perceptive.

  6. Martin Schönemann-LundSeptember 11, 2015 at 4:46 PM

    Was that a high sensitivity troponin coming back below LoD for the first blood sample?

    1. Martin,
      This is what I wrote:
      5. T: Troponin = 0 [first troponin (contemporary, not high sensitivity) was less than the level of detection).
      Total HEART score = 0-2. Risk of 30-day adverse events is less than 1.7%. Some might send her home.

      I always distinguish between the two.


  7. Hey Steve, great post and I like your evaluation by HEART score to highlight the danger of clinical scores. 1. They should never override clinic gestalt. 2. Using them correctly always requires some subjective input by the operator. With HEART score I find people underestimate atypical stories even though they often have some high risk features as in this case. The no "known" risk factor portion is also problematic. Patients who are "healthy" with no recent check-up make the "risk factor" portion indeterminate and perhaps the rule is not valid... Great case and will pass it on to our residents.

  8. Excellent case. Scary in fact given the transient nature of her symptoms.

    The strain images that are submitted are normal. They are however circumferential strain, which early in ischemia is often normal, even when there is significant impairment in longitudinal strain. This makes anatomic and physiologic sense when you consider the orientation of the myocardial fibers, with the subendocardial and subepicardial fibers being longitudinally arranged, and the mid-wall fibers being radially arranged. Thus, the impaired longitudinal strain resulting from ischemia affecting the subendocardial layers first, is the most sensitive strain finding in ACS.

    Great case.


Recommended Resources