This is the conclusion to Wednesday’s post; if you haven’t seen it, I highly suggest checking out the original case description and initial ECG.
Wednesday’s case introduced a 59 year old woman with a chief complaint of intermittent occipital head pain for about 10 days. Recently, it had started radiating into her neck and upper back/shoulders. This was her initial ECG and it is clearly abnormal:
Figure 1. EKG on arrival at the ED.
The computer identifies it as showing left ventricular hypertrophy (LVH) with secondary ST and T-wave abnormalities (which we call a “strain pattern” or just “strain”). LVH with strain is quite common—in fact, it’s the predominant source of false-positive cath lab activations—but is that really what we’re looking at here? That’s a vital question because, if it’s not LVH w/ strain, then we’re probably looking at an inferior STEMI.
Does this tracing show a STEMI mimic? Or a true STEMI mimicking a STEMI mimic?
To start our analysis, I’d first caution anyone against trying to attribute ST/T abnormalities to a “strain pattern” in the absence of obviously large voltages or an old ECG showing a similar pattern. While Fig. 1 does meet a couple of different voltage criteria for LVH (which the computer is excellent at calculating—one of the few things I rely on it for), the voltages really aren’t that impressive. If you’re going to call LVH with strain based on a single ECG, you want it to be something obvious—like the tracing in Fig. 2.
Figure 2. LVH with strain and massive QRS voltages.
On a side note, several commenters noted that they were swayed against the diagnosis of LVH by the absence of high voltage in the precordial leads of Fig. 1. While they were right to be skeptical of LVH, it was for the wrong reasons. Seeing high voltage in only one plane (large complexes in the limb leads but normal complexes in the precordials, or vice-versa) actually doesn’t do much to rule-out the diagnosis of LVH. Fig. 3 shows an ECG where the precordial leads demonstrate huge QRS complexes but the limb leads are perfectly normal.
Figure 3. LVH with huge precordial QRS complexes but normal limb leads; there is also 2:1 AV-block.
Back to our case…
There are a couple of other findings that also weigh against a strain pattern:
- Even if we believe there is LVH, the ST-deviations, especially in III and aVL, seem slightly excessive when considered in proportion to the size of QRS complexes. There is no good rule for determining excessive discordance with LVH (that 25% one is okay as a rule-in criteria and better than nothing, but terrible at ruling-out STEMI), but a seasoned eye will note that this seems like just a tiny bit too much elevation with a T-wave in III that is just a bit too tall for the QRS size.
- There is a convex morphology to the ST-elevation in III that is slightly atypical for LVH.
- There is a down-up morphology to the ST-depression in I and aVL. While sometimes normal with LVH, this has to be approached as abnormal and ischemic until proven otherwise.
- There is concordant ST-depression in V2 and V3.
Figure 4. Concordant ST-depression in V2 and V3.
I’ve highlighted that last point because I think it’s the most important and pretty-much seals the diagnosis of STEMI for me. Since we’re already concerned about an inferior MI (based on the findings in III and aVL), and since inferior STEMI’s are often accompanied by posterior involvement, seeing right-precordial ST-depression consistent with posterior STEMI leaves me very confident that the ST/T abnormalities we see in III and aVL are indeed caused by an infero-posterior STEMI.
When we discuss the right-precordial leads in the setting of LVH on this blog    , it’s usually because the strain pattern often produces ST-elevation in V1 and V2 (and sometimes V3) that mimics STEMI (see Fig. 2). In those cases the ST-elevation is appropriately discordant and direct opposite the negative QRS complexes. It’s less common to see concordant right-precordial ST-depression in those leads when there is LVH.
Figure 5. LVH w/ strain and concordant ST-depression in V3. This is not a STEMI.
Figure 6. LVH w/ strain and concordant ST-depression in V2 and V3. This is not a STEMI.
When we do sometimes see concordant ST-depression with LVH, most of the time is due to chronic diffuse subendocardial ischemia from multi-vessel coronary artery disease or demand ischemia (as in Fig. 5 & 6). The key in those cases is that it’s usually associated with ST-depression in most of the limb leads and ST-elevation in aVR. Looking back on Fig. 1, the limb lead ST-depression in our case is confined to I and aVL and there is no ST-elelvation in aVR.
So that right-precordial ST-depression, combined with the other findings listed above, confirms that we are very likely looking at an infero-posterior STEMI. But we have another issue…
Note that I said it’s “very likely” that we’re dealing with a STEMI, not “certain.” Despite my confidence in the ECG findings, this still isn’t an easy diagnosis because things get much trickier when you take the patient’s symptoms into account. While acute coronary syndrome (ACS) was present on the differential for her initial presentation (hence, why we did the ECG), before we did the test it didn’t seem very likely that her symptoms were due to myocardial ischemia. In other words, her pretest probability (our assumed likelihood that she was experiencing a STEMI prior to performing the ECG) was pretty low.
Now, her pre-test probability of ACS clearly isn’t 0%, but it’s also not very high. As a result, it’s going to take a pretty convincing test to diagnose a STEMI in this patient. That’s unfortunate since we already determined that her ECG, while suggestive of STEMI, wasn’t 100% conclusive.
Figure 7. If a patient came in with toe pain and this ECG they’d still go for emergent PCI despite an incredibly low pretest probabilty. The ECG is so strongly indicative of acute anterior STEMI that it is capable of swinging us all the way from a pretest probabiltiy of almost 0% to diagnostic certainty. In the case of our 59 year old woman though, even though her pretest probability of ACS was a bit higher than someone with “toe pain,” the ECG is less clear-cut. While we are mildly confident that we are looking at a subtle STEMI, we are not certain because we have a hazy clinical picture combined with a non-pathognomonic ECG.
We’ve moved ACS way-up on the differential, but we haven’t clinched the diagnosis. [On a side note: The ECG is a very subjective test, and while an expert like Dr. Steve Smith would assign a lot of confidence to his interpretation and could very well diagnose STEMI here in-spite of a low pre-test probability, most providers could and should not.] When faced with a non-diagnostic ECG in the face of possible ACS, it’s almost always a smart move to repeat the ECG.
In our case I didn’t see this patient’ EKG until an hour after she arrived, but as soon as I did I expressed my concerns to the treating physician and requested a repeat EKG. To refresh your memory, here’s that initial tracing again, followed by the repeat:
Fig. 1 (reprising its role).
Figure 8. Repeat ECG 1 hour later.
Well, that’s a bit different. There are several important findings in Fig. 8:
- The ST-elevation in lead III has disappeared (normalization of the J-point).
- There are now a terminal T-wave inversions in lead III (reperfusion T-waves).
- The ST-depression in I and aVL has mostly resolved (normalization of the J-point).
- The ST-depression in V2 has resolved (normalization of the J-point).
- The T-waves in V2 and V3 are slightly taller and more symmetric (posterior reperfusion waves).
The most important of those changes is the new terminal T-wave inversions in lead III—these are reperfusion T-waves. Many providers are familar with Wellens syndrome affecting the anterior leads, but they are not aware that you can see similar T-wave inversions in other distubutions.
Wellens syndrome is due to spontaneous reperfusion of an anterior STEMI. When the patient’s LAD is blocked, it results in a typical anterior STEMI pattern…
Figure 9. Subtle anterior STEMI.
But then if that previously closed LAD spontaneously re-opens, the leads that used to show ST-elevation sometimes develop these classic-looking T-wave inversions described by Dr. Wellens.
Figure 10. Wellens T-wave inversions indicative of spontaneous reperfusion.
What Dr. Wellens didn’t mention in his original paper—and what is still not well known—is that you can see these sorts of T-wave inversions in any sort of STEMI. If a lateral STEMI suddenly reperfuses, you might see T-wave inversions in the lateral leads. If an inferior STEMI reperfuses, you might see T-wave inversions in the inferior leads (like we do here). And finally, if a posterior STEMI reperfuses, you can see T-wave inversions in the posterior leads. These posterior reperfusion waves are reflected on the standard 12-lead as subtly taller, upright T-waves in V2 and V3 (which we also see here).
Note the other important finding, listed in big text at the top of the Fig. 8: The patient’s pain, while present at a 3 out of 10 during EKG #1, was absent during EKG #2. That confirms that we are looking at reperfusion T-waves in EKG #2, which in-turn means that EKG #1 must have been showing an infero-posterior STEMI!
At this point in the patient’s course I was convinced but no one else was buying that this patient with a “headache” was actually experiencing an intermittent and spontaneously reperfusing STEMI.
Thankfully her troponin-I (ref <= 0.04 ng/mL) came back mildly elevated at 0.80 ng/mL right around that time, buying her some stronger consideration for ACS. She received aspirin, nitro paste, metoprolol, clopidogrel, and enoxaparin in the ED and preparations were made to admit her for an unstable angina/NSTEMI workup.
About 90 minutes after EKG #2 the patient began to complain of slight pain again and another EKG was recorded, including posterior leads:
Figure 11. Repeat ECG showing subtle infero-posterior STEMI (again).
Figure 12. Posterior ECG (V7–V9) showing no ST-elevation in the posterior leads.
The T-waves in lead III are now purely upright again—this is a phenomenon known as pseudo-normalization. Rather than being normal and reassuring, it is actually an indication of re-occlusion of the culprit artery. When dealing with reperfusion T-waves, an inverted T-wave becomes a good thing because it means the previously blocked artery is now open. A sudden reversion of an inverted reperfusion T-wave to a normal, upright configuration is usually associated with re-occlusion of that temporarily unclogged coronary artery.
As I was the electrocardiograph wires after EKG #4, the patient mentioned that her pain had completely resolved again, so of course I reattached her and ran another tracing:
Figure 13. Infero-posterior reperfusion pattern… again.
This EKG suggest that she is in the process of reperfusion yet again, though there is still some mild residual ST-depression. The most important thing that that the T-wave in lead III has returned to its flipped “reperfusion” morphology—a good thing!
For a better visualization of changes across these last three EKG’s over the course of six minutes, check out the gif below.
Figure 14. The continuum between inferior STEMI and spontaneous reperfusion.
Follow each lead individually:
- Lead III transitions from mild ST-elevation with an upright T-wave (STEMI) to an upright T-wave with no elevation to finally an inverted reperfusion T-wave.
- aVL goes from showing ST-depression with a steep downslope (reciprocal changes) to an isoelectric J-point with very little downslope to the T-wave.
- V3 initially shows concordant ST-depression with a small T-wave (posterior STEMI), but eventually evolves to showing a isoelectric J-point with a T-wave that is slightly taller than before (mirror image of a posterior reperfusion wave).
Just because we can, here’s a similar animation of all five ECG’s over the patient’s ED course.
Figure 15. The process of intermittent injury and spontaneous reperfusion over the patient’s ED course.
The patient was admitted to telemetry where her troponin-I values (ref <= 0.04 ng/mL) every 6 hours trended as:
- 0.80 ng/mL
- 0.75 ng/mL
- 0.79 ng/mL
Echo the next morning showed normal left ventricular size and function with no regional wall motion abnormalities and a preserved EF of 55%. There was no echocardiographic evidence of LVH. That day she was transferred to the CCU at a nearby hospital where she experienced an uneventful course and underwent coronary angiography a couple of days later, as planned.
Figure 16. Unremarkable left coronary system.
Figure 17. The right coronary system shows a culprit lesion in the RCA.
Cath showed a culprit lesion in the right coronary artery (RCA), perfectly consistent with the pattern of infero-posterior injury and reperfusion we were seeing on the patient’s ECG’s! She received a single drug-eluting stent and had a good outcome with no significant loss of LV function or other sequelae.
Right coronary artery status-post PCI and a single stent.
It might seem crazy that this patient was experiencing anginal pain in the back of her head, but that’s actually a well documented (though not really well known) presentation of myocardial ischemia known as “cardiac cephalgia.” We’ll discuss this a bit more in a couple of days but I think we’ve covered more than enough ground for today.
Let me know if you have any questions in the comments or in response to our links to this case on Facebook and Twitter!