Conclusion to 83 Year Old Male: Shortness of Breath

Last week we presented the ECG of a patient experiencing progressively worsening shortness-of-breath over the course of a day and some marked ECG abnormalities. If you haven’t done so already, it would probably be a good idea to check out the original post first. Strap in, this is going to be a thorough discussion.

Here again is the patient’s initial ECG:

83yo M - SOB, Ill Appearing

Not a STEMI-equivalent.

This ECG shows:

  • Sinus tachycardia at a rate of 96 bpm.
  • First-degree AV-block (PRi of approx 240 ms).
  • Left anterior fascicular block (LAFB), resulting in…
  • Left axis deviation (mean frontal QRS axis approx -60 degrees), and…
  • Persistent large S-waves in V5 and V6.
  • Old anterior infarction (Q-waves in V1-V3, almost no R-wave in V4).
  • Moderate ST/T-wave abnormalities in a pattern of diffuse subendocardial ischemia.
    • ST-depression in I, II, aVL, aVF, V3-V6; ST-elevation in aVR, V1.
    • Frontal ST-vector of approx 200 degrees, towards the right shoulder.

The #1 take-away from this case is that this ECG is not a STEMI-equivalent. STEMI’s need immediate revascularization and, except in some specific cases, their first stop should be the cath lab. Diffuse subendocardial ischemia is quite the opposite in that it should usually be managed medically in the ED first. Only after initial stabilization and evaluation will select cases proceed to near-immediate or early catheterization.

It has been a popular topic of ECG teaching for the past few years but diffuse ST-depression with ST-elevation in aVR does not necessarily equate with left main coronary artery (LMCA) occlusion. It can be seen in the setting of significant obstruction of the LMCA or significant multi-vessel coronary artery disease, but those can be very different from true 100% occlusion of the left-main.

What you’re seeing when you encounter the above pattern is ischemia affecting the majority of the circumference of the left ventricle, except that it only affects the subendocardial aspect of myocardium, not its full thickness.

Going back to basic cardiac anatomy, there are three major layers that we consider when examining the myocardium in cross-section: the endocardium (inner layer, like the endothelial lining of blood vessels), myocardium (the actual cardiac muscle), and epicardium (outer layer, contiguous with the visceral pericardium, tough to differentiate the two).

Histology slide demonstrating the layers of the atrial myocardium. Click image for source.

Histology slide demonstrating the layers of the atria. The ventricles feature a much thicker myocardium with a thinner endocardium. Click image for source.

When we talk about “subendocardial ischemia,” what we are referring to is ischemia that affects only the heart muscle just below the endocardium, but not extending the full-thickness of the myocardium. This is a little confusing because, if you look at the above slide, you may get the impression that located “below” the subendocardium is the empty space in the cavity of the ventricle. What we are actually discussing is the myocardium that’s adjacent to the endocardium. Terms like “up” and “down” can get a bit confusing when dealing with a hollow and roughly spheroid-shaped organ like the heart.

If you were visualizing localized subendocardial ischemia in the distribution of a non-dominant circumflex artery you might imagine it affecting an area depicted by the yellow shading in the image below.

Localized subendocardial ischemia. Image modified from original by Vince DiGiulio. By Patrick J. Lynch, medical illustrator (Patrick J. Lynch, medical illustrator) [CC-BY-2.5 (], via Wikimedia Commons

Localized subendocardial ischemia.
Image modified by Vince DiGiulio. Original by Patrick J. Lynch, medical illustrator / CC-BY-2.5, via Wikimedia Commons.

 We don’t talk much about localized subendocardial ischemia because there’s no consistent pattern that it presents with on the ECG. As our site’s collective mentor Dr. Stephen Smith states so often: “ST-depression does not localize subendocardial ischemia.”

What he means is that even when ST-depression due to subendocardial ischemia appears to show a particular distribution on the ECG (inferior, lateral, etc…), it does not correlate well with actual findings of stenosis on cath. You might think you’re seeing “anterior ischemia” on the ECG but that doesn’t guarantee that there’s a significant stenosis in the LAD causing it.

From my own personal experience, I’ve very rarely seen cases of localized ST-depression that I’ve suspected were due to focal subendocardial ischemia. Perhaps focal subendocardial ischemia isn’t even a distinct electrocardiographic entity, but that’s a digression…

True ischemic ST-depression on the ECG really falls into two broad categories: ST-depression that is reciprocal to STEMI and ST-depression due to diffuse subendocardial ischemia. For more discussion on this and the five sub-categories that fall under those two headings check out this post from Dr. Smith.

So, if subendocardial ischemia affects only the inner portion of the heart muscle, what do we call it when the entire muscle thickness is involved?

Transmural ischemia.

Localized transmural ischemia.

Localized transmural ischemia. Image modified by Vince DiGiulio. Original by Patrick J. Lynch, medical illustrator / CC-BY-2.5, via Wikimedia Commons.

Acute transmural ischemia presents as a STEMI on the EKG and, unlike localized subendocardial ischemia, there is excellent correlation between the distribution of the ST-elevation and the region(s) of myocardium involved. Above shows the area of myocardium that would be ischemic during STEMI due to occlusion of a non-dominant left circumflex artery.

Which brings us back to the topic at hand: diffuse subendocardial ischemia.

Localized transmural ischemia. Image modified by Vince DiGiulio. Original by Patrick J. Lynch, medical illustrator / CC-BY-2.5, via Wikimedia Commons.

Diffuse subendocardial ischemia. Image modified by Vince DiGiulio. Original by Patrick J. Lynch, medical illustrator / CC-BY-2.5, via Wikimedia Commons.

This is what our patient in the scenario presented last week is experiencing: diffuse (or “circumferential”) subendocardial ischemia. The majority of the subendocardial region of his myocardium is ischemic but it’s not localized and it doesn’t affect its full thickness. What causes this to happen though?

The reasons why the inner portion of the myocardium is more prone to ischemia than that near the epicardium is a topic of much research with several prevailing theories but, in the most basic sense, the muscle there is a victim of two major factors:

  1. The subendocardium has a less robust blood supply than the subepicardium. Recall that the blood supply to the myocardium penetrates down from the major coronary arteries located at the epicardium, so the blood perfusing the subendocardium has to travel further through smaller arteries and arterioles.

    Image by Patrick J. Lynch, medical illustrator / CC-BY-2.5, via Wikimedia Commons.

  2. Compared with the subepicardium, the subendocardium experiences more pressure exerted on it by the blood in the left ventricle due to its greater proximity, reducing blood-flow through those already tiny arteries. This is especially true in the setting of high end-diastolic pressure since the myocardium is perfused during diastole.

Alright. So the subendocardium is more prone to ischemia, but why would it ever present as circumferential ischemia affecting the entire left ventricle? This is where the left main coronary artery finally comes into play.

If you fully occlude the left main coronary artery the patient will experience a global STEMI. Example of this are located here and here and here and here and here and here. Below is the region of myocardium experiencing transmural ischemia in the event of a true LMCA occlusion. In this illustration it’s the best-case scenario of a right-dominant coronary artery distribution, providing at least some perfusion to the infero-posterior wall of the LV.


Diffuse transmural ischemia. Image modified by Vince DiGiulio. Original by Patrick J. Lynch, medical illustrator / CC-BY-2.5, via Wikimedia Commons.

Patients rarely make it to the hospital alive with this degree of transmural ischemia and even fewer survive to PCI. True LMCA occlusion is a rare entity even in prehospital care.

When there is less than 100% occlusion, however, some blood-flow makes it through to the LAD and LCx and the left ventricle experiences diffuse subendocardial ischemia.  We see more of these cases because the patients have a better chance of survival thanks to the trickle of blood making it past the obstruction.

Now life would be easy if all subendocardial ischemia correlated with acute subtotal obstruction of the left-main. Unfortunately that’s not the case and the same exact type of ischemia can be produced in a number of settings.

The first is the “left-main equivalent” of multi-vessel coronary artery disease. It’s called that for a couple of reasons:

  • First, if you have a couple or three major vessels with CAD then the same area of myocardium (i.e. most of it) is going to experience subendocardial ischemia. Your myocardium doesn’t care if there’s one giant artery blocked or several of its subsidiaries; all it knows is that it’s not getting enough perfusion.
  • Second, like a single significant stenosis in the LMCA, multi-vessel stenoses causing ischemia are usually treated with bypass surgery. Neither scenario does well with attempts at stenting so in both cases patients end up needing CABG.

That’s not entirely straight-forward but at least it’s all issues primarily due to coronary artery disease, right? This is where “elevation in aVR” really falls apart: Anything that causes a global mismatch between oxygen supply and demand by the heart will produce diffuse subendocardial ischemia and the same exact ECG pattern as LMCA stenosis/multi-vessel CAD. Here are some examples in no particular order:

  • Anemia, causing decreased O2 delivery.
  • Hypoxia, causing decreased O2 delivery.
  • Sepsis, causing increased O2 demand.
  • Severe hypertension, causing increased O2 demand and decreased subendocardial perfusion.
  • PE, causing decreased O2 delivery and increased O2 demand.
  • Acute hypertensive pulmonary edema, causing decreased O2 delivery and increased O2 demand.
  • Tachycardia, most often AF or SVT, causing increased O2 demand.
  • Generalized inflamatory states, causing increased O2 demand and decreased O2 delivery.
  • COPD or CHF exacerbation, causing decreased O2 delivery and increased O2 demand.
  • Hypotension, due to decreased O2 delivery.

Throwing a significant chronic LMCA stenosis or diffuse CAD into the mix with any of the above will only worsen the supply/demand mismatch and increase the amount of ST-deviation.

Also, a somewhat similar pattern of ST-deviation is also seen in the setting of LVH with “strain” and in patients on digoxin, though with somewhat different T-waves.

As you can see, diffuse subendocardial ischemia really is fraught with pitfalls. In the case of our case presentation, almost everyone who commented fell for the misdirection and assumed the patient needed emergent cath. Andrew Merleman deserves a special shout-out for absolutely nailing the diagnosis and management in his comments on the EKG Club page. Rock on!

In reality the patient was experiencing and acute episode of bilateral pneumonia. The clue here was his increased temperature, gradual onset of the SOB, and productive cough. I also wouldn’t blame anyone for thinking it was a CHF exacerbation—the real key was just to avoid assuming that the EKG findings mandated immediate cath. The patient was admitted to the hospital, treated with BiPAP and antibiotics, and did well. Troponin-I levels (reference < 0.04 ng/mL) were trended and peaked around 1.0 ng/mL.

This patient did indeed experience ACS, but it was demand ischemia due to supply/demand mismatch; known as a type II MI. He certainly had coronary artery disease (as evidenced by his old anterior MI on the initial ECG and significant ST-depression in response to physiologic stress), and maybe even had a stable LMCA stenosis, but it was not the cause of his symptoms and did not mandate invasive investigation or treatment. Once his pneumonia and work of breathing got under control the ECG normalized quite as bit and returned to his baseline.

83yo M - SOB, Ill Appearing 02

ECG 2 Days Later. Not normal but normal-enough and matching his baseline.

If you can bear it, stay tuned for an upcoming post where we discuss just what qualifies as a “STEMI equivalent” and which patients with diffuse subendocardial ischemia need immediate cath. This is enough writing for one night…

If you want to read more about true LMCA occlusion vs. less severe obstruction, here’s the best resource.

Also, for those wondering just what the heck is going on with the R-wave in lead V2 in this patient, there will be a post on that as well. It’s due to the left anterior fascicular block, not old posterior MI.

Let me know in the comments if you have any questions or I wasn’t clear about. Subendocardial ischemia is a HUGE topic with many layers so it’s hard to do it justice in one blog post.



  • dawakhan says:

    A great case. Added a lot to my knowledge. I would like to ask about the sensitivity and specificity of st elevation in avr and lmca occlusion.

    • Vince D says:

      Life in the Fast Lane does a great review of the numbers available here:

      There’s two main problems.

      First, the term “occlusion” implies 100% blockage of the left-main; these studies only examined the presence of significant stenosis. They’re very different things and there’s actually a lot of cases of true 100% LMCA occlusion that show no ST-elevation in aVR or even franke ST-depression. The pattern of ST-deviations caused by true LMCA are very different from the diffuse subendocardial ischemia I’m discussing here. The former presents as a STEMI and this is an NSTEMI.

      Second, those studies are highly impacted by which patients they chose to include. If you only looked at patients with typical MI symptoms who experienced sudden-onset, crescendoing CP, then diffuse ST-depression with elevation in aVR would actually perform pretty well at identifying those with significant LMCA stenosis. It’s certainly a useful sign when used that way.

      The problem is that those presentations of diffuse subendocardial ischemia are few and far between and the utility of aVR elevation gets diluted out by all of the non-ACS causes I list above. For every case of true LMCA obstruction I’ve encountered I have had to wade through several dozen cases of non-LMCA-stenosis aVR elevation caused by those other causes.

      Hopefully that’s the take-home of this post: We shouldn’t abandon aVR but it’s certainly time to temper the enthusiasm. The field of emergency medicine is now too aware of aVR elevation andI saw dozens of comments yelling “LMCA occlusion!” or “cath lab now!” when I first shared this case on Facebook.

3 Trackbacks

Leave a Reply

Your email address will not be published. Required fields are marked *