Sunday, August 30, 2020

A 31 year old with Diabetes and HTN complains of bilateral arm tingling and headache

This ECG was texted to me with the message "A 31 year old with Diabetes and HTN complains of bilateral arm tingling and headache."

There is high lateral ST Elevation and inferior reciprocal ST depression.
There is also STE in V2.

The computer calls it a STEMI.

What do you think?

STE in I, aVL and V2 is a pattern associated with "Mid-anterolateral OMI," which is seen with OMI of the first Diagonal.  See more of Mid-anterolateral OMI

I wrote back: "I think this is a false positive due to LVH.  PseudoSTEMI.  I can't tell you exactly why.  It just looks like it.  ECGs are often like faces to me.  I recognize them." 

More explanation: It does not meet any criteria for LVH, but just has the "look."  As I think about it now, it is the STE and upright T-wave in V2, with T-wave inversions in V5 and V6.

Furthermore, one should always be suspicious when the symptoms are so non-specific, especially when the patient is so young.

I sent the ECG to Pendell and he answered: "It seems like very high chance of false positive for LVH. The history would allow me to proceed with that assumption until I do some more investigation.  If it was chest pain it would be more difficult to go with my gut on that."  So we had exactly the same interpretation.

Then my friend wrote back with more clinical info:

He had a persistent bad headache and his BP actually went up to 220/120. Negative head CT and negative troponin x 2. I tried metoprolol and oral hydralazine with no luck and finally nicardipine at 10 mg/hr got him to about 170/90 and pretty comfortable which I was happy with. Never chest pain but had to treat as hypertensive emergency. EKG exactly the same on repeat.

Comment: the fact of the very high BP is very supportive of LVH.  If he had texted me that, I would have been more certain of a false positive.  And then 2 "negative" troponins, although that wording allows for many different troponin patterns. 

In fact, severe hypertension by itself can lead to greatly increased oxygen demand and type 2 acute MI, sometimes with ST Elevation.  See this case of Type 2 STEMI due to severe hypertension.  Here are more Type II STEMI.

Follow up:

"Some follow-up on my scary EKG from last week - no change on next day EKG, significant LV hypertrophy on echo, equivocal abnormal stress test vs artifact so he got sent to cardiology and had a clean cath. All for want of a prior EKG. At least I got the definitive answer. Amazing how scary an EKG can seem in isolation."

One more comment, and this is a difficult concept to comprehend: Even if the troponins had returned with an acute MI profile, the absence of any evolution on the ECG from one day to the next would prove that these ECG findings are not a result of that MI.  Any ECG finding that is a manifestation of ischemia will change over time.  One would have to say that the MI did not manifest on the ECG at all.

MY Comment by KEN GRAUER, MD (8/30/2020):
The 12-lead ECG in today’s case was texted to Dr. Smith for his opinion. For clarity — I’ve reproduced this tracing in Figure-1. The patient was a 31-year old man with a history of diabetes and hypertension — who presented with “bilateral arm tingling and headache”. His BP in the ED was 220/120 mm Hg. No prior tracing was initially available.
  • HOW did YOU interpret ECG #1?
  • WHAT features about this case support Dr. Smith’s impression that ECG #1 is not indicative of an acute MI?
  • That said — WHAT features about this case are worrisome?

Figure-1: The initial ECG in this case (See text).

My THOUGHTS about ECG #1: The rhythm is sinus at 75-80/minute. Regarding intervals — the PR interval is normal and the QRS complex is narrow. The QTc may be of borderline duration. The frontal plane axis is normal (about +40 degrees). Voltage criteria for LVH are not satisfied. Regarding Q-R-S-T Changes:
  • There appears to be a narrow (albeit fairly deep) Q wave in lead aVL. (I thought this was most likely to be a normal septal Q wave).
  • R wave progression is normal — with transition (where the R wave becomes taller than the S wave is deep) occurring normally, here between leads V3-to-V4 (albeit the upper portion of the R wave in lead V4 appears to be cut off).
  • The most remarkable findings on this tracing relate to the appearance of ST segments and T waves. There is definite ST elevation in high-lateral leads I and aVL. There is also 1-2 mm of J-point ST elevation in lead V2.
  • There is coving of the ST segment in each of the inferior leads, and in leads V3-thru-V6. This is followed by fairly deep and symmetric T wave inversion in the inferior leads, and in leads V5 and V6 (with a suggestion of T inversion also in lead V4).

Putting the Findings in ECG #1 Together — IF this patient had presented with new-onset, worrisome chest pain — I would clearly be concerned about an acute evolving STEMI. After all, there is significant ST elevation in both high-lateral leads (leads I and aVL) — and, there is J-point ST depression with a near mirror-image reciprocal picture between leads III and aVL. ST coving with fairly deep T wave inversion is seen in each of the inferior leads, and in leads V5,V6 (with ST coving also in V3,V4).
  • BUT — the patient is a younger adult, and there is no mention of any chest pain.
  • That said — we should assume that the high-lateral ST elevation with potential reciprocal ST depression in the inferior leads is acute until we can demonstrate otherwise.

NOTE: There are some reasons to suspect that the findings in ECG #1 are not acute (Figure-2).

Figure-2: I’ve labeled some findings from Figure-1 (See text).

Among the reasons I thought ECG #1 might not reflect an acute MI were:
  • In addition to not having any chest pain — the clinical presentation in this case was malignant hypertension in a diabetic patient with longstanding hypertension. Many of these patients have marked LVH associated with ST segment coving and significant T wave inversion (not unlike what we see in ECG #1).
  • As noted earlier — voltage criteria for the ECG diagnosis of LVH in this 31-year old man are not seen in ECG #1. That said — sensitivity of the ECG for detecting LVH is imperfect at best (generally under 60%). False negatives are common. Moreover, despite not satisfying textbook criteria for LVH voltage — one gets the distinct impression that there is “extra QRS amplitude” in the form of surprisingly deep inferior lead S waves (given substantial height of the R waves in these leads). And clinically — many (if not most) patients with diabetes and longstanding hypertension who present to the ED with a BP = 220/120 mm Hg will have LVH.
  • As I’ve often emphasized in Dr. Smith’s ECG Blog — the finding of ST-T wave changes consistent with LV “strain” in a patient with “the right disease” (ie, longstanding hypertension, now presenting as malignant hypertension) — dramatically increases the likelihood of true LVH. This is true even when voltage criteria for LVH are not seen on ECG.
  • ST-T wave changes of LV “strain” most often manifest in one or more of the lateral leads (For more on LVH — See My Comment at the bottom of the page in the June 20, 2020 post). But instead of seeing ST-T wave changes of LV “strain” in lateral leads — some patients manifest a “mirror-image” of strain in anterior leads. I believe the unusual shape of the elevated ST-T wave in lead V2 in Figure-2 (within the BLUE rectangle) reflects LV “strain” in this patient with marked LVH. Doesn’t the mirror-image of this ST elevation in lead V2 (within the RED rectangle in Figure-2) look exactly like LV “strain” usually looks in lateral leads?
  • Finally — the small RED arrow in lead I of Figure-2 points to a most distinct J-point notch. While it clearly is possible for patients with acute infarction to manifest J-point notching — this notching, together with the shape of the gently upsloping, elevated ST segment we see in lead I resembles the picture commonly seen in repolarization variants.
  • BOTTOM Line: The patient in today’s case has malignant hypertension. He did not have chest pain. While fully acknowledging that I was not certain he was not having an acute event — I thought the above considerations could clearly explain the ST elevation in leads I, aVL and V2 — as well as the ST coving and T wave inversion in multiple other leads. Had this been acute lateral infarction — rather than the same ST coving with deep T wave inversion that we see in the inferior leads — I would have expected a different shape for the ST-T waves in lateral chest leads V5 and V6 (if not some ST elevation in these leads). I was pleased to learn that Echo confirmed significant LVH — and, that cardiac cath confirmed normal coronary arteries.
  • P.S. — In addition to malignant hypertension, the “bilateral arm tingling” this patient experienced suggested the possibility of hyperventilation (superimposed on his malignant hypertension). It is good to remember that hyperventilation itself is a cause of ST depression and/or T wave inversion! It turned out that this was not the case for this patient — because apparently, his ECG the next day was unchanged (whereas the T wave inversions should have resolved if hyperventilation was the cause).

Friday, August 28, 2020

3 days of shoulder and chest pain, and now cardiogenic shock

I was texted these ECGs.

"Bad chest pressure with severe left shoulder pain 3 nights ago.  Then SOB and nausea the next day.  Now appears to be in cardiogenic shock."

(Later review showed systolic BPs in the range of 55 to 83.  So she was quite hypotensive.)

First recorded at time zero:
There is sinus rhythm. Rate of only 70 suggests some beta blockade.
The QT is very long.
There is T-wave inversion in inferior leads, suggestive of reperfused or subacute MI.
There is a Q-wave in III, so this may be subacute
There is ST depression in V2-V4.
20 minutes:
Again, very long QT.
Now, T-waves are upright in inferior leads, inverted in aVL.
The QRS is the same, so this is not lead placement.
This is highly suggestive of re-occlusion (Pseudonormalization of T-waves).
The very long QT is strange.

They did a POCUS cardiac echo:

Anterior wall is contracting and thickening.
Lateral Wall (lower right) is not contracting or thickening.
Posterior wall (at bottom of screen) also is not contracting or thickening. 

Aside: Some insist on calling this posterior wall "lateral," even though it is clear that it is anatomically posterior.  This is because if you oriented the heart with the RV directly right and the LV directly left, then that wall would indeed be lateral.  But the heart is oriented with the RV more anterior and LV more posterior, so that wall really is closest to the back (posterior).  You also can see that this orientation of the heart also renders the anterior wall a bit lateral.

Whatever you call this wall, that it is posterior is important, because it means the ST Elevation vector of that wall will be directly opposite V2 and result in ST depression in V2, NOT ST elevation in lateral leads I, aVL, V5, or V6!!

This insistence on calling it lateral is due to a paper by Bayes de Luna --]

Case Continued

I could only see the ECGs, not the echo.

I learned that the patient is on Sotalol for control of PVCs.  This explains the long QT.  All electrolytes were normal.

Paraphrasing Smith response: "There are dynamic T-waves inferiorly and ST depression in V3.  This is ischemia until proven otherwise.  By the ECG, it is not necessarily subacute, but could be.  However, cardiogenic shock usually takes some time to develop, so it is probably subacute."

Then I was told that the troponin I returned at greater than 50,000 ng/L.  This can only be due to STEMI.

I said "activate the cath lab."

Another ECG was recorded at 70 minutes:
Inferior T-waves are again inverted, and upright in aVL

Here is the angiogram:
What do you see?
Here I annotate it:
This shows 100% occluded circumflex (red arrow) and a 90% stenosis of the LAD (Yellow arrow). The LAD was thought to be not thrombotic, but a chronic tight stenosis.

Both were opened:
Beautiful open arteries

Formal Echo
--Severely decreased left ventricular systolic function with an estimated EF of 20%.
--Regional wall motion abnormality--basal to mid anterolateral, basal to mid inferolateral, and apical lateral akinesis.
--Regional wall motion abnormality--mid to apical anterior hypokinesis.
--Elevated left-sided filling pressure based on Doppler parameters (grade 2 diastolic dysfunction.)

ECG from 2 days later:
Atrial Fibrillation now.
Inverted Reperfusion T-waves inferior

Learning Points:

1. Circumflex occlusion may have minimal ECG findings.

2. When one artery is occluded, then the heart depends on others (in this case, the LAD) to perfuse contracting myocardium.  A fixed stenosis in that other artery, especially in the context of hypotension from the occlusion of the first coronary artery, can lead to ischemia and very poor LV function and worsening shock.

3.  Cardiogenic shock and ACS is an indication for the cath lab, even if you don't think there is OMI.

4.  ST depression in V2/V3 is posterior MI until proven otherwise.

5.  A troponin I greater than 5000 ng/L is almost always due to OMI (there was some thought that this might be takotsubo, but takotsubo does not often have a troponin I greater than 5000 ng/L).

6. Sotalol prolongs the QT

7. There is new data showing better outcomes when bystander lesions (non-culprit) are stented.

MY Comment by KEN GRAUER, MD (8/28/2020):
Dr. Smith highlights a number of important lessons to be learned from today’s case. I limit my comments to some additional points of interest.
  • For clarity — I have consolidated the first 3 ECGs in this case in Figure-1.

Figure-1: The first 3 ECGs shown in today’s case (See text).

My THOUGHTS on ECG #1: We are told that the patient in today’s case had an episode of severe chest pain 3 nights prior to admission. This was followed the next day by increasing dyspnea — and, by the time the patient was seen in the ED (when ECG #1 was obtained) — the patient was in cardiogenic shock.
  • The rhythm in ECG #1 is sinus at ~65-70/minute.
  • The PR interval is normal and the QRS complex is narrow. As per Dr. Smith — the QTc is extremely long. Selecting lead aVL as that lead in which we clearly see the limits of the QT interval — I measure a QT ~540 msec. Correcting for a heart rate of ~65-70/minute suggests the QTc ~550-560 msec, which is markedly prolonged.
  • There is generalized low voltage. QRS amplitude does not exceed 5 mm in any of the 6 limb leads — and it is well under 10 mm in all 6 of the chest leads, despite normal standardization (See Figure-2 below).
  • There is no chamber enlargement (although given how reduced QRS amplitude is in all 12 leads — assessment of ventricular enlargement is probably not possible on the basis of ECG criteria).

Given the clinical presentation of chest pain 3 days earlier, followed by  development of cardiogenic shock — the KEY to interpretation lies with assessment of Q-R-S-T Changes:
  • Baseline artifact makes it problematic to interpret ST-T wave changes in leads I, II and III. We only see 2 complete complexes in these leads — and the 1st of these 2 complexes is completely distorted by artifact. Based on the appearance of the 2 complexes that we do see in lead aVF — it does look like there is a relatively large Q wave in lead III (ie, nearly equal in size to the tiny R wave amplitude in this lead) — with a smaller Q wave in lead aVF. My impression was that despite no ST elevation in the inferior leads — the ST segment was coved (best seen in lead aVF) with T wave inversion — and, a mirror-image oppositive (reciprocal change) picture for the ST-T wave in lead aVL was seen in lead aVF. As per Dr. Smith — this strongly suggests recent inferior MI with reperfusion changes (consistent with the history of CP onset 3 days earlier).
  • In the Chest Leads — as tiny as QRS complexes are, there is an rsR’ in lead V1, which together with narrow terminal s waves in lateral leads I and V6 qualifies as IRBBB (Incomplete Right Bundle Branch Block).
  • There is early transition — since the tiny QRS complex in lead V2 is already all positive! This is relevant — because early transition (with relative early increase in anterior lead R wave amplitude) is a sign consistent with posterior MI.
  • There is definite ST depression in lead V3. As Dr. Smith has commented on many occasions — maximal ST depression in leads V2,V3 indicates posterior MI until proven otherwise.
  • Depending on which of the 2 complexes we look at in lead V4 of ECG #1 — there may also be some ST depression in this lead.
  • Depending on which of the 2 complexes we look at in lead V5 — there may or may not be ST segment coving with T wave inversion ...
  • I thought there was ST segment coving with slight-but-real ST elevation in lead V6 (since both of the QRST complexes in lead V6 show this).
  • Putting This Together — In view of the history, and taking into account additional difficulties imposed by artifact, low voltage, and the greatly prolonged QTc — I thought the appearance and distribution of changes seen in ECG #1 were most consistent with recent infero-postero-lateral MI, that was probably now showing reperfusion changes. This anatomic distribution of ECG changes is consistent with the cath finding of acute LCx (Left Circumflex) occlusion.

Regarding the LOW Voltage: The list of potential causes of generalized low voltage on ECG is long (Figure-2).
  • One of the causes of low voltage that is easy to overlook is loss of myocardium (and of myocardial function). The dramatically reduced QRS amplitude in all 12 leads in today’s case provides yet one more clue to the presence of cardiogenic shock.

Regarding the Very Prolonged QTc: We are told that today’s patient was on Sotalol for control of PVCs. This apparently was a long-term antiarrhythmic medication for this patient — but we are not given further details.
  • Initial priorities in this patient were clearly to determine the anatomy — and reestablish coronary perfusion. This was superbly accomplished, as evidenced in the 2nd cardiac cath picture showing “beautifully open” coronary arteries.
  • Optimal management at this point in this patient’s course would entail finding out specifics regarding Sotalol antiarrhythmic therapy. This should include details regarding the dosing regimen — the nature of the ventricular arrhythmia being treated — and the patient’s “baseline” QTc at steady state for his/her current Sotalol dose.
  • A QTc of ~550-560 msec (as was seen on ECG #1) increases risk for Sotalol-induced Torsades de Pointes. This should be addressed.
  • Serum K+ and Mg++ values (as well as renal function) should be promptly checked — since low values of these electrolytes increases risk of Sotalol-induced Torsades.
  • The indication and dosing of Sotalol for this patient should be reconsidered prior to giving the next dose.

My THOUGHTS on Serial ECG Changes: To facilitate lead-by-lead comparison of the serial ECG changes in today’s case — I put the first 3 tracings that were shown together in Figure-1. Note that QRS size and morphology remains virtually the same for all 12 leads in each of the 3 tracings in Figure-1 — which tells us that any ST-T wave changes we might see must be real!
  • As per the astute clinical deductions by Dr. Smith — the principal change between ECG #1 and ECG #2 occurs in the limb leads. Specifically, the T wave inversion initially seen in lead aVF (and suggested in lead III) of ECG #1 — has been replaced by upright T waves. Given how tiny QRS amplitude is — the now upright T waves in leads II, III and aVF of ECG #2 look disproportionately tall (ie, hyperacute). As per Dr. Smith — these dynamic ST-T wave changes in the setting of cardiogenic shock suggest reocclusion.
  • Comparison of these same limb leads with their ST-T wave appearance in ECG #3 — shows return to T wave inversion in leads III and aVF that we saw in ECG #1, with the T wave again becoming upright in lead aVL. Perhaps there has been spontaneous reopening (and reperfusion) of the infarct-related artery?
  • In contrast to the above serial limb lead ST-T wave changes — there has been virtually no serial change in the chest leads between ECGs #1, #2 and #3.
  • NOTE: Isn’t it much EASIER to appreciate serial ECG changes when serial tracings are placed right next to each other?

Final Thought: I found the 2nd Learning Point highlighted above by Dr. Smith of special interest — since it can be all-too-easy to overlook the disproportionate effect on hemodynamics that occlusion of a single coronary artery might have when there is significant narrowing in one or more of the other coronary arteries.
  • I wonder how much the ST-T wave appearance in the 3 serial tracings of Figure-1 may have been influenced by some cancellation of forces? It would seem that this patient who presented in cardiogenic shock should have had ECG evidence of diffuse, subendocardial ischemia (ie, diffuse ST depression with ST elevation in lead aVR). Could some of the ST elevation from recent/acute LCx occlusion (in association with severe LAD narrowing) have been cancelled out by diffuse ST depression from diffuse subendocardial ischemia?

Figure-2: Causes of generalized low voltage on ECG (See text).

ADDENDUM = Correction! (August 30, 2020): Being an “expert” in ECG interpretation is sometimes very humbling. So I’d like to CREDIT gtob — who wrote us today. I’ve copied his comment below in Figure-3:

Figure-3: Comment from gtob questioning LA-LL Lead Reversal (See text).

Our THANKS to gtob — who I believe is completely correct in his comment to us! So, we all “live & learn”. Despite my love for detecting technical errors — missed this one. Let’s all learn from this.
  • In the July 28, 2020 post in Dr. Smith’s ECG Blog — I cited my favorite on-line “Quick GO-TO” reference for the most common types of lead misplacement, which comes from LITFL ( = Life-In-The-Fast-Lane). Simply put in, “LITFL Lead Reversal” into the Search bar — and the link comes up instantly!
  • In Figure-4 — I have put ECG #2 on TOP, with addition of a box listing the effects that LA-LL Lead Reversal has on the ECG.
  • In the MIDDLE of Figure-4 — I have inverted lead III — switched the position of leads I and II — switched position of leads aVL and aVF — and left aVR alone.
  • In the BOTTOM of Figure-4 — I have again shown ECG #1 — so that we can compare how correcting for the effects of LA-LL Lead Reversal results in a 2nd ECG that looks virtually identical to ECG #1. Therefore — Although clinical management would really not have been changed in this case (because this patient still needed prompt cath and reperfusion) — in retrospect, there were no “dynamic” ST-T wave changes. Instead (as per gtob) — there was LA-LL reversal on ECG #2.

Additional Learning Points from this Case:
  • Remember that we are all human and subject to missing certain findings. The KEY is to learn from our oversights — which is why I always like to go back and try to figure out why I may have missed a given ECG finding. That way, I won’t miss it again! This case presented a very difficult series of tracings to interpret because of very small QRS and P wave amplitude — LOTS of artifact on initial ECG #1 — minimal ST-T wave deviations — and a very long QTc.
  • As per gtob — the “tipoff” to LA-LL reversal in ECG #2, is that the P wave in lead I is clearly larger than the P wave in lead II (and that is unusual when there is sinus rhythm).
  • There is another subtle-but-real “tipoff” that I did notice, but I did not follow up on. In ECG #2 — we see a qRS complex. This is very different than QRS morphology was in lead III for ECG #1. And as we take another look at Figure-4 — it should now be obvious that morphology for the tiny QRS and ST-T wave complex in ECG #2 is virtually the mirror-image of what we saw in ECG #1. Next time I see this — I need to pay more careful attention to the relative size of P waves in leads I compared to lead II.
  • Live & Learn!

Figure-4: Showing the effects of LA-LL Lead Reversal — then taking another look at ECG #1! (See text — and See LITFL Lead Reversal).

Recommended Resources