Tuesday, July 31, 2018

An athletic 30-something woman with acute substernal chest pressure

Post by Smith, with short article by Angie Lobo (https://twitter.com/ALoboMD), a third year intermal medicine resident at Abbott Northwestern Hospital 

Case A 30-something woman with no past history, who is very fit and athletic, presented with 1.5 hours of substernal chest pressure.  It was non-radiating and without other associated symptoms except for nausea.  She had zero CAD risk factors.

Here was her ECG at time zero:
What do you think?














There is ST elevation in V2 with large fat T-wave.  There is ST depression in II, III, and aVF, and V3 to V6.

I saw this before any other information and knew immediately that it represented an LAD occlusion.  

There is 1 mm of STE in lead V2.  This must be explained, and normally would be explained by "normal variant," or "early repol."  However, normal variant never has associated ST depression, and here we see ST depression in II, III, aVF and V3-V6.  Moreover, there are large, fat T-waves in V2 and V3.  The T-waves in V3-V6 are subtle "de Winter's" T-waves (ST depression followed by a large upright T-wave).

The physicians told me that they "knew something was wrong" with the ECG, but couldn't express exactly what.

The pain was resolving after nitroglycerine when this ECG was recorded at time 44 minutes:
The ST depression is less
The T-waves are smaller

The physicians did not notice this either.  But this is what you are looking for if you do an ECG after pain resolution!!  You MUST look for it.  And then if you see it, it must make you realize that you missed the ST depression and hyperacute T-waves on the first ECG.

Next management step?

If you would not have activated the cath lab based on the first ECG, you really should strongly think about it now.  This patient clearly has an unstable LAD lesion that at one point was occluded or nearly so.  It is equivalent to a transient STEMI.  It was late evening and the patient will be in the hospital overnight with a potentially very unstable LAD lesion.

The first troponin I returned at 0.022 ng/mL (99% URL = 0.030 ng/mL)

Not having seen the ECG findings, the providers discussed her very low risk with her and whether she would like to go home and follow up if her 2nd troponin remained negative, or be admitted to observation. 

She felt more comfortable being admitted.

Subsequent events:

Later, before being taken to her room, the 2nd troponin returned at 1.01 ng/mL.  This is diagnostic of myocardial infarction.

Heparin was started, but there does not seem to have been discussion of cath lab activation.

About 3 hours after admission, just after midnight, she had a 30 beat run of non-sustained VT.

They recorded a 3rd ECG at time 7 hours:
Now you can see complete resolution of hyperacute T-waves.
There is no longer any STE in V2
There remains a bit of ST depression. 

About this time, the 4th troponin, drawn at 8 hours after onset of pain, peaked at 20.956 ng/mL.

(For those of you who are accustomed to the units of high sensitivity troponin (ng/L), this is equivalent to 20,956.00 ng/L.)

Now you have ECG and troponin evidence of ischemia, AND ventricular dysrhythmia, which means this is NOT a stable ACS.

Again, cath lab was not activated.


What does this troponin level mean?  Not much, but studies of STEMI and NonSTEMI show that about 70% of those diagnosed with STEMI have a peak troponin I above 10 ng/mL and that about 70% of those diagnosed with NonSTEMI have a peak troponin I below 10 ng/mL.  Troponin elevation is dependent on presence or absence of occlusion (remember many OMI receive a diagnosis of NSTEMI), duration of occlusion (which is dependent on rapidity of therapy or the luck of spontaneous reperfusion), area of myocardium at risk, collateral circulation, and more.

Case continued

Fortunately, the patient remained stable.  Next day, the patient was taken for an angiogram and found to have a reperfused LAD lesion with good flow that appeared to the angiographer as if it was a spontaneous coronary artery dissection.  It seems that there was some uncertainly about this.  The lesion was stented.


ECG at time 19 hours after cath:
Wellens' Pattern A T-waves are present (terminal T-wave inversion)
Pattern A is in contrast to Pattern B, which is a further evolutionary stage of Wellens and is more deep and symmetric.
These are reperfusion T-waves (the same thing as Wellens' waves)


Echocardiogram
Regional wall motion abnormality-distal septum and apex.
Regional wall motion abnormality-distal inferior wall.


ECG recorded at time 38 hours:
A further evolutionary stage of T-wave inversion.
This is beyond Pattern A but not quite Pattern B yet.
See examples of Pattern A and Pattern B at the very bottom of this post

All ischemic ST Elevation and/or hyperacute T-waves evolve.  The best evidence that ST elevation or large T-waves are ischemic comes from subsequent ECGs.  It they are static, then they are not due to ischemia.

Ischemic ST elevation will increase, decrease, or resolve.  Hyperacute T-waves will get larger, smaller, normalize, or invert.

This is better evidence for ischemia than any other data point.  Wall motion abnormalities may disappear with rapid reperfusion. Angiograms may be negative due to spasm or thrombus lysis or small vessel disease, or it may be a type 2 MI. Troponins may be negative with very rapid reperfusion, or measured too late, or chronically elevated due to cardiomyopathy or renal failure.  But the ECG, if it has a finding which is the result of acute ischemia, (and it does not always), will be dynamic.  If the ECG findings are due to a pre-existing chronic pathology, or to normal variant, they will almost always be relatively static over 48 hours.


What is Spontaneous Coronary Artery Dissection (SCAD)?

I asked Angie Lobo (https://twitter.com/ALoboMD), a third year intermal medicine resident at Abbott Northwestern Hospital (and Minneapolis Heart Institute) and an aspiring cardiologist, to write a couple paragraphs on SCAD.  She has done quite a bit of research on the topic.

Spontaneous Coronary Artery Dissection

Spontaneous coronary artery dissection (SCAD) is an infrequent but increasingly recognized cause of acute coronary syndrome (1- 4%) 1, most commonly affecting women (90%) 2. In middle age women, it accounts for 22-35% of all ACS presentations 1,3, and the reported incidence of ST-elevation myocardial infarction in this subset of patients is variable, but estimated to be between 24-50%4. SCAD is a non-atherosclerotic, non-iatrogenic and non-traumatic form of ACS, defined as a dissection occurring within the wall of one or more coronary arteries (LAD being the most commonly affected 2), leading to an obstruction of the lumen caused by formation of an intramural hematoma (IMH) or intimal disruption rather than atherosclerotic plaque rupture or intraluminal thrombus 2. SCAD occurs in patients with few or non-traditional cardiovascular risk factors. There has been multiple factors associated with SCAD that may predispose to a weaken arterial wall, fibromuscular dysplasia (FMD) and pregnancy being the most common 2,5.
            There are no randomized controlled trials for treatment strategies in SCAD. The current American Heart Association (AHA) consensus recommends a conservative approach (avoiding invasive therapies) in hemodynamically stable patients with good TIMI flow. Invasive therapies, such as percutaneous coronary intervention (PCI) or coronary artery bypass graft (CABG) should be reserved for patients with ongoing ischemia, left main artery dissection, or hemodynamic instability 6. It is of extreme importance to suspect SCAD in this population (specifically women) in order to deliver the adequate management (avoid management as atherosclerotic ACS, unnecessary invasive strategy) and decrease the rate of complication.


1.     Nishiguchi T, Tanaka A, Ozaki Y, Taruya A, Fukuda S, Taguchi H, Iwaguro T, Ueno S, Okumoto Y, Akasaka T. Prevalence of spontaneous coronary ar- tery dissection in patients with acute coronary syndrome. Eur Heart J Acute Cardiovasc Care. 2016;5:263–270. doi: 10.1177/2048872613504310.
2.   Tweet MS, Hayes SN, Pitta SR, Simari RD, Lerman A, Lennon RJ, Gersh BJ, Khambatta S, Best PJ, Rihal CS, Gulati R. Clinical features, management, and prognosis of spontaneous coronary artery dissection. Circulation. 2012;126:579–588. doi: 10.1161/CIRCULATIONAHA.112.105718.
3.     Nakashima T, Noguchi T, Haruta S, Yamamoto Y, Oshima S, Nakao K, Taniguchi Y, Yamaguchi J, Tsuchihashi K, Seki A, Kawasaki T, Uchida T, Omura N, Kikuchi M, Kimura K, Ogawa H, Miyazaki S, Yasuda S. Prog- nostic impact of spontaneous coronary artery dissection in young female patients with acute myocardial infarction: a report from the Angina Pecto- ris-Myocardial Infarction Multicenter Investigators in Japan. Int J Cardiol. 2016;207:341–348. doi: 10.1016/j.ijcard.2016.01.188.
4.     Saw J, Mancini GBJ, Humphries KH. Contemporary review on spontaneous coronary artery dissection. J Am Coll Cardiol. 2016;68:297–312. doi: 10.1016/j.jacc.2016.05.034.
5.     Saw J, Poulter R, Fung A, Wood D, Hamburger J, Buller CE. Spontaneous coronary artery dissection in patients with fibromuscular dysplasia: a case series. Circ Cardiovasc Interv. 2012;5:134–137. doi: 10.1161/ CIRCINTERVENTIONS.111.966630.
-->
6.     Hayes, S.N., Kim, E.S., Saw, J., Adlam, D., Arslanian-Engoren, C., Economy, K.E., Ganesh, S.K., Gulati, R., Lindsay, M.E., Mieres, J.H. and Naderi, S., 2018. Spontaneous coronary artery dissection: Current state of the Science: A scientific statement from the American Heart Association. Circulation, 137(19), p.e523.  doi:10.1161/cir.0000000000000564.




More on Wellens' Pattern A and Pattern B

From the original article:
At the time of this article, Wellens did not suggest that Pattern B was an evolutionary state from Pattern A. 
I don't believe anyone has ever written a peer reviewed article on this, but through long experience and through literature on reperfusion therapy for acute MI, I believe this to be true.  It is only established if there are multiple ECGs.
See this post:

Classic Evolution of Wellens' T-waves over 26 hours




This is one patient, with ECG recorded at time zero after pain resolution, then at time 2 hours, then at time 9 hours.  It was long ago that I realized that Wellens' is a reperfusion pattern and the pattern A is early and evolves into Pattern B.  


-----------------------------------------------------------
Comment by KEN GRAUER, MD (7/31/2018):
-----------------------------------------------------------
In my experience, the most instructive ECG cases consist of serial tracings with contemporaneous notation at each moment in time as the case evolves. KEY lessons to be learned in this case cannot be taught in isolation. For ease of discussion — I’ve put the first 4 sequential tracings in this case together in Figure-1.
  • The patient profile in this case is deceptive = a seemingly healthy, athletic and very fit 30-year woman with no risk factors and no significant prior medical history. The 1st “lesson” is, “All bets are off” — when an adult of any age presents to the ED with new-onset chest discomfort.
  • PEARL: When ECG findings are subtle — Look for subtle findings in multiple leads! Although by the time you are reading My Comment, you already know the outcome in this case — GO BACK and Take Another Look at the initial ECG @ Time = 0:00 (TOP tracing in Figure-1). How many leads show abnormal ST-T wave findings in ECG #1?
Figure-1: The first 4 sequential tracings in this case. Note that both frontal plane axis and the sequence of R wave progression is very similar in all 4 tracings — therefore, lead-to-lead comparison for assessment of ST-T wave changes between these serial tracings is much easier to accomplish than it would be if there was marked change in axis or in R wave progression across the chest leads (See text).
ANSWER: Except for aVR — every lead in ECG #1 shows abnormal ST-T wave changes:
  • Dr. Smith has already described the ST elevation in lead V2, in association with fatter-than-expected T waves in V2 and V3. He also noted the subtle pattern of deWinter T waves in leads V3-thru-V6 ( = ST depression in each of these leads, followed by rise to larger-than-expected T waves).
  • There is also ST coving and elevation in lead V1. While ST coving in V1 is not necessarily abnormal — the presence of ST elevation in association with ST-T wave abnormalities in V2,V3 in a patient with chest pain is clearly cause for concern.
  • As per Dr. Smith — each of the inferior leads (II,III,aVF) shows ST-T wave abnormalities. As is often the case when reciprocal ST-T wave changes are seen in the inferior leads — the most subtle changes tend to be seen in lead II. The amount of ST depression in lead II of ECG #1 is minimal — but — the ST segment in this lead “looks depressed” (because of the flat horizontal ST segment ledge, with abrupt transition from the end of the ST segment to the beginning of the T wave). If this picture of subtle-but-real abnormality in the ST-T wave in lead II of ECG #1 is not familiar to you — Memorize the shape of the ST-T wave in lead II!
  • There is ST elevation in lead aVL. Again, the amount of ST elevation in aVL is minimal — but the importance of recognizing deWinter-like T waves in the chest leads of a patient with new-onset chest discomfort and reciprocal ST-T wave changes in all 3 inferior leads, and at least some ST elevation in lead aVL — is that the combination of these findings suggests an ongoing acute event, and localizes the lesion to the proximal LAD.
  • Finally — the ST segment in lead I of ECG #1 is abnormally straight. This accounts for the rather abrupt transition between the end of the ST segment and the beginning of the T wave in this lead.
  • NOTE: None of the above findings in isolation is diagnostic. But taken together — the picture in ECG #1 of subtle-but-real ST-T wave abnormalities in 11/12 leads in an adult with new-onset chest discomfort should suggest acute coronary occlusion until you prove otherwise.
As emphasized by Dr. Smith — the best way to demonstrate acute ischemia is by identifying dynamic ST-T wave changes in association with change in the nature of chest pain. This often requires careful lead-by-lead comparison between serial tracings.
  • Take Another Look at Figure-1. How many leads show a difference when comparing the ST-T wave appearance in ECG #1 with that in ECG #2?
  • KEY POINTS: ST-T wave changes in serial tracings may be exceedingly subtle. First, assess if frontal plane axis and the sequence of R wave progression across chest leads is similar between the 2 tracings that you are comparing — since a significant difference in either of these factors may make accurate comparison of serial tracings much more difficult. Go lead-by-lead — comparing the appearance of lead I in both tracings; then lead II in both tracings; then lead III … and so on. Because differences may be so subtle — you are often looking for a “Gestalt” ( = overall) impression as to whether there is qualitative difference between the 2 ECGs.
----------------------------------------------------------------------------
Our IMPRESSION: Lead-by-lead comparison between ECGs #1 and #2 to me suggests that except for lead III (for which the shallow T wave inversion looks the same in both tracings) — there is slight improvement in virtually all other leads. Given that chest pain is resolving at the time ECG #2 was recorded — this supports the concept of dynamic ST-T wave changes in this patient.

Now compare ECG #3 with the first 2 tracings:
  • Looked at in isolation — the ST-T wave changes in ECG #3 are no more than nonspecific. But examined in context (ie, knowing this tracing was obtained 7 hours after ECG #1) — supports beyond doubt the concept of ongoing dynamic ST-T wave changes in this patient who presented with acute symptoms.
Cardiac cath performed the next day revealed a reperfused LAD lesion, that was thought to result from spontaneous coronary artery dissection. As per Dr. Smith — ECG #4 shows evolution to an ECG pattern consistent with Wellens’ Type A T waves (ie, ST-T waves marked by terminal T wave inversion in leads V2 and V3).
  • Since ECG #4 was obtained after the infarction occurred (and shortly after cardiac cath showed reperfusion of the “culprit” lesion) — we define the ST-T wave appearance in the chest leads of this tracing as consistent with reperfusion T waves.
  • But, IF the history associated with ECG #4 was different — clinical implications of these ECG findings might change dramatically. For example, imagine this patient had several hours of chest pain a day before she came to the ED — but that all chest pain had resolved at the time of presentation — and, that instead of ECG #1 being her initial tracing, that ECG #4 was her initial tracing. If this were the case, the appearance of the ST-T waves in the mid-chest leads of ECG #4 would be consistent with Wellens’ “Syndrome” — and would serve as an indicator of a tight, proximal LAD lesion in need of revascularization.

Sunday, July 29, 2018

Chest pain and T-wave inversion in lead V2

This is the ECG of a young man who complained of chest pain.

Here is the computer interpretation:
ATRIAL FLUTTER/TACHYCARDIA (this is obviously incorrect)
LEFT POSTERIOR FASCICULAR BLOCK [QRS AXIS greater than 109, INFERIOR Q]
MODERATE ST DEPRESSION [0.05+ mV ST DEPRESSION]
ABNORMAL ECG


I saw the ECG before seeing the patient, indeed before he was even in his room.  
I immediately saw that the computer was incorrect, but I found 2 abnormalities.  
What are they? 

I showed this to several physicians.  
They were worried about the T-wave inversion in V2.

One mentioned "Wellens' waves"

Another asked: "Are these juvenile T-waves?" (very good question!)
For more on this: Persistent Juvenile T-wave Pattern

What do you think?



I keep getting response tweets that T-wave inversion in V2 is normal.  It is not normal.  It is not necessarily pathologic, but it is not normal.  It occurs normally in approximately 1% of males over age 14 but younger than 30.  It is common in young women (these are the typical juvenile T-waves, link to relevant post above).

T-wave inversion in V2 occurs in many pathologies, including posterior MI, pulmonary embolism, Wellens' syndrome.  It is one of the minor criteria for arrhythmogenic right ventricular dysplasia ("Minor: Inverted T-waves in right precordial leads V1-V2")

Here is my list of Normal and Pathologic T-wave inversion:

• Normal t wave vector is leftward, inferior and anterior.
• T waves should be positive in lead II.
• Inversion in lead I in the presence of a positive QRS is always abnormal
• In a normal ECG, T waves in V5-V6 should always be upright.
• Inversion in lead III or aVL may be normal
• Lead aVF may be slightly negative
• Lead V1 is commonly negative
• Of leads V2-V4, normal inversion is rare in V2, more rare in V3 and most rare in V4
• QRS-T angle > 60 degrees











My interpretation was this: this is a normal ECG of a healthy young thin male.

But it looks abnormal because there are 2 different lead placement mistakes on the same ECG:

1) V1 and V2 were placed too high.  When you see T-wave inversion in lead V2, you should wonder if perhaps it is due to high lead placement.  How can you verify or refute that?  Look at the P-wave in V2: it should be upright.  Here it is negative.  Thus, V1 and V2 were placed too high.  The "ST depression" seen by the computer is in leads V1 and V2 and is a result of lead placement as well; see below that when they are recorded correctly, there is normal variant ST elevation.

2) Limb leads were reversed.  Look at the (limb lead) axis.  It is directly to the right, with tall R-waves in I and aVL.  Yet there are R-waves in V5 and V6.  Although V5 and V6 are slightly different than I and aVL because they are more inferior, they should not look this different.  The depolarization is to the left (V5 and V6), not to the right.  This is why the axis appears to be to the right and why the computer called it left posterior fascicular block.


The patient arrived in his room and still had his stickers on.  I obtained written permission to take a photo of his chest.  Look how high V1 and V2 are:
 Notice also he is a thin young man



I put the leads in the correct position (4th intercostal space) and recorded another ECG, also being certain that the limb leads were not reversed.

Here is the result:

Now it is simply a healthy looking ECG with high voltage (not due to LVH, but normal for a young thin healthy male)
The axis is normal now.
The T-wave is upright in lead V2, as is the P-wave.

The ST depression in V1 and V2 is replaced by normal variant ST elevation.




Interestingly, I forgot to put in the formal read on that first ECG.

Another physician did so, and simply confirmed the erroneous computer interpretation.

I have no idea how it failed to diagnose sinus rhythm.
It is hazardous to simply confirm the computer interpretation!


Here are K. Wang's comments:
1) The computer is calling Atrial flutter/atrial tachycardia, which obviously is wrong.  Computer is usually better than this. Where did the computer go wrong? I see that, in lead l, the negative P wave and the negative T wave occur at about same distance at a rate of about 160/m, which the computer is mistaking as ectopic atrial P waves or the negative component of the flutter waves.
2) Before one calls right axis deviation, always pay attention to the P wave in lead l. If it is inverted, it's either reversed arm leads or dextrocardia. One needs the precordial leads to sort this out. Obviously, a human error of reversing the arm leads is far more common than dextrocardia.
K. WANG.

From Sam Ghali:

Steve, Another great teaching case! It’s crazy how commonly V1 + V2 are placed high, precisely where you have shown. I think it’s mostly because it is easier to not remove the shirt, or the bra gets in the way in a woman. Regarding the limb lead reversal, the other key clue that the leads have been reversed is they leave an usual P-wave axis (+ in aVR, - in I & aVL) around 120-150 degrees, coinciding with the QRS axis. I presume this is why the computer did not call Sinus Rhythm! Sam



-----------------------------------------------------------
Comment by KEN GRAUER, MD (7/29/2018):
-----------------------------------------------------------
Short and simple case in which Dr. Smith reviews a “Must Know” Concept — which is ready recognition of Lead Misplacement. The fact that not one — but several physicians apparently missed both of the types of lead misplacement that occurred here — and, that another physician signed off on the grossly inaccurate computer interpretation of the 1st tracing, apparently without even looking at the ECG — is evidence that much work remains to be done in the area of ECG interpretation. Dr. Smith expertly covered the essentials of the ECG diagnosis in this case. I limit my comments to a few additional points. For clarity and ease of comparison — I’ve put both tracings together in Figure-1.
Figure-1: ECG #1 is the original tracing. ECG #2 is after limb lead and chest lead placement were corrected by Dr. Smith (See text).
  • There are many different types of limb lead misplacement. That said, the most common lead mix-up (by far!) — is LA-RA mix-up. When this is done — then the appearance of the QRS complex and ST-T wave in lead I will look like lead aVR typically looks — and vice versa. You can instantly suspect LA-RA mix-up has occurred if you remember that there should never be global negativity (of the P wave, QRS and T wave) in lead I. The only 2 conditions that cause this are: iLA-RA mix-up; and iidextrocardia (which is rare!). You can quickly rule out dextrocardia by looking to see if there is normal R wave progression (as well as making sure heart sounds are on the left). Note in Figure-1, in lead I of ECG #1 that there is global negativity! Even when there has been a huge lateral infarction that produces a large Q in lead I — you will not see global negativity in this lead.
  • The perspective of lead aVR is to look down at the heart from the right shoulder. As a result — this lead generally sees electrical activity as moving away from aVR, which often results in the global negativity normally seen in aVR. At times, you may see a component of positivity in aVR — but you should not see the positive P and QRS complex that is present in aVR for ECG #1.
  • Finally — there is NO upright P wave in lead II of ECG #1. This means that either the rhythm is not sinus — or — there is some kind of lead misplacement.
  • Note in Figure-1 that there is normalization of electrical activity in all 3 of these leads in ECG #2 after limb lead placement has been corrected ( = positive P, QRS and T wave in lead I; global negativity as it should be in aVR; and return of an upright P wave in lead II).
  • PEARL: My favorite “Quick GO-To” reference for the most common types of lead misplacement comes from LITFL (= Life-In-The-Fast-Lane). In addition to the superb user-friendly diagrams and charting in their write-up — I love the fact that you can instantly find this LITFL post by searching on-line for “lead misplacement litfl” — which I have done numerous times when I wanted to find quick answer to the question of which lead got mixed up with which other lead?
Our group surveyed the internet in search of assessing how prevalent chest lead misplacement is. Suffice it to say that even on internet sites purporting to demonstrate “correct chest lead placement” — errors in chest lead placement were disturbingly common. The most frequent of these is placing leads V1 and V2 one or two interspaces too high on the chest. For just one example of the potential clinical impact misplacement of leads V1 and V2 may have — you have no further to look than the ECG Blog post by Drs. Meyers and Smith from yesterday ( = July 28, 2018 You'll see my discussion of this error if you scroll down to the bottom of the page).
  • Recognition that leads V1 and V2 are probably placed too high on the chest is easy! Look for: iA significant negative component to the P wave in leads V1 and V2; iiT wave inversion isolated to leads V1 and V2; iiian rSr’ in V1 and V2; and ivAppearance of the QRST complex in V1 that closely resembles the QRST appearance in lead aVR. Just 2 or 3 of these findings are all that is needed to raise suspicion — which you can then easily check out by repeating the ECG after verifying lead placement. It should be mentioned that there may normally be a slight negative component to the P wave in lead V1 — and, that with left atrial abnormality, this negative component in V1 may become fairly large. But one does not usually see the deep, all negative P waves that appear in V1 and V2 of ECG #1. And one does not typically see a nearly identical appearance in V1 and V2, with negative P waves and T inversion that also looks nearly identical to what we see in lead aVR in ECG #2, once the limb leads were correctly placed.
  • PEARL: My favorite quick reference regarding placement of leads V1 and V2 too high on the chest, is the Letter to the Editor by Javier García-Niebla (Rev Esp Cardiol 61(10):1109-1110, 2008).


Recommended Resources