Atrial Flutter Mimicking ST Depression

A 65 year old presented with altered mental status and had an intracranial bleed:

One could be fooled into thinking this is sinus tachycardia (with a short PR interval) with diffuse ST depression.  But close inspection reveals flutter waves.  In particular, a totally upright p-wave in V1 is very unusual and should alert you to atrial flutter.  The fluttering baseline accounts for the apparent ST depression, although I cannot rule out some amount of true ischemic ST depression.

 Here is the ECG after cardioversion:

Now there is sinus.  Interestingly, this one also has an upright p-wave in V1 - so the rule is not universal!

Osborn Waves and Hypothermia

Case 1.  Temperature 30 degrees Celsius (86 degrees F) due to environmental hypothermia.

There is very slow atrial  flutter (rate = 167) with 4:1 AV conduction.  There is a wide QRS with a very large notch (in this case, a hump), or J-wave, at the end.  This is the classic Osborn wave of hypothermia.

click on ECGs for better resolution

Case 2. A young paraplegic presented confused.  Among the early tests performed was this ECG which was showed to me:

Sinus rhythm.  Long QTc [about 500ms; the computer read 180ms (!)].  There is ST elevation which alarmed the residents.  I did not think it looked like injury.  There are J-wave notches at the end of the QRS, particularly in V3, which are rather large for early repolarization, which should make one think they may be Osborn waves.

I asked about the temperature.

Here is the patient's previous ECG:

No ST elevation, and much smaller J-wave notching, seen best in V4.

A rectal temperature was 30.8 degrees Celsius.

Here is the ECG after rewarming:

J-wave notching persists but is much smaller.

Here they are side by side to better see the difference:

The J-waves are subtly larger in the hypothermic condition

The ECG in hypothermia

Rhythm: The most common rhythms in hypothermia are sinus bradycardia, junctional bradycardia, and atrial fibrillation.  Shivering artifact is common.  Atrial flutter is seen in case 1.  At temperatures below 30 C, the patient is at risk for ventricular fibrillation.   In this study of 29 humans cooled to 28-30 C for cardiac surgery, 19 developed atrial fibrillation and 2 ventricular fibrillation.

QRS: Osborn waves are thought to be pathognomonic of hypothermia, but can also be seen in normothermic patients.  "J-waves" or "J-point notching" is very common in early repolarization.   Very narrow Osborn waves were reported in severe hypercalcemia (level 16.3).  Sometimes a short ST segment of hyperCa can be misinterpreted as an Osborn wave (see image below); that is not the case in the aforementioned case report.   J-wave syndromes are proposed to give a unifying pathophysiology to Osborn waves of hypothermia and early repolarization, as well as Brugada syndrome.

Very large and wide J-waves, as in case 1, are almost exclusively due to hypothermia.  The etiology is beyond the scope of this blog, but may be read here.  

Hypothermia and pseudoinfarction patterns: MI or ischemia (either ST elevation or depression) may be mimicked either by 1) repolarization abnormalities (As in Case 2, with ST elevation) or by 2) confusing the J-wave with the ST segment, as in this case in JACC (full text) and this case in Archives of Internal Medicine (no full text).  This latter case also has ST segment depression as a repolarization abnormality.

 

Short ST segment (with resulting short QT interval) of hypercalcemia mimicking Osborn waves

 

This image courtesy of Dr. K. Wang from his Atlas of Electrocardiography.  The major difference between the Osborn wave and the example of hypercalcemia is that the Osborn wave is followed by an ST-T complex, while the wave directly following the QRS in hypercalcemia is the T-wave itself.

Tachycardia with Pericardial Effusion

A 52 year old man with a history of atrial fibrillation and prosthetic mitral valve replacement just 11 weeks prior presented with a complaint of a rapid regular heart rate; he could hear rapid clicking of his valve.  He was otherwise asymptomatic.  His medications included amiodarone for rhythm control of his atrial fibrillation.  This ECG was recorded:

The treating physicians diagnosed sinus tachycardia at a rate of 127.  They were worried about ST elevation in II, III, and aVF, with reciprocal ST depression in I and aVL.  With the Q-waves, they were not sure if this was old or new ST elevation.

They looked for a previous ECG and found this:

There was no ST elevation at baseline.  They were now worried about acute MI.

They did a bedside echocardiogram and found a large pericardial effusion.  BP was stable and normal at 110/83.  They infused a liter of normal saline, and the heart rate remained 127.  A repeat ECG was identical.


70 minutes after arrival, the first troponin I returned elevated at 0.160 ng/ml.  Now they were even more worried about MI. 


What is the diagnosis?

Slow Atrial Flutter with 2:1 conduction.  Slow because of the beta blocking effects of amiodarone. 

1.  Note the flutter waves in V1: one positive wave that appears to be a p-wave and another that immediately follows the QRS.  The one that looks like a p-wave is positive, whereas the p-wave on the previous ECG is negative!
2.  Any time the heart rate remains the same, in spite of time or fluids, it is almost certainly not sinus.
3.  There is no ST Elevation.  All of the apparent ST elevation is due to the baseline formed by the atrial flutter waves!

INR returned at 3.9.  The patient was given propofol and electrically cardioverted in the ED.  Here is the subsequent ECG:

Final diagnosis: Atrial flutter with pericardial effusion due to myocarditis due to postoperative Dressler's syndrome.  Troponin elevation from demand ischemia.  ST elevation from flutter wave.  The effusion disappeared with time; no surgery or drainage was necessary. 

Anterior ST elevation: is it STEMI?

I received a call from an outside hospital.  A 31 year old healthy auto mechanic had a prolonged exposure to Carbon Monoxide in his garage and presented with blunted level of consciousness and chest pain.  There were others who were also affected.

I requested a transfer so that he could undergo immediate hyperbaric oxygen therapy at our hospital.  After the ambulance left the other hospital, the physician called to state that the patient was having an anterior STEMI with "tombstones."

The CO level returned at 14, so I knew that any STEMI would be due to simultaneous and incidental acute coronary thrombosis, NOT due to the CO toxicity.  I asked him to immediately fax the ECG, which is shown here:

What is it?

There is an R'-wave in V1 with downsloping ST elevation and an inverted T-wave.  There is ST elevation in lead V2.  This is not STEMI.  It is Brugada pattern

Brugada is frequently mistaken for STEMI, although it can also mask anterior STEMI.  I was not worried about this ECG.  When the patient arrived, his Brugada pattern had resolved: 

There is minimal STE and poor R-wave progression, but the QTc is 384, so this is early repolarization.  The Brugada pattern is gone.

We sent him to the chamber.  He was admitted to the hospital and ruled out for MI.  The patient's CO level was only 14, but he had had a prolonged exposure, had enough toxicity to cause objective neurologic deficits, and the blood level had been measured after prolonged oxygen therapy.

Brugada pattern ECG is not Brugada syndrome, which requires more than simply an ECG.  It is important to note that the pattern and the risk for cardiac arrest may not always be present, but may be induced by fever or sodium channel blocking drugs.  This makes me wonder if CO poisoning could induce or unmask Brugada pattern on the ECG.

Also, V2 makes this Brugada pattern atypical.

The patient should be referred to an electrophysiologist.

Hyperacute T-waves, with a Twist

A 30 year old male complained of chest pain and then collapsed.  He was resuscitated from ventricular fibrillation.  He arrived at 0700.

These two ECGs were recorded at 17 minutes apart.

Which was first?

There are hyperacute T-waves in V1-V5, with some depressed ST takeoff in V3-V5.  There are also hyperacute T's in II, III, and aVF

There is ST elevation (injury) in V2-V4and II, III, aVF

The bottom one was recorded first, at 0719, the top one was recorded second at 0736.  The patient's artery had reperfused between the first and the second.

This illustrates nicely how hyperacute T-waves are present not only shortly after occlusion, but also shortly after spontaneous reperfusion, or, as I sometimes say: "both as the ST segments are on the way up, and on the way down."

As it turns out, the artery reoccluded, and at 0801, the angiogram showed a 100% occluded type III ["wraparound" (to the inferior wall)]  mid (after the second diagnoal) left anterior descending artery.  A large  thrombus was aspirated and the LAD was stented.

The EF later that day was 25%, with both inferior and anterior wall motion abnormalities.  However, as expected from the short duration of complete occlusion, the troponin I peaked at only 20 ng/ml.   A second Echo was done 4 days later:  the stunned myocardium had recovered, and the EF was 65%.

He underwent therapeutic hypothermia and in spite of some initial hypoxic encephalopathy, he completely recovered.

Click here to see the most popular post of all time, on hyperacute T-waves.

Atrial Repolarization Wave Mimicking ST Depression

A 25 year old woman presented with a caffeine overdose and chest discomfort.  This is her ECG:

There is sinus tachycardia (rate = 120) with what appears to be diffuse ST depression in leads II, III, aVF and V2-V6.  However, if you look closely, the PR segment is downsloping.  This is due to a pronounced negative atrial repolarization wave (atrial T-wave, or "Ta-Wave").  The wave is still negative at the J-point, and thus depresses the J-point.  

The ST segment is most commonly measured at the J-point and relative to the PR segment, but when there is a Ta-wave, this method is inaccurate.  K. Wang (my mentor) recommends measuring the ST segment relative to the end of the PR segment (this is also called the PQ junction, and the recommended location for measurement according to ACC/AHA)  However, in my experience, when there is a Ta-wave, the PR interval is still downsloping at this point and this method of measurement will underestimate the effect of the Ta-wave.  Also, in my experience, and contrary to research I outline below (and which contradiction I cannot explain), the greatest part of the Ta-wave is back to baseline by 60-80 ms after the J-point. In the case above, if you measure the ST deviation at 60-80 ms after the J-point and relative to the TP segment, you'll see that there is no ST depression.

Many textbooks recommend measuring the ST segment at 60-80 ms after the J-point and relative to the TP segment, presumably because it helps to avoid the issue of the Ta-wave.  However, especially in tachycardia, the TP segment may never come back to baseline after the T-wave; furthermore, the T-wave has often begun by 80 ms after the J-point, as in this case.

Below is a schematization of the Ta-wave:

The atrial  repolarization wave lowers the baseline, but its amplitude, if present at all, is not great (maximum 0.2 mV, or 2 mm at normal recording).  It is not finished until up to 180 ms after the J-point (see references below).   

How do you recognize the Ta-wave?  First, you have to be aware of it and look for it.  Then, you have to imagine a curve, like this drawing:

The Ta-wave inscribes a parabolic curve that can be imagined when viewing the ECG

Learning point: Beware diagnosing ST depression before considering the atrial repolarization wave as the etiology.

More Detail on the Ta-wave:

The Ta-wave is a mean of 320 ms after the end of the p-wave, with a duration of 2-3x that of the p-wave and a polarity always opposite of the p-wave.  The PTa duration (onset of p-wave to end of Ta-wave) is a mean 440 ms, though it varies with heart rate just like the QT interval.  Thus, if the PR interval is 160 ms, the Ta-wave ends about 280 ms later.  If there is normal conduction of 100 ms, the Ta-wave may still be present at 180 ms after the end of the QRS!  Even if the patient has Bundle Branch Block with a duration of 140 ms, then the Ta-wave may still be present 140 ms after the end of the QRS. 

Here is a short explanation of the atrial repolarization wave.

Here are two detailed articles measuring the Ta wave: one by Holmquist et al. and another by Debbas et al.

Left Bundle Branch Block (LBBB) with Chest Pain, concordant and excessively discordant ST depression V2-V6

A middle-aged male presented pain free after an episode of chest pain.  Here is the initial ECG (sorry some is cut off -- it is an iPhone shot from a friend):


There is LBBB with appropriate discordance of all ST segments.  Anterior ST elevation is appropriate, with highest ST/S ratio of 3.5/28 = 0.125 (mean normal = 0.11; normal up to 0.19).  There are concordant T-waves in V5 and V6.  This is a nonspecific sign of NonSTEMI.

 5 minutes later, the patient had crushing chest pain, and this ECG was recorded (again, some of limb leads are cut off):

Now there is concordant ST depression in V2 and V3.  This is a relative change of approximately 5 mm(!).   There is excessively discordant ST depression in V4-V6.   (V4 ratio is 2/6 = 0.33; V5 ratio = 2.5/6.5 = 0.38;  V6 = 2/6.5 = 0.31).  Thus, there is ischemic ST depression in V2-V6.   In normal conduction, ST depression from V2-V6 is often due to subendocardial ischemia, whereas when limited to V1-V4, it is usually posterior STEMI.  Either way, this is a patient with acute coronary syndrome with chest pain.  If you cannot control the symptoms with medical therapy, then the patient must go to the cath lab.   

I have written about excessively discordant ST elevation, but have not mentioned excessively discordant ST depression.  In our study of LBBB with and without coronary occlusion, just one lead with excessively discordant ST depression or ST elevation, as defined as a ratio of ST depression (or elevation) to the preceding R-wave (or S-wave), greater than 0.25, was very specific for ischemia (in our study, for occlusion).  More recent analysis of the data showed that 0.20 was probably a better cutoff.

The physician called the interventionalist, who did not agree there was ischemia on the ECG.  The patient was started on nitroglycerine IV and the pain subsided, as did the ECG findings. 

The patient was admitted pain free on nitro and no immediate cath was done.  The troponin I peaked later at 0.18 ng/ml. 

The next AM, the patient had another episode of pain that could not be resolved with maximal medical therapy.  He went for emergent cath, which showed a proximal lad 95% stenosis with deep ulcer and a 90% mid lad stenosis.  Both were stented.

Later, the troponin peaked at 5.6, and echo showed anteroseptal hypokinesis with EF <40%.

So this was LBBB with concordant and excessively discordant ST depression, representing ST depression in leads V2-V6, completely consistent with subendocardial ischemia due to profound LAD ischemia.