Monday, December 18, 2017

Are These Wellens' Waves??

This is a repost.  I've received a few questions like this, so wanted to re-inforce the idea of down-up vs. up-down T-waves


One of our residents texted me this ECG and was worried about Wellens' waves.

A middle-aged male presented with vomiting.  Here was the initial ED ECG:
What do you think?

Here is my response:  "What is the Potassium?"

Wellen's waves are always Up-Down T-waves, not Down-Up T-waves as here.  Down-Up T-waves in V2 and V3 have only 3 causes, as below.

Moreover, the QT interval is impossibly long at about 560 ms, with a corrected QT of 593 ms (Fridericia correction, which seems to be the best; QTc(Bazett) = 611 ms.

When the QT is impossibly long, think that the "T-wave" is actually a U-wave, not a T-wave.

Here the U-waves are highlighted:
1. In lead V6, one can discern discrete T-waves (red arrow) and U-waves (black arrow).
2.  Go down to rhythm strip lead II across the bottom, identify T- and U-waves there.
3. Go 3 beats earlier in lead II to identify the U-wave there (green line).
4. Draw the green line up to show the U-waves in leads V1-V3

So these are indeed U-waves, not T-waves

1) Posterior MI with some reperfusion (reciprocal to Up-Down T-waves of the posterior wall, analogous to Wellens' of the posterior wall as recorded from the anterior wall). 

See case below
2) Hypokalemia (in which case the upright component is really a U-wave).  In this case, V6 is pathognomonic: you can see a clear large U-wave following the T-wave.  It must be hypokalemia.  Notice also the very long QT, which is really a long QU-wave.

Other down-up T-waves:

3) Reciprocal findings between leads III and aVL in post-ischemic (Wellens' type) T-waves:
             In lead III: reciprocal to up-down (Wellens' type) post-ischemic T-waves in aVL.  
             In lead aVL: reciprocal to up-down (Wellens' type) post-ischemic T-waves in III.

The K was 2.0 mEq/L.

For contrast
Here are classic Wellens Pattern A (biphasic) waves:
Notice they are biphasic Up-Down.

Clinical Course

The patient had all serial troponins below the level of detection.  Potassium was repleted.  Here is the ECG after normalization of K at 3.5 mEq/L:

Here, again, the previous ECG at K = 2.0 mEq/L:

Here is an example of a Down-Up T-wave from Reperfusing Posterior MI.
It comes from this fascinating post:

Series of Prehospital ECGs Showing Reperfusion of Inferior-posterior STEMI:

See this too:

Learning Points

1. Wellens' waves (Pattern A) are biphasic Up-Down.  
                            (Pattern B is deep symmetric inversion)
2. Down-Up waves in V2, V3 should make you think of reperfusing posterior MI or hypokalemia. 
3. A very long QT (really a QU) should make you suspect hypokalemia.
4.  Look for clear U-waves in other leads.
5.  Finally, Wellens' syndrome is a SYNDROME that requires 1) typical anginal chest pain 2) Resolution of the chest pain 3) ECG recorded after resolution.


  1. Nice case that clearly shows what Wellen’s ST-T waves are not! As per Dr. Smith — the remarkable finding here is the obviously markedly prolonged QT interval (the QT is almost 2/3 of the R-R interval). In the absence of ischemia/infarction or bundle branch block — the finding of a decidedly long QT interval should immediately prompt a “List” = i) Drugs; ii) Serum Electrolytes (low K+/low Mg++/low Ca++); and/or iii) some CNS catastrophe (stroke; bleed; trauma; tumor; coma, etc) as the most likely contributing causes to a prolonged QTc interval. As per Dr. Smith — electrolyte disorder (low K+/low Mg++) is most likely here, given diffuseness of changes with prominent U waves. In addition to U waves — hypokalemia typically produces ST-T wave flattening and/or slight ST depression. It generally doesn’t produce the deeper T wave inversion that is suggested here in leads V3,V4 — so one couldn’t rule out concomitant ischemia on the initial tracing in this case. That said, the follow-up tracing (when K+ = 3.5 mEq/L) no longer shows T wave inversion — so true ischemia is unlikely to have been a factor. I’ll add a few additional thoughts: i) Although serum K+ in the final tracing = 3.5 mEq/L, which is at the lower limit of “normal” — given that this patient was initially so profoundly deficient of K+, it is likely that body potassium (the vast majority of body K+ resides in the intracellular, not extracellular space) is still depleted. I suspect that the residual ST-T wave flattening that we see in this follow-up ECG might further improve once body K+ stores are fully restored. ii) Rather than “QTc” prolongation — the initial ECG showed a long “Q-U” interval. I suspect in the initial ECG that we have a biphasic T wave, with the U wave fusing imperceptibly with the terminal portion of the positive part of this biphasic T wave.

  2. Thank you for posting this case. As a paramedic I’ve had ED M.Ds tell me that the up down version were U waves, much to my dismay. Thank you so much for your continued contribution to the free education movement Dr. Smith and all of your contributors.

    1. Brandon, thanks for the nice words.
      Steve Smith


Recommended Resources