Friday, January 31, 2014

Paroxysmal SVT (PSVT) that repeatedly recurs in spite of successful conversion with adenosine

An elderly male complained of dyspnea.  His pulse was regular at just under 150.  BP was 110/70.  There was no evidence of shock or pulmonary edema.  Here is his ED ECG:
There is a very rapid, wide complex tachycardia (QRS = 150 ms).  There are no P-waves before the QRS's.  There is a definite Right Bundle Branch Block and Left Anterior Fascicular Block pattern, so this is not VT.  
The bifascicular block was new.  It could be a new block, or a rate-related BBB
 If you look closely, you can see the inverted (retrograde) P-waves.
(See annotated ECG below, with arrows pointing to inverted P-waves) 
It is clearly SVT with aberrancy.
See arrows pointing to retrograde P-waves

He was given 6 mg of adenosine, and converted to sinus rhythm for about 20 seconds, then reverted back to PSVT.

This was repeated with the same results.

What would you do?

Farther below is what we did (skip the rationale for now, if you already know the pathophysiology of PSVT).  The rationale for management cannot be fully understood without the following.  Here is a good review article in the New England Journal (free full text).

Anatomy of PSVT: Most PSVT is AVNRT (AV nodal re-entrant tachycardia), with the re-entrant circuit in the AV node ("dual AV nodal pathways).  (The remainder is due to an accessory pathway such as WPW and to atrial tachycardia).    In AVNRT, there are two pathways within the node which have different conduction speeds.  In addition, the fast pathway has a long refractory period and the slow pathway has a short refractory period.

Sinus Rhythm: Normally, a sinus beat gets to the dual pathways and travels down both: it gets down the fast pathway quickly and then proceeds both down the HIS bundle and up the slow pathway.  As it travels up the slow pathway, it meets the impulse coming down the slow pathway from above and they both stop each other.

Initiation and propagation of PSVT: 

The common form of the arrhythmia is known as "slow-fast AVNRT".  It is initiated by a premature beat, usually a premature atrial beat (PAB, or more commonly known as PAC), and much less often due to a premature junctional or ventricular beat.

If that premature beat arrives at the AV node when the fast pathway is still refractory, it terminates there. And if it arrives at just the right time, when the slow pathway is no longer refractory (the refractory period is shorter than that of the fast pathway), then it can proceed down that slow pathway.  It then will travel to the bundle of HIS and activate the ventricles.  But it will also go retrograde up the fast pathway.  (The fact that it made its way down the slow pathway gave the fast pathway time to recover from its refractory period).

If you happen to witness the PAB, you will notice that it has a longer PR interval than the sinus beats because it is conducting through the slow pathway.

Now the beat goes back up the fast pathway and when it gets to the top, it proceeds BOTH up to the atrium (resulting in a retrograde P-wave that is usually buried in the QRS), and down the slow pathway (which is no longer refractory because it has a short refractory period).  Every time it gets to the inferior junction it conducts to the ventricles, an every time it gets to the superior junction it conducts to the atria.  It goes around and around the two limbs.

To terminate this re-entrant rhythm, all you need to do is stop conduction in the pathways until the sinus node can take over again.  This can of course be done with vagal maneuvers, adenosine, verapamil, or electrical cardioversion.

However, if the SVT recurs, then brief interruption of conduction in the AV node, such as with adenosine, will only be a temporary solution.  Same with electricity.

Thus, in this case, we need to prevent initiation of the SVT, or use a drug that maintains prolonged AV blockade.  To do so, one can try to prevent the formation of PABs.  Beta blockers are pretty good at preventing PABs, and also at slowing AV conduction.

Further management

The plan was to load with a beta blocker, then convert again with adenosine if necessary, with the expectation that the beta blockade would prevent another PAB and the rhythm would remain sinus.  Additionally, it is also possible for beta blockade to convert the rhythm (in addition to preventing PABs) because it slows AV node conduction. 

Therefore, we started esmolol (500 mcg/kg) by bolus and drip (50 mcg/kg/min).  We chose esmolol because it is short acting and can be turned off if there are hypotensive complications.  There were no adverse effects (and no response either), so another 500 mcg/kg bolus was given, with increase in the drip to 100mcg/kg/min.  No response, so another bolus and drip was increased to 150mcg/kg/min.  There was still no response.

At this point I was sure that if we gave adenosine, the patient would convert and remain in sinus.

So we gave adenosine 6 mg.  He converted.  10 seconds later there was a PAB and her reverted to PSVT again.

What would you do now?

Now the patient needs either:

1) a drug that will block AV nodal conduction for longer than adenosine, and better than esmolol did at the given dose.  A calcium channel blocker such as verapamil or diltiazem would be good (In the era prior to adenosine, we used to give verapamil  regularly for PSVT, with good results).  There is some potential for significant negative inotropic events with these medications.  A word of caution: avoid use of both beta blockade and calcium channel blockade simultaneously.


2) a drug other than a beta blocker to prevent a PAB

After reading the New England journal article above (while managing the patient), I was considering  either verapamil or ibutilide.  I called our consulting cardiologist and he suggested amiodarone, followed by adenosine to convert again.  Like beta blockers, amiodarone prevents the initiating beat.

We gave 150 mg of amiodarone over 10 minutes, then started an infusion.  The rhythm stayed  in SVT but slowed gradually to 130 then 120 beats per minutes, then converted spontaneously to sinus.

Here is the post conversion ECG:
There is now sinus rhythm.  The BBB is still present, so this was not rate related.


The patient did well.  He had an episode of atrial fibrillation, but spontaneously converted.   Ultimately, the SVT was controlled with metoprolol.


1. SVT that recurs needs therapy to either block the AV node for a prolonged period (adenosine's effect is for seconds only), or therapy to prevent the PABs that initiate any recurrent SVT.


  1. Why do you want to avoid the use of beta blockers and calcium channel blockers simultaneously?

    1. I should say you must just be very careful because they can cause a lot of negative inotropy together, hypotension.

  2. The heart rate often adds to the differential diagnosis of a tachyarrhythmia. It is interesting here, as 150 bpm is slow for an AVnRT. The rate fits VT even moreso than it does AVnRT. Moreover, VT can also be accompanied by retrograde p waves. The key is really that the rhythm converted with adenosine, prior to that, it is difficult to absolutely rule out VT (I think...)

    Help me understand this statement: "There is a definite Right Bundle Branch Block and Left Anterior Fascicular Block pattern, so this is not VT"--is that 100% correct? A Bundle Branch Reentrant VT can also have this pattern, or am I wrong?

    1. Meghan,

      Yes, PSVT is usually faster, but not always, especially in the elderly. Response to adenosine by itself does not make the diagnosis: there is RV outflow tract VT which is adenosine sensitive. But it has a different morphology with an inferior axis. See here: RVOT VT is also rare.

      It is possible to have RBBB and LAFB morphology in VT if it originates in the posterior fascicle (posterior fascicular VT), as in this case: So on first glance this could be posterior fascicular VT, but it is relatively very rare, so I would only hold that possibility in the back of my mind. Also, posterior fascicular VT does not respond to adenosine, but does respond to verapamil!

      Finally, we know the diagnosis is SVT with aberrancy because we see the same QRS morphology when the rhythm is sinus.

      Thanks for the great question.

      Steve Smith

  3. Hi Steve - You say there is "potential for significant negative inotropic effects" with diltiazem and verapamil. Are there not also negative inotropic effects with beta blockers? You used in this case without mention of concern for negative inotropy? I ask because it is often said to avoid calcium channel blockers for rate control in tachycardias among heart failure patients. I have never been able to find solid grounding for this statement. Why would beta blockers be okay for tachycardia but not diltiazem in CHF? UpToDate says of rapid afib "both diltiazem and verapamil have negative inotropic effects...should be used with caution in heart failure." Yet it gives no reference. Is there any clinical grounds for this? Or are these statements based on animal models? Why the concern for CCBs but not BBs? Thanks in advance.

  4. Jason,
    I use diltiazem freqently in atrial fib with RVR. It is relatively safe, but does cause hypotension a significant amount of the time. You must be careful with it. Verapamil even more so. Both Ca blockers and beta blockers equally cause negative inotropy and hypotension in these patients. That is why I used esmolol. If there are adverse effects, you can turn it off and its effect is gone in less than 10 minutes.
    Does that answer the question?

  5. Yes, thanks. Your answer squares with my practice. I, too, use CCBs and BBs in afib with RVR. Not sure why some literature seems to urge caution with CCBs but not beta blockers. Thanks for your terrific blog.

  6. Great post as usual!

    There may be a typo in the esmolol dose though: I guess your loading dose is 500mcg/kg, not 500mg/kg.

    I sometime use CCB for adenosine-refractory SVT, rapid AFIB or CCB-sensitive VT. I give it by slow (20 minutes) infusion, as described in the below paper and have yet to observe a single episode of hypotension.

    Best regards.

    Lim, S. H., Anantharaman, V., Teo, W. S., & Chan, Y. H. (2009). Slow infusion of calcium channel blockers compared with intravenous adenosine in the emergency treatment of supraventricular tachycardia. Resuscitation, 80(5), 523–528.

  7. Thanks for this Great explanation. I had similar case before where I thought of Atrial tachycardia due to the recurrence of SVT after adenosine and electrical cardioversion but responded to Verapamil later. I kept in mind if SVT converted to sinus after adenosine (temporarily) then SVT recurs, its against re-entry mechanism and I should think of Ectopic atrial tachycardia !! is that true (narrow complex only, excluding VT just to make it simple)

    1. No, that is not true. All Re-entrant tachycardias need to be initiated, and if they are converted, they can always initiate again. Many PATs are re-entrant as well (some automatic, some re-entrant). MJ Perrin (see below) thinks this might be PAT. I don't think so.

  8. do you ever use calcium gluconate before ccbs to prevent hypotension?

  9. Interesting case. I wonder if the P wave is actually positive – the contour of the ST/T wave in lead III and aVF suggests this. In tachycardia the segment is predominantly positive in aVF, and clearly negative in sinus. Similarly lead III looks like it has a positive wave superimposed on the ST/T when comparing between sinus and tachycardia. If positive, it can't be AVNRT or VT. I think the arrow is pointing to late activation of the QRS – often see that with RBBB due to late RV activation.

    Then the differential is between AT and AVRT (accessory pathway). The latter is not so common in the elderly - there is a natural attrition rate for accessory pathways with each passing decade. So AT most likely? Response to amiodarone and adenosine would be consistent.

    Would be interesting to study! I think finding the P wave is the most difficult aspect of arrhythmia differentiation. I've been surprised in a couple of cases to find out at EP study that what I thought was the P wave was just a notch of the QRS or T wave, and vice versa. Perhaps I'm wrong again, but just my 2 cents.

    1. MJ,
      It's certainly possible. I'm skeptical. Thanks for the great input! I guess we need a EP study to know for sure.
      Steve Smith

    2. Good eye! I think the buried P-waves are indeed later than indicated, though the contour of the T-wave in III and aVF might give them a bit more size than they are due. I'm most convinced by the tiny pseudo-bifid T-wave in II whose first peak marches out in every single lead.

      It's a tough call (and I'm rubbish at telling which way is "up" with buried P-waves), but it does look to me like the P-axis is roughly around 120 degrees and seemingly inconsistent with AVNRT.

      An interesting case gets even more interesting!

    3. I don't see it. But fortunately it is academic.

  10. When I gave verapamil long ago, I always used calcium pretreatment. However, there is a good article showing no effect of calcium prior to diltiazem.


DEAR READER: We welcome your Comments! Unfortunately — due to a recent marked increase in SPAM — we have had to restrict commenting to Users with a GOOGLE Account. If you do not yet have a Google account — it should not take long to register. Comments give US feedback on how well Dr. Smith’s ECG Blog is addressing your needs — and they help to clarify concepts of interest to all readers. THANK YOU for your continued support!

Recommended Resources