Amiodarone: Push It, Push It Real Good


Amiodarone is one of the most pimpable drugs for pharmacy residents/students. From its pharmacology to excipients, there are so many nuances, important points, and pharmacy trivia. In terms of administration, we’re all taught to dilute and filter while administering amiodarone intravenously. In almost all circumstances, this is referring to administration of amiodarone to patients with a pulse. But when we’re at the bedside of a patient with pulseless ventricular tachycardia or fibrillation (pVT/VF) the question becomes should amiodarone be pushed undiluted, or further diluted with D5W?
Naturally, there are conflicting answers and practices. According to the prescribing information, amiodarone MUST be diluted with D5W to concentrations between 1 and 6 mg/mL (for a 300mg load, that would be between 50mL to 300 mL).1 Furthermore, anything more concentrated than 2 mg/mL (that would be 300mg in 150mL) should be administered via a central venous catheter. There are obvious problems to doing this in a cardiac arrest scenario.

What’s the risk of undiluted amiodarone?
Rapid administration of amiodarone is associated with hypotension, based on animal and some human data.2-4 There are three leading theories as to the cause of the hypotension:

1) Excipients in intravenous forms of amiodarone, which are polysorbate 80 and benzyl alcohol (for every 50 mg of amiodarone there is 100mg of polysorbate 80, and 20.2 mg benzyl alcohol).
2) Amiodarone’s beta-blocking/calcium channel blocking properties.
3) Local allergic reaction to one or many of amiodarone’s components.

While the overall quality of the data describing this problem is poor, there are some interesting findings that may support a more rapid, undiluted administration.

A commonly cited paper that supports the undiluted administration of amiodarone, is unfortunately, often misquoted. This study examined the hemodynamic effects of rapid, undiluted amiodarone bolus in pulseless patients was this paper from Finland.5 This was a retrospective study investigating the occurrence of side-effects and outcomes of the patients who received undiluted amiodarone during resuscitative efforts in Helsinki during a 2-year period. The local protocol had been updated to no longer stipulate further dilution of amiodarone (300mg) prior to administration in patients who after three ineffective shocks, one sequence of CPR, and epinephrine 1mg. The protocol included a rapid bolus of approximately 200 mL of Lactated Ringer’s immediately following the amiodarone bolus.

This study showed no difference in blood pressure between patients who received amiodarone versus those who did not. However, that’s not really an appropriate comparison since the populations were too heterogeneous (pVT/VF vs PEA vs asystole) and fails to answer the clinical question at hand. It would have been better to compare a diluted amiodarone versus an undiluted amiodarone bolus in patients with pVT/VF.

Since there is no data that directly answers this question, extrapolating other data sheds some light on the magnitude of the risk.

Munoz et al looked at 20 patients undergoing coronary arteriography who received 5 mg/kg IV of either amiodarone or a polysorbate 80 (aka Tween 80) free amiodarone product.6 Looking at the left ventricular systolic pressure three minutes before and three minutes after administration, the authors found that the polysorbate 80-free version was associated with a significantly smaller decrease in this outcome (amiodarone: 110 + 11 to 86 +/- 11 mmHg; polysorbate 80 free amiodarone: 114 +/- 22 to 106 +/- 19, P = 0.01).  A drop in SBP to the 80s is certainly concerning, however, the dose in this study is much larger than used in routine clinical practice for pVT/VF. For the average 80kg adult this dose would be 450mg. For the average Texan I see, the dose would be 500mg. While this evidence supports the animal models suggesting a hemodynamic effect of rapid undiluted amiodarone, it still does not describe the risk to the patient population I've been referring to.2-4

While the evidence is unclear regarding potentially clinically significant vasodilation/hypotension in patients in pVF/VT, it seems as though further dilution and slower infusions do nothing to reduce the incidence of hypotension. In a retrospective cohort analysis, patients that received IV loading doses of conventional amiodarone (polysorbate 80/benzyl alcohol) were compared with patients who received amiodarone with cyclodextrin.7  These patients did not receive “code dose” boluses, but rather, received the initial infusion load (1 mg/min for 6 hours, followed by 0.5 mg/min for 18 hours). Despite the lower rate of infusion, and dilution, there was still a statistically significant difference in incidence of hypotension across all measurements: 0-6 hours, 12-18 hours, 24 hours and requirement of fluid boluses.

The more recent PROCAMIO study, best described elsewhere, also used 5 mg/kg dosing for amiodarone administered over 20 minutes (unclear if the dose was diluted or not).8 Compared to procainamide, amiodarone was associated with more severe hypotension requiring immediate electrocardioversion (amiodarone, 6 patients versus 2 receiving procainamide).  Although, there were more patients in the procainamide arm experiencing hypotension not requiring cessation of infusion (procainamide, 5 patients versus 2 in the amiodarone group).

So where do I land on this? Since we have the backing of AHA, in a pVT/VF scenario, it is reasonable to continue to administer the 300mg bolus of amiodarone undiluted.9 However, in any other case, consider dilution/filter/slow(er) rate of administration. Alternatively, consider stocking non polysorbate 80 containing formulations since they appear to have a smaller hemodynamic impact.

(Side note, cyclodextrin containing amiodarone may not contain polysorbate 80, but must still be filtered during infusion).

  1. "Amiodarone injection [prescribing information].", Irvine, CA: Teva Parenteral Medicines, Inc., 2008
  2. Platou ES, Refsum H. Acute electrophysiologic and blood pressure effects of amiodarone and its solvent in the dog. Acta Pharmacol Toxicol (Copenh) 1986;58:163-168
  3. Gough WB, Zeiler RH, Barreca P, El-Sherif N. Hypotensive action of commercial intravenous amiodarone and polysorbate 80 in dogs. J Cardiovasc Pharmacol 1982;4:375-380
  4. Somberg JC, Cvetanovic I, Ranade V, Molnar J. Comparative effects of rapid bolus administration of aqueous amiodarone versus 10-minute cordarone I.v. infusion on mean arterial blood pressure in conscious dogs. Cardiovasc Drugs Ther. 2004 Sep; 18(5):345-51
  5. Skrifvars MB, Kuisma M, Boyd J, Määttä T, Repo J, Rosenberg PH, Castren M. The use of undiluted amiodarone in the management of out-of-hospital cardiac arrest. Acta Anaesthesiol Scand. 2004 May;48(5):582-7
  6. Munoz A, Karila P, Gallay P, et al. A randomized hemodynamic comparison of intravenous amiodarone with and without Tween 80. Eur Heart J 1988;9:142-8
  7. Lindquist DE, Rowe AS, Heidel E, Fleming T, Yates JR. Evaluation of the Hemodynamic Effects of Intravenous Amiodarone Formulations During the Maintenance Phase Infusion. Ann Pharmacother. 2015 Dec;49(12):1317-21
  8. Ortiz M et al. Randomized Comparison of Intravenous Procainamide vs. Intravenous Amiodarone for the Acute Treatment of Tolerated Wide QRS Tachycardia: the PROCAMIO Study. Eur Heart J 2016
  9. Kleinman ME, Brennan EE, Goldberger ZD, Swor RA, Terry M, Bobrow BJ, Gazmuri RJ, Travers AH, Rea T. Part 5: adult basic life support and cardiopulmonary resuscitation quality: 2015 American Heart Association guidelines update for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation. 2015;132(suppl 2):S414–S435

    Practicing emergency medicine in New Zealand: A Canadian’s perspective

    Dr. Rob Woods is a well known Emergency Physician and program director for the FRCPC Emergency Medicine Program at the University of Saskatchewan, who recently practiced emergency medicine in New Zealand on a one-year sabbatical. Here, he provides us some with answers to many questions he has been getting, insights into the country, lifestyle and medicine from abroad.  Why did you go to New Zealand? We also thought it would be great for ...

    The post Practicing emergency medicine in New Zealand: A Canadian’s perspective appeared first on CanadiEM and was written by Rob Woods.

    I was texted this ECG: "Middle-aged male with Prehospital Cath Lab Activation"

    A physician caring for his new arrival, a 50-something with acute chest pain and dyspnea and syncope, texted me this initial ED ED ECG. 

    The computer read was "Marked ST Elevation, ***Acute MI***" 

    No previous ECG was available.
    What do you think?
    Computerized QTc = 399 ms
















    My response was: "Normal variant"

    Question: "De-activate cath lab?"

    Answer: "It does not look like myocardial infarction".

    He sent the prehospital ECG.  This also was read by the computer as "***Acute MI*** and was the basis for the prehospital activation.

    Here it is:
    My response: "Normal Variant"







    I added: "One must always remember that there can be a coronary occlusion in the presence of a normal ECG.  So I cannot tell you that this patient does not have a coronary occlusion, but I can say that it does not show on the ECG."

    He de-activated the cath lab.

    The patient ruled out for MI by 4 serial troponins below the level of detection.



    Let's look at the ED ECG more closely:
    Sinus rhythm
    There is ST Elevation in anterior leads: V2-V6
    There is ST elevation in I, II, aVL

    Why is this not STEMI?

    First, I prefer to use my formulas and expertise to recognize coronary occlusion when it was not previously diagnosed or suspected.  I am reluctant to reverse a diagnosis of ***STEMI*** because I would rather there be a false positive cath lab activation than a false negative.

    However, with some practice and expertise, this can be done with very high accuracy.

    First, let's address the precordial STE:
    1. The T-waves are not tall: the T/ST ratio is low (T-wave not much taller than ST segment)
    2. High R-wave voltage (the single best predictor of whether subtle precordial STE is due to LAD occlusion or not!)
    3. Very well-formed J-waves in V4-V6

    You could consider using the formulas for differentiation of subtle LAD occlusion from normal variant STE.

    One problem with this: there is a non-concave ST Segment in lead V2.  This is rarely seen in early repolarization (normal variant), and such cases were excluded from our study.  However, in this case, I happen to know from simple expert recognition that this ECG is an exception to the rule.

    Using QTc = 399
    RAV4        = 19
    QRSV2     = 23
    STE60V3  = 2.5

    3 variable = 20.34 (very low, most accurate cutoff is 23.4)
    4 variable = 14.84 (very low, most accurate cutoff is 18.2)

    Second, let's address the limb lead STE

    Inferior MI?  No
    1. The inferior STE is limited to lead II, and has significant PR depression
    2. There is no reciprocal ST depression in aVL, which is present in 99% of inferior MI.

    High lateral MI?  No
    1. The STE in I and aVL are associated with flat T-waves
    2. There are well formed J-waves
    3. There is no reciprocal ST depression in lead III.


    ECG Diagnosis:

    Normal variant ST Elevation vs. Pericarditis.  No evidence of MI (which is different than saying that the patient does not have MI)

    Final Diagnosis:

    No MI.  Uncertain whether there is pericarditis or normal variant. 

    I favor normal variant for several reasons
    1. No active chest pain
    2. No objective signs of pericarditis (no rub, no effusion, no positional pain)
    3. Normal variant is far more common
    4. It simply looks to me like normal variant
    5.  I believe that in this age of highly sensitive troponins that any ST elevation caused by pericarditis should be associated with some release of troponin.  Here all were below the level of detection.