Amiodarone

This chapter is relevant to Section G8(iv)  of the 2017 CICM Primary Syllabus, which asks the exam candidate to "understand the pharmacology of antiarrhythmic drugs".  Specifically, amiodarone is on the menu, because it keeps coming up in the CICM Part One exams. 

• Question 11 from the second paper of 2021
• Question 2 from the second paper of 2018
• Question 11 from the first paper of 2016
• Question 13 from the second paper of 2014
• Question 22 from the second paper of 2010
• Question 5 from the second paper of 2008 

This abundant attention is easily explained by the ubiquitous applications of this popular broad spectrum antiarrhythmic. Amiodarone hits every note one the Vaughan Williams scale, and is effective in the management of both atrial and ventricular rhythm abnormalities. Additionally, it has  it is remarkably safe in the haemodynamically unstable patient, and it is available as a parenteral infusion, given slowly which offers the chance to change one's mind. In short, it has all sorts of favourable characteristics. It also has properties which are somewhat less charming, and which have made it somewhat unpopular in the management of chronic rhythm problems out in the community. These will also be discussed, as CICM examiners are fond of asking about them. 

In summary:

Amiodarone is popular, and literature describing it is abundant.  For a detailed explanation of its electrophysiological effects, Kodama et al (1997) is the best free option, and for clinical use, Hamilton et al (2020) has a thorough review of the evidence. Overall, if one needed a single reference, Kowley et al (1997) is free and just short enough that you can get everything you need quickly, in the froth of last-minute pre-exam panic. Unfortunately, it is written rather well and has excellent references, which means one really needs to concentrate on remaining incurious, lest one become overwhelmed with the desire to read more.

Chemical structure and chemical relatives of amiodarone

"Chemical relatives" In terms of pharmacological relatives, amiodarone has several among Class III antiarrhythmic agents, with which it is inexplicably grouped (because how would you choose one class in which to put it?). Dronedarone, another benzofuran molecule, is a similarly multi-class antiarrhythmic, with the advantage of being non-iodine-containing - but unfortunately, with the disadvantage of increasing mortality in virtually every clinical application. Other less lethal Class III drugs include ibutilide, dofetilide, vernakalant and bretylium. 

Chemically, amiodarone is an iodinated benzofuran derivative (benzofuran being a fuzed double-ring molecule made of a six-carbon benzene and a five-carbon furan ring). In case you are wondering, benzofuran is an oily liquid which is naturally found in coal tar, and its properties have absolutely nothing to do with those of amiodarone per se, but it sure does have numerous useful derivatives. One merely needs to reach for one's dog-eared copy of Comprehensive Heterocyclic Chemistry (1996, the 2nd volume) to find multiple examples in the excellent review by Keys & Dibble. Without listing every possible permutation on the benzofuran theme, one can summarise by saying that the family contains numerous pesticides and herbicides, dyes, photographic chemicals, polymer plastics, as well as more familiar macrolides (ivermectin). These are all distant cousins. First degree relatives include benzbromarone (a uricosuric), and brofaromine, a weird nootropic MAO inhibitor used in the management of dementia. Unlike amiodarone, their preferred halogen is bromine.

Amiodarone went with iodine. In fact it has two iodine atoms per molecule. That might not seem like much, but one must recall that iodine is far down along the periodic table, and in amiodarone those two atoms iodine account for 37.5% of its total molecular weight. That's probably also unimportant.

What, then, might be an important chemical tidbit to remember about its structure? That would probably be the uncanny resemblance to thyroid hormones. These molecular structures were "borrowed" from van Erven & Schalij (2010), because they illustrate this best:

That can't be good, you might think. Indeed, it is not. Apart from directly interfering with TSH and T₃/T₄ activity, this impostor molecule also gets metabolised, liberating iodide ions. As one might imagine, this iodide excess is as vast as it is unwelcome.  You only need about 150-200 μg of iodide per day to enjoy a goitre-free life. In contrast, a 900mg infusion of amiodarone might liberate about 30mg of iodide (and we often might give around 1-2g of this drug to people over the course of a couple of days)> Thus, the treatment dose of amiodarone is well in excess of the daily recommended iodine intake, by a factor of perhaps 200. Obviously that can't be benign, if you pause even briefly to think about the chemical properties of elemental iodine. But more on this later. 

Pharmacokinetics of amiodarone

The pharmacokinetics of amiodarone are sufficiently unique to constantly attract the attention of examiners, from numerous disciplines and particularly from critical care specialities. 

Routes of administration

Amiodarone is only available as an oral or IV formulation. There does not seem to be any merit of giving it in any other way, nor is there any literature out there on any more adventurous methods of administration (for example, nobody seems to have ever had an amiodarone enema). In case it needs to be given down a finebore tube, it is nicely shelf-stable as a syrup. 

Intestinal absorption

Amiodarone has slow and erratic GI absorption. Maximum plasma levels in healthy volunteers were achieved about 4.5 hours after the pills were swallowed (Pourbaix et al, 1985). Oral bioavailability is variably reported as anything between 20% and 80%, give or take; some drug is lost to a very random first-pass effect. 

Diluent, excipients and compatibility

Intravenous amiodarone was difficult to formulate. Andreasen et al (1981) recorded their travails, when they tried to create an IV mechanism of delivery and found that the molecule of amiodarone was inherently unstable in aqueous solution. "When 5 μg/ml was added to a protein-free physiological buffer solution pH = 7.4, after 12 min at room temperature we found that the concentration had fallen by 20%", the authors lamented. After four days in the fridge, only 10% of the drug remained intact. They resorted to storing it in 50% ethanol for the purpose of their experiment, but of course that was not going to be popular with clinicians (for one, it would eat its way out of IV tubing). To reduce the alcohol content for the market, they eventually settled on polysorbate 80 as a stabiliser. In every ampoule, there is now about 100mg polysorbate 80 and 20.2 mg of benzyl alcohol per every 50mg of amiodarone. Though polysorbate 80 is safe and ubiquitous (ice cream, eye drops, etc), it is not exactly an inert noble gas, and has specific biological effects. Gough et al (1982) found that it is responsible for the hypotension one can sometimes see when giving amiodarone quickly (their specific dog experiment delivered 5mg/kg over 5 minutes, which is probably faster than most human applications).

Amiodarone is incompatible with a great many things. For one, it adsorbs onto the surfaces of PVC bags and IV tubing. It could theoretically be expected to play nice with a range of IV drugs, but most of the time local pharmacy protocols insist that it be administered via a dedicated lumen, and not mixed with anything. Classically, it is administered with 5% dextrose as a diluent, as there is a widespread belief that it is incompatible with saline. Early studies determined that the chloride ions in saline interfered with its solubility, increasing the risk of precipitation. This may be something of a myth, as Campbell et al (1986) tested dextrose and saline solutions of amiodarone over 24 hours and found that active drug was well preserved in both. However, judging by the fact that forty years later we are still using dextrose suggests that nobody read that issue of J Cardiovasc Pharmacol.

The importance of central venous administration

Though in extremis one tends to give IV amiodarone via a peripheral cannula, for longer term infusions central venous access is preferred, because amiodarone is a serious acidic irritant. Even a generous dilution (900mg in 500ml of 5% dextrose) ends up with a pH of 4.14-4.65. If you insist on giving it peripherally, you should expect a 40% incidence of phlebitis (Norton et al, 2013). Extravasation injuries can be horrific and disfiguring (made worse by the fact that often the patient is receiving the amiodarone for a problem which also makes them a poor anaesthetic candidate for skin grafting). In short, one should feel anxious to transition the patient to oral dosing wherever central access is not available.

Distribution

Amiodarone is poorly soluble in water, but highly soluble in nonpolar solvents. Its pKa is 6.56.  Once in the bloodstream, it is highly protein-bound (96%), which is generally reflective of its tendency to bind to everything everywhere.

Amiodarone distributes extremely widely, with authors quoting VOD values in the range of 66-144 L/kg. Where does it go? An excellent rat-themed study from Latini et al (1985) presented this chart, apparently from human data:

As you can see, fatty tissue is by far the richest source of amiodarone, and can become significantly saturated. If one's normal-shaped human body contains about 10kg of fat, that 100 μg/g of drug would translate to about 1g of amiodarone, hanging around in the fatty tissue.  In some cases, this reservoir is much higher. Harris (1983) reported a liver concentration of 1020 μg/g in the biopsy sample of a patient with deranged LFTs; following the ratios described in the graph one can imagine that it would have been ten times higher in their adipose tissue. As one might imagine, following the withdrawal of long term therapy, this cache takes a long time to empty, giving amiodarone an elimination half-life as long as a hundred days. 

Metabolism

Virtually none of the administered dose escapes through the body unchanged, and none is eliminated in the urine. Amiodarone is metabolised by CYP3A4 through oxidative deethylation. The desethylamiodarone formed by this process is actually active to about the same extent as the parent molecule, except even more tenacious (with a longer half life and greater tissue binding). Apart from this deethylation, numerous possible pathways of metabolism are listed in the literature, but none seems clinically important. CYP3A4 does have importance, because it is inhibited and activated by numerous substances, which would obviously affect amiodarone clearance. More on that later. 

Clearance and half-life

It is somewhat difficult to discuss half-life for this drug. Obviously, depending on how and when it is measured, one could come up with very different numbers, and it is difficult to say which number is actually clinically relevant. There is a very short initial distribution half-life, where the drug disappears from the bloodstream by distributing into tissues. You would probably say that this is an irrelevant variable as the plasma levels of amiodarone are not really connected to its magnitude of effect.  This initial distribution is followed by an impossibly long elimination half-life, as it leaches out of all the tissues and is slowly scrubbed out by the liver. Again, this would not be representative of anything clinically important, as the redistributed amiodarone would only achieve a very low concentration. To illustrate, this image is borrowed from Holt et al (1983):

So, what do you do, when they ask you "what is the half-life of amiodarone"? There is no way to weasel out of it, especially as CICM examiners clearly demand precision. In Question 13 from the second paper of 2014, their one-liner comment chided lackadaisical candidates who "lost marks for being too approximate on the pharmacokinetics". But, how can you be precise, if the literature offers a massive range of half-lives? Fortunately, back in Question 5 from the second paper of 2008, the examiners gave us a half-life which - now that it is printed in the college answers - could not possibly be marked wrong. 

It's 29 days.

Pharmacodynamics of amiodarone

Amiodarone is all things to all people. As mentioned above, it is classified along with sotalol and bretylium in Vaughan Williams Class III, but realistically it could have been listed as a Class I, Class II or Class IV agent just as easily. It does not help that it has markedly different effects depending on how it is administered, and at what dose. What follows is an attempt to unravel this mess of effects in a way which, if it does not bring clarity to the forum, will at least leave the reader with enough references to scratch their head over.

Molecular target and mechanism of effect

Amiodarone hits numerous receptors, which would be best categorised in order of Vaughan Williams class:

• Class I effect: voltage-gated sodium channels are blocked by amiodarone, with what appears to be some selectivity for the inactivated channels, and particularly for those in ischaemic tissue. According to Mason et al (1984), it markedly decreases the upstroke velocity of Phase 0, and has intermediate dissociation kinetics which result in use-dependent block. It also produces a prolonged action potential, mainly by its Class III effect. This combination of features would have to make it a Class Ia agent, along with quinidine and procainamide. The reader is reminded that this subclassification is heretical, as it is not mentioned anywhere in the literature, and so it should not be repeated in the exams. 
• Class II effect: Amiodarone is desribed as a beta-blocker, but it does not "block" the beta-receptor per se. In fact it also has alpha-"blocker" effects,  but neither phenomenon involves a classical interaction, where it competes with endogenous ligands for the catecholamine recognition site on the receptor. According to Ghuran & Camm (2000), it is a non-comptetitive blocker, which also happens to markedly reduce the number of catecholamine receptors on the myocardial muscle, as well as interfering with the intracellular regulatory unit of adenylate cyclase. The resulting effect resembles beta-blockade, but really is nothing like it (for one, the cAMP effect renders amiodarone overdose victims immune to the positive inotropic effects of insulin and glucagon). The positive upshot of this is a suspected synergy between amiodarone and beta-blockers, particularly where rate control is called for.
• Class III effect: Like sotalol, amiodarone is classified as a Class III agent, if you had to commit it to any single category. This effect is mediated by potassium channel blockade, which slows the rate of repolarisation and prolongs the refractory period. The longer action potential and delayed repolarisation are reflected in the prolonged QT interval, which is often seen with these drugs. Interestingly, the refractory period prolongation is said to happen only with medium-to-long-term use, where you basically saturate every myofibril with amiodarone. Acute IV dosing of amiodarone really only seems to affect the refractory periods of AV nodal cells and other sensitive conducting tissue. Thus, reliable sources indicate that "the class III action of the drug may be the least identifiable during the early phases of intravenous therapy".   As such, you could say that most of the clinical effects of an IV amiodarone infusion for rapid AF are related to its Class II and Class IV effects.
• Class IV effects: Amiodarone appears to have a nonselective L-type calcium channel blocker effect, analogous to that of verapamil and diltiazem. Lubic et al (1994) found that it also displaced nitrenidipine from its binding sites, explaining its mild peripheral vasodilator effect in high doses.

Clinical effects

In summary:

• With rapid infusion, excipient effects:
  • Hypotension due to decreased contractility (polysorbate 80)
• With acute use, Class II and Class IV effects:
  • Delayed AV conduction
• With chronic use, Class Ia and Class III effects
  • Prolonged repolarisation and QTC duration
  • Prolonged action potential duration
  • Decreased reentry, as the result of this
  • Increased risk of early afterdepolarisations
  • Minimal effect on QRS

Clinical use of amiodarone

Armed with a bucket of amiodarone, you mount the parapet against a horde of premature ectopic complexes. But is it useful? Let's look through some of the accepted indications for amiodarone, and their supporting evidence.

Cardiac arrest

Amiodarone is still recommended for the management of refractory VF in cardiac arrest. We say "still", because this guideline is tenuous, and other antiarrhythmics (eg. lignocaine) have been eliminated from ALS protocols. This recommendation comes from a 1999 clinical trial (Kudenchuk et al), in which the amiodarone group appeared to have about 10% better survival to admission. That is to say, amiodarone is credited with the finding that more of these patients lived long enough for the ward clerks to process their admission paperwork. Later data has not demonstrated any actual meaningful survival benefit, but the AHA still favours it with a Class IA recommendation.

Ventricular tachycardia

In most other countries, for the treatment of not-yet-dead VT, amiodarone is second line and procainamide is the first choice. Procainamide has compared so favourably with amiodarone in the PROCAMIO study (Ortiz et al, 2017) that it has become the first choice for many people. In Australia of course for some unknown reason procainamide is not available and we resort to using amiodarone. For chronic prevention, it is the antipodes' first line for patients prone to ventricular arrhythmias who have decreased LV function (so that these people can go five minutes without getting slammed by an AICD discharge).

Atrial fibrillation

In the ICU, where we cannot afford the haemodynamic effects of beta-blockers and where the vagotonic effects of digoxin will be ineffective, amiodarone is often the agent of choice for control of acute onset AF. We are also often less concerned about cardioverting the patient, as the AF is often witnessed, and stroke is unlikely. Moreover, it is not contraindicated in patients with severe structural heart disease or heart failure, like many other agents. The Australian tendency to throw amiodarone at critically ill patients may be a completely non-evidence-based feature of local ICU folklore, and there appears to some sort of a geocultural heterogeneity in its popularity. For instance, Hamilton et al (2020) report that in the United States, "CCBs are administered as initial treatment most frequently (36%) for patients with AF and sepsis, followed by BBs (28%), digoxin (20%). and amiodarone (16%)". In contrast, over 80% of British ICU physicians confessed to choosing amiodarone as their first line agent. To emphasise again, there is no RCT level evidence to guide management here. For chronic AF, where evidence is bountiful, other agents are more suitable, and amiodarone tends to be viewed as second or third line therapy, for patients refractory to everything else.

​​​​​​Contraindications

Apart from some sort of genuine iodine allergy, there are no absolute contraindications to its use. A few relative contraindications could be conjured up, in case anybody ever gets asked to produce a list. For example:

• An already prolonged QT
• AF of unknown duration, or , ever worse, 
• A known atrial appendage thrombus
• Second or third degree heart block
• Severe hyperthyroidism or hypothyroidism

Toxicology of amiodarone

Adverse effects of amiodarone with short-term use in the ICU are virtually unknown. Its mild vasodilator and negative inotrope effects tend to get lost in the positive haemodynamic consequences of controlling rhythm and rate, and we never tend to give it so quickly as to see that polysorbate-related hypotension. Chronic use, on the other hand, has a massive host of toxicities, which the college have asked about on multiple occasions. It is therefore surprising that SAQs like Question 11 from the second paper of 2021 seem to take exam candidates by surprise (in that specific instance only 17% passed).

Adverse effects with acute and chronic use

It is probably better to create your own system for this, so as to make it easier for yourself to remember them, but in case you don't want to, van Erven et al (2010) have this table with some incidence values, which has been somewhat expanded here:

• Cardiac:
  • Bradycardia (5%)
  • Torsades de pointes (<1%)
  • Hypotension with rapid IV administration (due to polysorbate 80)
• Pulmonary
  • Pulmonary fibrosis (1.6%)
• Dermatological
  • Photosensitivity (25-75%)
  • Blue skin discolouration (<10%)
• Ophthalmic
  • Corneal deposits and cataracts (100%)
  • Optic neuritis (<1%)
• Thyroid
  • Wolff-Chaikoff effect: iodine-induced hypothyroidism (1-23%)
  • Hyperthyroidism can also occur (1-32%)
• Hepatic
  • LFT derangement (15%)
  • Hepatitis and cirrhosis (<3%)
• Neurological
  • Ataxia, tremor, peripheral neuropathy
  • Sleep and memory disturbances
• Urogenital
  • Epididymitis
• Pharmacokinetic
  • By binding to albumin and other protein, amiodarone can displace other highly protein-bound drugs, and increase their free fraction (increasing their effect)
• Obstetric
  • Penetrates the placenta, causing iodine overload and hypothyroidism of the foetus, causing neurodevelopmental abnormalities
  • Excreted into breast milk

Drug interactions

Amiodarone has predictable physiological interactions with other antiarrhythmic drugs, which one can basically describe as an undesirable combined effect. Your rate and rhythm might end up being controlled too well. Additionally, it is an inhibitor of P-glycoprotein, an efflux pump involved in the elimination of digoxin, which means that people on combined digoxin/amiodarone therapy will have unexpectedly elevated digoxin levels (Fenster et al, 1985). It also acts as an inhibitor for CYP 3A4, which means the following common drugs will have impaired clearance:

• Warfarin
• Atorvastatin and simvastatin
• Flecainide
• Quinidine
• Phenytoin
• Cyclosporin

Toxicity in outrageously vast overdose

The slow gastric absorption and rapid distribution of amiodarone is something of a blessing if one ever decides to take an unreasonably large amount of it. However, accidents do happen. One might expect these to have a  rather unpleasant cardiovascular effect. Additionally, the therapeutic range is relatively narrow. For example Haas et al (2008) reported a case of rescue ECMO being required to save a newborn boy who accidentally received 15mg/kg of amiodarone (only three times the normal dose). It is not clear from the article whether he arrested or not, but multiple boluses of adrenaline and several periods of CPR followed, until finally ECMO was commenced. He was decannulated within 36 hours, and discharged home 9 days later. In case of intentional self-poisoning with oral amiodarone, activated charcoal appears to be rather effective, reducing amiodarone bioavailability by 50% even after 90 minutes.