Adrenaline

This chapter is relevant to Section G8(i)  of the 2017 CICM Primary Syllabus, which asks the exam candidate to "understand the detailed pharmacology of inotropes and vasopressors". Adrenaline is an endogenous catecholamine, a sympathomimetic drug that has less receptor selectivity than the others. It has appeared in three CICM primary exam papers, and promises to appear in the future. The typical trend of these questions has been "compare adrenaline with some other inotrope":

• Question 1 from the first paper of 2021 (adrenaline on its own)
• Question 2 from the first paper of 2018 (adrenaline vs. milrinone)
• Question 18 from the second paper of 2012 (adrenaline vs. the "ideal" inotrope)
• Question 8 from the first paper of 2012 (adrenaline s. levosimendan)

Of these, the most recent SAQ had a pass rate of 90%, which suggests this question has now lost some of its discriminatory properties, and may not be seen again for a while.

In short:

That "single best reference for further information" this is rather misleading, as the PI really does not inform you of anything useful. For a substance which has been around for 120 years and which is found in every arrest trolley operating theatre and dental surgery,  it is surprising that no definitive and all-encompassing monograph is available. It is ridiculous that a search for "pharmacology of adrenaline" should yield no single authoritative source from a reliable author.  Dalal (2020) from StatPearls is one quick overview, but not nearly enough to answer CICM questions. Gorain et al (2020) is neither quick, nor really an overview, and mainly focuses on adrenergic receptors. In short, there is no single definitive reference for what follows. An extensive bibliography is appended at the end, mainly so everyone can see how widely this information is scattered. 

Chemical structure and chemical properties of adrenaline

Adrenaline is a catecholamine, that is to say, a phenol ring with two hydroxyl groups and an ethylamine tail which is just one carbon atom longer than noradrenaline. Here are some molecular structures for comparison, where for some reason dopamine is also included:

The chemical and pharmacological relatives of adrenaline and the relationship of its structure to its function are discussed elsewhere. In short, that extra methyl group at the end of its amine tail takes away its alpha-selectivity, grants some beta-selectvity, and makes it a broad-spectrum sympathomimetic.

Adrenaline vs epinephrine

As (according to some unreliable corporate data) a vast dominant proportion of the readers come from America, a reasonable person would need to make some space for an aside regarding nomenclature here. Specifically, an aside to address the disagreement regarding what we should call this catecholamine, which has developed historically between Britain and one of its former colonies. Yes, sure, the nerds would invoke its true name (4-(1-Hydroxy-2-(methylamino)ethyl)-1,2-benzenediol) but for obvious reason that's not a name you could conveniently shout during a cardiac arrest. A punchy moniker is called for. Apparently, several of them.

One might look for some sort of historical precedent, i.e. let the discoverer name the drug, but that is somewhat fruitless. Napoleon Cybulski was the first to get some sort of haemodynamically active substances out of the adrenal gland in 1985, but his pureed gland extract was firstly not pure adrenaline, and secondly he decided to call it nadnerczyna, a word which in English is somewhat less than mellifluous, which in Polish means "a thing which comes from the adrenal glands", nadnerczy being the word for "adrenals".  Still, raw untreated nadnerczyna would probably pass for adrenaline in undiscerning hands, and thought chemically there was more work to be done, one might conclude that this name should have primacy.

As is typical for Eastern European scientists of the period, Cybylski's work went unrecognised while George Oliver and Edward Schäfer ended up being credited with the discovery of the haemodynamic properties of adrenal extract. The reason for this is probably just some good old fashioned colonial Anglocentrism. Oliver published in The British Medical Journal (September 14, 1895), whereas Cybulski published in the somewhat less-renowned Gazeta Lekarska (June 20th, 1895) which is probably the only reason we are not yelling about 10ccs of nadnerczyna during our cardiac arrests. This is something of a pity, as Cybulski was by all accounts a highly principled researcher, whereas Oliver's experiments would have never passed any human research ethics committee approval process. He basically injected his own son with some animal-derived adrenal extract:

"Oliver... applied an instrument for measuring the internal diameter of the radial artery to his young son. Having given his son an injection of an extract of the suprarenal gland, Oliver thought that he detected a change in the diameter of the artery."

In any case, Oliver and Schäfer never referred to this stuff as anything other than "extract" or "tincture", and the origins of it they referred to as "the suprarenal gland". Within the next twelve month, two more authors published on the same topic, who were much more inclined to give it a resonant name. Sigmund Fraenkel, working in Vienna, published an account of him extracting some kind of vasoconstrictive goop from adrenal glands which he called sphygmogenin, and Otto von Furth produced something he called suprarenin in Strasbourg. According to Ishida (2018), who wrote an entire book documenting this exact issue, the subsequent years yielded a completely unhinged orgy of nomenclature (mainly used in trade names) including adnephrin, adrenaline, adrin, supracapsulin, suprarenalin, atrabilin, chelafrinum, epirenan, hemostasisn, ischemin, paranephrin, renoform, suprenaphran tonogen and vasconstrictin.

So, how did we end up with Adrenaline and Epinephrine? Turns out, it was just lazy capitalism. Adrenaline (or "Adrenalin") was the commercial name of an adrenal extract marketed in Europe by  Parke, Davis & Co.  Though "epinephrine" was a well-known alternative name for a while, the institutional adoption of this term comes from the 1926 edition of the US Pharmacopeia, and takes its origin from a letter written by O. W. Smith, the president of Parke, Davis & Co, to Jokichi Takamine (who was the first to synthesise chemically pure adrenaline). In his letter, Smith encouraged the publication of "epinephrine" as a means of protecting his trademark:

"...We have suggested to the sub-committee... that the word "Adrenalin" is registered, is a valid trademark, and that under the circumstances the committee would probably use the word "Epinephrin" in the Pharmacopeia. We are afraid that if the word "Adrenalin" is used it may encourage manufacturers to use it also, whereas so far all of them have kept off the grass."

Thus, Adrenaline was the trade name of generic epinephrine in America during this formative period. In Europe, adrenaline was adopted because the prevailing view of eminent scientists was that "epinephrine" was in fact a name given to a physiologically distinct substance, i.e they contested the assertion that American epinephrine and British adrenaline were the same. Technically, that was true, because the first mention of epinephrine is traced to Abel (1896), who gave that name to the inactive benzoyl derivative he mistakenly extracted. Some might say the US Pharmacopeia was therefore wrong to use that term for their "proper" adrenaline, which was fully active and chemically distinct. However, this error has propagated far and wide, to the point that the official IUPAC generic name for this drug internationally is now "epinephrine". IUPAC is basically an American body, and it appears now that this term cannot be dethroned from its position of dominance. So, should we relabel all the drug cupboards and ampoules in Australia? At this late stage to change the name of such a well-known and critically important drug would present a serious public health risk and though "we shall, if we must, even get used to epinephrine ... eventually", writes Aronson (2000), "the dangers in changing the name from adrenaline to epinephrine will far outweigh any other problems during the lengthy changeover period". 

Pharmaceutical presentation

Adrenaline presents as an ampoule of either low volume highly concentrated solution (1:1000) or as a more dilute high-volume solution (1:10,000). These numbers are a shorthand for the dilution ratio of adrenaline (in grams) in water (in ml). Thus, 1:1000 is 1g in 1000ml, or 1mg/ml. The 1:10,000 ampoules are therefore 100µg per 1ml, or 1mg in 10ml. 

The ampoule usually contains sodium metabisulfite as an antioxidant as well as sodium chloride, sodium citrate, citric acid monohydrate and hydrochloric acid - titrate to a pH of 2.5-3.5, all to preserve this inherently unstable molecule. Without the extra acidification, it is practically insoluble in water. The ampoule glass is usually amber or brown, to protect the contents from ultraviolet light which tends to photodegrade adrenaline into adrenochrome-melanin (Polewiski & Slawinska, 1991). When it is administered, 6mg of the drug tends to be mixed with 100ml of 5% dextrose for reasons already mentioned elsewhere; even though this practice appears to be ritualistic rather than scientific, as adrenaline remains stable in saline for over ninety days without any loss of potency.

Stereoisomers and racemic adrenaline

Adrenaline is chiral, and therefore, when adrenaline is manufactured, one should expect to end up with a mixture of two enantiomers, dextro and levo.  The natural human form is levo-adrenaline, which has greater activity; whereas the dextro-rotary form is about fifteen times weaker (Rice et al, 1989). In general, the adrenaline (and noradrenaline) that you get from the shop will be the levo-isomer, with dextro-adrenaline present as a 1% impurity, and you'd have to specifically call for "racemic epinephrine" if you really wanted it for some reason.

Pharmacokinetics of adrenaline

Adrenaline can be given by a surprisingly large number of routes:

• Intravenously
• Intramuscular (eg. for anaphylaxis)
• Subcutaneous (eg. with local anaesthetic)
• As a nebulised inhaled agent (eg. for croup or post-extubation stridor)
• Topically, as a spray vasoconstrictor for bronchoscopy
• As eye drops, for glaucoma
• Squirted down the ETT, in a cardiac arrest (when no other route is available)

What if I drink it?

Like noradrenaline, adrenaline is not absorbed very well, and has a tendency to get metabolised by brush border COMT and MAO. This has not stopped people from trying to find an enteric means of administering it, particularly in the early 20th century. The results were broadly disappointing. Menninger (1927) records an entire history of this, and his opening paragraph reads like a list of different ways to say "it doesn't work":

"Sollmann states that "it is entirely ineffective by oral administration" and that "the blood pressure response is practically absent" when given by mouth. In "Useful Drugs"  it is stated that, "when given by mouth, it produces no evident effect on the general circulation." Cushny states that "when given by mouth, it has no effect on the blood pressure or heart." According to an article in the "Pharmacology of Useful Drugs," "epinephrine is one of the few alkaloids in the materia medica which are not absorbed from the gastrointestinal canal with such rapidity as to induce any appreciable effect." In the recent series of articles on glandular therapy in The Journal of the American Medical Association, it was stated, regarding epinephrine, that "when the drug is administered by mouth, it is quickly destroyed, either by the gastric or by the intestinal secretions, and its systemic effects are consequently not obtained."

Central and peripheral administration

Potent vasoconstrictors such as adrenaline and noradrenaline should probably be given via a central line wherever such a line is available. Though given the choice most people would choose the CVC, peripheral adrenaline does still happen in CVCless circumstances. Most reasonable people would agree that this is something you do while waiting for central access. The same reasonable people would also agree that the data which describes the complications from peripheral adrenaline is scarce and observational, whereas the data describing line-related sepsis is abundant, which means the case for peripheral vasoconstrictors may be stronger than we think. As usual, one's practice should be determined by one's department policy, which in turn will be determined by whoever shouts loudest during the department meetings. For the purposes of exam answers, CICM trainees should write something noncommittal, such as "peripheral catecholamines may be acceptable temporarily,  under appropriate supervision and with appropriate monitoring".

Distribution and protein binding.

Studies of adrenaline distribution are frustrated by its extremely short half-life, which makes it hard to study. Where data are available, it appears to be confined to the circulating volume, something like 0.1-0.2 L/kg. It is about 12% protein-bound, and has a pKa of 9.69.

Metabolism and clearance

Like noradrenaline, adrenaline has a shockingly brief half-life (2-3 minutes) and is metabolised so rapidly by COMT and MAO that venous adrenaline concentrations are stable even as you vary the infusion rate, i.e. most of it is gone before you have a chance to sample it (Best & Holter, 1982). Judging from some dog studies by Chu et al (1999), the vast majority of this metabolism occurs in the liver and gut, where 86-93% of the delivered catecholamines are cleared on first pass.

Pharmacodynamics of adrenaline

Adrenaline is generally said to be a nonspecific adrenergic agonist, as if it binds to all of the receptor subtypes with equal affinity.  This is not entirely true. It would probably be more accurate to call it a potent full β₁ and β₂ agonist with a somewhat lower affinity for α₁ and α₂ receptors.

Here is a table of inhibitory constants (K_i), where the smaller the K_i number, the greater the binding affinity of the drug. The β₁,  β₂ and β₃ data come from  Hoffmann et al (2004).

Thus:

• β₁ effect of adrenaline is similar to that of noradrenaline, i.e. the receptor affinity is similar between the two substances. 
• β₂ effect: adrenaline has a higher affinity than salbutamol but less than isoprenaline.
• β₃ effect: adrenaline has very poor affinity for these receptors
• α₁ and α₂ effects of adrenaline and noradrenaline are similar

That might seem weird, as one has become accustomed to adrenaline being treated as an inotrope, whereas noradrenaline is generally not regarded as one. The reason for the discrepancy between receptor affinity data and clinical experience seems to be due to the added β₂ effect which adrenaline exerts at the myocardium. Kaumann and colleagues (1989) demonstrated that this added β₂ stimulation increases intracellular cAMP by a substantial degree, such that adrenaline can increase the force of contraction via β₂-adrenoceptors by up to 60% of its maximal β₁ response. Noradrenaline, because of its lower β₂ affinity, does not have this effect untill you start using truly insane doses.

Cardiovascular effects of adrenaline 

Whenever haemodynamic effects of inotropes and vasopressors are mentioned, one often finds this old diagram trotted out. It is from a paper by Allwood et al (1963), the authors of which had cobbled it together out of historically published data for the purpose of representing the effects of these substances in a simplified way:

So, with adrenaline, systemic vascular resistance drops? To get to the bottom of that, more detail is required. Stratton et al (1985) tortured some normal healthy humans with adrenaline at different infusion strengths. Grainy scans from the original paper are available, but are poor in quality. Fortunately they also presented their raw data, which could be plugged into a spreadsheet. This clearly demonstrates that adrenaline is actually an "inodilator":

The systemic vascular resistance basically halved in this study. Ehringer & Konzett (1962) looked at this and concluded that the main reason for this was β₂ receptor activation in skeletal muscle, where blood flow increased. MAP did not drop much, and the systolic actually increased, because the cardiac output nearly doubled (helped by the increased heart rate, stroke volume, and decreased afterload). 

Now, one should be reminded that these were healthy normal humans (as much as can be said of random guys who volunteered for medical experiments). Adrenaline in these people works the way nature intended, out there on the Olduvai savannah. In the ICU, these are not going to be representative of our subjects. How does adrenaline perform in septic patients, or those in cardiogenic shock?

Moran et al (1993) was able to answer this question with a dose-finding study performed in real live sepsis patients (eighteen of them, with varying pathologies). With only a minor amount of spreadsheet fiddling, their data was remixed and arrayed in a table below. Note that the dose range extends a bit further than the healthy patient data (even in the 80s, Stratton et al could not bring themselves to give their volunteers a brain-exploding infusion of 0.25 μg/kg/min). Similar data are unfortunately not available from the cardiogenic shock population, because Levy et al (2011) were probably appropriately reluctant to expose their critically ill cardiogenic shock patients to either preposterously high or laughably low adrenaline doses. However, they did record the doses they used, and the change from baseline, which means their data can be extrapolated and plotted along with the septic group. None of this is especially scientific and the reader is encouraged to take it with a grain of salt.

So: in septic shock adrenaline acts mainly as a vasoconstrictor, and the cardiac output remains relatively stable (though admittedly it is already raised). The systemic vascular resistance remains low with modes doses as the skeletal muscle beds are probably already rather vasodilated and the additional β₂ effects go unnoticed.  At higher doses, α₁ begin to manifest. In contrast, in cardiogenic shock, the peripheral vascular resistance is already very high at the baseline, and it is the α₁ effects that are lost. In that scenario adrenaline acts as an inodilator: it increases cardiac output and decreases SVR. 

Adrenaline in right heart failure and pulmonary hypertension

Mainly for Question 2 from the first paper of 2018, which asked for a comparison of adrenaline and milrinone, some small amount of space must be devoted to the effects of adrenaline on the right side of the circulation. The same can probably be said about Question 8 from the first paper of 2012 (adrenaline s. levosimendan). In short:

• Adrenaline increases pulmonary arterial pressure by increasing pulmonary vascular resistance, though this is generally not a very significant rise. In an old study by Witham & Fleming (1951), the PA pressures of healthy volunteers increased by about 4.4 mmHg on average.
• Adrenaline increases RV contractility, because it is an inotrope and the RV is a bunch of cardiac myocytes (i.e. it is not surprising that it should do this). Le Tulzo (1997) found that this remains the case even in patients with significant RV failure.

On the basis of this, one would have to say that adrenaline would not be one's first choice for the management of the failing right heart, particularly where high PA pressure is involved. A drug with RV afterload-reducing features, such as milrinone or levosimendan, would be better.

Metabolic effects of adrenaline 

All this extra cardiovascular horsepower is not free. Somewhere along the line you pay for it with increased oxygen consumption, which is where adrenaline tends to lose marks among intensivists. Consider what might happen in a scenario where the flow in your coronary arteries might not be entirely sufficient. Günnicker et al (1995), for example, found that 0.1 μg/ kg/min of adrenaline was enough to increase the myocardial oxygen consumption by about 40% in their perioperative CABG patients. This increase in oxygen expenditure bought them about 60% more cardiac output.

Apart from this crude metabolic bean-counting, adrenaline (and β₂ receptor agonists in general) have several other inconvenient metabolic effects, as follows:

Hyperglycaemia due to adrenaline

Adrenaline interferes with glucose homeostasis in a number of ways which can broadly be summarised as "everything you can think of to liberate glucose for uptake from the circulation". In general, the sympathetic nervous system works to increase blood glucose in times of stress (as a part of the stereotypical stress response to critical illness), and adrenaline plugs into that process by acting on exactly the same receptors as would synaptic noradrenaline. Specific mechanisms, induced by catecholamines more generally, are identified and summarised in an excellent review by Fagerholm et al (2011). Of particular interest is Table 1. Its most useful content has been digested and included in this list of hyperglycaemia-inducing functions of adrenaline:

• Suppression of insulin secretion (α₂-mediated)
• Increased insulin resistance
• Release of glucagon (β, α₂ mediated)
• Increased lipolysis (α₁, β mediated)
• Increased glycogenolysis (α₁, β mediated)
• Increased gluconeogenesis (α₁, β mediated)

The hyperglycaemia can be reasonably serious, and can occur with relatively trivial doses of adrenaline. Jensen et al (2011)  gave their healthy volunteers only 0.05 μg/kg/min and found that the glucose content of their blood increased by 50% over the first 2 hours.

Hyperlactataemia due to adrenaline 

Adrenaline makes the lactate increase, which is also seen with the other β-agonists (notably, salbutamol). In the presence of adrenaline, there is a suspected increase of N+/K+ ATPase activity, leading to an increased demand for ATP which drives an increase in aerobic glycolysis. Specifically, binding to β₂-adrenoreptors leads to an increase of cAMP production, which is what increases Na⁺-K⁺ATPase activity. The relationship between adrenaline-induced skeletal muscle lactate production and its Na+/K+ ATPase activity has been convincingly demonstrated: well-oxygenated skeletal muscle produced tons of lactate when adrenaline was administered, and an infusion of ouabain (an Na⁺/K⁺ ATPase inhibitor) completely abolished this adrenaline-driven lactate production. Similarly, beta-blockade reversed this effect and caused lactate to decrease. 

But... that does not answer the very reasonable question, why would it do that? The increased Na⁺/K⁺ ATPase activity could conceivably have some beneficial effect in allowing repeated or sustained contraction by maintaining the sodium and potassium gradients essential for aforementioned contraction, but it is not as if those would just collapse without adrenaline. After all, sustained and repeated muscle contraction happens all the time, not just during the times of critical fight-or-flight. So: this liberation of lactate appears to be some sort of stress-specific phenomenon. Some authors theorise that this release of lactate benefits those tissues which can make use of the Cori cycle mechanism, i.e. adding another energy source in case somehow the 50% extra glucose is not enough. None of these assertions are supported by any evidence whatsoever, and trainees are encouraged to treat them with hostile suspicion. 

Hypokalemia due to adrenaline

Like salbutamol, which is used as a rapid rescue therapy for severe hyperkalemia, adrenaline can drop the potassium levels by a β₂-mediated mechanism (Reid et al, 1986). Again, this is due to the activity of Na⁺/K⁺ ATPase. The ATPase pushes sodium out, and pulls potassium in. The resulting intracellular potassium shift can be rather marked, particularly where the dose is high. Of course, that potassium has not left the body, and one needs to be reminded that with the withdrawal of the adrenaline infusion it will come right back again. The rebound hyperkalemia can also be impressive. Veerbhadran et al (2016) report a case of hyperkalemic cardiac arrest which - they believe - was solely attributable to this phenomenon.

Arrhythmias due to adrenaline

Adrenaline is a potent proarrythmic. There are several mechanisms for this, which are discussed and explored directly by Darbar et al (1996):

• Hypokalemia, as already mentioned
• QTc prolongation (at high doses), mainly due to increased calcium currents
• β₁-mediated increase in ectopic pacemaker activity

Catecholamines are generally known to promote both early afterdepolarisations and late afterdepolarisations, so all of  this is hardly surprising. The effect appears to be enhanced by some of the older halogenated ethers, eg. halothane. 

Indications for the use of adrenaline

For the CICM primary exam, a sound grasp of the evidence is probably not essential, as the questions will mainly revolve around basic physiology and pharmacology. The most one can expect is to be asked to list the accepted indications for the use of adrenaline, which are as follows:

• As an inotrope (inodilator?), for shock
• As a systemic vasoconstrictor, in anaphylaxis
• In cardiac arrest
• As a neb, for bronchospasm or airway oedema
• For open angle glaucoma, as eye drops
• As an adjunct to local anaesthesia
• For the injection of gastric ulcers

This, of course, is far from exhaustive, and if anybody comes up with others, they are encouraged to submit them in the contact form below.

Contraindications and interactions

There are no "hard" absolute contraindications to the use of adrenaline. As an endogenous catecholamine, it would be surprising to find anybody who is allergic to it, but reactions to its excipients (particularly the sodium metabisulfite) are totally within the realms of possibility. That would be fairly rare. More common would be "cautions" or "caveats" to the use of adrenaline, some of which are presented here in no particular order.

• Cardiac arrest in the post-cardiac-surgical patient: the usual 1mg doses of adrenaline are not recommended here. You can still use adrenaline, but you would have to use small quantities, and carefully. Often, cardiovascular collapse in this group can be rapidly reversed by (for example) unblocking a mediastinal drain or by pacing. At this stage, one suddenly gets one's circulation back, and realises that their afterload is massively increased. The resulting spike of blood pressure can do mechanical harm (avulse fragile coronary anastomotic sites, break aortic root repairs), to say nothing of the increased myocardial workload. None of the current guidelines specifies any particular dose, but all agree that smaller doses are called for.
• Hypertrophic obstructive cardiomyopathy: Adrenaline, and all the other rate-increasing inotropes, would not be well tolerated by patients with left ventricular outflow tract obstruction. These people, with increased contractility, will suffer from more obstruction, and will have a paradoxical decrease in their cardiac output. 
• Takotsubo cardiomyopathy:  As a phenomenon characterised by excessive catecholamine sensitivity of the left ventricular base and septum, this cardiomyopathy will logically worsen if presented with even more catecholamines.
• Catecholaminergic polymorphic ventricular tachycardia is a freakishly rare congenital condition which usually presents as exercise-induced cardiac arrest (Bellamy et al, 2019). Giving more catecholamines during such an arrest will usually be counterproductive.

Beyond that, one may be able to find a few "site-specific" contraindications to adrenaline (eg. conventional wisdom recommends that you shouldn't ever inject it into the penis), but these would probably be more in the realm of emergency medicine rather than ICU, and would be unlikely to score marks in the CICM First Part Exam