Metaraminol

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". Metaraminol only appears in one SAQ from the CICM past papers, where in Question 18 from the second paper of 2019 it is compared to noradrenaline. This is unfortunate, as it is ubiquitous in Australian Intensive Care practice, and probably deserves a bit more mention. It is an old weak direct and indirect sympathomimetic with only one redeeming feature, that being its low potency and the relative safety this confers upon its peripheral administration. 

In summary:

For something which is found in every anaesthetic drug cabinet in the authors' country, it is surprising how little is written about metaraminol. The literature is so scarce that anaesthetics trainees turn to Reddit for help in despair. As a manifestation of this neglect, the Wikipedia entry for this drug is literally one paragraph and four references, upon the reading of which one could easily walk away with the impression that its clinical utility is limited to the off-label treatment of priapism. Without trivialising the impact of epidemic priapism in Australia, the readers must be made aware that metaraminol has other valid applications in critical care, which will be explored here in more detail so that at least one semi-serious non-Reddit resource is available to the public.

Chemical structure and chemical properties of metaraminol

Metaraminol was first described in 1951, in a paper by Peck et al which has a promising title but which is not available electronically. It is not clear whether it was initially marketed for critical care indications; its first applications were as a nasal decongestant (Taylor, 1952). It is a levorotary molecule which looks like this:

Yes, that is a familiar phenylethylamine structure, and yes this drug structurally resembles catecholamines. Unlike them, it does not have a hydroxyl group in the 4th position on the phenol ring, so you can't call it a catecholamine per se (i.e. that's not a catecholamine ring); moreover this subtraction greatly reduces its potency as a direct receptor agonist.  The amine group at the alpha carbon position on the ethylamine makes this drug insusceptible to metabolism by MAO, as well as giving it indirect sympathomimetic properties.

Pharmaceutical presentation

Being a highly water-soluble drug, metaraminol does not need any special carrier solvent or weird excipient. It comes in saline, together with a bit of sodium bisulfate as a preservative. Its presentation is only really remarkable for one reason, which is that it comes as an ampoule that contains something other than a single normal dose of a drug. Each 10mg ampoule contains twenty normal 0.5mg doses. In this, metaraminol resembles the inotropes and vasopressors which are intended to be given as infusions.

Pharmacokinetics of metaraminol

For whatever reason, the pharmacokinetics of metaraminol are not particularly well reported. Perhaps this drug is too old and too rare in the American pharmacosphere to be of interest to investigators, or perhaps all the studies are buried deep in the unscanned archives of discredited journals. Either way, finding any reliable source describing the metabolism and distribution of metaraminol was quite difficult. 

Administration and absorption

Surprisingly, metaraminol has some oral bioavailability. In fact Weil (1957) wrote some "observations on its oral use in the treatment of hypotension". Specifically, he observed that it is well absorbed after oral administration, though apparently doses 5-6 times higher were required to achieve the same pressor effect. This, generally, is not how it is used worldwide. Moreover, the reasons for this higher dose are not very clear: by all accounts this drug does not undergo any hepatic metabolism whatsoever, so there should be no first pass effect to speak of. 

Safety of peripheral administration

Bryant & Howard (1957) were disappointed by the necrotising effects of accidentally extravasated noradrenaline but pleased with the relatively benign consequences of subcutaneous metaraminol injection. "Sloughing of tissue did not result", they coldly noted over a pile of dead experimental dogs. This is reflected in modern practice, where metaraminol is administered peripherally, often through a tiny little cannula. Because of the low potency of its direct effects, extravasation injuries resulting in necrosis are rare. Pelner & Waldman (1960), in a letter to the editor of JAMA, reported their extensive complication-free experience with doses that in modern days would be considered unthinkable:

"We have treated a large number of patients in coronary and surgical shock with intravenous injections of metaraminol (as much as 300 mg. in 1,000 cc. of fluid)...  without one incident of ischemic necrosis."

In short, in the context of this world gone mad where people were routinely giving their patients peripheral noradrenaline and whole litres of concentrated metaraminol,  the modern practice of administering cautious 0.5mg aliquots seems like a completely safe and normal thing to do.

Protein binding and distribution

Metaraminol is highly water-soluble and approximately 45% protein-bound, according to PubChem. Following administration, it appears to distribute rapidly into tissues, where it is taken up into catecholamine storage vesicles. And then it just stays there.   Shore et al (1964) determined that there was a surprisingly large amount of metaraminol remaining in the tissues even long after its clinical effects are gone ("persists for days in heart and brain", they said). According to work by Anton & Berk (1977), the heart spleen and adrenal glands take up most of it, according to the density of aforementioned vesicles in those organs. "All organs, except for brain, concentrated the drug relative to whole blood", the authors noted.

From this, one might surmise that the volume of distribution of metaraminol is very large. Unfortunately, there is no directly available data on this in humans, but at least Anton & Berk had the decency to report their doses and serum concentrations, allowing us to calculate the VOD ourselves. With a dose of 1mg/kg, and the normal Sprague Dawley rat weighing something like 200g, we can estimate a per-rat dose of 0.2mg. At its peak, the metaraminol concentration in the blood of these rats was around 25 μg/100ml, or 0.25mg/L. Crudely, that gives a volume of distribution of 0.8L, or around 4L/kg. 

Metabolism and elimination

Bizarrely, what exactly happens to metaraminol remains unclear. Again, review of this subject is severely hampered by the lack of original source availability in electronic form. Snippets of information for this can ultimately be found in various secondary sources, but they occur with the sparsity of interstellar gas. For example, in Volume 1 of Foreign Compound Metabolism in Mammals (Hathaway, 1970), we find this:

"The fact that metaraminol... remained unchanged & bound to tissues for 1-2 wk after ingestion prompted in vitro studies, which showed that the drug resists biotransformation by liver enzymes which metabolise structuraly related compounds"

What studies? Where are they published? Infuriating 1970s conventions which left out the title of the papers (referencing them only as "C. J. Furst, Brit. J. Pharmacol., 1968, 32, 57") had made it somewhat difficult to track this down. It was ultimately possible to find Gram & Wright (1965), who indeed demonstrated pretty convincingly that rabbit tissues have no interest in processing metaraminol. Rabbit liver enzymes made short work of tyramine and amphetamine, but metaraminol concentrations remained stable after an hour of incubation. The same was found in lung and kidney preparations. "Enzymatic inactivation does not account for termination of the biologic effects of metaraminol", the authors were forced to conclude. 

So. The half life of metaraminol was about 48 minutes in the rat study by Anton & Berk (1977), but this was only the distribution-related α-half life. After that, it is followed by a much longer (5-6 hour) redistribution half-life, as it gradually returns from cellular vesicles and into the circulating volume. Both of these are clearly totally unrelated to the duration of its clinical effect, which may only last 10 minutes. 

 Pharmacodynamics of metaraminol

This aspect is probably known better than the pharmacokinetics. An excellent old paper by Harrison et al (1963) is available at the Annals of Internal Medicine, and features some of the most charming old textbook fonts (any readers who manage to get a hold of the original will immediately recognise that its typography hits all the same aesthetic notes as Deranged Physiology). Livesay et al (1954) is also a solid account.

In short, the pharmacodynamic effects of metaraminol need to be separated into direct and indirect effects. The indirect effects are basically the same as the effects of noradrenaline, because metaraminol stimulates noradrenaline release from storage vesicles by acting as an impostor, displacing it into the synapse and into the circulation. That's its main mechanism of haemodynamic effect. Additionally, metaraminol has some direct effects, which are related to some low-affinity α₁ and β₁ receptor binding, and which do not really manifest unless you use truly insane homicidal doses.

Direct sympathomimetic effects

Metaraminol is generally described as a direct α₁ and β₁-adrenergic agonist. This assertion must come from somewhere, but it is hard to find its origin. One can infer things from findings by Stone et al (1966), who discovered that it has a lipolytic action on adipocytes which is not related to noradrenaline release, and which is blocked by phenoxybenzamine (so it must be α-receptor-mediated). The concentration required to produce this effect was about one thousand times greater than the concentration of noradrenaline required to produce the same effect, which makes plenty of sense (i.e. we expect it to be a weaker α-agonist, as we use 500 μg boluses of metaraminol to achieve the same pressor effect as 5 μg/min of noradrenaline). 

Yes, it is also supposed to have a β₁-agonist effect. No, it's not easy to track down the basis for this. Yes, you are right, this is not an effect we ever see clinically. The β₁-agonist effect of metaraminol definitely appears more often in the pages of textbooks then in the cardiovascular system of actual patients.  If this is a real effect, exerted weakly upon real receptors, then it is completely drowned by the reflex decrease in heart rate which occurs in response to an increase in blood pressure.

So where does this story come from? Perhaps from case reports like Smith et al (1960), who blamed metaraminol for the near-fatal arrhythmia of an otherwise healthy housewife. She was fine until some well-meaning bystander innocently injected her with 10mg of metaraminol to treat post-operative hypotension. Did the metaraminol cause her pulse to become thready and irregular, or was it something else? And if it was metaraminol, then was this due to a β₁-agonist effect?

To attribute everything going wrong in this case report to β₁-agonist effects requires one to suspend their disbelief. However, it appears that in truly heroic doses metaraminol actually causes rhythm disturbances, increases the heart rate and increased cardiac contractility. This is supported by Meijler & Meerschwam (1962), who gave metaraminol directly into a bunch of disembodied rat hearts. "Amounts of aramine surpassing 100 μg always gave rise to serious and often irreversible rhythm disturbances", they noted. A dose of 100 μg into a 250g laboratory rat (400 μg/kg) would multiply up to a bolus of 28mg for a normal 70kg human adult, which even in the 1960s would have been insane. A lower dose (25 μg, or a 7mg bolus for a 70kg human) caused some increased contractility, an effect which was described as similar to that of noradrenaline (i.e. probably a low potency β₁-agonist effect). With doses below 5 μg (1.4mg boluses in the human), this effect was not seen, which explains why one does not encounter it with responsible use in the modern era.

Indirect sympathomimetic effects

By all reasonable accounts, metaraminol exerts most of its effect on the circulatory system by acting as a "false neurotransmitter". It is selected by VMAT reuptake proteins instead of noradrenaline, and displaces it from storage vesicles. The released noradrenaline then plays the main role in the haemodynamic effects of metaraminol, which explains why they are so noradrenaline-like. 

This was revealed elegantly by Harrison et al (1963), who demonstrated that the action of metaraminol was basically dependent on the presence of noradrenaline in tissues. Animals whose noradrenaline stores were depleted by the use of reserpine had minimal response to metaraminol, as might be expected from its trivial direct effects. Also, the use of metaraminol in normal dogs increased serum catecholamine concentrations, as noradrenaline was released from storage.

This effect can't possibly occur only at noradrenergic synapses, as VMAT-2 proteins are ubiquitous (and they also transport adrenaline, dopamine and serotonin). However, it appears that metaraminol only displaces noradrenaline from these vesicles, and all the other neurotransmitter amines are not affected. The clinical effects of metaraminol are therefore just the effects of systemic noradrenaline release.

Tachyphylaxis to metaraminol

Tachyphylaxis, or "a rapid decrease in response to repeated doses over a short time period", is often mentioned as one of the characteristics of metaraminol. Supposedly, with sustained use, metaraminol displaces so much noradrenaline from the vesicles, that no further displacement is possible, and no additional benefit will be achieved from further doses.

This is definitely seen with ephedrine, another indirect sympathomimetic.  Interestingly, ephedrine seems to be more susceptible to it than metaraminol. In a study by Ngan Kee et al (2001), comparing ephedrine and metaraminol during spinal anaesthesia, it was ephedrine that had the tachyphylaxis problem. For metaraminol, there is less data for this, and in fact some prolific users of the drug (eg. Dr E. M. M. Besterman, Senior Medical Registrar at the Middlesex Hospital of London in 1959) swear that one of its advantages is that "it may be administered by repeated injections without the development of tachyphylaxis". 

In contrast to Dr Besterman's assertion, other hard-using contemporaries observed that "after the prolonged administration of metaraminol to patients in shock, the response to the drug diminished, at a time when the pressor response to the administration of norepinephrine persisted" (Harrison et al, 1963). These authors also went on to demonstrate that, with sustained use of metaraminol as infusion, its indirect effects decrease in magnitude. Over two hours, the noradrenaline concentration in the circulation of metaraminol-infused dogs had halved. At the same time, the effect of the infusion also decreased:

"Although the infusion of metaraminol initially produced striking increases in arterial pressure and contractile force, toward the end of the 2-hour infusion period these variables had returned toward control levels."

If that's not "a rapid decrease in response to repeated doses over a short time period", reader, then what is. Harrison et al truly hammered the point home by also giving tyramine at the end of that two hour infusion. Tyramine is known to force the release of noradrenaline from storage vesicles. But after two hours of metaraminol, a tyramine bolus only produced the most anaemic of responses. Ergo, noradrenaline vesicles were depleted, the authors reasoned. They then actually biopsied the dog hearts, pureed the tissue, and analysed its noradrenaline content directly, demonstrating without a shadow of a doubt that there was less of it. 

Superstitious belief in some kind of sustained effects

The duration of its effect is said to be 20-60 minutes in many sources (eg. in this FDA data sheet). However, in addition to this, there is also said to be some sort of sustained effect that continues for many hours after the initial administration of the drug. You might hear a senior intensivist describe this now and then. Give a few more doses, they say. Eventually, it will just keep working for a while, and you won't need to keep giving more. This might sound plausible, as metaraminol does hang around in the tissues for many days. On the other hand, we have literally just discussed a beautiful experiment describing and explaining how the effect of metaraminol decreases with sustained use. 

So. Does metaraminol have some kind of persistent effect, or does it not?  Prescribing guideline documents seem to think that it does:

"sustained metaraminol use or overly frequent dosing can result in cumulative effects that persist even when therapy is discontinued"

It was impossible to find any direct support for this in the literature. Anecdotes from senior colleagues certainly abound, but these are always tainted by the suspicion of confounding circumstances. Was is the metaraminol infusion that kept working, or was the patient's shock improving on its own, did you finally give them the right amount of fluid, did you correct the acidosis, achieve source control, etc etc? These are questions without answers. No randomised controlled trial is to be expected. For now, we must work with reports like Weil (1960), who at least offers a well-documented anecdote for our review:

The graph is poorly labelled; the lower line is heart rate and the upper line is the systolic blood pressure. As you can see, this patient's haemodynamics had stabilised following the administration of 24mg of metaraminol over 6 hours. The prolonged duration of action of metaraminol may have contributed to this, as the authors gave a big dose over a short period. On the other hand, where was all the tachyphylaxis? Could it be that the patient just got better, and the reason the blood pressure stayed stable after the metaraminol was ceased is all because it wasn't working anyway by that stage? Hard to say. Fortunately for the CICM exam candidates, it appears that either the college examiners don't hold these weird backward views, or they are keeping them to themselves. Confronted with a radicalised senior colleague espousing this sort of rhetoric, trainees are advised to nod and smile.

Clinical effects of metaraminol

In contrast to drugs like levosimendan, where one feels silly discussing its effects on healthy volunteers who would never ordinarily require it, metaraminol is often used on healthy people during surgery, and so studies of its effects in healthy normal individuals feel relevant (and Malmcrona et al (1964) did exactly this). Their data, from the pharmacodynamics discussion above, should be predictable. Blood pressure went up, heart rate went down by a compensatory baroreceptor mechanism, and cardiac output remained basically the same. See:

Haemodynamic effects compared to noradrenaline

Meteraminol exerts most of its effect by stimulating noradrenaline release, and so logically one might expect its clinical effects to be essentially identical to those of noradrenaline. A study by Giuseppe et al (2005) tested this assertion by comparing noradrenaline and metaraminol infusions in septic patients. Their table of results is offered here because there's basically no better way to display these data. The doses of the drugs were pretty decent: 0.30 μcg/kg/min for noradrenaline and 2.5 μcg/kg/min for metaraminol (10.5mg/hr, or 20ml/hr of the standard infusion).

 

So. Not completely identical, but very close nonetheless. Metaraminol was more of a pulmonary arterial vasoconstrictor than noradrenaline (good to know) but in every other parameter these drugs were indistinguishable. Posed with this table without column labels, nine out of ten cardiac anaesthetists wouldn't be able to tell which drug is which.