This chapter is related to Section U2(i) and U2(v) from the 2017 CICM Primary Syllabus, which expects the exam candidate to "understand the pharmacology of glucocorticoids" and "understand the pharmacology of mineralocorticoids". This is obviously a big deal for the ICU, where nobody ever dies without a trial of steroids. A strong attachment to steroids is reflected in the historical exam papers, where the most common demand is to describe hydrocortisone, or to compare it with other steroids:
From this, it follows that the CICM First Part Exam candidate should be very familiar with the pharmacokinetics and pharmacodynamics of hydrocortisone, and should have a dim awareness of dexamethasone and methylprednisolone. Beyond these substances there are a million other steroid molecule variations which the critical care practitioner will basically never encounter, and which are deprioritised in the text below, unless there is some educational reason to mention them. '
Name Hydrocortisone Dexamethasone Methylprednisolone Class Glucocorticoid Glucocorticoid Glucocorticoid Chemistry Steroid Steroid Steroid Routes of administration IV, oral, s/c, IM, intranasal, topical, intra-articular, etc Oral or IV IV or topical Absorption 97% oral bioavailability 70-100% oral bioavailability 80% oral bioavailability Solubility pKa=12.59; lipid soluble; available as water-soluble salts for IV injection pKa=12.42; lipid soluble; available as water-soluble salts for IV injection pKa=12.61; lipid soluble; available as water-soluble salts for IV injection Distribution VOD = 0.5L/kg ; 90% protein-bound VOD=0.65L/kg; 65% protein-bound (mainly to albumin) VOD= 1.38L/kg; 77% protein-bound (mainly to albumin) Target receptor Glucocorticoid receptor, which is a cytoplasmic and nuclear receptor, that regulates gene transcription and protein synthesis (but some actions are also attributed to membrane-bound receptors and nongenomic pathways) Metabolism Metabolised mainly in the liver Metabolised extensively in the liver Metabolised extensively in the liver Elimination Free drug is reabsorbed in the distal tubule; there is minimal renal excretion of active molecules. Inactive metabolites are eliminated in the urine Inactive metabolites are eliminated in the urine; there is minimal excretion of unchanged drug (~1%) Time course of action Half-life 60 minutes, maximum 2 hours (with high doses) Half-life 4-5 hours (though biological activity persist for 36-72 hours) Half-life 2-3 hours Mechanism of action A combination of genomic effects and nongenomic effects, where some (medium and long term) activity is mediated by the regulation of protein synthesis, and some more immediate effects are mediated by the interference in cell membrane function, intracellular second messenger systems and membrane-bound glucocorticoid receptors Clinical effects - Immunosuppression (decreased granulocyte and lymphocyte activity)
- Reduced airway oedema, bronchodilation
- Sensitisation to catecholamines, increased cardiac output
- Neuropsychiatric effects (euphoria, mania, insomnia, psychosis)
- Metabolic effects (hyperglycaemia, hyperlipidemia, decreased insulin sensitivity)
- Fatty tissue redistribution, osteoporosis, proximal myopathy
- Hypernaremia, hypokalemia, water retention
- Adrenal suppression
- Increased risk of opportunistic infections (Aspergillus, Pneumocystis, Strongyloides)
Williams (2018) is the best single reference for this topic, available for free from Respiratory Care. If it has some gaps, it is in the comparison of potencies, which seems like an important and examinable topic, and for this the reader can refer to Scherholz et al (2019).Czock et al (2005) is also excellent, but extremely long (38 pages), making it unsuitable for last-minute revision.
The word "steroid" ("sterol-like") is one of those Greek word roots, so prevalent in medical nomenclature from the earlier 20th century when students of science and medicine were still impressed by Greek and Roman classics as something endowed with an innate dignity and authority. The word originates from "sterol", which in turn comes from the Greek στερεός, meaning "strong" or "stiff". According to Aronson (2018), it was applied to cholesterol originally, mainly because of it being found in gallstones (they called it cholesterin, because chole, bile, and sterin, solid chunk). The -ol came later, when it was discovered that these substances are actually secondary alcohols. From this, a whole range of substances got named ergosterol, lumisterol, phytosterol, et cetera, to the point where authors began to refer to the whole chemical class as "the sterols" in the early 1900s. To add "-oid" as a means of representing that something resembles the sterols was a more modern decision, consciously taken by Callow & Young in 1936 to describe "the group of compounds comprising the sterols, bile acids, heart poisons, saponins and sex hormones". This nomenclature has existed without much alteration until today, though the definition IUPAC have agreed on has been sanitised somewhat:
"Steroids are compounds possessing the skeleton of cyclopenta[a]phenanthrene or a skeleton derived therefrom by one or more bond scissions or ring expansions or contractions."
The chances of anybody being asked to draw this structure in a CICM exam are basically zero. More importantly, the trainees are reminded that, if they decide to spontaneously reproduce this structure in a written answer, it will take up half a page and elicit some kind of unpredictable emotions from the examiners, but certainly score no marks. Still, it is depicted below (using the image from the IUPAC document) mainly to demonstrate that the carbon atoms and rings in the molecule are numbered and lettered, so as to impose some kind of order on the process of naming them.
The basic elements of structure here are:
All the elaborate multitudes of different steroid molecules vary mainly by the stuff that's hanging off this four-ring structure, but the seventeen carbon core remains the same for all of them. For example, the addition of a hydroxyl group to carbon number 3 produces an alcohol (conventionally referred to as a sterol rather than a steroid), such as cholesterol.
Steroids and sterols are ubiquitous in biology, and so it would be ridiculous trying to list their "relatives" in any sort of comprehensive way. It will suffice to say that the pharmacologically important steroids we use are members of a tremendously huge group that is present in all living things (including viruses and Archaea), of which the vast majority of members are essential structural cellular components, contributing for example to the exterior cell membranes or the envelopes of organelles. Steroids and sterols are precursors for myelin, bile acids, yeast cell walls and toad venom. Among this diverse group of molecules, there is a minority that happen to function as signalling molecules, and of these signalling molecules a minority are recognisable as hormones (because they enter the circulation). There are synthetic drug versions of these natural signalling substances that impersonate them functionally by imitating their structure, and there are drugs which have completely different pharmacological effects but which happen to accidentally have a steroid core (such as, for example, aminosteroid neuromuscular junction blockers). Of the former and latter, there are such countless multitudes that nobody could possibly derive any meaningful benefit from listing them, let alone from revising that nightmarish inventory for exam purposes. Still, it is impossible to rule out the possibility that some readers may develop or already harbour some deep weird need for steroid chemistry. Those people are referred to the nine-hundred-page The Biology of Cholesterol and Related Steroids by Myant (1981), where the former director of the MRC Lipid Metabolism Unit of Hammersmith Hospital "tried to assemble as much as possible of the information likely to be needed by anyone beginning to study sterols in relation to living organisms."
The casual reader will have no interest in yeast membranes. The chemical relatives of steroids they are interested in would have to be the steroid hormones. Even this is difficult: Hans Selye, the father of the modern steroid classification system, writing in 1943 complained that "the manifold and overlapping activities of hormonal steroids tend to give the impression that there is no orderly pharmacological correlation among them". He was right (there is none) but we make do with his classification framework, made popular mainly because it is familiar and not because it is accurate or useful:
There are other ways (eg. by chemical structure, eg. the methasones and the acetonides), but within these systems there is no relationship between the taxa and their clinical characteristics, which makes them easily forgotten. The best reference to treat this subject pragmatically actually turned out to be the steroid chapter from the sensationally named Synthesis of Best-Seller Drugs by Vardanyan & Hruby (2016). It was also the resource that contained the best most consistent molecular structures for every possible steroid molecule on the market. And what a market it is. Looking at just the array of available corticosteroids can produce a profound depression in the student of pharmacology, particularly if their exams give them the impression that they might have to learn them all in detail. Benedeck (2011) lists some of the usual suspects that the CICM trainee is likely to encounter, and they are listed here in order of their appearance on the pharmacy shelves.
|Bovine or porcine adrenal extract (mostly cortisol)||1930s|
|Synthetic cortisol (i.e. hydrocortisone)||1952|
|Prednisone and prednisolone||1954|
Yes, reader, you could have purchased adrenal cortex extract off the shelf in the 1930s, and it would have probably been Eschatin (Parke, Davis & Co). Though reports of its utility were being published intermittently during that decade, most authors admitted to using it unscientifically (eg. for "asthenia" and "toxemia"), as well as for Addisonian crises (though without anybody examining the reason for its activity there). To quote Rogoff (1934), "Not much has been added to knowledge of the adrenal cortical hormone since our demonstration of its availability in extracts of the gland". The matter remained unresolved until 1949, when Kendall & Hench demonstrated the clinical activity of cortisol (ultimately being awarded the Nobel prize for medicine in 1950).
For a group of substances in constant widespread use, which are made of basically the same molecular core with a few optional added groups, one might expect there to be a clear relationship between the structural variations and the differences in function. Such a relationship does exist, but is remarkably difficult to distil from the various published sources - for some reason, there is no single reference to clearly summarise all their features. What follows had to be patiently reconstituted from Kasal (2010), Williams (2018) and Vinson (2011).
There is probably not a lot of reason to learn this in any sort of detail for the CICM exams, but it is nice to be vaguely aware of the underlying mechanisms which explain the differences in steroid function. Because of the various intentional and accidental manipulations of steroid molecular structure, we now have a range of substances that all surpass the potency of the original endogenous cortisol. The overall trend in steroid selection has been a movement towards greater glucocorticoid receptor selectivity (and lesser mineralocorticoid activity) as well as greater lipophilicity, greater molecular stability, and transsuppressive rather than transactivating gene transcription effects (as the latter are more responsible for undesirable effects such as cataracts and weight gain, according to He et al, 2014).
But before we move on:
It would be fair to say that, if you can think of a route of administration, somebody at some stage has toyed with the idea of administering hydrocortisone that way. It is conventionally given as an IV dose, but can be subcutaneous, oral, anal, vaginal, topical, inhaled, intranasal intraocular and buccal, intra-arterial and intra-articular, or just . As far as one can tell, so far nobody has tried to roll it up and smoke it, but apart from that every possible route has been taken. With a broad enough search strategy and time, something similar can be discovered for virtually every steroid in the field of animal studies, i.e. they have been tried every which way. Having said this, each has pharmacokinetic and pharmacodynamic peculiarities which make a certain route more favourable as compared to others, under specific circumstances. Ultimately the route of administration is determined mainly by the primary disorder being targeted, and by the desire to avoid adverse effects. These factors could be summarised as follows:
With these factors taken into account, specific routes of the application of steroids have developed as rituals and traditions, which probably vary geographically and culturally among the scattered tribes of medicine. The ophthalmologist is obviously going to do something different with dexamethasone than the medical oncologist or ENT surgeon. From the vantage point of the intensivist, which is the only position natural for the author, the following distribution of use patterns has become convention:
|Steroid drug||Route of administration|
|Hydrocortisone||Oral, IV or topical|
|Methylprednisolone||IV or topical|
|Dexamethasone||Oral or IV|
On the whole, these highly lipophilic drugs are well absorbed, but some have decreased oral bioavailability because of extensive first pass metabolism, which makes them poor choices for systemic oral therapy (or, looking at it differently, excellent choices for topical and inhaled administration).
|Steroid drug||Oral bioavailability||pKa|
|Hydrocortisone||97% oral bioavailability||12.59|
|Prednisone||80% oral bioavailability||12.59|
|Prednisolone||70% oral bioavailability||12.59|
|Fludrocortisone||70-100% oral bioavailability, but very variable||12.55|
|Triamcinolone||23% oral bioavailability||11.75|
|Methylprednisolone||80% oral bioavailability||12.61|
|Dexamethasone||70-100% oral bioavailability||12.42|
|Betamethasone||65% oral bioavailability, but this is rat data||12.42|
|Fluticasone||<1% oral bioavailability (extensive first pass metabolism)||12.19|
|Budesonide||9-21% oral bioavailability (extensive first pass metabolism)||13.75|
|Mometasone||<1% oral bioavailability (extensive first pass metabolism)||13.84|
Water solubility of unmodified steroids is extremely poor. Observe: the pKa range occupied by this group of molecules is generally north of 12, making them fully un-ionised at physiological pH. Ther pharmaceutical preparation, however, can modify these properties to make them favourable for the specific route of administration. For example, esters of steroids are poorly water-soluble and require hydrolysis by esterases to liberate the active drug into the circulation - which would be terrible if you tried to inject them into the blood, but potentially quite beneficial if you wanted to create a slow-release depot or a regional anti-inflammatory effect, for example after injecting the steroid into an inflamed joint. On the other hand, steroid salts are usually highly water-soluble and can be administered intravenously (such as dexamethasone sodium phosphate). In short, though steroids are ostensibly lipophilic and hydrophobic, this is not a barrier to marketing them as aqueous solutions.
Once they have gained access to the circulation, most corticosteroids tend to bind to proteins, for example to the corticosteroid-binding globulin, for which they have a variable affinity. Cortisol itself is only about 90% protein-bound (mainly to CBG), with the rest available as a free fraction. Other steroids often bind to albumin instead, for example dexamethasone and methylprednisolone. One might expect this pattern of protein-binding to result in a small volume of distribution (confined to the circulating plasma) but because these drugs are highly lipophilic they do tend to distribute widely into tissues and the VOD values can be quite large.
|Steroid drug||Volume of distribution||Total protein binding|
The last point about differences in tissue penetration for different corticosteroids requires some further discussion, as in this specific aspect of their performance lays the foundation of the different indications for corticosteroids in clinical use. Why do you see methylprednisolone given as a pulse, or specifically indicated for lung disease, whereas dexamethasone seems to be preferred for cerebral oedema? Well, reader, the answer is mainly related to the differences in their tissue penetration. This sort of thing probably falls more under "indications" rather than "pharmacokinetics, and is discussed far below.
For the vast majority of steroids, the rate of plasma clearance is fairly rapid, taking 2-3 hours. They are all metabolised mainly in the liver, usually in stages, and with each stage progressively becoming less biologically active and more water-soluble. The rate of hepatic metabolism can be quite quick, as one can see below; for example, hydrocortisone has a half-life of only one hour, maximum two (with high doses). This, of course, does not matter, because the action of corticosteroids is dependent on genetic transcription, and the effect last much longer than the actual drug persists in the circulation.
|Steroid drug||Half-life (hrs)||Route of elimination|
Being steroids, which is to say being children of cholesterol, one could reasonably expect these substances to have at least fractional biliary excretion, and maybe even a little enterohepatic recirculation, just like bile salts. This is not a feature for all steroid molecules; Adlercreutz (1979), though focusing on mainly steroid derivatives used for oral contraception, noted substantial enterohepatic recirculation only for steroids with a minimum molecular weight of about 500 Da, a polar carboxyl or sulfate group, and a few other structural features. It does certainly seem to happen more in the lesser mammals (Berliner et al, in 1962, confirmed that this sort of thing was going on in the rat), but in humans, the dominant pathway of elimination appears to be renal.
The physiology of corticosteroid and mineralocorticoid activity is already well covered in the chapters dealing with Section U1(vi), "describe the control, secretions and functions of renal and adrenal hormones". The interested reader can also see Nicolaides et al (2020) for the sort of molecular detail that most people would agree is totally unnecessary for the CICM primaries. What follows is a brief summary of these physiological effects of activating this receptor, which can be rapidly regurgitated under exam conditions:
On a more clinically relevant macro scale, the physiological effects of corticosteroid therapy are:
In reading about steroids, one often comes across the discussion of their relative potency, and yet never the discussion of how that potency is measured and compared. Considering that medicine has such a rich weird history of measuring the potency of drugs using bizarre surrogates (hamster uterus, chilled cat's blood, etc), one might expect something similar to be true for steroids. Unfortunately, it is not. It is easy to measure the potency of heparin, because it only really does one thing, and of vasopressin which only does two things, but steroids - they perform so many different useful functions, some of which are extremely difficult to quantify.
Arguably the most useful function of steroids to measure would be their anti-inflammatory effects. Valiant attempts to measure this objectively have been undertaken by heroic investigators and their victims. For example, one approach would be to treat steroids as a "bacteriostatic" antibiotic of sorts, except where the "bacteria" are the inflammatory cells of the human immune system. In this fashion one may be able to test the effect of steroids on preventing the proliferation of leukocytes in response to a normally immunogenic stimulus, and produce an "MIC" value of sorts. Mager et al (2003) did exactly this, exposing six adult males to different steroids and then testing their lymphocytes' capacity to proliferate to determine the minimum inhibitory concentration. They were able to rank steroids in terms of potency, in descending order as follows:
This counterproliferative effect probably has a direct relationship to the effect of steroids on lymphoma, but if lymphoma is not your problem, it is hard to draw a line of direct comparison between these data and - let's say - lupus. As the result, steroid potency is known across each medical specialty in a different way. For example, the relative potency of topical steroids is assessed by the human vasoconstrictor assay (Mckenzie, 1962), which measures their antiinflammatory effect indirectly. It is therefore perhaps better to regard the relative potencies of steroids in a fairly qualitative way, rather than trying to memorise a list of values.
Unfortunately, the CICM trainee should be prepared to memorise a list of values, as the table of relative steroid potency is a high yield target for CICM examiners. What follows is a representative table concatenated from chunks of this reference and that reference, which the author does not even regard as important enough to list, mainly because the equivalent doses and relative potency multipliers are so consistent across sources. Some might vary trivially (eg. the LITFL steroid table and the identical MDApp online calculator list the relative antiinflammatory potency of dexamethasone as 25 instead of 30), but otherwise these are all very similar, and no author emerges as an outlier.
|Potency relative to hydrocortisone|
|Steroid drug||Equivalent dose (mg)||Anti-inflammatory||Mineralocorticoid|
It was impossible to find the relative potency values for inhaled and topical agents mainly because they are infrequently discussed in terms of direct comparison; people simply don't try to bring this up in their paper because nobody is ever going to need to dose convert between IV dexamethasone and topical mometasone. Similarly, the magnitude of relative systemic mineralocorticoid effect for non-systemic steroids is not reported anywhere, potentially because there is no human data (as none of these drugs are absorbed systemically to any appreciable extent when they are being used in their normal way).