Systemic corticosteroids for specific indications

This chapter is only very vaguely 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". Taking an extremely flexible and relaxed approach to the definition of "understand", one could stretch these syllabus goals to include a discussion of the reasons behind our choices for a specific steroid agent in specific circumstances. The importance of this can be explained in the form of points:

This has never been tested in CICM exams, and is probably rather unlikely to be scrutinised in any other basic science exam, mainly because these choices are often not entirely scientific and so it would feel wrong to put them into questions and pretend like there is a truly correct answer. Still, the author felt that this is a subject that does not get very much oxygen, and there does not appear to be a single resource available anywhere to explain the main points. 

Different indications for different corticosteroids

Though their pharmacokinetic characteristics (lipid solubility, protein binding, molecular weight and diffusability) are all roughly similar, clear differences are observed in the tissue distribution of steroids which seems to affect their anti-inflammatory potency in specific conditions. To be clear, we are discussing the tissue distribution of steroids administered systemically, excluding those ones which you can inhale directly into the inflamed lung, or smear directly on the inflamed skin. The following is a series of broad statements made about some specific organs of interest:

Choice of steroids for severe pulmonary disease

Methylprednisone is the drug most commonly recommended for the sort of pulmonary inflammatory conditions that the intensivist might encounter in their daily life (severe ARDS, COVID, pulmonary vasculitis, organising pneumonia). 

This use pattern seems to date back to the 1980s, whereas prior to this prednisolone and hydrocortisone seem to have been the favourites.  Specifically, everything seems to be supported by a series of studies, starting probably with Braude & Rebuck (1983), which found that the distribution of methylprednisolone into alveolar fluid is greater than that of other steroids. That specific experiment involved giving seventeen patients methylprednisolone and prednisolone, and then measuring the concentration of each drug in bronchoalveolar washings. Plenty of methylprednisolone washed out, but barely any prednisolone, and in fact the concentration of methylprednisolone in the BAL fluid was related to the plasma concentration in a fairly linear fashion. A later series of experiments by Vichyanond et al (1989) and  Greos et al (1991) confirmed and refined these findings. 

This suggested that methylprednisolone was the best option if one were interested in having a strong regional steroid effect in the alveoli. However, hydrocortisone concentration also seems to rapidly equilibrate between alveolar fluid and plasma (again, an experiment by Braude & Rebuck, but a year earlier, in 1982). So why did hydrocortisone not become the drug of choice? Well; according to some steroid propaganda by Meduri et al (2018), the preference for methylprednisolone is mainly related to its higher affinity for the glucocorticoid receptor than hydrocortisone, which should logically translate into greater potency of effect. Thus, even though the same number of molecules might end up in the lung tissue, methylprednisolone will do more anti-inflammatory work per molecule. 

But, one might still counter, surely in that case it would be better to use dexamethasone, generally agreed to be the steroid with the highest affinity for the glucocorticoid receptor? A fair comment, considering that recently a large proportion of the population have been able to directly experience the pulmonary anti-inflammatory efficacy of dexamethasone in treating COVID-19 pneumonitis. Looking through the literature, no satisfactory explanation presents itself. Authors include promising subheadings in their papers, like "the rationale for choosing dexamethasone", but then they waffle under them. Reading between the lines of Mehta et al (2022) and Wandroo (2020) it appears the decision to choose dexamethasone was mostly made in terms of access and cost, as well as the possibility that, at such low doses, dexamethasone has a better safety profile (causing less adrenal suppression, for example). In other words, probably factors other than just  pulmonary tissue penetration were taken into account. In terms of efficacy, some authors have suggested methylprednisolone may still be better (for example Ranjbar et al, 2021, and Pinzón et al, 2021).

Choice of steroids for central nervous system disease

Dexamethasone is the drug most frequently recommended for severe inflammatory CNS disease (eg. cerebral vasculitis), to prevent inflammatory complications of CNS infections (eg. in meningitis), or to reduce cerebral oedema. It would be wonderful to quote some strong high-level evidence to support the use of dexamethasone in these scenarios, but close examination reveals these practices are guided more by convenience and tradition. Everything seems to date back to Galicich & French (1961), who injected 40mg of dexamethasone directly into the carotid artery of a comatose patient with a large glioblastoma, and were so impressed with the results that they continued to administer it to neurosurgical patients (4mg every 6 hours). Neurosurgical interns in 2022 might find that this approach remains essentially the same, albeit no longer intraarterial, and with occasional minor dose variations.

Exisiting literature (including RCTs) certainly supports this practice, but does not explain the choice of steroid particularly well. Still, there is some established rationale if you dig around. When Balis et al (1987) found that free prednisone and dexamethasone concentrations had reached equilibrium between the blood and CSF, they concluded that protein binding would be the most important determinant of CNS penetration. At normal therapeutic doses, prednisolone is about 90% protein bound, whereas dexamethasone is only 77% bound, which results in a better penetration into the brain. There are two other important factors: because of its longer half-life in the CSF, the duration of action for dexamethasone in the CNS is longer, and it has no mineralocorticoid effects which means theoretically a high dose can be given and the whole patient (and therefore also their brain?) will be less oedematous. That might sound silly, and probably is, because the total water content of the brain (as measured by MRI) definitely does not change, i.e. dexamethasone is not like mannitol. Lastly, there is a perception (at least among neurooncologists) that dexamethasone, of all steroids, is the least likely to cause delirium (Ryken et al, 2009), though the importance of this is questionable in patients already delirious from cerebral oedema, or those already comatose in the ICU. 

What about other steroids? Well, they also help decrease cerebral oedema. For example, Yamado et al (1983) demonstrated this with methylprednisolone, and in fact experimental animal  models suggested it has some sort of superiority, but these studies never really progressed beyond the stage of comparative biology. For all of the steroid drugs, the penetration into brain tissue is clearly related to the size of their free unbound fraction, and so conceivably one could give a high enough dose to saturate all the protein binding sites and increase the free fraction, making the drug more effective at CNS penetration. However, some steroids have less than 1% free fraction, which means one would have to give truly preposterous doses for this effect to become relevant. On this basis, some might say it's just better to take a lesson from seventy years of neurosurgical experience, and just give the dexamethasone. 

Steroid penetration into kidney tissue and other solid organs

Prednisolone is the steroid most often recommended for the treatment of acute interstitial nephritis. Fernandez-Juarez et al (2018) reported that most specialist centres use about 0.8mg/kg daily prednisolone for acute interstitial nephritis, for example. At the same time, this review revealed that half of all studies also "pulse" steroids, which is usually methylprednisolone, and usually in tremendous doses (around 1g per day). What is the scientific foundation underlying these choices? Unfortunately nothing specific comes from reading the literature describing this application of pulse steroids, other than generic justifications for the use of pulse steroids in a broader sense. It is possible that the choice of oral prednisolone for interstitial nephritis might be guided by the tendency of these patients to be reasonably well, i.e. they probably represent a ward-based or community population.

Steroid choices for cortisol hormone replacement

For those of us unlucky enough to have dysfunctional adrenal glands, some sort of long-term glucocorticoid replacement is necessary. This role is usually handled by administering hydrocortisone as an oral formulation. This is logical: hydrocortisone and cortisol are the same molecule, so it makes sense to replace like with like. Moreover it has excellent bioavailability, making it a convenient choice for oral administration, and the short half-life matches the cyclical pattern of natural secretion (i.e. you need higher levels in the morning, but you don't need high levels while you sleep). A high potency is not required, which means one does not need to resort to prednisolone or dexamethasone, and a mineralocorticoid effect is necessary which means dexamethasone is not a good choice.

Steroid choices for long term immunosuppression

Immunosuppression with steroids is a popular choice in the context of solid organ transplantation, or in the battle against autoimmune disease, or to brutally suppress the uprising of some insurgent lymphocytes in leukaemia or lymphoma.    Immunosuppression is the objective here, rather than a regional anti-inflammatory effect, which means that affinity for the receptor is the most important characteristic (rather than the penetration into any specific tissue).

For organ transplant recipients,  prednisolone is the most common long-term steroid choice, and it appears that the convenience of administering it orally is the main deciding factor. Solid organ transplant recipients are typically expected to carry on their immunosuppression at home, and prednisolone is a  convenient oral formulation. Hydrocortisone is also oral, but hasn't enough potency, and dexamethasone is not a good choice for long-term administration because it has an unpleasant aldosterone-suppressing effect (discussed below).

Steroid choices for short-term immunosuppression

To crack down on a gang of rebel B-cells, dexamethasone is usually called for, and in truly massive doses. Four days of 40mg per day had at one stage been the standard in the induction therapy for multiple myeloma, replaced more recently by a comparatively gentle 40mg weekly. Thymic atrophy and lymphocyte apoptosis is demonstrated (at least in animal models), antibody production plummets, and with these events the symptoms of lymphoma or ITP subside, replaced with (presumably, milder) side-effects of high dose steroids.

Why dexamethasone, and why so much? The rationale seems to have some biological foundations. Specifically dexamethasone is used because to give a high dose of another steroid would likely produce significant mineralocorticoid effects, and the leukaemia patient does not need those. Also, when tested against other agents, dexamethasone had a lower minimum inhibitory concentration for suppression of lymphocyte proliferation (Mager et al, 2003), i.e. it is a more potent suppressor of lymphocytes than other systemically available steroids. Lastly, though the antiproliferative and counter-humoural effects of long term low dose steroids and short term high dose steroids appear to be similar, the side effect profile appears to be better with the short courses and high doses, perhaps because short term side effects (insomnia and mood disturbance) are more difficult to measure and quantity than long term side effects (like fluid retention and weight gain).

"Pulse" methylprednisolone and other mega-dose steroids

Sometimes, you just need to drop a piano on the whole immune system. For this, typically 1g of methylprednisolone is administered intavenously, daily, for three to five days. Daily 100mg doses of dexamethasone, or about 600-700 mg of prednisolone, are considered equivalent. This is often referred to as a "pulse" of steroids, presumably because the terms "crisis" and "storm" were already taken by other endocrinological processes. This immunoablative artillery barrage was popularised in the 1970s by articles such as this editorial in the Lancet, where the advantages of high and short (instead of low and slow) were touted. The purpose of the "pulse" dosing, as with the other high immunosuppressant doses, was supposed to be the avoidance of long-term side effects - theoretically the short course of high doses would produce the desired immunosuppressant effect without the long-term problems of diabetes and weight gain. Moreover, from the experience of some dermatology authors, it appeared that there was "an unusually high rate of complete remission (after treatment withdrawal)", i.e. the steroid effect persisted for long enough that a longer weaning course of steroids was not required.

The specific choice of methylprednisolone over other steroids is, unfortunately, not explored especially well in the literature, except to mention that it was "especially popular". From the largely retrospective and case-controlled nature of most of the studies describing this practice, no additional information can be gleaned. From what few direct comparisons have been made, it certainly appears that pulses of dexamethasone are at least as effective (for Graves ophthalmopathy, anyway). It certainly seems that, if the objective is to quickly occupy each and every glucocorticoid receptor, then any systemically administered steroid would ultimately achieve this goal, provided it is administered in a large enough dose.

Though there is no evidence, and it might appear that methylprednisolone has become a fashionable choice for "pulses" though pure accident, some papers reference an old review by Melby (1974) which suggests that its main advantage over dexamethasone is its duration of action. Methylprednisolone, Melby argues, has a very short half-life, and therefore only acts briefly on the regulatory receptors of the hypothalamic-pituitary axis, thereby causing less adrenal suppression than dexamethasone. From this, it follows that you should use the shortest-acting steroid you can, if you plan to give massive doses.