Evidence-Based Corticosteroid Therapy in Critical Illness
World Small Animal Veterinary Association World Congress Proceedings, 2014
Johan P. Schoeman, BVSc, MMedVet, PhD, DSAM, DECVIM-CA
Professor of Small Animal Internal Medicine, Head of the Department of Companion Animal Clinical Studies, Faculty of Veterinary Science, University of Pretoria, South Africa

There is a general tendency in veterinary medicine towards the noncritical acceptance of new advances, often because of the pressure to emulate human medicine. Corticosteroid therapy for critically ill patients represents no exception to this general trend. Besides, there are some cogent theoretical arguments for the use of corticosteroids in critical illness - a state usually characterized by an exuberant proinflammatory response, where several factors, such as infectious agents, trauma or tissue inflammation challenge the immune system and alter the HPA axis. Indeed, corticosteroids induce the synthesis of Annexin 1, a binding protein, which is responsible for inhibiting phospholipase A2 and eicosanoid formation. Furthermore, it blocks leucocyte migration, increases IL-10 release and induces apoptosis of inflammatory cells. Moreover, glucocorticoids play a major role in regulating the activity of nuclear factor kappa B, which plays a crucial role in cytokine gene transcription after exposure to an invading pathogen. Lastly, corticosteroids have a myriad of other antiinflammatory actions, such as the activation of endothelial nitric oxide synthetase. All of the above contribute to a justification for corticosteroid use in the face of severe and overwhelming inflammation.

Following the erroneous attribution of hypotensive shock to inherent adrenal failure rather than to drug-induced reduction of cortisol at the Glasgow Royal infirmary in 1982,1 a plethora of studies have addressed this issue in human critical care. Even after the flawed original reports were shown to have been subsequent to the use of the cortisol-inhibiting sedative, etomidate,2 the fervour in search of the golden bullet in critical care was undimmed. Eventually, a condition coined relative adrenal insufficiency entered the literature, purporting to represent a state of normal-to-elevated basal cortisol with a blunted response to ACTH. Subsequently, numerous studies have investigated the diagnosis of this condition and the risks and benefits of corticosteroid supplementation in critical illness. Yet, despite more than 30 years of investigation and more than 20 meta-analyses, the use of glucocorticoids in patients with critical illness remains extremely controversial with resultant conflicting recommendations.3 Nevertheless, there is at least consensus that a condition exists that manifests clinically as systemic hypotension, refractory to fluid loading and vasopressor use, yet eminently responsive to corticosteroid therapy. This systemic hypotension may be due to down-regulation of smooth muscle adrenergic receptors, the expression of which is modulated by glucocorticoids. The above condition has later been termed critical illness-related corticosteroid insufficiency (CIRCI), and its definition was somewhat refined to represent a state of insufficient corticosteroid-mediated down-regulation of inflammatory transcription factors.4 However, its aetiology remains only partially elucidated and several cytokines have been implicated in this reversible dysfunction of the HPA axis. In this regard, TNF-alpha has been shown to impair both pituitary CRH-mediated ACTH release and ACTH-stimulated cortisol synthesis in the adrenal.

Lately, however, several of the assumptions regarding a so-called "adequate" cortisol response to critical illness and its diagnosis are in the process of being debunked:

1.  The body is able to regulate intracellular glucocorticoid concentrations, irrespective of circulating concentrations, thus rendering circulating plasma levels highly problematic as indicators of glucocorticoid action.

2.  There is high variability in hourly cortisol concentration in critically ill patients.

3.  There is a large variation in the time points of assessment after onset of illness in the different studies (from 8 hours up to 61 days).

4.  There are marked gender differences in response.

5.  There is high variability in the currently available cortisol assays.

6.  Many studies had pharmacological confounders, such as etomidate.

Consequently, yet another name has lately been proposed for this condition, analogous to the sick euthyroid syndrome - namely, sick euadrenal syndrome, rather than a sick adrenal indicating adrenal insufficiency.5


The diagnosis of this adrenal insufficiency in critical illness, with the aid of an ACTH stimulation test is very problematic. Firstly, because the dose of ACTH and the appropriate response to ACTH in critical illness is controversial and almost impossible to predict, given the ill-defined magnitude of stress and inflammation and the nature and scope of individual response to critical illness. Secondly, the test's reproducibility is very poor in critically ill human patients. Thirdly, interpretation of the test is generally based on total plasma cortisol concentrations, which has been shown to correlate very poorly with free cortisol and even less with interstitial cortisol. Lastly, since we have no clinical test that quantifies corticosteroid activity at tissue level, the diagnosis of CIRCI remains elusive at this time. To get any closer to a definitive answer, future studies should measure markers of glucocorticoid receptor nuclear density, annexin 1, PLA 2 or NF kappa B levels and correlate them with survival and other clearly defined outcome measures. In sum, corticosteroid therapy should not be based on the results of an ACTH stimulation test, because its mechanism of hastening shock reversal has very little to do with adrenal insufficiency and much more with vascular hyporeactivity.


Fortunately, high-dose steroid therapy in critical care has been sufficiently discredited to allow for its discontinuation, yet the new "low-dose corticosteroid" regimens found immediate appeal and piqued the interest of clinicians desperate to improve the lives of their critically ill patients.6 The problem when deciding on appropriate treatment, in essence, is one of chasing moveable set points; since effective regulatory systems are capable of adjusting set points according to changing demands and hence the appropriate dose of corticosteroid therapy for each individual patient at a given time point will remain elusive for some time to come.

However, a degree of consensus has emerged regarding low-dose corticosteroid therapy in humans:

1.  It is part of the standard of care or so-called "treatment bundle" in septic shock that is unresponsive to fluid and vasopressor therapy (i.e., only those patients whose systolic blood pressure does not increase to 90 mm Hg or above after 1 hour of aggressive fluid resuscitation and vasopressin or therapy).

2.  It is associated with a more rapid recovery from nonresolving severe acute respiratory distress syndrome (ARDS) of > 7 days duration; but not in early ARDS and not in ARDS that has been ongoing for > 14 days.

3.  It should not be used in H1N1-induced ARDS.

4.  In bacterial meningitis, its use decreases hearing loss and other neurological sequelae, but not mortality.

5.  There is not enough evidence to support its use in severe community-acquired pneumonia, nor in acute spinal cord injury.

So, what does this all mean for the critically ill small animal patient?

Currently, there are no evidence-based guidelines for the treatment of CIRCI, or indeed whether the condition exists in critically ill veterinary patients; however, the following can be deduced after carefully extrapolating from human medicine and learning from their mistakes.

1.  It seems reasonable to use supplemental doses of corticosteroids in volume-resuscitated patients in which it is difficult to maintain their blood pressure, despite the use of high doses of vasopressors.

2.  The optimal dosage, duration and type of corticosteroid therapy are currently unknown, but a 3–5 mg/kg/day (or 200–250 mg/day for 7 days) equivalent of hydrocortisone seems to be the dose and duration settled on for humans. Since dogs tend to have roughly half of the basal serum concentration of humans, one would tend to suggest that a dose of 1–2.5 mg/kg/day of hydrocortisone equivalent be used. Since prednisolone and methylprednisolone both have approximately 5 times the glucocorticoid potency of hydrocortisone, they should be used at 0.2–0.5 mg/kg/day. Since dexamethasone is about 6 times as potent as prednisolone, it should be used at 0.03–0.08 mg/kg.

3.  There is no evidence in veterinary medicine showing treatment benefit after basing the decision on the results of an ACTH stimulation test and this approach has been sufficiently debunked in human critical care.5,7

4.  Lastly, treatment is mainly based on clinical diagnosis and confirmed by subsequent response to corticosteroid therapy.

5.  Abruptly stopping corticosteroid therapy will likely result in rebound of proinflammatory mediators with a resultant recurrence in the features of sepsis and tissue injury. The initial dosage should be continued for 7 days and then reduced every 3 days and stopped after 14 days.

In sum, it is important to highlight what we do not know concerning glucocorticoid therapy in critical illness, so that we as clinicians can realise the limitations and constraints under which we attempt to practice evidence-based medicine:

Apart from the fluid and vasopressor unresponsive septic human subgroup, we do not know which patients with other forms of critical illness should be treated with glucocorticoids.

We do not know what the optimal treatment window is. We are unsure of the optimal dosing strategy. We do not know which glucocorticoid should be used.


1.  Finlay WE, McKee JI. Serum cortisol levels in severely stressed patients. Lancet. 1982;1:1414–1415.

2.  Fraser R, Watt I, Gray CE, Ledingham IM, Lever AF. The effect of etomidate on adrenocortical function in dogs before and during hemorrhagic shock. Endocrinology. 1984;115:2266–2270.

3.  Marik PE. Glucocorticoids in sepsis: dissecting facts from fiction. Crit Care. 2011;15:158.

4.  Marik PE. Critical illness-related corticosteroid insufficiency. Chest. 2009;135:181–193.

5.  Venkatesh B, Cohen J. Adrenocortical (dys)function in septic shock - a sick euadrenal state. Best Pract Res Clin Endocrinol Metab. 2011;25:719–733.

6.  Annane D, Sebille V, Charpentier C, et al. Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock. J Am Med Assoc. 2002;288:862–871.

7.  Sprung CL, Annane D, Keh D, et al. Hydrocortisone therapy for patients with septic shock. N Engl J Med. 2008;358:111–124.


Speaker Information
(click the speaker's name to view other papers and abstracts submitted by this speaker)

Johan P. Schoeman, BVSc, MMedVet, PhD, DSAM, DECVIM-CA
Department of Companion Animal Clinical Studies
Faculty of Veterinary Science
University of Pretoria
Pretoria, South Africa

MAIN : Critical Care : Corticosteroid Therapy
Powered By VIN