Synthetic glucocorticoids were first introduced in 1949 after the development of a purified preparation, known as cortisone, and became a revolutionary treatment for patients with primary adrenal failure and other acute-chronic inflammatory and autoimmune diseases. In anesthesiology, it is widely used to treat reactive airway diseases, acute nerve injury, nausea or vomiting, inflammatory diseases, and excessive immunosuppression during organ transplantation or cardiopulmonary bypass [1]. Shortly after the development of synthetic cortisone, there were two case reports of perioperative secondary adrenal crisis. The two young patients on chronic cortisone therapy stopped steroids before the surgery; they suddenly died after the surgery, and the result of the autopsy showed bilateral adrenal atrophy [2,3]. Since then, it has become common practice to perioperatively administer glucocorticoids at a supra-physiological dose, the so-called stress-dose, to patients on steroid therapy for long duration and in suspected adrenal failure cases.

The number of patients exposed to steroid therapy prior to surgery is increasing as corticosteroids are widely used for systemic administration and as inhalation and topical drugs [4]. Accordingly, clinical trials and retrospective analyses have been conducted regarding stress-dose glucocorticoids administration in the perioperative period for patients on steroid therapy [5,6]. However, many clinical trials had a low level of evidence, including unclear criteria for adrenal failure with a limited number of patients, resulting in various proposed recommendations for the administration of glucocorticoids in the perioperative periods [7]. Therefore, this article reviews the physiology of adreno-cortical hormones and indications and applications of stress-dose glucocorticoids in the perioperative periods based on recently published recommendations [7,8].

Adrenocorticosteroids are steroid derivatives produced in the adrenal cortex and include three endogenous hormones: glucocorticoid, mineralocorticoid, and androgen. All of them are synthesized when cholesterol is converted to pregnenolone by the cytochrome P450 enzyme [1]. Of these, glucocorticoids are secreted from the zona fasciculata of the adrenal cortex and the most important glucocorticoid is cortisol. Cortisol is an essential hormone for maintaining life. It mediates carbohydrate and protein metabolism, fatty acid transfer, electrolyte and fluid balance, and anti-inflammatory reactions. Cortisol enables the synthesis and release of catecholamines and contributes to normal vascular permeability, vascular tone, and myocardial contraction by regulating β-receptor synthesis and regulation [1].

The secretion of adrenal cortical hormones is regulated by the hypothalamic-pituitary-adrenal (HPA) axis. The corticotropin-releasing hormone (CRH) secreted from the hypothalamus stimulates the secretion of adrenocorticotropic hormone (ACTH) in the anterior pituitary gland, and ACTH stimulates the adrenal gland. This positive feedback for cortisol secretion and negative feedback for inhibiting the secretion of CRH and ACTH due to increased cortisol concentration regulates the secretion of cortisol [1].

The secretion of cortisol changes depending on the pulsatile secretion of CRH and ACTH according to the circadian rhythm. Serum cortisol concentration reaches the highest concentration of about 15 μg/dl around 6–9 am, then drops to the lowest concentration of below 2 μg/dl around 11 pm to 1 am. The median value of the 24-h period is approximately 5.2 μg/dl [9]. In normal adults, the adrenal glands produce approximately 5 to 10 mg/m²/day (body surface area per day) of cortisol, which is equivalent to 5 to 7 mg of oral prednisone or 20 to 30 mg of hydrocortisone [10,11]. In plasma, approximately 90% of circulating cortisol binds to corticosteroid-binding globulin (CBG), an α₂-globulin binding protein synthesized in the liver [12]. The remaining 5–10% binds to albumin or circulates freely and exerts an effect on target cells. When plasma cortisol concentration exceeds 20–30 μg/dl, CBG is saturated, and the concentration of free cortisol increases rapidly [12].

Sudden physiological and mental stress such as trauma, burns, major surgery, hypoglycemia, high fever, low blood pressure, severe exercise, and cold exposure activates the HPA axis and increases blood ACTH and cortisol levels. In a normal response to stress, the blood cortisol concentration increases to 18–20 μg/dl and adrenal cortisol secretion increases up to 30–45 μg/dl during moderately stressful situations and about 260 μg/dl in a highly stressful life-threatening situation. Increased cortisol levels normalize within approximately 24 to 48 h after the stress is resolved [11].

Adrenal insufficiency (AI) is the inability of the adrenal glands to produce adequate amounts of corticosteroids in response to various pathophysiologic states; this condition can be classified into primary, secondary, and tertiary AI depending on the cause [11]. Primary AI is an abnormality in the adrenal gland itself and is caused by the destruction of the adrenal cortex due to autoimmune diseases, viral and tuberculosis infection, hemorrhage, metastatic cancer, and sepsis. Secondary AI is rare but is caused by impaired production of ACTH or CRH due to damage or dysfunction caused by diseases of the pituitary gland or hypothalamus. Tertiary AI is the most common form, widely included in secondary AI, and is typically caused by inhibition of the hypothalamus or pituitary due to iatrogenic corticosteroid therapy; the degree of adrenal dysfunction is variable and sometimes reversible. In these patients, mineralocorticoid secretion is not affected and only cortisol production is reversibly inhibited [13]. Tertiary AI rarely occurs when oral prednisone dosage is less than 5 mg or when steroid is taken for a short period of less than 2 weeks, regardless of the dosage. However, AI has been reported to occur in more than 60% of patients taking high-dose oral prednisone (> 40 mg) even after only about 6 days [11,14].

Surgery causes a stress response with a wide range of endocrine, immune, and cardiovascular effects. During major surgery, proinflammatory cytokines, CRH, ACTH, and cortisol levels increase proportionally, resulting in an increase in cortisol secretion up to approximately 5–10 times the normal secretion, that is, 75–150 mg/day [16,17]. After surgery, the diurnal secretion of cortisol malfunctions temporarily, and the serum concentration of cortisol rises due to surgical stress [11].

In a recent meta-analysis, the change in serum cortisol concentration before and after surgery in patients without steroid therapy was analyzed in 71 studies since the 1990s [18]. In this study, the invasiveness of surgery was divided into three stages from grade 1 to 3 according to the modified Johns Hopkins surgical criteria [19]. Minor to moderately invasive procedures with less bleeding (potential blood loss < 500 ml) were included in grade 1; for example, breast biopsy, removal of minor skin or subcutaneous lesions, myringotomy tubes, hysteroscopy, cystoscopy, vasectomy, circumcision, fiberoptic bronchoscopy, diagnostic laparoscopy, dilatation, and curettage. Moderately to significantly invasive procedures (potential blood loss 500–1,500 ml) were included in grade 2; for example, thyroidectomy, hysterectomy, myomectomy, cystectomy, cholecystectomy, laminectomy, hip/knee replacement, nephrectomy, and major laparoscopic procedures. Highly invasive procedures (potential blood loss > 1,500 ml) were included in Grade 3, for example, major reconstruction of the gastrointestinal tract, major genitourinary surgery, cardiothoracic procedures, and intracranial procedures. In this review, it was found that the grade of surgery significantly affected cortisol secretion [18]. Patients undergoing grade 1 surgery did not show an intraoperative cortisol peak, and postsurgical cortisol concentrations were similar to those at baseline. Nevertheless, when compared to published data on healthy, unstressed adults, the mean cortisol output over the first 24 h after grade I surgical procedure was approximately doubled. Patients undergoing grade 2 and 3 surgery had 3.5–4 times higher cortisol output than that of healthy, unstressed individuals within the first 24-h postoperative period. Moreover, in both grade 2 and 3 surgeries, mean cortisol values remained elevated in comparison with the baseline measurements up to postoperative day 7.

Due to ethical issues, only a few studies have investigated the change in cortisol concentration and the incidence of AI after discontinuation of steroids in patients taking steroids. In 1973, Kehlet and Binder [16] investigated the occurrence of acute AI after steroid discontinuation in 73 patients undergoing major surgery, including splenectomy and colon resection. Patients took 5–80 mg of prednisone for various periods in this study. As a result, about 10% of patients developed perioperative hypotension, but only 3 patients showed low blood cortisol levels, and most of them were treated by fluid administration. They also measured cortisol concentration in patients undergoing major abdominal surgery and minor procedures such as hand surgery or uterine curettage, and estimated cortisol secretion due to surgical stress was found to be 75–150 mg/day and 50 mg/day for major and minor surgeries, respectively. The results of this study formed the rationale for subsequent recommendations of perioperative glucocorticoid supplementation [7,20].

De Lange and Kars [5] and Khazen and El-Hussuna [6] investigated the incidence of AI following administration of stress-doses or conventional maintenance doses of glucocorticoids based on prospective and retrospective studies. The randomized controlled trials included in their meta-analysis mostly targeted minor to moderate surgery with a limited number of patients, and both meta-analyses concluded that the incidence of perioperative AI was very low [5,6]. These studies reported that patients taking less than 5–10 mg/day of prednisone did not have adverse side effect of AI, even if only the daily dose was maintained during relatively less invasive surgery [21–23].

Despite the low incidence of surgery-induced AI, several major pathophysiologic mechanisms support the necessity of glucocorticoid administration in the perioperative period.

All synthetic glucocorticoids are derivatives of cortisol, an endogenous glucocorticoid. Drugs used as therapeutic glucocorticoids include hydrocortisone, prednisolone, and dexamethasone (Table 2) [1]. Among these drugs, hydrocortisone, which has the same structure as cortisol, is the most commonly used synthetic glucocorticoid. Prednisone is an inactive prodrug that is activated to prednisolone by 11β-hydroxysteroid dehydrogenase after administration [31]. Synthetic corticosteroids have different glucocorticoid and mineralocorticoid activities. Table 2 shows the relative efficacy of commonly used corticosteroids compared to that of hydrocortisone [1]. When adrenal gland function is reduced, corticosteroids are used to replace both glucocorticoids and mineralocorticoids. Therefore, in patients with primary AI in which mineralocorticoids are not secreted, dexamethasone is not appropriate and synthetic mineralocorticoids, fludrocortisone, and hydrocortisone should be administered [7]. Secondary and tertiary AI are caused by a lack of glucocorticoids, so the administration of drugs with mineralocorticoid activity may cause side effects such as dose-dependent edema, fluid retention, and hypokalemia [7]. Therefore, when a high dose of hydrocortisone of 100 mg or more is required in patients with secondary and tertiary AI, switching to a drug with a higher glucocorticoid activity ratio than that of mineralocorticoid such as methylprednisolone should be considered [24].

Glucocorticoids should be administered during the perioperative period because cortisol secretion in response to surgical stress may fail, resulting in AI, hemodynamic instability, and adrenal crisis. Therefore, the dose of glucocorticoid should be administered at the same level as that of normal physiological response to the surgical stress after evaluating the invasiveness of surgery and inhibition of the HPA axis [24]. If there is no suppression of the HPA axis or the requirement due to surgical stress does not exceed the maintenance dose of glucocorticoids already being taken, a perioperative stress-dose of glucocorticoid is not required unless the patient shows signs of AI [7]. However, when glucocorticoid requirement increases rapidly due to surgical stress, and the inhibition of the HPA axis is clinically important, the administration of stress doses should be considered [7,8].

Based on recent studies, recommendations were published in Anesthesiology in 2017 and Anaesthesia in 2020 [7,8]. As presented in Table 3, Liu et al. proposed recommendations based on estimated daily cortisol secretions according to the invasiveness of surgery [7,16,17].

Recently, the Royal College of Anaesthetists and the Endocrinology Society of the United Kingdom also published guidelines in Anaesthesia [8], but they are different from those of Liu et al. [7]. They have similarity in that they recommend administering stress-dose steroids to patients with primary and secondary AI and HPA axis suppression. However, they recommended 100 mg of hydrocortisone for all patients undergoing minor procedures as well as major surgeries [8]. After the publication of this recommendation, when questions arose about administering the same dose for a relatively simple operation [40], the authors responded that this dose may not be appropriate for simple procedures [41]. Therefore, we can only refer to their recommendations for major surgery. According to this guideline, for patients undergoing major surgery, hydrocortisone 100 mg or dexamethasone 6–8 mg should be administered at time of induction of anesthesia, followed by immediate initiation of a continuous infusion of hydrocortisone 200 mg/day until the patients can be administered double the pre-surgical oral dose [8].

Cortisol has a variety of critical physiological actions, and an increase in concentration due to surgical stress is important for maintaining hemodynamic stability during surgery. Although studies with more appropriate evidence are still required, evaluation of patients with possible AI and glucocorticoid administration according to surgical stress is crucial and can improve prognosis.