Nearly a century of using glucocorticoids in patients with active infection

GLUCOCORTICOIDS USED TO TREAT INFECTIONS

In the first descriptions of adrenal insufficiency by Dr. Thomas Addison in the mid-19th century, death was considered an inevitable outcome of the disorder.³ This was echoed by Dr. Charles-Édouard Brown-Séquard,⁴ who removed the adrenal glands in animals, all of which subsequently perished. However, it took decades of work to understand why the adrenal glands are critical to life.

Early experiments showed that animals that had their adrenal glands removed were less likely to overcome exposure to various organisms such as streptococcus, leading researchers to erroneously conclude the adrenal glands were important for the immune response.⁵ In the days before the widespread use of antibiotics, this observation spurred attempts at giving adrenal extract to patients with various infections.

The Perla and Marmorston⁵ paper from 1940 reports on a series of cases in which cortisol was given to patients with pneumonia. The introduction states the following: “It is well established that the suprarenal cortex plays a significant role in the mechanism of resistance to intoxications, bacterial and protozoan infections and to secondary shock. Removal of the suprarenal glands in rats is followed by a profound drop in the natural resistance to toxins, poisons, bacterial and protozoan infections and a disturbance in the capacity to form antibodies. The resistance to anaphylactic shock is decreased.”

The paper goes on to describe the experience of treating multiple cases of bronchopneumonia and malaria and 6 cases of the “grippal infection” with cortical extract. The following observations were made⁵:

• Maintenance of normal blood pressure with prevention of collapse

• Decrease in the evidence of toxicity

• Increase in the sense of well-being

• Apparent shortening of the period of convalescence

Multiple papers from the 1930s through 1940s highlight the use of adrenal extract in various infections, mostly without a clear worsening of symptoms. While no randomization or blinding occurred and the outcomes were subjective, many authors thought the adrenal extracts improved outcomes.^5–7

Over time other functions of the adrenal glands were elucidated, antibiotics were discovered, and enthusiasm for the importance of this organ in the immune response waned.

OUR UNDERSTANDING EVOLVES

In the 1940s, the anti-inflammatory effects of exogenous glucocorticoids were recognized quickly as the pain and swelling of joints in patients with rheumatoid arthritis rapidly resolved after administration of glucocorticoids, allowing for near-normal ambulation and function of the extremities. Decades of work went into elucidating the myriad mechanisms contributing to the anti-inflammatory effect of glucocorticoids, including inhibiting neutrophil migration and phagocytosis, among other functions.⁸ At the same time as the understanding of anti-inflammatory effects of glucocorticoids improved, immunoassays helped demonstrate the high cytokine production that occurs in severe infections like septic shock, again making glucocorticoids a potential tool to reduce an overabundant immune response in active infections.

Another important feature of glucocorticoids is their positive hemodynamic effect. In the 1950s, adrenalectomized rats were recognized to develop “vascular collapse” secondary to uncontrolled vasodilation and hypotension mimicking shock.⁹ Studies showed that cortisol, acting in concert with norepinephrine, aids in vasoconstriction.¹⁰ In a 1951 study, intravenous norepinephrine was administered to 3 patients before they were started on glucocorticoids and 24 hours after glucocorticoids were started; the largest increases in blood pressure were seen when the patients were receiving concomitant glucocorticoids.¹⁰

Over the decades after their introduction, glucocorticoids were recognized to have pleiotropic physiologic effects, with potential to dampen an overactive immune response and improve hemodynamic stability during an infection.

SEPTIC SHOCK

It wasn’t until the advent of flow-directed pulmonary catheterization and cytokine assays in the 1960s to 1980s that we fully recognized the vasodilatory state leading to significant hypotension and downstream ischemia in septic shock, as well as the abundant immune response (elevated interleukin-1 among others) it triggers.¹¹ As the anti-inflammatory and hemodynamic impact of glucocorticoids became better understood, the potential to use glucocorticoids, in combination with antibiotics, to treat septic shock became more attractive.

In 1963, the first prospective, randomized, double-blind, placebo-controlled trial of glucocorticoids (oral hydrocortisone 100 mg) vs placebo showed no survival benefit but no clear worsening of outcomes either.¹² In 1976, a combined prospective and retrospective analysis of 2 steroid regimens showed a significant survival advantage in the steroid group, leading to the widespread practice in the 1970s and 1980s of administering high-dose steroids in shock.¹³

This practice was supported by an animal study in 1980 in which 14 adult baboons were injected with Escherichia coli alone; E coli plus antibiotics; or E coli, antibiotics, and methylprednisolone.¹⁴ Not surprisingly, the baboons who received only E coli died, but more surprisingly the baboons receiving E coli and antibiotics also died. The only baboons that survived were the ones given glucocorticoids as well as antibiotics.

This study sparked multiple large multicenter studies evaluating the use of glucocorticoids in the setting of septic shock in the 1980s. The results were mixed, with multiple studies showing no improvement, and one study even finding a mortality increase in the glucocorticoids arm. A 1995 meta-analysis of 10 trials studying glucocorticoids in septic shock found a non-statistically significant trend toward harm with the use of steroids.¹⁵ These data led to waning enthusiasm for glucocorticoids in the setting of sepsis in the 1980s and 1990s.

In 2002, the pendulum swung again, as a large multicenter French study showed a mortality benefit in the steroid arm, reigniting the debate about glucocorticoid use in septic shock.¹⁶ Since then, many larger trials have been conducted. A more recent meta-analysis of 45 randomized controlled trials concluded that glucocorticoid use likely has some benefit in septic shock.¹⁷

BACTERIAL MENINGITIS

The concern for collateral damage from the immune response during an infection perhaps is best illustrated by bacterial meningitis, an infection in a closed space where increased immune cells and lysed bacteria can increase pressure, rapidly leading to permanent brain damage. Interest in the use of glucocorticoids in bacterial meningitis was sparked by a paper published in 1985 that described using different types of glucocorticoids in an animal model of pneumococcal meningitis.¹⁸ In the study, methylprednisolone or dexamethasone in combination with antibiotics was compared with antibiotics alone in rabbits infected with Streptococcus pneumoniae. Impressively, the rabbits with meningitis receiving dexamethasone demonstrated a substantial decrease in the cerebrospinal fluid pressure compared with those that received methylprednisolone and antibiotics alone.

This led to a 2002 study of dexamethasone in combination with antibiotics in adults with bacterial meningitis, which showed improvement in outcomes based on the Glasgow Outcome Scale in the dexamethasone and antibiotics group compared with the antibiotics alone group.¹⁹ The benefit of dexamethasone in bacterial meningitis was less clear in a more recent meta-analysis of trials of corticosteroids in children with bacterial meningitis, which found a clear decrease in hearing loss but no clear benefit in neurologic outcomes.²⁰

SEPTIC JOINT

Septic joint is a medical emergency requiring prompt antibiotics and potential surgical intervention. Both animal and human data exist on the use of exogenous glucocorticoids in the setting of a septic joint.

In the animal study, rabbits with a monoarticular septic joint were given either intra-articular glucocorticoid injections in the septic joint in addition to systemic antibiotics or antibiotics alone. The rabbits were euthanized 2 weeks after treatment, and the steroid group had less histologic joint destruction.²¹

In 2 studies, children with infected joints were given oral glucocorticoids in addition to standard antibiotic therapy; one study demonstrated a benefit in range of motion in the affected joint after 1 year, and the other study showed no difference.^22,23

Glucocorticoids are not considered standard of care for the use in septic arthritis, but data exist showing their potential beneficial use. Critically, in the animal and pediatric studies, the diagnosis of septic joint was known, meaning fluid was drawn from the joint and appropriate antibiotics were given concomitantly with the glucocorticoids.

DISCLOSURES

Dr. Brown has disclosed consulting and teaching and speaking for Amgen and Chemocentryx.