ARTICLE
Venkatesh B, Finfer S, Cohen J, Rajbhandari D, Arabi Y, Bellomo R, Myburgh J. Adjunctive Glucocorticoid Therapy in Patients with Septic Shock. N Engl J Med. 2018;378(9):797-808.
OVERVIEW
- Sepsis has been identified by the World Health Organization (WHO) as a global health priority with reported death rates among hospitalized patients ranging from 30-45%.
- Prior randomized, controlled trials (RCTs) have showed conflicting results regarding the mortality benefits of steroids in sepsis.
- Despite lack of consistent evidence, clinical practice guidelines recommend the use of hydrocortisone in patients with septic shock refractory to fluid resuscitation and vasopressors.
- The ADRENAL trial was designed to test the hypothesis that hydrocortisone results in lower mortality than placebo among patients with septic shock.
STUDY DESIGN
- Investigator-initiated, international, pragmatic, double-blind, parallel-group randomized, controlled trial
- Intention to treat, pre-defined subgroups, and sensitivity analyses
- Inclusion Criteria
- Adults > 18 years undergoing mechanical ventilation in a participating intensive care unit (ICU) who had been treated with vasopressors or inotropic agents for at least 4 hours at the time of randomization
- Strong clinical suspicion of infection
- Fulfilled > 2 systemic inflammatory response syndrome (SIRS) criteria
- Exclusion Criteria
- Age < 18 years
- Expected to receive systemic glucocorticoids for an indication other than septic shock
- Received etomidate during the current hospital admission
- Expected to die from pre-existing disease within 90 days after randomization
- Treatment limitations in place
- Met all inclusion criteria for > 24 hours prior to randomization
- Randomization
- Between March 2013 and April 2017, 3800 patients in 69 medical and surgical ICUs in 69 countries underwent randomization
- Randomization was stratified according to participating center and medical or surgical admission
- Patients were assigned to receive a continuous IV infusion of 200 mg per day of hydrocortisone or matching placebo over a maximum of 7 days or until ICU discharge or death, whichever occurred first
- Patients, treating clinicians, and trial personnel were unaware of trial group assignments
Outcomes Measures
- Primary: death from any cause 90 days after randomization
- Secondary: death from any cause at 28 days after randomization, time to the resolution of shock, recurrence of shock, length of ICU stay, length of hospital stay, frequency and duration of mechanical ventilation, frequency and duration of treatment with renal-replacement therapy, and the incidence of new onset bacteremia or fungemia between 2 and 14 days after randomization
Population Characteristics
- 3800 patients were enrolled at 69 medical-surgical ICUs
- 114 (3.0%) withdrew
- 28 (0.8%) lost to follow-up
- 1832 randomized to hydrocortisone group, 1826 randomized to placebo group
- No significant differences in baseline characteristics between the groups
- Mean age: 62 years
- Percentage male: 61%
- Median APACHE II Score: 23.5
- Percentage of patients with surgical admissions: 31%
- Primary site of infection: pulmonary origin for medical admissions; abdominal origin for surgical admissions
KEY RESULTS
- Primary Outcome
- No significant difference in 90-day mortality: 511 of 1832 patients randomized to hydrocortisone (27.9%) and 526 of 1826 patients randomized to placebo (28.8%) died
- No significant difference in 90-day mortality observed in the 6 prespecified subgroup analyses (sex, admission type—medical/surgical, catecholamine dose, site of sepsis, APACHE II score, and time from shock onset to randomization)
- Secondary Outcomes
- Patients randomized to glucocorticoid therapy had significantly:
- Shorter time to resolution of shock (median 3 days vs. 4 days)
- Shorter time to discharge from ICU (median 10 days vs. 12 days)
- Shorter duration of initial episode of mechanical ventilation (6 days vs. 7 days)
- Fewer blood transfusions (37.0% vs. 41.7%)
- There were no statistically significant differences in: mortality at 28 days, time to hospital discharge, number of days alive and out of the ICU, number of days alive and out of the hospital, recurrence of mechanical ventilation, recurrence of shock, duration and rate of the use of renal-replacement therapy, and new-onset bacteremia or fungemia
- Patients randomized to glucocorticoid therapy had significantly:
- Adverse Events
- There was a total of 33 adverse events reported, 1.1% in the hydrocortisone group vs. 0.3% in the placebo group (P=0.009)
|
Adverse Event |
Hydrocortisone |
Placebo |
|
Hyperglycemia |
6 |
3 |
|
Hypernatremia |
3 |
0 |
|
Hyperchloremia |
1 |
0 |
|
Hypertension |
3 |
0 |
|
Bleeding |
2 |
1 |
|
Encephalopathy |
3 |
0 |
|
Leukocytosis |
2 |
0 |
|
Myopathy |
3 |
0 |
|
Septic arthritis |
1 |
0 |
|
Ischemic Bowel |
1 |
0 |
|
Abdominal-wound dehiscence |
0 |
1 |
|
Circulatory shock |
1 |
0 |
|
Thrombocytopenia |
1 |
0 |
|
Miscellanious |
0 |
1 |
Study Strengths
- Primary outcome (mortality at 90 days) was both clinically relevant and patient centered
- A high proportion of eligible patients received the trial intervention as planned, and few enrolled patients were lost to follow-up
- Strong external validity given inclusion of 69 sites in 5 countries
Study Limitations
- Adverse event data was collected only on events judged by the treating clinicians to be related to the trial regimen
- There was limited data collected on secondary infections—only bacteremia and fungemia were recorded
- The appropriateness of antibiotic therapy was not evaluated
- Unclear how this data should be extrapolated to clinical settings where etomidate is frequently used, given that its use served as an exclusion criterion
Conclusions
A continuous infusion of hydrocortisone did not result in lower 90-day mortality than placebo; however, steroid use led to a significantly more rapid resolution of shock, shorter time to ICU discharge, and lower incidence of blood transfusion.
Discussion
Many studies regarding the effect of steroids in patients with severe sepsis and/or septic shock primarily look into the evidence of mortality benefit. The ADRENAL trial looked at a large cohort of sepsis/septic shock patients. Although mortality was not reduced for this cohort, contributors to overall mortality such as number of days on a ventilator, transfusion requirements, number of days in ICU, and time to reversal of shock were reduced.
The ADRENAL trial, along with prior trials, have showed conflicting results regarding the mortality benefits of steroids in sepsis. Given the heterogeneity of sepsis, steroid treatment may not benefit all sepsis patients. Individualization of treatment plan rather than application of a generalized treatment strategy should be considered. For instance, should we be looking into the rate and process of steroid metabolism in the critically ill patient? Is the etiology of sepsis more important than the overall picture of sepsis, when considering steroid intervention, or by the time someone has entered refractory septic shock, is the etiology less important and the overall inflammatory process the focal point? Perhaps the reason why we have not seen a consistent answer from these trials is because we have not defined the specific septic cohort which would benefit from steroids.