By Drayton A. Hammond, PharmD, MBA, BCPS, BCCCP

Clinical Pharmacy Specialist, Adult Critical Care, Rush University Medical Center, Chicago

Dr. Hammond reports no financial relationships relevant to this field of study.

SYNOPSIS: In this systematic review and meta-analysis of randomized, controlled trials comparing administration of corticosteroids with placebo or standard supportive care in sepsis, corticosteroids were associated with reduced 28-day mortality.

SOURCE: Fang F, Zhang Y, Tang J, et al. Association of corticosteroid treatment with outcomes in adult patients with sepsis: A systematic review and meta-analysis. JAMA Intern Med 2019;179:213-223.

The current Surviving Sepsis Campaign guidelines provide a weak recommendation based on low-quality evidence against the routine use of IV hydrocortisone in patients with septic shock if adequate fluid resuscitation and vasopressor therapy can restore hemodynamic stability.1 However, the guidelines suggest its use if hemodynamic stability cannot be restored. No recommendation is provided for or against its use in sepsis without shock. Because of the low-quality evidence and potential benefits alongside minimal adverse effects, two randomized, controlled trials (RCTs) with the largest sample sizes to date were conducted.2,3 Fang and colleagues sought to critically evaluate all relevant literature for corticosteroids in sepsis to better inform contemporary practice. Their study was conducted in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols study (PRISMA-P) and registered a priori in the PROSPERO database.4 They hypothesized that relevant clinical outcomes would differ between those who did or did not receive corticosteroids and could be affected by other factors, including corticosteroid dose, year of publication, illness severity, and quality of evidence. Articles indexed in Medline, Embase, and/or the Cochrane Central Register of Controlled Trials by March 20, 2018, were considered for inclusion, resulting in 37 RCTs that included 9,564 patients. Random effect models were used for all outcomes, although fixed effect models were evaluated as a sensitivity analysis. Corticosteroids were found to reduce the risk of the primary outcome of 28-day mortality compared to placebo or standard supportive care (26.3% vs. 29.2%; risk ratio [RR], 0.90; 95% confidence interval [CI], 0.82-0.98; I2 = 27%) in the full analysis. Similar results were observed when relevant sensitivity analyses were performed, such as excluding studies published earlier than 2000, those reporting only ICU or in-hospital mortality, those with a non-low risk of bias, and those with fewer than 200 patients. Although a funnel plot analysis suggested some asymmetry between positive and negative trials and sample sizes, a trial sequential analysis confirmed the required information size was met to make confident statements regarding 28-day mortality. Lower doses (< 400 mg hydrocortisone/day) and longer course ( 4 days) of corticosteroids demonstrated mortality benefit at 28 days (RR, 0.82; 95% CI, 0.85-0.98 and RR, 0.92; 95% CI, 0.85-0.98, respectively). Additionally, patients with septic shock (RR, 0.91; 95% CI, 0.82-1.02) may show reduced 28-day mortality rates, but those with sepsis do not (RR, 0.89; 95% CI, 0.61-1.31). Secondary outcomes for which corticosteroids were favored included ICU and in-hospital mortality (RR, 0.85; 95% CI, 0.77-0.94; I2 = 0% and RR, 0.88; 95% CI, 0.79-0.99; I2 = 38%, respectively), time to resolution of shock (mean difference [MD], -1.35 days; 95% CI, -1.78 to -0.91 days), vasopressor-free days (MD, 1.95 days; 95% CI, 0.80-3.11 days), shock reversal at day 7 (MD, 1.95 days; 95% CI, 0.80-3.11 days), and ICU length of stay (MD, -1.16 days; 95% CI, -2.12 to -0.20 days). Outcomes were similar between groups for 90-day mortality (RR, 0.94; 95% CI, 0.85-1.03; I2 = 27%). Corticosteroids were associated with higher incidences of hyperglycemia (RR, 1.19; 95% CI, 1.08-1.30) and hypernatremia (RR, 1.57; 95% CI, 1.24-1.99).

COMMENTARY

Despite the use of corticosteroids in sepsis for more than 50 years, clinical equipoise has persisted because studies evaluating corticosteroids have been of inadequate sample size and design to confidently and precisely make statements of benefit and harm. Supportive care for sepsis has become protocolized and evolved to reduce the clinical impact of most individual interventions. Nevertheless, two well-designed trials2,3 were conducted recently. The authors of a meta-analysis have incorporated the outcomes from those trials into their results.5

The trials with the greatest weight in this and another meta-analysis were the Adjunctive Corticosteroid Treatment in Critically Ill Patients with Septic Shock trial (ADRENAL) and Activated Protein C and Corticosteroids for Human Septic Shock (APROCCHSS) trial.3,4 In the ADRENAL trial, 3,713 patients with septic shock who were mechanically ventilated were randomized to hydrocortisone or placebo.2 Those who received hydrocortisone experienced multiple benefits, including faster resolution of shock and shorter duration of the initial episode of mechanical ventilation, but exhibited similar 90-day all-cause mortality and adverse effect profiles. A subgroup analysis of patients randomized to an intervention within six to 12 hours of shock onset suggested those randomized to hydrocortisone were at lower risk for 90-day mortality. In the APROCCHSS trial, 1,241 patients with septic shock who had a similar or greater degree of critical illness as those in the ADRENAL trial initially were randomized to one of four groups: drotrecogin alfa or placebo paired with hydrocortisone and fludrocortisone or placebo.4 The investigators paused their work when drotrecogin alfa was removed from the market; later, investigators continued their work with two arms: hydrocortisone and fludrocortisone or placebo. The trial sponsor terminated the study after 97% of the expected sample size had been achieved because the trial agents expired. Nevertheless, 90-day all-cause mortality was lower in the corticosteroid arm. Adverse events were grossly similar between groups, except hyperglycemia (defined as at least one episode of blood glucose 150 mg/dL), and occurred more frequently in the steroid arm. While both trials had their limitations, their overall strengths were plentiful. Including these trials in an updated meta-analysis was both logical and needed.

The meta-analysis by Fang et al suggests particular nuances to providing corticosteroids in sepsis that may help guide clinicians’ decisions. Providing hydrocortisone at lower doses (< 400 mg/day, most commonly provided as 200 mg/day in bolus or continuous infusion dosing) and for longer durations (four or more days) appears most beneficial. Additionally, patients with septic shock may benefit most from their use. Notably, the authors only considered one long-term mortality endpoint, 90-day mortality, which was similar between the groups. However, another contemporary meta-analysis considered mortality between 60 days and one year as long-term mortality, which was subjective but defensible, and revealed a lower relative risk of mortality with corticosteroids (relative risk, 0.94; 95% CI, 0.89-1.0).5 This may be the endpoint of greatest importance for evaluating the efficacy of a therapy in sepsis and warrants further consideration. The authors also evaluated serious adverse effects and found a greater relative risk for hyperglycemia and hypernatremia with corticosteroids; however, the definitions for these effects varied widely between studies. Regardless of their rates of occurrence, these adverse effects are relatively benign and controllable in most critically ill patients. Overall, corticosteroid use in septic shock appears reasonable and potentially advantageous.

REFERENCES

  1. Rhodes A, Evans LE, Alhazzani W, et al. Surviving Sepsis Campaign: International guidelines for management of sepsis and septic shock: 2016. Intensive Care Med 2017;43:304-377.
  2. Venkatesh B, Finfer S, Cohen J, et al. Adjunctive glucocorticoid therapy in patients with septic shock. N Engl J Med 2018;378:797-808.