New Evidence of Potential Benefits of Dexamethasone and Added on Therapy of Fludrocortisone on Clinical Outcomes of Corticosteroid in Sepsis Patients: A Systematic Review and Meta-Analysis

The aim of this study is to investigate clinical outcomes of corticosteroid treatment in patients with sepsis or septic shock. An electronic keyword searches of PubMed, EMBASE, and Google Scholar were conducted per PRISMA guidelines. The pooled analyses on the corticosteroid impact on mortality, adverse events, and clinical outcomes were performed. Subgroup analyses on the clinical outcomes in relation to corticosteroid dose, duration, and agents were performed. Pooled analyses of 21 randomized control trials revealed substantially reduced mortality (RR 0.93, 95% CI 0.88–0.99, p = 0.02) and length of stay in intensive care unit (SMD −1.66, 95% CI −1.91–−1.40, p < 0.00001) without increased risks of adverse events (RR 1.04, 95% CI 0.96–1.12, p = 0.38). No significant improvements of other clinical outcomes were observed. Subgroup analyses demonstrated substantially reduced mortality with short-term (≤7 days) low-dose (<400 mg/day) corticosteroid treatment (RR 0.91, 95% CI 0.87–0.95, p < 0.0001). Moreover, dexamethasone (RR 0.40, 95% CI 0.20–0.81, p = 0.01) and combined hydrocortisone and fludrocortisone treatment (RR 0.89, 95% CI 0.84–0.94, p < 0.00001) provided substantial reduction of mortality whereas hydrocortisone alone did not reduce the mortality risk in sepsis patients. Thus, further controlled studies on the clinical outcomes of potential corticosteroid options on sepsis-related clinical outcomes are warranted.


Introduction
Sepsis is classified as systemic inflammatory responses to infection manifested by innate immune system activation, which subsequently induces life-threatening organ dysfunction or septic shock [1]. The number of incident sepsis steadily increases each year, affecting approximately 48.9 million patients worldwide [1]. The deleterious features of sepsis involving responses from vascular, immune, platelets, and plasma protein substantially increase the risk for mortality, which is estimated to be 30-50% [2]. Sepsis is considered as the leading cause of in-hospital death, and 11.0 million sepsis-related mortality was reported in 2017 [2]. However, underestimation of sepsis-related mortality is anticipated as data on incidence and mortality of sepsis in low-and middle-income countries are limited [1].
The major pathophysiological components of sepsis include cytokine-mediated inflammation, endothelial injury, vasodilation, and hypercoagulability; nevertheless, the recommended treatment modalities mainly include broad-spectrum antibiotics for infection, fluid resuscitation, along with vasopressor to ameliorate hemodynamic imbalance [2,3]. The current guideline, Surviving Sepsis Campaign, published by Society of Critical Care Medicine and European Society of Intensive Care Medicine, encourages intravenous antibiotic initiation within an hour of recognition of sepsis or septic shock, as early antibiotic administration reduces infection-mediated inflammation, thereby improving survival [4]. A previous study also revealed 7.6% reduction in survival for every hour of delay in antibiotic initiation [4,5]. Nonetheless, the sepsis-related mortality still remains high despite appropriate treatment, implying the need for discovery of therapeutic agent that may improve clinical outcomes in sepsis patients.
The current guideline recommends low dose (<400 mg, typically 200-300 mg/day) hydrocortisone only in sepsis patients with adrenal insufficiency or refractory hypotension defined as systolic blood pressure <90 mmHg notwithstanding appropriate fluid resuscitation and vasopressor treatment [4,6,7]. However, previous studies evaluating corticosteroid-related clinical outcomes in sepsis patients provided controversial results regardless of strong immunosuppressive anti-inflammatory activity [8][9][10]. Moreover, these studies recruited critically-ill patients with diagnosis other than sepsis, which may impede clinical applications of these results. Therefore, the objective of this study is to evaluate clinical outcomes of corticosteroids in sepsis and septic shock patients by performing pooled analyses of double-blinded, placebo-controlled randomized trials investigating efficacy and safety of corticosteroids.

Search Strategy and Study Selection
This study was prepared according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [11]. A systematic literature search of PubMed, Embase, and Google Scholar was performed to identify randomized clinical trials evaluating clinical outcomes including efficacy and safety of corticosteroid in sepsis patients (from inception to October 2020). The methods of initial database search include a combination of keywords and Medical Subject Headings including 'corticosteroids', 'steroids', 'sepsis', and 'septic shock' in title/abstracts. Reference lists of studies eligible for full-text review were further screened to identify eligible studies. Two reviewers (JS and YJC) searched electronic databases and identified eligible clinical trials, and any disagreements regarding study selection were resolved by the third person (Shin). The eligibility of studies was determined by prespecified inclusion criteria: (1) patients aged >17 years who had primary diagnosis of sepsis or septic shock, (2) double-blinded, placebo-controlled randomized controlled trials (RCTs) comparing clinical outcomes of corticosteroid (intervention) over comparator (placebo), (3) studies that assessed outcomes of interests, and (4) studies published in English. Review articles, meta-analyses, duplicate studies, conference abstracts, proceedings, case reports, editorials, studies without full-texts, and studies written in languages other than English were excluded. Additionally, any studies that recruited patients with primary diagnoses other than sepsis such as acute respiratory distress syndrome (ARDS) or systemic inflammatory response syndromes (SIRS) were excluded. The primary outcomes of interest include mortality defined as death after randomization and adverse events (AEs) including gastrointestinal bleeding, hyperglycemia, and secondary infection after corticosteroid treatment. The secondary outcomes of interest include duration of mechanical ventilation, organ failures, respiratory failures, length of stay in hospital or intensive care unit (ICU), and reversal of shock. Two reviewers extracted study characteristics including first author; publication year; intervention regimen, duration, and dose; comparator; patient inclusion criteria; and outcomes of interest, and the doses of corticosteroids were converted into hydrocortisone equivalent dose.

Risk of Bias Assessment
The risk of bias assessment of included studies was evaluated by Cochrane Risk of Bias [12], and studies were scored as low, unclear, or high in the following features: randomization sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other potential bias such as differences in baseline characteristics. Any disagreements on the study quality assessment were discussed until a consensus was reached. The funnel plots and Egger's test were utilized to detect publication bias: a symmetric funnel plot and p > 0.05 from Egger's test imply a low risk of publication bias.

Statistical Analysis
Pooled analyses of the outcomes of interest were conducted using RevMan (Review Manager Version 5.4, The Nordic Cochrane Center, The Cochrane Collaboration, Copenhagen, Denmark, 2020). The effect size of continuous variables such as length of stay in hospital or ICU and duration of mechanical ventilation were presented as weight standard mean differences (SMD) with 95% confidence intervals (CIs). The dichotomous variables including mortality, organ failure, respiratory failure, reversal of shock, and AEs were evaluated with relative risks (RR) and 95% CIs. Studies that measured mortality at multiple time points were analyzed as separate mortality cases. I 2 index was utilized to determine heterogeneity across the studies, and Mantel-Haenszel fixed-effect model was used to analyze outcomes with low heterogeneity (I 2 < 50%) while the random-effect model was performed to analyze outcomes with high heterogeneity (I 2 > 50%) [13]. Subgroup analyses were performed to identify factors affecting corticosteroid-related clinical outcomes and 2 factors were analyzed: (1) treatment regimen in consideration of treatment duration (≤7 days or >7 days) and corticosteroid dose (hydrocortisone equivalent dose <400 mg/day or ≥400 mg/day) and (2) corticosteroid regimens including hydrocortisone, hydrocortisone with fludrocortisone, dexamethasone, methylprednisolone, and prednisolone. Any studies not meeting the criteria of subgroup analyses were excluded from the analyses. p-values were estimated by two-sided tests and any p-values < 0.05 were considered statistically significant.
The current guideline recommends low-dose (<400 mg/day) hydrocortisone treatment for ≥ 3 days (typically 200 to 300 mg/day for 5 to 7 days in real practice) in sepsis or septic shock patients who have refractory hypotension despite appropriate fluid resuscitation and vasopressor administration, and routine corticosteroid therapy is not suggested [4,7]. Our study results demonstrated similar results of previous meta-analyses which demonstrated reduced 28-day mortality with low-dose corticosteroid (<400 mg/day) [8,9], and another study also displayed a linear relationship between mortality and corticosteroid dose administered in the first 24 h after study enrollment [36], implying high-dose corticosteroid administration may result in more harm than benefits. Moreover, our study suggested that longer corticosteroid treatment (>7 days) does not guarantee improved survival, as cytokine-mediated inflammations are most substantial during the early phase
The current guideline recommends low-dose (<400 mg/day) hydrocortisone treatment for ≥ 3 days (typically 200 to 300 mg/day for 5 to 7 days in real practice) in sepsis or septic shock patients who have refractory hypotension despite appropriate fluid resuscitation and vasopressor administration, and routine corticosteroid therapy is not suggested [4,7]. Our study results demonstrated similar results of previous meta-analyses which demonstrated reduced 28-day mortality with low-dose corticosteroid (<400 mg/day) [8,9], and another study also displayed a linear relationship between mortality and corticosteroid dose administered in the first 24 h after study enrollment [36], implying high-dose corticosteroid administration may result in more harm than benefits. Moreover, our study suggested that longer corticosteroid treatment (>7 days) does not guarantee improved survival, as cytokine-mediated inflammations are most substantial during the early phase of sepsis [1]. However, caution is advised with the determination of the actual duration of short-term (≤7 days) corticosteroid therapy at this point despite the recommendation of the guidelines (≥3 days) because a previous meta-analysis revealed improved survival with corticosteroid treatment duration ≥ 4 days [9].
This study demonstrated interesting results in relation to corticosteroid agents. Similar to the results of previous studies [8,9], combination therapy of hydrocortisone and fludrocortisone substantially reduced mortality of sepsis or septic shock patients. However, the current guideline recommends hydrocortisone as the glucocorticoid of choice in sepsis patients because, in addition to glucocorticoid effects, hydrocortisone also provides sufficient mineralocorticoid activity [4]. Notably, our meta-analysis findings also revealed that hydrocortisone combined with fludrocortisone, a major mineralocorticoid agent, improved survival outcomes in sepsis patients. Furthermore, we observed markedly improved survival with dexamethasone treatment (RR 0.40, 95% CI 0.20-0.81, p = 0.01), whereas hydrocortisone did not reduce mortality (RR 0.99, 95% CI 0.94-1.05, p = 0.76) in sepsis patients. The advantageous aspects of dexamethasone in sepsis patients may include the greatest glucocorticoid potency and anti-inflammatory activity among corticosteroids and dexamethasone has even higher anti-inflammatory activity than non-steroidal anti-inflammatory drugs [37]. The interest in dexamethasone has skyrocketed recently because of a study which demonstrated that 10-day treatment of low-dose dexamethasone (6 mg daily, hydrocortisone equivalent dose of 160 mg/day) substantially reduced 28-day mortality in hospitalized COVID-19 patients who required oxygen treatment [38], implying low-dose dexamethasone may play a strong drug candidate for improving survival of critically-ill patients. However, a question on the safety of dexamethasone in sepsis or septic shock patients must be answered because all studies evaluating dexamethasone effects administered a high hydrocortisone-equivalent dose (≥400 mg) in sepsis patients, and comprehensive analysis of clinical outcomes including efficacy and safety associated with low-dose (<400 mg/day) dexamethasone administration in sepsis patients is warranted due to the limited number of studies.
As far as we know, this is the first meta-analysis investigating corticosteroid impact on sepsis-related clinical outcomes from double-blinded, placebo-controlled RCTs. Moreover, another distinct feature of this study from previous meta-analyses is the inclusion of study populations with primary diagnosis of sepsis or septic shock [8,9]. Previous meta-analyses demonstrated controversial results on corticosteroid impact on mortality [8,9]. A meta-analysis of clinical outcomes of corticosteroid in pediatric and adult sepsis patients displayed insignificant short-term (28-31 day) mortality (RR 0.93, 95% CI 0.84-1.03, p = 0.15) [8], whereas a meta-analysis of adult sepsis patients revealed significantly improved 28-day mortality (RR 0.90, 95% CI 0.83-0.98, p = 0.02) [9]. However, these meta-analyses included studies that recruited patients with primary diagnoses other than sepsis or septic shock such as ARDS, SIRS, and community-acquired pneumonia [8,9], and data from patients without sepsis or septic shock diagnosis were also analyzed, which may impede clinical application of the study results in sepsis patients. This study demonstrated substantially reduced sepsis-related mortality with corticosteroids, however, subgroup analyses provide variable results in relation to treatment duration, dose, and corticosteroid agents, providing evidence that a hydrocortisone and fludrocortisone combination regimen and dexamethasone can be promising agents for improving survival in sepsis patients. However, further controlled studies on the clinical outcomes including efficacy and safety of dexamethasone in sepsis or septic shock patients are warranted.
This study possesses some limitations. First, heterogeneity across the studies may hinder clinical applications of this study. The included studies had variable study designs and outcome measurements. Moreover, although the included studies recruited patients with primary diagnosis of sepsis or septic shock, the conventional treatment may vary among the patients because of diverse underlying causes of sepsis including etiologic microorganisms. Additionally, sepsis may induce multiple organ failures and patient man-agements are guided based on the clinical presentations. To minimize the potential issues with heterogeneity and publication bias, our study team only analyzed the double-blinded, placebo-controlled RCTs. Additionally, inclusion of studies only published in English may raise concerns pertaining to limitation on study selection and strength of outcomes. To prevent the loss of evidence, our study group performed additional screening for the studies written in languages other than English and identified no eligible studies for the analysis, implying the strong validity of our study design and outcomes. Another concerning aspect is related to unassessed patient-related risk factors for poor prognosis of sepsis and corticosteroid-related AEs in the original studies, which may cause treatment response variabilities in these patients. Moreover, this study did not show any improvements in other clinical outcomes such as length of stay in hospital, organ failures, respiration failures, duration of mechanical ventilations, and reversal of shock due to limited number of studies. Nonetheless, this study possesses significant implication as survival is a critical indicator of sepsis-related clinical outcomes considering mortality rate of 30-50% in these patients, and our study team demonstrated improved survival with corticosteroid treatment in sepsis patients and identified factors that may benefit corticosteroid therapy. However, further studies on patient-specific factors related to variations in corticosteroid responses in sepsis patients to promote clinical benefits of corticosteroid therapy are needed.

Conclusions
Corticosteroid significantly reduced mortality in sepsis or septic shock patients. The pooled analyses revealed markedly reduced mortality with short-term (≤7 days) low-dose (<400 mg/day) corticosteroid treatment. Combination therapy of hydrocortisone and fludrocortisone and dexamethasone can be a promising therapeutic option to improve survival outcomes in sepsis patients.

Supplementary Materials:
The following are available online at https://www.mdpi.com/article/10 .3390/jpm11060544/s1, Table S1: Quality assessment of included studies of randomized controlled trials, Figure S1: Funnel plots of the study outcomes; (a) mortality and (b) adverse events.