Corticosteroids in Pediatric Septic Shock: A Narrative Review Authors: Immacolata Rulli, Angelo Mattia Carcione, Federica D’Amico, Giuseppa Quartarone, Roberto Chimenz, Eloisa Gitto Published: 2024 ABSTRACT Objective: A controversial aspect of pediatric septic shock management is corticosteroid therapy. Current guidelines do not recommend its use in forms responsive to fluids and inotropes but leave the decision to physicians in forms refractory to the first steps of therapy. Data Sources: Review of literature from January 2013 to December 2023 from online libraries Pubmed, Medline, Cochrane Library, and Scopus. Study Selection: The keywords “septic shock”, “steroids” and “children” were used. Data Extraction: Of 399 articles, 63 were selected. Data Synthesis: Regarding mortality, although the 2019 Cochrane review supports reduced mortality, benefits on long-term mortality and in patients with CIRCI (critical illness-related corticosteroid insufficiency) are not clear. Yang’s metanalysis and retrospective studies of Nichols and Atkinson show no difference or even an increase in mortality. Regarding severity, the Cochrane review claims that hydrocortisone seems to reduce the length of intensive care hospitalization but influences the duration of ventilatory and inotropic support, and the degree of multi-organ failure appears limited. Further controversies exist on adrenal function evaluation: according to literature, including the Surviving Sepsis Campaign guidelines, basal or stimulated hormonal dosages do not allow the identification of patients who could benefit from hydrocortisone therapy (poor reproducibility). Regarding side effects, muscle weakness, hypernatremia, and hyperglycemia are the most observed. Conclusions: The literature does not give certainties about the efficacy of corticosteroids in pediatric septic shock, as their influence on primary outcomes (mortality and severity) is controversial. A subgroup of patients suffering from secondary adrenal insufficiency could benefit from it, but it remains to be defined how to identify and what protocol to use to treat them. 1. Introduction Septic shock is a major cause of both mortality and morbidity in children worldwide, particularly in low- and lower middle-income countries (LICs and LMICs), which account for 80% of cases and deaths globally [1,2]. Early treatment generally involves fluid resuscitation to ensure adequate organ perfusion [3]. Additionally, hemodynamic support with vasoactive agents is a critical component in managing patients with fluid-refractory septic shock (FRSS). For these patients, the Surviving Sepsis Campaign guidelines [4] recommend considering corticosteroids as an adjunctive therapy. However, the use of adjunctive corticosteroids varies significantly in clinical practice, with recommendations often described as “may or may not be used”. Gene expression analyses in children with sepsis have identified endotypes associated with either increased or decreased mortality risk with corticosteroid treatment. Despite these findings, the true effect of corticosteroid therapy on sepsis-related mortality and morbidity remains uncertain due to the lack of large, well-designed pediatric interventional trials, although some are currently underway (e.g., NCT0340139). Corticosteroids may improve hemodynamics in patients with septic shock through various mechanisms, including enhancing myocardial contractility and vasomotor tone by increasing calcium availability, increasing β-adrenergic receptor sensitivity and expression, reducing norepinephrine reuptake, and decreasing the production of nitric oxide and prostacyclin. Additionally, corticosteroids may stimulate intracellular adhesion factors, which could play a role in reducing capillary leak. In the literature, two opposing perspectives on corticosteroid use in pediatric septic shock have emerged, supported by both clinical and pathophysiological arguments. Proponents of hydrocortisone highlight its pleiotropic effects, which contribute to hemodynamic stabilization through both immediate nongenomic mechanisms—such as decreased norepinephrine reuptake, increased β-adrenergic receptor sensitivity, and enhanced calcium availability in myocardial and vascular smooth muscle cells, resulting in inotropic and vasoconstrictive effects—and delayed genomic mechanisms, including reduced production of prostacyclins and nitric oxide, increased vascular tone, and stimulation of intercellular adhesion factors, which reduce capillary leak and enhance inotropic effects [5]. Conversely, critics of hydrocortisone in septic shock raise concerns about its immunosuppressive effects, particularly on an already compromised adaptive immune system in pediatric sepsis. Cortisol can impair T-cell receptor signalling, T and B lymphocyte function, and antigen presentation. Furthermore, potential corticosteroid side effects, such as hyperglycemia, delayed wound healing, muscle weakness, and increased risk of infections, are particularly significant in critically ill children. Exogenous steroids may also exacerbate the sepsis-induced stress response, increasing the risk of metabolic derangements and progression to multi-organ failure (MOF) [6]. Given these conflicting perspectives, we conducted a systematic review and meta-analysis in children with FRSS to evaluate the effects of specific vasoactive agents on all-cause mortality and other clinically important outcomes. 2. Materials and Methods The research question has been resumed in a PICO (population, intervention, comparison, outcomes) format. The population is represented by children affected by septic shock and subjected to steroid therapy to evaluate the outcome of mortality, morbidity, and side effects compared to those who have not practiced this kind of therapy. 2.1. Inclusion Criteria (a)Studies on patients (age less or equal to 18 years) affected by refractory septic shock who received steroid treatment. Studies on patients (age less or equal to 18 years) affected by refractory septic shock who received steroid treatment. (b)Literature published from January 2013 to December 2023. Literature published from January 2013 to December 2023. (c)To clear up such controversial questions, we decided to select studies from different levels of evidence (RCT and observational cohort studies included). To clear up such controversial questions, we decided to select studies from different levels of evidence (RCT and observational cohort studies included). 2.2. Exclusion Criteria a.studies not pertaining to septic shock studies not pertaining to septic shock b.studies just on adult patients studies just on adult patients c.studies on preterm newborns studies on preterm newborns d.not in the English language not in the English language e.abstracts abstracts f.book chapters book chapters g.conference abstracts conference abstracts h.preprint articles preprint articles 2.3. Search Strategy A narrative review of the literature was conducted, searching the online libraries Pubmed, Cochrane Library, Medline, and Scopus. The keywords “septic shock”, “steroids”, and “children” were used for the research. Figure 1. 2.4. Outcomes Our primary outcome was to determine if steroid therapy in refractory pediatric septic shock was associated with a reduction or an increase in mortality. Secondary outcomes were shock reversal time, severity and duration of MOF, duration of inotropic support, duration of ventilatory support, PICU, and hospital length of stay. 2.5. Selected Articles We identified 401 eligible studies through the online database search strategy. After removing duplicates, we read titles and abstracts to assess eligibility for the review; the potentially relevant ones were read in full text. At the end of this screening, 63 studies were selected. 3. Results 3.1. Mortality The 2019 Cochrane review, including studies on both adults and children, states that subgroup analysis based on participant-related factors suggests that age (children versus adults) does not affect the patient’s response to corticosteroids (hydrocortisone or equivalent dose of another steroid). Corticosteroids probably reduce 28-day mortality compared to placebo or standard therapy (relative risk RR 0.91, with 95% confidence interval between 0.84 and 0.99, 11.233 participants, 50 studies, evidence of moderate certainty), with little/no difference in long-term mortality (RR 0.97, CI 95% from 0.91 to 1.03, 6236 participants, seven studies, evidence of poor certainty). They also seem to reduce hospital mortality (RR 0.90, CI 95% 0.82–0.99, 8183 participants, 26 studies, moderate degree of certainty). Children with septic shock may also gain more survival benefits from steroid therapy than those with less severe sepsis. Patients with critical adrenal cortical insufficiency (CIRCI) seem to have no significant reduction in the 28-day risk of death [7]. Yang et al. conducted a meta-analysis including 12 RCT studies (701 children recruited, suffering from shock caused by sepsis or dengue fever). The included studies were divided into a first group (1975–1996) and a second (2009–2017). Older studies were conducted in Asia and Africa, most recent instead in high resource countries like the UK and Canada. Steroids considered were hydrocortisone, methylprednisolone, and dexamethasone. The start dose of hydrocortisone was between 2 and 37.5 mg/kg. The length of therapy was between 1 and 7 days. Regarding the primary outcome, the fixed effect model showed reduced mortality in patients treated compared to the controls (odds ratio: 0.67; 95% confidence interval: 0.46–0.98; p = 0.041), but applying a random effect model results exhibited a negative result (odds ratio: 0.69; 95% confidence interval: 0.32–1.51; p = 0.252). Dose-response analysis showed also that a low dosage (0.6–5 mg/kg) could reduce mortality, a high dosage (5–37.5 mg/kg) could increase it instead. The authors suggested against corticosteroid application for septic shock in children under current medical conditions. However, several limitations affected this study: the study level, the low number of septic shock patients, the analysis of just about mg/kg dose and not about total dose, and the limited data about the follow-up period [8]. Nichols et al. conducted a retrospective chart review recruiting 70 children with refractory septic shock. Those were subjected to a dosage of total random serum cortisol and treated with stress-dose hydrocortisone in a retrospective study. Children treated had a higher mortality in PICU and a more severe degree of disease control (all p < 0.05). In children with serum cortisol > 18 mcg/dl, corticosteroid therapy was associated with increased mortality [9]. Atkinson et al. conducted a retrospective analysis of an ongoing multicenter database. Using PERSEVERE, a validated stratification tool based on biomarkers, 496 patients were stratified into three groups of initial mortality risk (low, intermediate, and high). Among corticosteroids, 78% received hydrocortisone, 16% methylprednisolone, and 6% dexamethasone. The median duration of corticosteroids was 5 days. Patients receiving corticosteroids in the first 7 days after hospitalization (n = 252) were compared with those who did not receive them (n = 244). Children treated had a higher mortality rate than those not treated. In the entire cohort, corticosteroids have been associated with an increased risk of mortality (OR 2.3, IC 95% 1.3–4.0, p = 0.004). In each group, corticosteroids did not show better outcomes. Similarly, they did not show better results among patients without comorbidities or in groups of patients stratified by PRISM. The authors concluded that the stratified risk analysis failed to demonstrate the benefit of any type of corticosteroid in this cohort [10]. Alkhalaf et al., in a retrospective cohort study, assessed that children treated with steroid corticosteroids showed a higher sequential organ failure assessment (SOFA) score and required longer ventilator and inotropic support. However, in-hospital mortality was not significantly different between groups. Patients treated showed a minor length of stay in the PICU. The authors also stated that after adjustment for baseline characteristics, severity scores and medical intervention, no association was found between hydrocortisone and increased mortality (p = 0.492) [Table 1] [11]. 3.2. Shock Severity The 2019 Cochrane review states that corticosteroids reduce length of stay in PICU for all the participants (average difference—1.07 days, 95% CI from −1.95 to −0.19, 7612 participants, 21 studies, high degree of certainty) and length of hospital stay in general for all participants (average difference—1.63 days, 95% CI from −2.93 to −0.33, 8795 participants, 22 studies; evidence of a high degree of certainty). The dose and duration of therapy did not affect the response to corticosteroids. In the study conducted by Atkinson et al., previously cited, patients treated with corticosteroids had a higher organ failure degree, a greater severity of disease, and a greater demand for vasoactive drugs than those who did not receive corticosteroids. For the entire cohort, corticosteroids have been associated with a complicated course (OR 1.7, IC 95% 1.1–2.5, p = 0.012). Nichols et al. argue that children with serum cortisol < 18 mcg/dl exhibit a lower severity of the disease than those with >18 mcg/dl serum cortisol (all p < 0.05). In children with serum cortisol < 18 mcg/dl, stress dose hydrocortisone therapy was associated with a higher length of stay in the intensive care unit and the hospital and fewer days without ventilatory support (all p < 0.05). Children with serum cortisol > 18 mcg/dl treated with hydrocortisone therapy had an increased need for ventilatory and/or inotropic support (all p < 0.05). Alkhalaf et al., in a retrospective cohort study, state that patients receiving corticosteroids had a lower risk of prolonged stay in the intensive care unit than those not receiving them. 3.3. Evaluation of Adrenal Function Weiss and other authors of current guidelines cited several pediatric and adult studies that attempted to use random serum concentrations of cortisol and/or cortisol stimulated with ACTH to identify patients who could benefit more from hydrocortisone therapy, but reliable limit values have not been clearly identified. This is due to various factors, including the variability of the cortisol test itself, cortisol metabolism (11-beta-hydroxysteroid dehydrogenase) during sepsis, corticosteroid binding globulin concentrations, tissue factors (for example, elastase and anti-glucocorticoid compounds) and cellular (for example, glucocorticoid receptor). Therefore, they concluded that the use of random cortisol or stimulation tests to guide the prescription of hydrocortisone in pediatric septic shock cannot be recommended. Menon et al. instead concluded their study by stating that random levels of free and total cortisol are strongly correlated with the severity of disease in children with septic shock [12]. Vila Perez et al. recommend the administration of hydrocortisone in patients with adrenal insufficiency confirmed by the stimulus test or suspected in cases of shock refractory to inotropic treatment. However, limitations of this test for the diagnosis of corticosteroid insufficiency related to critical diseases and the benefit of corticosteroids both in patients who responded and in those who did not respond suggest that this test should not be used to select patients who will likely benefit from corticosteroids [13]. Singh et al. evaluated adrenal function by dosing basal salivary cortisol and stimulated by ACTH in 51 children with nonresponsive septic shock (30 and 60 min) and basal salivary cortisol in 79 controls. Basal salivary cortisol among patients was higher than in healthy children. Survivors and those with catecholamine refractory shock had higher basal cortisol levels but with a statistically insignificant difference. Absolute adrenal insufficiency (basal salivary cortisol < 1.3 nmol/L) was diagnosed in 15.7%. A relative adrenal insufficiency (increased cortisol above basal value after stimulation < 25 nmol/L) in 68.6% of all patients, 71.9% in nonsurvivors and 71.4% in patients with refractory shock [14]. Nichols et al., at the end of the study results set out in paragraphs 1 and 2, concluded that hydrocortisone in catecholamine-refractory pediatric septic shock is related more to the severity of the disease than the levels of serum cortisol and has been associated with worse outcomes, regardless of the levels. 3.4. Early Versus Late Therapy El Nawawy et al., in a randomized prospective trial, included three groups of patients. The first group received corticosteroids in phase three of treatment, according to the current international guidelines. The second group was managed like the first but underwent stimulus testing with ACTH. The third group received corticosteroids at the beginning of fluid therapy. A fourth group was then created, adding patients who needed corticosteroids in the third phase of therapy from the first and the second group. Serum cortisol and plasmatic ACTH concentrations were assessed in all patients. Data showed a shorter shock reversal time in patients receiving steroids at the beginning of treatment compared to those treated at the third stage of treatment (p = 0.046). Mortality resulted in nonstatistical differences between groups [15]. 3.5. Septic Shock and Immunoparalysis Bline et al. conducted a study involving 102 children, defining immunoparalysis as a reduced TNFα response (p < 0.0001). Thirty-one of the participants received hydrocortisone, and these children were more likely to be older (p = 0.04), immunocompromised at baseline (p = 0.006), have higher PRISM III scores (p = 0.0003), higher vasoactive inotrope scores (p = 0.0002), and experience more days of multiple organ failure (MOF) (p = 0.002). A total of 33 children exhibited immunoparalysis. Hydrocortisone treatment was associated with a longer duration of MOF in children with immunoparalysis after adjusting for covariates (adjusted relative risk [aRR] 3.7, p = 0.0006). No such association was found for children without immunoparalysis (aRR 1.2, p = 0.67). Additionally, hydrocortisone was linked to a longer duration of MOF in children with both sepsis and immunoparalysis, but not in those with a more intact immune system. The authors concluded that further well-designed studies utilizing immunophenotyping are needed to optimize hydrocortisone strategies for this patient population [16]. Alder et al. compared all patients with septic shock in their study and found that those who had a complicated course exhibited lower levels of glucocorticoid receptor expression in peripheral blood leukocytes. Further analysis suggested that the combination of low glucocorticoid levels and high cortisol was associated with a higher rate of complicated courses (75%) compared to other combinations [17]. 3.6. Side Effects Side effects were extensively explored in the 2019 Cochrane review: The authors state that corticosteroids determined increased risk of muscular weakness (RR 1.21, 95% CI 1.01 to 1.44; 6145 participants; 6 studies; high-certainty evidence), but probably not the risk of superinfections (RR 1.06, 95% CI 0.95 to 1.19; 5356 patients enrolled; 25 studies; evidence of moderate certainty). The risk of hypernatremia (high certainty) and probably hyperglycemia (moderate evidence) also increased. Also, moderate evidence shows that there is probably little or no difference in the risk of gastrointestinal bleeding, stroke, and cardiac events. Low-certainty evidence shows that there is little or no difference in the risk of developing neuropsychiatric disorders. No evidence of increased risk of superinfection was found also in the RCT by El Nawawy et al. [15]. 4. Discussion 4.1. State of the Art The efficacy of steroids in pediatric septic shock lacks definitive evidence and consensus. The pathophysiological reasoning on sepsis and secondary adrenal insufficiency in critically ill patients and the need to stabilize hemodynamics in the event of failure to respond to inotropes would in themselves be valid reasons for the use of steroid therapy, but current evidence in literature is not convincing for the primary outcomes. In fact, regarding mortality, although the 2019 Cochrane review highlights a reduction in 28-day and in-hospital mortality, benefits on long-term mortality and in patients with CIRCI are not as clear. On the contrary, several other studies show no difference between treated and untreated patients or even an increase in mortality. Results regarding the severity of the shock are also equally controversial; although in some studies, steroid therapy seems to reduce the length of stay, the influence on factors more relevant to the severity appears to be reduced, such as the duration of ventilatory and inotropic support and the degree of multiple organ failure. Further controversies exist regarding the evaluation of adrenal function; in fact, according to the studies analyzed, basal, or stimulated hormonal doses do not allow us to identify with certainty the patients who could benefit most from hydrocortisone therapy due to the various factors that contribute to altering reproducibility; there are also studies such as that of Nichols et al. who find more risks than benefits of steroid therapy regardless of serum cortisol levels. Another factor that some studies call into question with regard to steroid therapy is immunoparalysis, as children who are more weakened from an immune point of view have a worse response (higher share of MOF) compared to children with more intact immune function without taking into account the role that low serum expression of glucocorticoid receptors has in determining failure of steroid therapy. Finally, regarding side effects, the Cochrane review leads us to be more vigilant, as they are the only ones supported by a good level of evidence on muscle weakness, hypernatremia, and hyperglycemia. 4.2. Knowledge Gap and Further Directions Recent studies unanimously agree on the necessity of conducting a randomized controlled trial to definitively assess the efficacy of steroid therapy in the management of sepsis. A large-scale trial, the “Stress Hydrocortisone in Pediatric Septic Shock” (SHIPSS study, NCT03401398), is currently underway, aiming to enroll 1032 pediatric patients. This study seeks to clarify the role of hydrocortisone in the treatment of sepsis. The trial protocol involves an initial intravenous hydrocortisone bolus of 2 mg/kg, followed by 6 mg/kg every six hours for a maximum duration of seven days, or until inotropic support is no longer required for at least 12 h. The primary outcomes include 28-day mortality and a reduction of more than 25% in the quality-of-life score. Preliminary results from the study are expected to be released in September 2025. Future research strategies should focus on developing rapid diagnostic tests with high sensitivity and specificity to identify individuals with FRSS and corticosteroid deficiency, who could benefit from corticosteroid treatment. Furthermore, exploring the genomic phenotype of patients could allow for early prediction of severe disease expression in individuals with septic shock before the development of refractory forms. This would help identify those at higher risk for FRSS progression, potentially guiding the early initiation of corticosteroid therapy in this targeted population. Early treatment initiation in septic shock patients at risk for FRSS may yield better outcomes compared to delayed therapy. 4.3. Key Messages What are the practical considerations for clinical use? Fluid resuscitation, inotropic support and antibiotic therapy remain the standard treatment for managing septic shock in pediatric patients. Fluid resuscitation, inotropic support and antibiotic therapy remain the standard treatment for managing septic shock in pediatric patients. In cases of refractory septic shock, the use of corticosteroids as a third-line therapy should be tailored for each patient. In cases of refractory septic shock, the use of corticosteroids as a third-line therapy should be tailored for each patient. Adrenal insufficiency may occur in cases of shock refractory to catecholamines; although diagnostic tests for adrenal insufficiency may not be sufficiently accurate, the presence of hypoglycemia and persistent hypotension can guide the clinician toward the use of hydrocortisone. Adrenal insufficiency may occur in cases of shock refractory to catecholamines; although diagnostic tests for adrenal insufficiency may not be sufficiently accurate, the presence of hypoglycemia and persistent hypotension can guide the clinician toward the use of hydrocortisone. The administration of corticosteroids should be carefully considered in immunocompromised patients. The administration of corticosteroids should be carefully considered in immunocompromised patients. 4.4. Strength and Limitations The main limitation of the review is that we incorporated studies of all qualities, so this affects the robustness of our findings. The narrative nature of this review provides a flexible and broad approach to this topic but does not include an exhaustive search of all possible evidence. Different criteria of selection and analysis could also offer other valid points-of-view. 5. Conclusions There are several key issues surrounding the use of hydrocortisone in pediatric septic shock that continue to spark debate. While current guidelines acknowledge its potential role in managing refractory septic shock, there has yet to be a randomized controlled trial (RCT) that provides conclusive evidence. The existing literature does not offer definitive answers regarding the effectiveness of steroid therapy in pediatric septic shock, as its impact on primary outcomes such as mortality and morbidity remains uncertain and controversial. While it is possible that a subset of patients, particularly those with secondary adrenal insufficiency, may benefit from this treatment, the methods for identifying these patients and determining the most appropriate treatment protocol are still unclear and require further exploration.