The efficacy of steroids in reducing morbidity and mortality from extreme hyperthermia and heatstroke—A systematic review

Abstract Severe hyperthermia from classical or exertional heatstroke, or from drug ingestion or other noninfective pyrogens, is associated with a high mortality and morbidity. A systemic pro‐inflammatory response occurs during heatstroke, characterized by elevated cytokines with endotoxemia from elevated lipopolysaccharide (LPS) levels. Corticosteroids reduce LPS and cytokine levels, suggesting that they may improve outcome. A systematic review searching Embase, MEDLINE, and PubMed from the earliest date available until September 2019 was conducted, according to the PRISMA guidelines, with five papers identified. In four studies, systemic steroids administered before or at the onset of heat stress improved mortality or reduced organ dysfunction. Survival time was greatest when steroid administration preceded heat stress. In one study, a nonsignificant increase in mortality was seen. A dose response was observed, with higher doses extending survival time. Animal studies suggest that steroids improve mortality and/or organ dysfunction after an episode of heat stress or extreme hyperthermia.

as a T CORE above 40.5°C with neurological dysfunction. It is the third commonest cause of death in athletes. 4 Classical heatstroke (CHS) presents with similar symptoms to EHS but in the absence of exercise or exertion. CHS is often seen in meteorological heat waves and has a 28-day mortality rate of 58%, increasing to 71% at 2 years. 5 The clinical, biochemical, and physiological similarities between other noninfective hyperthermic states, for example, after drug ingestion, suggest that the pathological changes are at least partly due to hyperthermia per se, irrespective of the cause. A large number of drugs appear to have hyperthermic properties 6,7 ; common categories include serotonergic (eg, antidepressant agents and opioids), anticholinergic (eg, antihistamine and antipsychotic agents), volatile anesthetic agents, and neuroleptic medications. A number of mechanisms are thought to be responsible for the hyperthermia, including reduction in heat dissipation, and changes in uncoupling protein function 8 ; the latter appears to allow diversion of protons normally involved in the production of ATP to instead be dissipated as heat. 9 Current treatment options for heatstroke and severe hyperthermia remain limited. The priority remains rapid cooling, to a T CORE below 38.6°C, ideally at a rate greater than 0.16°C·min −1 , 10,11 and supportive treatment. While a specific treatment may be efficacious in a particular condition, (eg, dantrolene in malignant hyperthermia, and cyproheptadine in serotonin syndrome), drug treatment is not generally recommended or of benefit in most cases. 6,11 Hyperthermia is associated with the development of organ dysfunction (see below), which may require supportive treatment.
Hyperthermia is directly cytotoxic, affecting membrane stability and transmembrane transport protein function with electrolyte homeostasis, protein, and DNA synthesis disrupted. The nuclear matrix shows damage at lower temperatures than other parts of the cell, with significant endothermic changes observed at 40°C. 12 Direct cell death in humans occurs at temperatures of around 41°C, with the rate of cell death increasing markedly with further temperature increases, primarily due to protein denaturation. 13,14 The microvasculature is affected rapidly during hyperthermia, with capillary dilation, vascular stasis, and extravasation into the interstitium at a temperature of 40.5°C. 15 Renal glomerular filtration rate reduces after a T CORE increase of 2°C and worsens further with increasing temperature. 16 Acute kidney injury (AKI) affects one in six hospitalized patients with EHS 17 and has been reported in 53% after CHS. 18 Renal failure sufficient to require renal replacement therapy has also been described after hyperthermia due to neuroleptic malignant syndrome, 19 malignant hyperthermia, 20 and recreational drug use. 21 Hepatocellular dysfunction is common at temperatures above 40°C, 22 and coagulopathy occurs with a reported incidence of 45% in CHS. 18 The integrity of the blood-brain barrier is disrupted, 23 allowing translocation of systemic toxins into the cerebral circulation.
The gastrointestinal (GI) barrier is composed of physical factors such as enterocyte membranes and tight junctions between enterocytes, along with an immunological defense system, to minimize translocation of toxic substances from the intestinal lumen to the internal environment. Systemic hyperthermia increases the permeability of the gastrointestinal tract increasing the rate of gut bacterial translocation. Exposure leading to a T CORE exceeding 41.6°C-42.0°C, even after 60 minutes, induces a rapid sloughing of intestinal epithelial surface and an increase in intestinal permeability, including to large molecules up to a molecular weight of 4000 Da. 24 More modest increases in temperature to 39°C-41°C have been shown in vitro to cause an increase in paracellular permeability. 25 The changes to permeability happen early, within a few hours, but are reversibleparacellular permeability returns to normal even if the hyperthermia is maintained for 24 hours 25 In addition to these direct effects of hyperthermia, blood flow to the intestinal tract is reduced at temperatures above 40°C. 26 This exacerbates the loss of the GI barrier integrity and increases the potential for endotoxemia and systemic lipopolysaccharide (LPS) increases, which initiates the release of pro-inflammatory cytokines. 27,28 Lipopolysaccharides are large molecules forming part of the outer membrane of gram-negative bacteria. LPS contains a hydrophobic domain, known as endotoxin, 29 which stimulates release of pro-inflammatory mediators if they enter the systemic circulation.
A pro-inflammatory response is a well-developed defense mechanism, triggered by infective pathogens and toxic insults, such as trauma, and removes injurious stimuli and initiates tissue repair. Proinflammatory cytokines, for example, tumor necrosis factor-alpha (TNF-α) and interleukin-1 (IL-1), are released primarily from macrophages and mast cells; cytokines and cell migration to the site of injury promote neutralization of the antigen and tissue repair and healing. In one study, LPS was elevated in 100% patients admitted to hospital after heatstroke, coinciding with a T CORE of 42.1°C. 30 Significantly increased LPS levels have also been found in the portal circulation of heat-stressed rats at a T CORE of 41.5°C. 26 Heat stress in primates increases portal and systemic LPS concentration. 31 LPS is thought to be responsible for some of the deleterious effects of hyperthermia-administration of purified LPS produces diffuse endothelial injury, tissue hypoperfusion, and refractory shock, 28 and attenuation of systemic LPS by anti-LPS antibodies 32 improves survival after heat stress. Administration of antibiotics against intestinal microorganisms prevents endotoxemia from occurring 33 and appears to improve mortality, 34,35 suggesting that endotoxemia occurring with heat stress is gut derived. Endurance exercise also produces endotoxemia, 36,37 but whether this is primarily related to T CORE increases is not certain. In one cross-over study, an increase in the intensity of exercise increased the T CORE and intestinal permeability, 38 and the cytokine profile of EHS and CHS is similar to that produced in endotoxemia, 39 suggesting that irrespective of the initial stress, the resulting gut-derived endotoxemia may be similar. Abolition of endotoxemia significantly reduces cytokine production. 39 The role of cytokines in heat stress is unclear with an inconsistent response to thermal stress. A number of pro-inflammatory and anti-inflammatory cytokines are elevated at the time of hyperthermia from heatstroke. Acute phase reactants may also increase. (IL-6) may be elevated in all patients. 40 There is some correlation with outcome; the rise in IL-6 and the duration of the increased expression is related to mortality, independent of the maximum core temperature obtained. 41 Mice pretreated with IL-6 before exposure to heat take longer to reach a T CORE of 42.4°C, showing less organ damage, and attenuation in the increase of other cytokines. 42 Antagonism of IL-1 also improves survival. 43 Development of other hyperthermic states may also be associated with inflammatory mediators. Neuroleptic malignant syndrome (NMS) may be at least partly driven by an acute phase response; acute phase response mediators are reported to rise, and peak at 72 hours.
Conversely, levels of anti-inflammatory agents such as serum iron and albumin initially decline then return to the normal range, coinciding with clinical improvement. 44 IL-6 and TNF-α levels have also been found to be significantly increased in NMS, 45 and IL-6 in malignant hyperthermia. 46 Glucocorticoids inhibit many of the initial events in an inflammatory response, promoting the resolution of inflammation. 47 Acutely, glucocorticoids inhibit the vasodilation and increased vascular permeability that occurs following an inflammatory insult and decrease leukocyte migration to the site of injury. 48 Most of the anti-inflammatory and immunosuppressive actions of glucocorticoids are attributable to alterations in the genetic transcription in leukocytes. 48 While the precise role of the inflammatory response in heatstroke is unclear, reducing the inflammation appears to be of benefit. Corticosteroids reduce levels of the majority of cytokines, 47 and the administration of prophylactic glucocorticoids prevents heatstroke-induced LPS rise in an animal model. 49,50 These data therefore suggest that administration of corticosteroids may have a beneficial role in the treatment of heatstroke.

| ME THODS
Evidence for the clinical effectiveness of steroids in the acute treatment of hyperthermia and heatstroke was assessed by conducting a systematic review of published research evidence. The review adhered to the PRISMA checklist (Appendix A). 51

| Identification of studies
RCTs were identified by searching three electronic medical databases (Embase, MEDLINE, and PubMed), from the earliest date until September 2019. In addition, the EU Clinical Trials Register and the Cochrane library were searched. Further attempts to identify studies were made by examining the reference lists of all retrieved articles and review articles identified by the original searches. "Cancer" terms were excluded from searches, to exclude the studies on hyperthermia as a treatment modality.
The search terms used for the three searches are summarized as follows, and detailed in Appendix B. The Embase database was searched from 1974 to September 2019, using the "explosion" search terms of "heat stress," "heat injury," "hyperthermia," and "steroid." The MEDLINE database was searched from 1950 to August 2019, using the "explosion" search terms of "heat stress disorders" and "steroids." The PubMed database was searched from 1966 to August 2019, using the search terms "steroid*," "heat illness," "heat stroke," and "heat stress" in the titles or abstracts. No limits on any searches were set.

| Inclusion and exclusion criteria
Two reviewers independently screened all titles and abstracts. Fulltext papers of any titles and abstracts that were considered relevant by either reviewer were obtained where possible. The relevance of each study was assessed according to the inclusion criteria stated in Table 1. Studies that did not meet the criteria were excluded. Any discrepancies were resolved by consensus.

| Types of intervention
Published randomized controlled trials (RCTs) where steroids were administered to animals or humans before or after the onset of heat stress, compared against placebo, were sought.

| Outcome measures
Trials where the steroids were assessed against survival data or evidence of organ dysfunction were included. These outcomes represent the most clinically relevant responses. Survival data were taken as the primary outcome; organ dysfunction as the secondary outcome.

| Assessment of bias
For preclinical or animal studies, the reports were assessed for bias by using the Syrcle risk of bias tool. 52

| Measures of treatment effect
We included any measure of mortality, including survival time, temperature at which death occurred, and absolute survival numbers TA B L E 1 Study design and characteristics to be identified from the search

Population(s) Animal or human studies
Intervention(s) Administration of steroid before or after exposure to hyperthermia or heat stress

Comparators
The intervention will be compared with the control group Outcomes Survival or organ dysfunction after cessation of heat insult or trial. A P value of less than .05 was taken as statistically significant.

| Exclusion criteria
Studies were excluded in part or in total if the steroids were given in combination with another treatment. Results were only included if the effect of steroids alone was compared with the control group, either in a subgroup or the study as a whole.

| Subgroup analysis
Additional statistical analysis between the intervention and control groups was undertaken where appropriate if not reported in the study. Statistical analysis performed by the review authors is highlighted in the text; all other analyses were extracted from the study.
The Student's t test was used for continuous outcome data; the chisquared statistic for discrete outcome data.

| RE SULTS
Electronic searches identified 8553 citations. Hand searches revealed no further studies. Titles and abstracts were assessed for relevance to the review (stage 1 screening), and duplications were identified, resulting in 63 potential citations being retained. The full texts of these citations were obtained. After applying inclusion criteria to these full-text papers (stage 2 selection), 58 citations, which did not meet the inclusion criteria, were excluded. Five citations were therefore included in the systematic review (Appendix C). No studies were found that investigated the secondary outcome that did not also investigate mortality.
Five studies were found which met the criteria (Tables 2 and 3).
Of the five studies, three used rats [53][54][55] and two used primates. 50,56 No human studies were found. Four of the studies used dexamethasone 53-56 and one methylprednisolone. 50 In three studies, the steroid was given after the onset of heat stress. [53][54][55] With the exception of one study, 56 all studies reported improved survival (three reaching statistical significance) and markers of organ dysfunction.
All included studies were assessed for risk of bias (see Table 4 and Figure 1), and none were excluded after consideration of bias impact. All the papers stated that animals were allocated at random, but none described the allocation process in detail. None of the papers described or compared characteristics of the intervention and control groups separately or were randomly housed, but there was not any indication that there were differences between the groups.
None of the papers stated that the investigators were blinded to the allocation, for example that the caregivers were separate to the investigators, but the reviewers considered that given that objective data were being collected in all cases, the reported outcomes are unlikely to have been affected by any lack of blinding. In four of the five papers, all study animals were accounted for, but in all papers, the results of all the proposed outcomes were reported.
The studies were considered too heterogeneous in their methodology and outcome measures to enable completion of a metanalysis. A descriptive summary was therefore completed.

| Mortality
Administration of 4 mg kg −1 of dexamethasone to rats, either before or after the onset of heatstroke (defined as the time a decrease in peak mean arterial pressure (MAP) and cerebral blood flow (CBF) in the striatum occurred), improved survival time from 101 ± 3 minutes (control) to 250 ± 9 minutes and 122 ± 3 minutes, respectively. 53 Administration of a higher dose (6 mg kg -1 ) before or after heatstroke onset further improved survival time to greater than 450 minutes, and 321 ± 5 minutes, respectively. These data highlight that the administration of steroids prior to the onset of heatstroke lengthens survival time compared with administration after the onset, and that the higher dose (6 mg kg −1 ) had greater benefit than the lower dose (4 mg kg −1 ) (statistical analysis for this review). Bilateral adrenalectomy was performed in a further subgroup; MAP, CBF, and time to death were significantly lower in this group, but these changes were attenuated by dexamethasone.
A later study 54 observed incremental doses of dexamethasone at the onset of heatstroke increased survival time to 104 ± 9 minutes (4 mg kg −1 ), 204 ± 25 minutes (6 mg kg −1 ), and 268 ± 27 minutes (8 mg kg −1 ) compared with untreated controls (24 ± 3 minutes) (statistical analysis for this review). analysis since it did not meet the inclusion criteria for the review.

| Cardiovascular
All studies documented the effect of steroid administration on mean arterial blood pressure (MAP). Three studies 53-55 demonstrated that heat stress reduced MAP compared with nonheated controls; four studies 50,53,54,56 showed that this hypotension was improved at specific time points with steroid administration. One study 50 showed an improvement in MAP at specific core temperatures, which became statistically significant above 42°C. The heart rate was recorded in one study, 50 which was lower at all temperatures compared with untreated but heated controls. Of the three papers recording cerebral blood flow (CBF), two 53,54 document an improvement after the onset of heatstroke compared to untreated but heated controls.

| Neurological damage
Two studies investigated neuronal damage by histological changes against a previously defined score, 53,55 with significant worsening compared with unheated controls and an improvement in the steroid-treated group in both groups, one reaching statistical significance. 53 Cerebral levels of three biochemical markers of neuronal damage (glutamate, glycerol, and lactate/pyruvate ratio) were shown to increase with heat stress, and were reduced significantly in the steroid-treated group 54 ; this reduction in the heat stress-induced increase in cerebral biomarkers was not, however, seen in the study by Yang. 55 The difference in effect across studies may be explained by the dose difference: the study by Yang used a dose of 4 mg kg −1 , compared with the higher dose (8 mg kg −1 ) in the study by Lui, where a difference was seen.

| Inflammatory cytokines
Administration of steroids in one study prevented a detectable increase in serum LPS, 50  Other biases SelecƟve outcome reporƟng (reporƟng bias)

Random group allocaƟon (selecƟon bias)
Groups similar at baseline (selecƟon bias) Blinded group allocaƟon (selecƟon bias) Number of studies study are recorded after 4 mg kg −1 , compared with the higher dose of 8 mg kg −1 in the study where a difference was seen.

| Hepatic, renal, and coagulation system
Two studies detailed the effect of steroid administration on biomarkers of renal and hepatic function, and on clotting factors. The first 54 showed a statistically significant improvement in biomarkers in all three organ systems after steroid treatment compared with the heated control group; the second study conversely showed a deterioration in markers of liver and renal function and coagulopathy, some reaching statistical significance. 56

| D ISCUSS I ON
Administration of corticosteroids improved survival time and organ dysfunction due to heat stress, and a reduction in endotoxin and pro-inflammatory mediators in 80% of the studies included in the review. In addition, administration of another anti-inflammatory agent in combination with a corticosteroid in one study improved outcomes more than with one agent alone. 55 The deterioration in the condition of the rats following adrenalectomy with improvement after the addition of dexamethasone 53  The definition of heatstroke also differed across the studies: in one study, heatstroke was defined as the time that systolic blood pressure fell below 90 mmHg, 56 and in others, the time at which MAP and CBF fell. [53][54][55] However, heatstroke in humans may be present without changes in blood pressure.
No human studies were identified for the review, and the applica- and central neurological dysfunction, 10 which would not be possible to discern in an anesthetized animal model. Furthermore, anesthesia has more recently been observed to affect the inflammatory response. 59,60 All the subjects were anesthetized in these studies, but the implications are unclear. In three of the studies, 53-55 the subjects were anesthetized using urethane, which is known to cause immunosuppression. 61 The other two studies 50,56 used ketamine, which is associated with reduction in the pro-inflammatory TNF-α and IL-1, 62 higher levels of which are associated with adverse outcomes, but also a reduction of IL-6 62 ; which in turn is associated with improved outcome upon systemic increase.
Further evidence that steroids may be effective in human heat stress come from a number of recently published case reports. In one, persistent cardiovascular failure and high serum cytokine levels, associated with a worse outcome, 39,[41][42][43] improved after the administration of hydrocortisone, and the patient was subsequently discharged home. 63 Similarly, out of five patients admitted with classical heatstroke, three were treated with blood purification therapy, who subsequently survived, while the two who only received conventional therapy died. 64 The study authors propose that the improved outcome was due to the removal of pro-inflammatory cytokines, 64 suggesting that glucocorticoids may have a similar effect. 47,48 The administration of steroids before onset of heat stress in three of the studies made these data less relevant to clinical practice.
Two of the remaining studies, where steroids were given after onset of the heat stress, showed an improvement in mortality and organ dysfunction, although the effect of steroids administered after the insult was lower. 53 The optimal dose of corticosteroid from these studies is also uncertain. Two studies 53,54 showed a dose-dependent improvement. In the study by Yang 55 where the effect of dexamethasone on cytokine levels and neuronal damage was not significant, the dose used was 50% of the dose used in the study with similar methodology where significant differences were observed.
The optimum glucocorticoid and duration of treatment were not addressed in any of the studies, and remain to be determined. lower than that used in the one other primate study, where a reduction in endotoxemia and mortality was observed. 50 The successful treatment study administered 30 mg kg −1 methylprednisolone, equivalent to 5.6 mg kg −1 dexamethasone, whereas 2 mg kg −1 was used in the study describing the worsening response following steroids. The animals were anesthetized with ketamine, the influence of which on the immune function as discussed above is uncertain.
Treatment of severe hyperthermia irrespective of the cause currently remains limited to rapid cooling and supportive measures in the majority of cases. Development of new treatments to reduce the associated morbidity and mortality is urgently required.
Administration of corticosteroids appears promising and warrants further investigation.
Gastrointestinal permeability and a pro-inflammatory response appear to occur as a consequence of increased thermal load, irrespective of the cause 14,[24][25][26][27]39 ; whether corticosteroids are efficacious in hyperthermia of any etiology is unclear but would also warrant further investigation. Identifying a particular cause of a raised temperature is often difficult, but may be noninfectious in up to two-thirds of cases, 66 suggesting that steroid administration might prove to be beneficial even when the cause cannot be identified.

| CON CLUS ION
Heat stress is associated with a profound pro-inflammatory response. Steroids appear to improve morbidity and mortality in most animal studies, but their relevance to humans in clinical practice is uncertain. Further studies examining dose responses to corticosteroid administration in humans are warranted, notably where delivery occurs after the onset of heat stress.

ACK N OWLED G EM ENTS
The authors are grateful to the staff of the Royal Surrey County Hospital library for their assistance with the literature search and obtaining articles for review.

CO N FLI C T O F I NTE R E S T
The authors declare that there is no conflict of interest.

AUTH O R CO NTR I B UTI O N S
Both authors were involved in the conceptualization, data collection and analysis, and the writing of the paper. Both authors have seen and approved the final version.

DATA AVA I L A B I L I T Y S TAT E M E N T
All available data can be obtained by contacting the corresponding author. Provide an explicit statement of the question(s) the review will address with reference to participants, interventions, comparators, and outcomes (PICO)

Eligibility criteria 8
Specify the study characteristics (such as PICO, study design, setting, time frame) and report characteristics (such as years considered, language, publication status) to be used as criteria for eligibility for the review Describe the mechanism(s) that will be used to manage records and data throughout the review Selection process 11b State the process that will be used for selecting studies (such as two independent reviewers) through each phase of the review (ie, screening, eligibility and inclusion in meta-analysis) Data collection process 11c Describe planned method of extracting data from reports (such as piloting forms, done independently, in duplicate), any processes for obtaining and confirming data from investigators

Data items 12
List and define all variables for which data will be sought (such as PICO items, funding sources), any preplanned data assumptions and simplifications Outcomes and prioritization 13 List and define all outcomes for which data will be sought, including prioritization of main and additional outcomes, Describe anticipated methods for assessing risk of bias of individual studies, including whether this will be done at the outcome or study level, or both; state how this information will be used in data synthesis

Data synthesis 15a
Describe criteria under which study data will be quantitatively synthesized 15b If data are appropriate for quantitative synthesis, describe planned summary measures, methods of handling data and methods of combining data from studies, including any planned exploration of consistency (such as I 2 , Kendall's τ) 15c Describe any proposed additional analyses (such as sensitivity or subgroup analyses, meta-regression) 15d If quantitative synthesis is not appropriate, describe the type of summary planned Meta-bias(es) 16 Specify any planned assessment of meta-bias(es) (such as publication bias across studies, selective reporting within studies) Confidence in cumulative evidence 17 Describe how the strength of the body of evidence will be assessed (such as GRADE)