Heat stroke

For a time, Question 23 from the first paper of 2015 had been the only CICM Part II  question asking about heat stroke. For trainees from a country composed almost entirely of beach and desert, Australian ICU trainees were surprisingly unfamiliar with heat stroke, and only 31% of them passed that question. In an effort to be counted among their number, the author has put together this summary. Subsequently, the college had also asked about heat stroke in Question 22 from the second paper of 2018, ahead of the record-breakingly hot summer of that year. This SAQ was completely different to the 2015 version, and asked for details which were probably more relevant to routine clinical practice.

As is typical of these chapters, this summary it is far from brief. For brevity, the time-poor exam candidate is referred to LITFL, where the pages on hyperthermia and heat stroke offer concise exam-oriented information. Locally, another chapter deals with the physiological response to dehydration. Malignant hyperthermia is a completely different issue, and is discussed elsewhere, as are the various different infectious and non-infectious causes of fever.

In short:

In addition to the summary above, the interested reader can review some published articles on this topic:

• A NEJM overview (Bouchama et al, 2002)
• An anaesthetic/ICU perspective (Grogan et al, 2002)
• An emergency physician's perspective (Glazer, 2005)
• The experience of army personnel (Bricknell, Part 1 and Part 2, 1996)
• Biochemical exploration of an animal model (Bouchama et al, 2005)

Definition of heat stroke and other thermal stress disorders

Heat syncope

This is the loss of consciousness due to peripheral vasodilatation in a high ambient temperature.

Heat cramp

This is the muscular cramping occurring during exercise in a high ambient temperature. It is related to salt deficiency and is usually benign.

Heat exhaustion

This is a state of dehydration and weakness which accompanies exposure to extremes of heat. Salt depleted heat exhaustion is what happens when unacclimatized personnel exercise and replace only water losses; water-depleted heat exhaustion affects well-acclimatised people who fail to replace water losses. In either case, the problem is a combination of electrolyte derangement and hypovolemia. The clinical features of this thermal stress disorder are nausea, vomiting, extreme weakness and lethargy, headache and dizzyness. The characteristic feature discriminating between heat stroke and heat exhaustion is the preservation of a normal level of consciousness in the latter.

Heat stroke

This is a condition characterised by severe hyperthermia, dry skin and a decreased level of consciousness. Most authors offer 40.6°C as the threshold temperature. For some reason, it is specifically a rectal temperature. The college answer in 2018 gives the following definiton:

"a core body temperature usually in excess of 40ºC with associated central nervous system dysfunction in the setting of a large environmental heat load that cannot be dissipated."

The defining feature of heat stroke is the rise of core body temperature in the context of a failing thermoregulatory system. Somewhat artifically, it is divided along aetiological boundaries into exertional and non-exertional heat stroke.

Exertional heat stroke

Well, it occurs following exertion. The usual setting is a healthy young person exercising in the heat. The characteristic feature of such heat stroke (Adams et al, 2012) is increased body thermogenesis due to exercise, and the failure of otherwise normal healthy thermoregulatory mechanisms. These are people who have wet, sweaty heat stroke.

Non-exertional heat stroke

As the name suggests, exertion is not required. The usual setting is an unhealthy elderly person with impaired thermoregulation, exposed to heat. This is the "classical" dry heat stroke of the person who is behaviourally incapable of finding water or escaping the sun.

Physiology of thermoregulation

An excellent overview of thermoregulation is afforded by an old BJA article (Buggy et al, 2000). Without going into excessive detail, it is possible to say that human thermoregulation is both behavioural and autonomic.

The autonomic component of thermoregulation is coordinated centrally by the hypothalamus, and involves some combination of sweating and vasodilatation. Afferent type C nerve fibers conduct the sensation of heat to the hypothalamus, reaching their maximal firing rate at 45-50°C. The anterior hypothalamus then receives this information, and the posterior hypothalamus coordinates an autonomic response. The homeostatic set point is usually between 36.7°C and 37.1°C: across this temperature range, the hypothalamic effector neurons do not fire. There is a diurnal variation in this (by about 1°C) and it varies even more during the menstrual cycle. The total range also seems to be extended by anaesthesia (from 0.4°C to 4.0°C). The effector response to hyperthermia is sweat production and cutaneous vasodilatation, leading to increased heat exchange and convective heat loss. Cardiac output increases by up to 20L/min, and changes in cutaneous vascular resistance can increase blood flow to the skin by up to 8L/min (Rowell et al, 1983). The evaporation of 1.7 ml of sweat will consume 1 kcal of heat energy, and at maximal efficiency sweating can dissipate about 600 kcal per hour.

The behavioural component of thermoregulation is by far the most important, quantitatively. No amount of vasodilation or sweating can ever be as effective as the pursuit of a nicely air-conditioned sports bar. At maximum, up to 2L of sweat can be produced per hour, and it is important to replace this by drinking voluntarily.  As sweat contains sodium and chloride, salt replacement must accompany rehydration. Salt depletion and dehydration are both important to the pathogenesis of heat stroke; Deschamps et al (1989) demonstrated that during heavy exercise the replacement of saline improved the rate of heat dissipation. Ergo, the combination of beer and salted snacks is the ideal replacement strategy.

Sustained exposure to excessive heat will eventually derail both behavioural and autonomic effector responses. As you dehydrate, the conservation of fluid prevents sweating. At this stage, the vasodilated skin becomes your enemy, allowing better heat exchange with the hot outside world (i.e. one's core temperature equilibrates with the ambient temperature).

Pathophysiology of heat stroke

Pathological processes associated with extreme hyperthermia

Failure of thermoregulation

• Dehydration leads to an impairment of the cardiovascular response to heat; the cardiac output does not increase sufficiently
• As a result, less sweat is produced
• The skin heats up
• Heat damage occurs in sweat glands, and sweat production is slowed to nil
• The hot dry skin fails as a heat exchange organ
• Hyperthermia worsens

Direct cellular damage from heat

•  At extreme temperatures (49°C to 50°C) cellular necrosis occurs in less than five minutes (Buckley et al, 1972)
• The threshold for cellular damage seems to be 41.6-42°C for 8 hours or so. Bynum et al (1978) performed a series of hideous experiments on sedated humans, and found that the consequence of heating them to that temperature for an hour at a time was still fairly safe.

Systemic inflammatory response

• Cytokine release due to heat stress
• Bacterial translocation and endotoxaemia
• Endothelial cell injury, which causes DIC and coagulopathy

The college, in their answer to Question 23 from the first paper of 2015 also mentioned uncoupling of oxidative phosphorylation, failure of enzyme systems, increased "Membrane permeability" (presumably increased cell membrae permeability?) and increased sodium leak into cells.

Organ system dysfunction resulting from heat stroke

The Grogan article (2002) is a gold mine of carefully organised information on this specific topic, ordered by organ system. The list offered below is a summary based largely on this article.

Respiratory failure with ARDS

• This is a consequence of the systemic inflammatory process
• Pulmonary oedema of cardiogenic origin is also not out of the question

Cardiovascular compromise:

• Tachyarrhythmia (electrolyte derangement and the effect of heat on the conducing system)
• Hypotension (due to volume depletion, circulatory diversion of blood into the vasodilated extremities, and excessive nitric oxide production)
• QT interval prolongation and ST segment changes
• Myocardial infarction

Neurological disturbance:

• Delirium, lethargy, coma, seizures
• The cerebellum is apparently the most susceptible to damage.
• The neurological sequelae are long-lasting in the majority of patients (about 75%)

Electrolyte and acid-base disturbances

• Lactic acidosis
• Respiratory alkalosis
• Hypophosphataemia and hypocalcemia due to deposition in injured muscle (Knochel & Caskey, 1977)

Renal failure

• Rhabdomyolysis
• Myoglobin-induced acute tubular encrosis

Acute hepatitis

• LFT derangement is seen due to direct tissue injury and splanchic blood flow diversion

Disseminated intravascular coagulation

• DIC occurs in some unknown proportion of heat stroke patients
• Consumption coagulopathy is exacerbated by hepatic failure

Intestinal bacterial translocation and endotoxaemia

• The diversion of blood flow away from the gut results in poorly perfused hollow viscera.
• The resulting bacterial translocation is to blame for much of the SIRS seen with heat stroke

History, examination and investigations of heat stroke

Risk factors for heat stroke

Environmental factors:

• High temperature (obviously)
• High humidity
• Low air movement

Patient factors:

• Increased thermogenesis (eg. exercise, hyperthyrodism, sepsis)
• Decreased volition or mobility (eg. dementia, delirium, physical disability)
• Failure of normal thermoregulatory mechanisms (eg. use of anticholinergic drugs leads to impaired sweating).
• Male gender (women are weirdly protected from heat stroke, particularly exertional heat stroke)

Acclimatisation: features protective against heat stroke

•  The process of acclimatization to heat is the gradual adaptation to increasing amounts of work performed in an uncomfortably hot environment. It usually takes several weeks.
  • Increased cardiac output
  • Activation of the renin–angiotensin–aldosterone axis
  • salt conservation by the sweat glands (i.e. the sweat of acclimatised people is less salty)
  • sodium retention by kidneys
  • an increase in the capacity to secrete sweat,
  • expansion of plasma volume,
  • an increase in the glomerular filtration rate to resist exertional rhabdomyolysis.

Classical features of clinical examination in heat stroke

• Raised body temperature
• Neurological dysfunction - restlessness, delirium, coma
• Hyperdynamic circulation; distributive shock (Shahid et al, 1999)
• Dry skin (usually)
• Seizures

Characteristic laboratory findings in heat stroke

• ABG: acidosis, probably mixed metabolic; as well as respiratory alkalosis and hypoxia
• BSL: elevated (catecholamines)
• FBC: haemolysis, thrombocytopenia, anaemia, raised white cell count
• EUC: renal failure, hyperkalemia
• CMP: hypophosphataemia and hypocalcemia (Knochel & Caskey, 1977)
• LFTs: raised transaminases and bilirubin. Specifically, AST and LDH will be raised.
• CK: elevated
• Urinary myoglobin
• Coagulopathy (DIC): raised PT and APTT
• Raised acute phase inflammatory markers (CRP, ferritin)

Factors which affect prognosis following heat stroke

In their answer to Question 23 from the first paper of 2015, the college only say that the prognosis depends on how hot, and for how long. Are there any genuine prognostic features? The best resource for this was Grogan et al (2002); that paper identified the  first three of the below-listed features. The others were found at the LITFL page about heat stroke; they in turn seem to originate from an audit of mortality which followed the 2003 heat wave in France (Misset et al, 2006).

• LDH, CK and AST levels (when extremely high) were predictive of non-survivors in a study of heat-stroked Haj pilgrims (Alzeer et al, 1997)
• Failure to decrease the core body temperature to below 38.9° within the first 30 minutes of presentation.
• A hyperdynamic circulation is protective, but a sluggish hypodynamic circulation is associated with a poorer survival
• Found collapsed at home (as opposed to public place or care facility)
• Preexisting cardiac disease
• Use of diuretics
• High body temperature
• Low Glasgow Coma Score
• Low platelet count
• Prolonged prothrombin time
• High serum creatinine
• High SAPS II score
• Use of vasoactive drugs within the first 24 hrs in the ICU

At 1 year follow-up, about one third were dead and another third had moderate or severe functional impairment.

Management of heat stroke

Goals of therapy for heat stroke

• Early, aggressive cooling
• Support of multiple failing organ systems

Cooling is the priority. The faster you get them under 39°C, the better their outcome. This goal of care is actually far from scientific. We know that if you take too long to cool them, their outcome is worse; but we don't know what the right number is. We aim for under 39°C. Not quite arbitrarily, but mainly because the early achievement of this endpoint was associated with improved survival in some of the studies.

More cooling techniques can be found in the chapter on the many methods of inducing therapeutic hypothermia. Here, only a few notable options will be listed. The rst of the management can only be described as "supportive" and aims to ameliorate the adverse effects of being undercooked.

Options for cooling the heat stroke patient

• Evaporation of cold water sponges
• Ice packs
• Immersion in ice water
• Contact cooling by blankets and jackets
• Iced gastric, colonic, bladder, or peritoneal lavage
• Infusion of cold intravenous fluids
• Invasive technique such as cooling of the dialysis circuit, or ECMO

There is not specific approach which is thought to be more effective than other approaches. For isntance, in a letter to Intensive Care Medicine, Hadad et al (2005) pointed out that in the Israeli Defence Forces, with tap water and a fan one is able to achieve a core temperature rate drop of 1°C every 9 minutes. Costrini (1990), looking at different ways of cooling down overheated athletes, suggested ice water immersion to be the best method. A more detailed discussion of cooling methods is carried out in the chapter on inducing therapeutic hypothermia. The college, in their answer to Question 22 from the second paper of 2018,  mention alcohol sponge baths as a discredited alternative.  This practice has been discredited since the 1960s, when it killed children (Senz et al, 1959) and adults (Wise, 1969) by producing a surprising amount of alcohol absorption (they were using mainly isopropyl "rubbing" alcohol). On the other hand, if your objective is to achieve heroic levels of intoxication, percutaneous obsorption is a valid method (Puschel et al, 1981).

A word about dantrolene in heat stroke

Dantrolene is a skeletal muscle relaxant that reduces muscular heat produced during abnormally sustained contraction. In heat stroke, there is no such abnormally sustained contraction, but the goal of reducing the muscular production of heat is laudable, as it may speed up the cooling process. Some might offer a standard muscle relaxant for this purpose. Anyway, two RCTs tested dantrolene in heat stroke. of the two, the smaller poorly designed one showed some false promise, whereas the larger better designed one confirmed that dantrolene has no role to play (Bouchama et al, 1991).

Supportive ICU management for the heat stroke patient

1. Intubate to protect the airway, if unconscious
2. Ventilate with lung protective ventilation, anticipating ARDS
3. Manage haemodynamic instability aggressively, with a mixture of cold IV fluids and vasopressor agents
4. Protect from seizures (no specific evidence to recommend benzodiazepines or any other conventional agents)
5. Control hyperkalemia and hyperphosphataemia of rhabdomyolysis
6. Consider early dialysis. Watch for myoglobinuria
7. Early trophic feeds to maintain gut integrity
8. Correct the coagulopathy of DIC