THERMOREGULATION

Control of body temperature, like most complex biological systems, is maintained by a complicated system of sensors and controls. The target set-point varies by <1°C per day on a circadian basis, and by <½°C monthly in women. However, at any given time the core temperature is within a few tenths of 1°C of the set-point.

The three major components of the thermoregulatory system are:

Temperature is sensed by A-δ fibres (most cold signals) and unmyelinated C fibres (most warm signals). These sensors are distributed throughout the body but the largest contribution is from the thermal core (deep abdominal and thoracic tissues, and neuraxis).

Signals from these sensors ascend via the spinothalamic tracts in the anterior spinal cord to the thermoregulatory centre located in the preoptic region of the hypothalamus near the floor of the third ventricle. This region contains heat-sensitive neurones but also receives neural input from other thermoreceptors. The preoptic region receives afferent information from peripheral thermoreceptors, determines the thermoregulatory set-point and coordinates appropriate responses.

Temperature is then regulated by a variety of central structures that compare integrated thermal inputs from skin, neuraxis and deep tissues with reference temperatures for each thermoregulatory response.

The most important effector response in humans against extreme environments is behavioural, and outweighs autonomic changes. At extremes of age, hypothalamic temperature regulation is impaired and less effective.

Humoral mediators from the circulation act to alter temperature primarily via the organum vasculosum of the lamina terminalis (OVLT), an area of fenestrated capillaries in the hypothalamus that permits cytokine access to neuronal receptors. Cytokines appear to be the endogenous pyrogens, with interleukin-6 (IL-6) and prostaglandin-E₂ (PGE₂) being a final common pathway. In addition to elevating body temperature, several cytokines also reduce the thermoregulatory set-point, and are known as endogenous cryogens.¹

Patients in the ICU are likely to have disturbance of both heat production and heat loss. Heat is produced as a result of metabolic activity and energy expenditure. Inflammation or infection as part of an acute phase response results in an increase in body temperature and energy expenditure. Pyrexia in turn results in an increased metabolic rate; however, reducing activity with sedation and muscle relaxation reduces energy expenditure.

Regulation of heat loss, which is the predominant effector of thermoregulation, is usually disturbed in the ICU population. Patients are usually nursed semi-naked, bed-bathed frequently, sedated and sometimes paralysed, and infused with drugs and fluids at ambient temperature. This effect is particularly pronounced during renal replacement therapy. Peripheral blood flow may be affected by vasopressors and the ability to shiver abolished by muscle relaxants. Behavioural defences may be compromised by sedative use. All of these effects play a variable part and, together with the method of temperature measurement, should be considered when evaluating fever in the ICU.

FEVER IN THE ICU

Fever is defined by a regulated hyperthermia, that is, it is a regulated elevation in the preoptic set-point temperature. Endogenous pyrogens as well as other mediators inhibit warm-sensitive neurones that normally facilitate heat loss and suppress heat production. This elevates the set-point temperature for all thermoregulatory responses and activates cold defences such as vasoconstriction and shivering, which decrease heat loss and increase metabolic heat production respectively. The set-point temperature returns to normal when pyrogen concentrations decrease, triggering heat loss by vasodilatation and sweating.²

Fever may reflect a wide variety of pathological processes including infection, inflammation, trauma, malignancy and connective tissue diseases (Table 74.1), necessitating a systematic and comprehensive diagnostic approach.³ It is often assumed that a patient presenting with a fever should be treated, regardless of the presence or absence of other symptoms. However, the evidence that anti-fever treatments lead to an improvement in morbidity or mortality, or even patient comfort, is lacking.⁴

Table 74.1 Causes of fever in the ICU

The development of fever in response to infection may be a protective adaptive response, and appears to be a phylogenetically preserved evolutionary response because of its survival value.⁵ In mammalian models, increasing body temperature results in enhanced resistance to infection. In humans, retrospective clinical trials have shown a positive correlation between maximum temperature on the day of bacteraemia and increased survival in patients with Gram-negative bacteraemia and spontaneous bacterial peritonitis.⁶ Also, septic patients with hypothermia have a poorer outcome than those who develop fever, although this causality is less clear. Both local and systemic hyperthermia has been used to facilitate cancer treatment. The protective effects of fever result from increased immune and cytokine functions.^7–10

Temperature elevation has been shown to enhance:

In addition, elevated temperatures inhibit some pathogens, such as Streptococcus pneumoniae.¹¹

Moderate fever is a common occurrence in ICU patients, but approximately half of these are non-infectious in origin.^12–14 The presence of fever frequently results in the performance of diagnostic tests and exposes the patient to unnecessary invasive procedures and inappropriate use of antibiotics.¹⁵

Whilst very high fevers (> 40°C) are dangerous, it is less clear whether moderate elevation of body temperature is detrimental and, indeed, may be protective.⁴ Moreover, artificially lowering the temperature of a febrile patient may mask the signs of infection and make diagnosis and monitoring more difficult. Any decision to adopt anti-fever measures, physical or pharmacological, must take into consideration the variable response by this patient population. Antipyretics may be ineffective.The usual concern about external cooling measures inducing peripheral vasoconstriction, reducing heat loss and making the pyrexia worse by shivering and hypermetabolism, may not be observed in sedated ICU patients.¹⁶ The most likely cause for this response is the drugs used to maintain sedation.^17,18

Pyrexia is associated with a number of deleterious physiological effects. Cardiac output, oxygen consumption, carbon dioxide production and energy expenditure are all increased, particularly in the presence of shivering. Oxygen consumption is increased on average by 10%/°C.² These changes are poorly tolerated by patients with limited cardiorespiratory reserve, and this group of patients would probably derive benefit from cooling measures. Other patient groups that require special consideration include those with immunosuppression, prosthetic implants and acute brain injury. Recent trials of therapeutic moderate hypothermia and traumatic brain injury indicate that hypothermia is a complicated treatment that is likely to benefit only a subgroup of patients with traumatic brain injury.^19–21

HEAT STROKE

A diagnosis of heat stroke is suggested when hyperthermia is associated with neurological abnormalities after exposure to high ambient temperature and/or vigorous exercise. Rectal temperature is usually greater than 42°C. Two distinct forms are recognised, and the spectrum of injury includes milder forms of thermal injury often termed heat stress.

Exertional heat stroke is a consequence of prolonged, intense exercise in warm humid environments, often seen in athletes and military recruits. Classic heat stroke is commonly seen in sedentary, elderly patients with underlying illnesses during heat waves. Factors predisposing to heat stroke are listed in Table 74.2. About 80% of heat stroke deaths occur in people aged 50 years and older, because of the diminished ability of the older body to compensate for increased core temperatures. Heat stroke is estimated to be the cause of approximately 1700 deaths each year in the USA.²² The European heat wave of 2003 was responsible for > 14 000 excess deaths within 2 weeks in France alone, of which a third were attributed to heat stroke, hyperthermia or dehydration.^23,24 A high mortality rate of > 62% was reported for this cohort, which is higher than that for leading killers in ICUs such as acute respiratory distress syndrome (ARDS) and septic shock.²⁵ Furthermore, there is a late mortality contributed by survivors who have sustained neurological injury. A study of former heat stroke patients suggests that susceptible individuals have a poorer physiological response to heat stress in terms of core temperature, heart rate and sweat response.

Table 74.2 Predisposing factors to heat stroke

There are two autonomic responses to heat stress: sweating and active precapillary vasodilatation. Sweating is extremely effective and can dissipate up to 10 times the basal metabolic rate, provided that environmental conditions such as ambient temperature, humidity and wind speed are optimal. The resemblance between heat illness and the effects of antimuscarinic drugs, which produce a central anticholinergic syndrome, is explained by the postganglionic, cholinergic sympathetic innervation of sweat glands. Vascular responses to heat stress include vasodilatation of peripheral vascular beds and vasoconstriction of splanchnic and renal beds. During severe heat stress, blood flow through the top millimetre of skin can be equal to the entire resting cardiac output.

DRUG-INDUCED HYPERTHERMIAS

In contrast to fever, the thermoregulatory set-point during hyperthermia remains unchanged at normothermic levels; however, body temperature increases in an uncontrolled fashion and overrides the ability of effector mechanisms to dissipate heat. Although a raised body temperature is not necessarily due to increased heat production but rather due to an imbalance between heat production and loss, most hyperthermias result as a consequence of net heat gain. Hyperthermia can result in dangerously high core temperatures by two mechanisms:

The numerous causes of hyperthermia are listed in Table 74.4. This section will review the relatively common causes of drug-induced hyperthermias, including malignant hyperthermia, neuroleptic malignant syndrome, and the sympathomimetic and anticholinergic syndromes.

Table 74.4 Causes of hyperthermia