Matt Harmon

110 Chapter five normothermia, as we found larger than expected individual variation in cytokine levels. Also, sedation due to pharmacological adjuncts may have partially negated the effects of cooling by reducing hyperventilation and resultant hypocapnia. Hyperventilation has previously been shown to lower cytokine levels in an endotoxemia model. 16 Induced normothermia lowered heart rate while maintaining perfusion as measured by MAP and lactate, suggesting cooling did not adversely affect tissue perfusion. Previously, cooling of febrile sepsis patients reduced the amount of vasopressor dose needed to maintain an adequate MAP. 4 Reducing heart rate during sepsis could directly benefit patients, by reducing myocardial oxygen consumption and improving diastolic relaxation time which improves coronary perfusion. 17 Interestingly, in patients with septic shock, treatment with esmolol decreased mortality. Esmolol reduced heart rate and endogenous norepinephrine levels while increasing stroke volume index, 8 and microcirculatory flow. 17 Thereby, cooling may be of benefit in patients with supposedly high adrenergic activity. However, not all studies show hemodynamic benefits from cooling to normothermia. In previous studies in patients with septic shock and fever, cooling to normothermia decreased MAP, cardiac output and increased lactate. 6,7 These apparent differences with our results may in part relate to timing and intensity of cooling. 7 Although these studies are difficult to compare due to the use of different target temperatures and lack of a control group, the results indicate that during cooling, some form of tissue perfusion monitoring is warranted. In general, clinical or biological markers to assess the risks and benefits of fever in individual septic patients is lacking. Subsequently, treating all septic patients with induced normothermia will likely result in a heterogeneity of the treatment effect. Future clinical studies should focus on finding markers to determine a patient’s appropriate body temperature in sepsis, in order to determine which patients may benefit from cooling. Importantly, cooling was well tolerated in the awake volunteers. Cooling in awake volunteers without pharmacological adjuncts to reduce shivering results in an increased metabolic rate and actually aggravates fever due to compensatory autonomic mechanisms. 11 Several different pharmacological interventions have been previously used to overcome shivering during cooling of awake volunteers. 13,18 Clonidine, an a2-agonist used in this study, is effective in preventing shivering in healthy individuals undergoing external cooling, possibly by impairing central thermoregulatory control. 12 MgSO 4 has been shown to increase the rate of cooling, likely in part by preventing peripheral vasoconstriction. 19 In addition, MgSO 4 can mitigate thermal discomfort due to surface cooling. 19 Buspirone is a serotonin 5-HT1A partial agonist and directly lowers body temperature. In a previous