The hypokalaemia that came from the cold

ABSTRACT

While electrolyte disorders are common in nephrologists’ clinical practice, hypothermia is a condition that nephrologists rarely encounter. Hypothermia can induce several pathophysiological effects on the human body, including hypokalaemia, which is reversible with rewarming. Despite growing evidence from animal research and human studies, the underlying mechanisms of hypothermia-induced hypokalaemia remain unclear. Boubes and colleagues recently presented a case series of hypokalaemia during hypothermia and rewarming, proposing a novel hypothesis for the underlying mechanisms. In this editorial, we review the current knowledge about hypothermia and associated electrolyte changes with insights into the effects of hypothermia on renal physiology.

INTRODUCTION

Electrolyte disorders represent a common issue in most nephrologists’ clinical practice, usually when the cause is not immediately obvious or when the electrolyte disturbance is severe or refractory to treatment. In comparison, hypothermia is not a condition nephrologists deal with on a regular basis. Hypothermia results in several pathophysiological effects on the human body, such as hypokalaemia, which improves with rewarming. Despite the considerable evidence from experimental animal research and findings in humans, this phenomenon is still not completely understood, but many pathways have been proposed. These include K⁺ transcellular shifts, sympathetic drive with increased β-adrenergic activity and cold-induced diuresis with probable kaliuretic effect. In this issue of the journal, Boubes and colleagues [1] report an interesting case series of hypokalaemia during hypothermia and rewarming, and propose an interesting new hypothesis for the underlying mechanisms. Although the study had a limited number of subjects (n = 4), they showed a strong correlation between serum K⁺ level and changes in body temperature, with correlation coefficients that are >0.8 (P < .001). They discuss possible explanatory mechanisms and postulate that serum K⁺ changes are explained primarily by transcellular K⁺ shifts that are mediated through temperature-sensitive K⁺ exit channels [1]. In this article, we aim to put their findings into perspective. We discuss hypothermia and the commonly associated electrolytes changes, shed light on mechanistic pathways of hypothermia-associated K⁺ changes and describe the effects of hypothermia on renal physiology.

POTASSIUM AND OTHER ELECTROLYTES DISTURBANCES IN HYPOTHERMIA

Therapeutic hypothermia

Therapeutic hypothermia is an essential aspect in the treatment after cardiac arrest as well as a well-established therapeutic option in perinatal asphyxia [2–4]. In cardiac surgery, especially in operations on the aortic arch, hypothermia is used to minimize the risk of neurological damage to the greatest extent [5]. Considering the level of evidence, the decision to initiate therapeutic hypothermia for traumatic brain injury or subarachnoid haemorrhage is a case-by-case decision [6–8]. During therapeutic hypothermia, various clinically relevant dysregulations of serum electrolytes take place, especially potassium. In the context of controlled moderate hypothermia (32–34°C), a regular decrease in serum potassium usually ensues. Once deep hypothermia (<32°C) is reached, a much more pronounced decline in serum potassium takes place. While this is a consistent observation, it is still clinically difficult to predict changes in serum potassium in relation to temperature changes, and studies examining the magnitude effect of hypothermia on serum potassium are scarce.

In a study of 41 patients with severe traumatic brain injury, Polderman et al. investigated the effects of therapeutic hypothermia with a target temperature of 32°C [9]. Compared with the control group, the group of 21 subjects had a statistically significant decrease in serum magnesium, phosphate, calcium and potassium concentrations within the first 6 h of hypothermia (K⁺ dropped from 4.2 ± 0.59 to 3.6 ± 0.7 mmol/L, P < .01). The authors postulated that the underlying cause was the cold diuresis observed during the induction of hypothermia and the associated electrolyte losses. In a study with 500 cardiac surgery patients undergoing extracorporeal circulation and a target temperature range of 32–34°C versus 250 patients undergoing noncardiac surgery, electrolyte changes were comparable to those observed in the group of patients with traumatic brain injury. Similarly, the authors postulated that the underlying mechanism was a combination of hypothermia-induced increased diuresis with consecutive electrolyte loss and an intracellular electrolyte shift [10]. In a recent retrospective study of 310 patients with out-of-hospital cardiac arrest in whom targeted temperature treatment was initiated with a target temperature of 33°C for up to 48 h, Kirkegaard et al. observed temporary changes in potassium, magnesium, calcium and phosphate concentrations [11]. However, because the observed changes were minor and within reference ranges, the authors concluded that electrolyte changes in critically ill patients with targeted temperature treatment are less critical than previously thought. MacLaren et al. examined changes in serum electrolyte concentrations in a retrospective study of 91 patients with therapeutic hypothermia after cardiac arrest [12]. During hypothermia, there was a decrease in potassium and phosphate concentration and an increase in magnesium concentration. During the cooling or rewarming phases, significant electrolyte variations occurred. The authors attributed the electrolyte changes in the cooling phase to a shift of electrolytes to intracellular or extracellular and transcellular space and to an increased diuresis. In the examined study population, more than half of the patients developed an acute but reversible renal failure during hypothermia. It should be noted that in patients with impairment of the integrity of the central nervous system, there may be a disturbance of renal function with consecutive electrolyte imbalances unrelated to hypothermia (brain–kidney crosstalk) [13]. The administration of intravenous anaesthetics or hypnotics necessary for the induction and maintenance of hypothermia must also be considered in the interpretation of related electrolyte disturbances. In a study of patients with acute subarachnoid haemorrhage who were given barbiturates for sedation, Seule et al. observed concomitant hypernatraemia [8].

One additional possibility to explain cold-induced transcellular K⁺ flux is an activation of brown adipose tissue. While limited in quantity in adults, this type of fat is activated by way of adrenergic nerves following cold exposure. Brown adipose tissue functions to dissipate energy through the production of heat due to the presence of an uncoupling protein called UCP-1 located on the inner mitochondrial membrane. Acute cold exposure is associated with increased circulating aldosterone, a hormone that regulates the internal distribution of K⁺ and activates brown adipose tissue [14, 15]. It is interesting to speculate that sympathetic nerves and aldosterone following cold exposure may contribute to acute sequestration of K⁺ within depot of brown adipose tissue.

Thus, for electrolyte shifts during therapeutic hypothermia, it can be stated that hypokalaemia usually occurs during the cooling phase, presumably due to redistribution. In rewarming, the risk of hyperkalaemia with consecutive severe cardiac arrhythmias is substantial. The clinical relevance of changes in serum concentrations of sodium, phosphate, calcium and magnesium is uncertain because of the inconsistent evidence. A further important indication of therapeutic hypothermia is in neonates with moderate to severe hypoxic encephalopathy which is associated with rather complex changes owing to the unique pathophysiological properties of the newborn [16–18].

Accidental hypothermia

Accidental hypothermia mostly arises from accidents, intoxications or near-drowning situations. Secondary hypothermia, such as hypothyroidism-related hypothermia, is less common, usually occurs in the winter months and often affects the elderly [19].

Accidental hypothermia may cause hypokalaemia at its onset due to β-adrenergic stimulation, which is aggravated by an intracellular redistribution during its progression [20]. According to the European Resuscitation Council (ERC), hypokalaemia can be classified as mild ([K⁺]: 3.0–3.4 mmol/L), moderate ([K⁺]: 2.5–2.9 mmol/L) or severe ([K⁺]: <2.5 mmol/L) [21]. In moderate to severe hypokalaemia, the potassium deficit may range from 300 to >500 mmol, and must be replaced. Since potassium can only be substituted via an indwelling venous cannula in dilute solution up to approximately 80 mmol/24 h, the insertion of a central venous catheter should be considered. Twenty to 40 mmol/h can be substituted easily with close monitoring of serum potassium concentration. If there is only a sluggish increase in potassium concentration, concomitant hypomagnesaemia must be considered, since Na⁺/K⁺-ATPase is magnesium dependent. In this context, a bolus of 20 mmol Mg²⁺ over 2–4 h followed by 10–20 mmol/day via continuous infusion has proven effective. In any case, a rapid rise in serum potassium must be expected during rewarming, as well as concomitant acidosis.

As in therapeutic hypothermia, the body reacts to decreasing core temperature with, among other things, hypothermia-induced increasing diuresis. Up to a core body temperature of 25°C, serum electrolyte concentrations usually remain constant [22]. However, one of the major challenges in managing patients with accidental hypothermia seems to be hyperkalaemia with very high serum potassium concentrations, sometimes >10 mmol/L [23, 24]. While this could arise from metabolic acidosis, which affects up to 30% of hypothermic patients, extreme hyperkalaemia can also be an indication of irreversible cell damage [24–26]. However, the underlying pathophysiological changes have not yet been adequately elucidated. For example, metabolic acidosis can occur with hypokalaemia, as reported in subjects after immersion in cold water [27]. It is possible that as yet unidentified ion pumps that exchange potassium against intracellular H⁺ and are temperature dependent could be the underlying pathophysiological explanation [1].

Severe hyperkalaemia from the outset of resuscitative efforts in accidental hypothermia usually indicates a poor prognosis. Current ERC guidelines suggest a potassium limit of 8 mmol/L for avalanche victims and 12 mmol/L for other causes of acute hypothermia as thresholds beyond which survival is unlikely [21]. Those thresholds were based on case reports of survivors without neurologic impairment [23, 28]. It is important to note that according to ERC recommendations, hyperkalaemia should not be used as a sole parameter when determining whether to discontinue therapy in severely hypothermic patients [29–31].

Hyperkalaemia can be classified as mild ([K⁺]: 5.5–5.9 mmol/L), moderate ([K⁺]:6.0–6.4 mmol/L) and severe ([K⁺]: ≥6.5 mmol/L). Similar to hypokalaemia, hyperkalaemia can also be associated with life-threatening tachycardic or bradycardic cardiac arrhythmias. The indication to treat even a moderate hyperkalaemia usually arises from any cardiac arrhythmias or muscle weakness [32]. In the treatment of hyperkalaemia, a distinction is made between short-term emergency treatment and treatment aimed at eliminating excessively high potassium concentrations. Emergency management includes the administration of calcium gluconate to prevent life-threatening cardiac arrhythmias. Glucose-insulin infusion causes a redistribution of potassium from extracellular to intracellular. Inhaled or intravenous β2-sympathomimetics have the same effect, but they may induce undesirable tachycardic arrhythmias in the hypothermic critically ill patient due to β1-adrenoceptor agonism which occurs with higher doses. Accompanying metabolic acidosis can be treated with sodium bicarbonate. However, interventions to eliminate excessive high potassium are more effective. This can be achieved by administering loop diuretics such as furosemide, supplemented, if necessary, by infusion of isotonic saline. Cation exchangers such as polysulfonate resins must be administered orally or rectally and, therefore, may not be appropriate in patients with accidental hypothermia. The most effective, but also the most invasive, is hemodialysis. This can also be used for active and controlled warming of the hypothermic patient [33]. In patients with concomitant rhabdomyolysis, medium cut-off haemofilter systems can be integrated into the dialysis system to prevent and treat resulting acute renal injury [34].

In general, electrolyte alterations in patients with accidental hypothermia are similar to those in patients with therapeutic hypothermia. However, patients with accidental hypothermia have a higher risk of severe, outcome-relevant hyperkalaemia due to concomitant injuries.

HYPOTHERMIA EFFECTS ON RENAL PHYSIOLOGY

Generally, when hypothermia ensues, a number of circulatory changes take place, resulting in redistribution of blood flow. Cardiac output drops, but vascular resistance goes up, resulting in a significant reduction in renal blood flow even with maintained mean arterial pressure [35, 36]. Increased blood viscosity due to hypothermia appears to augment vascular resistance [37, 38]. The notable decline in renal blood flow is believed to result from afferent vasocontraction, resulting in a decrease in glomerular filtration rate (GFR). This reduction can reach up to 50%, as shown in animal models [39]. It is conceivable that the decline in GFR is mediated through increased sympathetic activity, given the rise in catecholamine levels and sympathetic activity during the initial phase of hypothermia [40, 41]. However, experiments on rats with bilateral adrenalectomy and denervated kidneys continue to manifest a reduction in GFR, suggesting a direct effect of hypothermia on renal vascular bed and GFR [42].

Interestingly the reduction of GFR is not coupled with a reduction in urine output. In fact, an increase in urine flow is usually observed and often referred to as ‘cold-induced diuresis’ [35]. In humans, this is demonstrated not only in studies of post–cardiac arrest patients receiving therapeutic hypothermia with an increase in urine output at the induction phase of hypothermia but also in studies of healthy volunteers [43–45]. Cold-induced diuresis is believed to be related to both solute and water diuresis. Numerous animal experiments showed that hypothermia increased the fractional excretion of Na⁺ and K⁺, indicating a considerable loss in renal tubular absorptive capacity. This could be more pronounced for K⁺ than Na⁺ [35, 39, 46]. However, clinical studies on human subjects do not show an increased urinary K⁺ excretion but rather urinary Na⁺ excretion [45, 47]. This effect may also be explained by the observed increase in atrial natriuretic peptide, resulting in salt diuresis irrespective of the subject's volume status [45, 48, 49]. Another possible interplaying factor is the effect of antidiuretic hormone (ADH). Hypothermic rats were found to have lower ADH levels than controls [50]. There is also a possibility that distal tubules become less responsive to ADH, resulting in higher free water excretion [51]. However, human studies reveal conflicting results on levels of ADH in hypothermia [47, 52].

Even though the majority of studies showed an increase in urine production, there is not much evidence to back this up as a cause of hypokalaemia. In fact, hypothermia-induced hypokalaemia can be demonstrated even in anephric animals, underscoring the role of transcellular K⁺ shifts [53].

CONCLUSION

John le Carré’s 1960s espionage novel ‘The Spy Who Came in from The Cold’ is often characterized as sophisticated and known for its unexpected twists and turns and leaves room for interpretation to the present day. To some extent, the same applies to hypokalaemia associated with hypothermia. Boubes and colleagues [1] have added to the intrigue with their current paper and present food for thought for nephrologists who are interested in unusual electrolyte disorders. Acute physicians, intensivists and nephrologists should be more aware of this unusual association when they see patients with hypothermia of whatever cause.

CONFLICT OF INTEREST STATEMENT

M.E. and A.W. are members of the CKJ editorial board.