Description of the condition
Hypotonic hyponatraemia is a common condition, occurring in up to 60% of people admitted to hospitals, depending on the definition of hyponatraemia, the types of patients who are studied and the healthcare facility to which these patients are admitted (Upadhyay 2009). Hypotonic hyponatraemia is usually defined as a serum sodium concentration < 135 mmol/L with an osmolality < 285 mOsm/kg (Reynolds 2006). It develops when the body retains an excess of water relative to the amount of sodium. It can be caused by intrinsic kidney disease but usually results from incomplete suppression of vasopressin activity despite decreased tonicity of the plasma. In situations of decreased circulating blood volume, vasopressin release is increased in a physiologic response to maintain haemodynamic homeostasis. This occurs both with true volume depletion and with reduced effective arterial circulating volume, as seen in heart failure, liver cirrhosis or nephrotic syndrome. In the syndrome of inappropriate antidiuretic hormone secretion, the increased release of vasopressin is non-haemodynamic and can have multiple causes including ectopic production of vasopressin by a variety of tumours (Verbalis 2007).
When plasma tonicity is low, water tends to enter the cells and causes them to swell. If blood sodium concentrations drop rapidly (within a 48 hour period), the swelling of brain cells may lead to brain oedema, brain stem herniation and eventually even death. Fortunately, when blood sodium concentrations drop more gradually, brain cells adapt to their hypo-osmolar surroundings and prevent swelling by the transport of solutes from the intracellular to the extracellular compartments. As a consequence, immediate symptoms attributable to chronic hyponatraemia are usually less severe (Reynolds 2006). Nevertheless, people with chronic hyponatraemia have reduced attention and less stable gait than those without hyponatraemia (Renneboog 2006). They fall more often and have increased risk of osteoporosis and bone fractures (Arampatzis 2013; Hoorn 2011; Kinsella 2010; Renneboog 2006; Verbalis 2010). Finally, they stay in hospital longer and have an increased risk of death, even when sodium concentrations are only mildly decreased and underlying or comorbid conditions are taken into account (Wald 2010).
Description of the intervention
It is accepted that acute hyponatraemia requires an immediate increase in serum sodium concentration to prevent severe neurologic complications (Ellison 2007). What to do with chronic hyponatraemia is less clear. Firstly, chronic hypotonic hyponatraemia has been treated under the assumption that increasing the sodium concentration improves important health outcomes; that patients live longer, feel better and are hospitalised less frequently. Although several observational studies have indicated an association between hyponatraemia and undesirable outcomes, it is still unclear whether treatment improves them (Upadhyay 2009; Wald 2010). Secondly, once brain cells have adapted to their hypo-osmolar environment, they become vulnerable to osmotic demyelination. Although rare, osmotic demyelination is a devastating neurologic complication that may occur when the myelin sheath around pontine and extrapontine neurons breaks down after increases in serum sodium concentration exceed 8 to 12 mmol/L/24 h and 20 mmol/L/48 h (Adrogue 2012; Ellison 2007; Reynolds 2006). Treatment for chronic hyponatraemia must balance the uncertain benefit of increasing the sodium concentration against this risk of overly rapid correction.
How the intervention might work
Whatever the underlying cause, hyponatraemia usually results from urine being insufficiently dilute to maintain serum osmolality within the normal range (Adrogue 2000). Several treatment strategies can be used to try and overcome this (Adrogue 2012; Ellison 2007; Verbalis 2007).
Restriction of fluid intake aims to decrease the amount of free water needing excretion.
Urea and mannitol improve electrolyte-free water clearance by increasing urine osmolarity and creating osmotic diuresis (Lindner 2012).
Loop diuretics, such as furosemide, bumetanide and ethacrynic acid, impair free-water absorption in the collecting duct by reducing the hypertonicity of the renal medulla.
Corticosteroids with a mineralocorticoid effect increase renal sodium retention by active reabsorption of sodium in the principal cells of the cortical collecting tubule.
Demeclocycline, lithium, phenytoin and vasopressin receptor antagonists act by pharmacologically inhibiting the effect of antidiuretic hormone on the principal cells of the collecting duct, thereby limiting insertion of water channels in the luminal membrane and thus preventing free water reabsorption.
As hyponatraemia with true volume depletion is treated by restoring volume with water and salt, it will not be covered in this review.
Why it is important to do this review
The benefits and harms of treatments for chronic non-hypovolaemic hypotonic hyponatraemia have not been formally evaluated in a systematic review. Two systematic reviews have explored the efficacy and safety of vasopressin receptor antagonists (e.g. conivaptan, lixivaptan, satavaptan, tolvaptan) versus placebo, no treatment or fluid restriction (Jaber 2011; Rozen-Zvi 2010), but to our knowledge there has been no attempt to compare them with any other intervention or to compare any of the other interventions versus placebo or against one another.
Both systematic reviews have found an early increase in serum sodium concentration, but no improvement in outcomes important to patients. Indeed, most randomised controlled trials (RCTs) have evaluated short-term and surrogate outcomes only, making it difficult to adequately asses any expected benefit in the long-term. Since the most recent systematic review was published, at least three additional RCTs comparing vasopressin receptor antagonists versus placebo have been completed, increasing the total sample size by at least 60%. Although outcomes are still mostly surrogate and short-term, the largest study was terminated early due to a numeric imbalance in the number of early deaths in the experimental group (FDA 2012). We believe it justifies reanalysis including in-depth investigation of possible heterogeneity at this point.