Hyponatremia With Intracranial Disease: Not Often Cerebral Salt Wasting

The degree to which hyponatremia occurs primarily as a result of natriuresis has remained controversial for many years. Cerebral salt wasting (CSW) was first proposed by Peters et al (1) in 1950 as an explanation for the natriuresis and hyponatremia that sometimes accompany intracranial disease, particularly subarachnoid hemorrhage (SAH), in which up to one-half of patients often develop hyponatremia. After the first clinical description of the syndrome of inappropriate antidiuretic hormone secretion (SIADH) in 1957, such patients were generally assumed to have hyponatremia secondary to nonosmotic arginine vasopressin (AVP) secretion with a secondary natriuresis (2). However, both clinical and experimental data have suggested that some patients with SAH and other intracranial diseases may actually have a primary natriuresis leading to extracellular fluid (ECF) volume contraction rather than SIADH (3, 4), in which case the elevated measured plasma AVP levels may actually be physiologically appropriate for the degree of volume contraction present. The major clinical question as to the frequency of CSW as a primary or contributory cause of hyponatremia is critically dependent on the criteria used to assess the ECF volume status of these patients; opponents argue that there is insufficient evidence of true hypovolemia despite ongoing natriuresis (5), whereas proponents argue that the combined measures that have traditionally been used to estimate ECF volume do, in fact, support the presence of hypovolemia in many such cases (6). Noticeably lacking from this ongoing debate has been carefully done clinical research that critically assesses the cause of hyponatremia in neurological diseases such as SAH. The article by Hannon et al (7) in this issue of the JCEM addresses this deficiency by carefully studying 100 patients with acute nontraumatic aneurysmal SAH with serial assessments of clinical ECF volume status by a single experienced clinician and measurements of plasma cortisol, AVP, and brain natriuretic peptide (BNP). Their results demonstrated the expected high incidence of hyponatremia in 49% of the patients; but when all analyses were considered, the cause of the hyponatremia was attributable to SIADH in 71.4% of patients and acute glucocorticoid deficiency in 8.2%, with the remaining cases caused by incorrect iv fluid administration or hypovolemia. Most significantly, no cases were found that met historically accepted criteria for a diagnosis of CSW. Thus, as opposed to the current trend to diagnose CSW with a relatively high frequency in neurological cases, this careful analysis indicates that CSW is an exceedingly rare cause of the hyponatremia in SAH and, by analogy, likely in other intracranial disorders as well. Since the original cases studied by Schwartz and Bartter, increased renal sodium excretion has been viewed as one of the cardinal manifestations of SIADH, indeed one that later became embedded in the requirements for its diagnosis (8). However, next to the use of the term “inappropriate,” probably no other aspect of SIADH has been so widely misinterpreted. Demonstration that the natriuresis accompanying administration of antidiuretic hormone is not due to AVP itself, but rather is due to the volume expansion produced as a result of water retention, was unequivocally shown by Leaf et al (9) even before the description of the clinical occurrence of SIADH. Subsequent metabolic balance studies demonstrated that excess urinary sodium excretion and a negative sodium balance occurred during the development of hyponatremia in patients with SIADH, but eventually urinary sodium excretion simply reflected daily sodium intake (10). Patients appear to exhibit renal sodium “wasting” because they continue to excrete sodium despite being hyponatremic, but in reality they have simply achieved a new steady state in which they are in neutral sodium balance, albeit at a lower serum sodium concentration ([Na+]). Although this interpretation is now supported by abundant clinical and experimental evidence, several important questions remain unanswered regarding natriuresis and hyponatremia: What physiological and/or pathophysiological mechanisms underlie the natriuresis? Is natriuresis in SIADH always secondary to AVP-induced water retention or is hyponatremia sometimes caused primarily by sodium losses? Even when natriuresis is secondary to water retention, can the natriuresis further aggravate the hyponatremia? Studies of long-term antidiuretic-induced hyponatremia in both dogs and rats have indicated that a larger proportion of the hyponatremia is attributable to secondary sodium losses rather than to water retention (11, 12). However, it is important to appreciate that in these models the natriuresis actually did not worsen the hyponatremia, but rather allowed volume regulation of blood and ECF volumes to occur. Therefore, over long periods what begins as a “purely” dilutional hyponatremia from water retention becomes a mixed hyponatremia in which urinary solute losses allow maintenance of equivalent levels of hyponatremia but with lesser degrees of volume expansion due to water retention. Much of the past difficulty in consistently demonstrating expanded plasma or ECF volumes in patients with SIADH using radioisotope tracer dilution techniques can probably be ascribed to this process. It has become clear that intrinsic renal mechanisms are capable of producing both diuresis and natriuresis in response to increases in renal perfusion pressures; this mechanism has been shown to underlie the renal escape from antidiuresis produced when AVP-infused animals are continually fluid loaded (13). However, it has not yet been proven whether this mechanism is sufficiently sensitive to detect the relatively mild degrees of volume expansion that accompany dilutional hyponatremias. Another, not mutually exclusive, possibility is that the natriuresis is mediated via increases in circulating natriuretic peptides such as atrial natriuretic peptide (ANP) and BNP. Most cases of SIADH have been shown to have elevated levels of these peptides into ranges that are capable of promoting renal sodium excretion (14). Both plasma and cerebrospinal fluid ANP and BNP levels are clearly elevated in many patients with SAH (15) and have been found to correlate variably with hyponatremia in patients with intracranial diseases (16). However, because SIADH also is frequently associated with elevated plasma ANP and BNP levels, this finding alone does not prove causality, which is nicely demonstrated by the studies of Hannon et al (7) in which the serum BNP levels found in patients with SIADH were equivalent to the other etiological groups. Ample precedent certainly exists for hyponatremia due to sodium wasting with secondary antidiuresis in Addison's disease, as well as diuretic-induced hyponatremia. Characteristic of these disorders, normalization of ECF volume with isotonic saline infusions restores plasma osmolality to normal ranges by virtue of shutting off secondary AVP secretion. If hyponatremia in patients with SAH occurred via a similar mechanism, it should also respond to this therapy; however, studies have indicated that it does not (17). These other results raise the possibility of disordered AVP secretion as well as a coexisting stimulus to natriuresis in many such patients. It therefore seems possible that SAH and other intracranial diseases might represent a mixed disorder in which some patients have both inappropriate AVP secretion and exaggerated natriuresis; which effect predominates in terms of the clinical presentation will depend on their relative intensities as well as the effects of concomitant therapy. In patients with CSW, the elevated ANP and BNP levels could act simply to further exacerbate the secondary natriuresis produced primarily by AVP-induced water retention and ECF volume expansion. The second notable result of the study by Hannon et al (7) is the finding of evidence of putative cortisol deficiency in a surprisingly high percentage of patients with SAH, with normalization of serum [Na+] after steroid replacement of the hyponatremic SAH patients. Nonosmotic AVP secretion has been strongly implicated in the impaired water excretion of glucocorticoid insufficiency because elevated plasma AVP levels have clearly been documented in animals and patients with hypopituitarism. That these elevated AVP levels are causally related to the impaired water excretion was suggested by studies using an AVP V2 receptor antagonist, which demonstrated near normalization of urinary dilution in adrenalectomized mineralocorticoid-replaced rats (18). AVP-independent mechanisms have also been suggested to play a role in the impaired water excretion of glucocorticoid deficiency, but AVP-independent effects of glucocorticoid insufficiency remain poorly defined at the present time. Experimental data have suggested participation of both hemodynamic stimuli and direct pituitary effects on AVP secretion with glucocorticoid deficiency. Both clinical and experimental studies have shown a modest but significant effect of glucocorticoids to inhibit pituitary AVP secretion. Alternatively, in the absence of glucocorticoid feedback inhibition of the parvocellular AVP neurons that project to the median eminence rather than to the posterior pituitary, AVP content increases markedly in this area. This presumably reflects increased secretion of AVP into the pituitary portal blood system in order to stimulate pituitary ACTH secretion. Because the pituitary portal blood eventually drains into the systemic circulation, increased levels of AVP released from the median eminence could therefore increase plasma AVP levels sufficiently to produce some degree of inappropriate antidiuresis; it is important to remember that such levels need not be very high, but simply inappropriate for the plasma osmolality. There are several important implications of the results of Hannon et al (7) for the management of hyponatremia in patients with intracranial diseases. First, all hyponatremic patients require careful assessment of ECF volume status; patients assessed as euvolemic should be assumed to have SIADH and be treated accordingly pending additional laboratory confirmation. Typical of SIADH, infusion of isotonic saline will generally not correct the hyponatremia and may even worsen the serum [Na+] in some cases because the infused water is retained whereas the infused sodium is excreted, a process that has been called “desalination” (19). However, fluid restriction is relatively contraindicated in patients with SAH due to an observed increased incidence of adverse outcomes from cerebral vasospasm (20). Therefore, in patients with neurological disease in whom the hyponatremia is felt to be symptomatic or to compromise standard therapies, consideration should be given to infusing hypertonic (3%) saline or combining isotonic saline with aquaretic therapies such as vasopressin V2 receptor antagonists (vaptans) and urea or, in carefully selected cases, loop diuretics. Although vaptans are contraindicated in hypovolemic patients, which would include patients with true CSW, they can be safely utilized in most hyponatremic patients with SAH who have SIADH and therefore are by definition euvolemic, as long as care is taken to maintain the patient's overall fluid balance neutral or positive. Secondly, endocrinologists must remain vigilant about assessing the status of the hypothalamic-pituitary adrenal axis in all patients with intracranial disease, including SAH. Hyponatremia in a patient with sellar and suprasellar disease should always be considered to represent glucocorticoid deficiency until proven otherwise. As pointed out by Hannon et al (7), cortrosyn stimulation tests are not valid in patients with potential acute ACTH insufficiency, and insulin tolerance tests cannot be employed safely in patients with active neurological disease. Consequently, the approach of assessing serial morning cortisol levels in hyponatremic patients and empirically replacing patients below critical levels, at a minimum < 300 nmol/L (11 μg/dL), is a sound approach until patients can be re-evaluated more rigorously at a later date. The insights provided by the study of Hannon et al (7) therefore provide the basis for a more rational approach to the evaluation and treatment of patients with hyponatremia and intracranial disease and serve to demonstrate the utility of well-constructed prospective cohort studies for understanding clinically important disorders of fluid and electrolyte metabolism. Disclosure Summary: The author is a consultant to Cornerstone Therapeutics, Inc, Ferring Pharmaceuticals A/S, and Otsuka America Pharmaceutical, Inc, and has received research grants from Otsuka America Pharmaceutical, Inc. atrial natriuretic peptide arginine vasopressin brain natriuretic peptide cerebral salt wasting extracellular fluid subarachnoid hemorrhage syndrome of inappropriate antidiuretic hormone secretion.