Glucocorticoids and Water Balance: Implications for Hyponatremia Management and Pituitary Surgery Open Access

Water balance is fundamental to all homeostasis. The body’s exquisite maintenance of tonicity by the centers of osmoregulation in the ventral hypothalamus is well established; however, the contribution of glucocorticoids to water balance is less clear. Here, we explore the complex interplay of glucocorticoids on osmoregulation, renal free water clearance, and systemic hemodynamics to establish their role in the maintenance of water balance and in the pathophysiology of hyponatremia, with special reference to the postoperative period after pituitary surgery.

The body independently regulates plasma tonicity (effective osmolarity) and plasma volume through water and sodium balance, respectively. Disorders of volume (hypervolemia and hypovolemia) are disorders of sodium balance, while disorders of tonicity (hyponatremia and hypernatremia) are disorders of water balance. Volume disorders may develop independently or coexist with disorders of tonicity.

The body regulates plasma tonicity through arginine vasopressin (AVP) (antidiuretic hormone) release, which controls the concentration of urine and the thirst response. AVP secretion is stimulated by multiple factors, but the two primary drivers are tonicity and the effective intravascular volume state. Tonicity is sensed by the circumventricular organs and, together with volume sensors (baroreceptors) in the great vessels, controls the secretion of AVP from the posterior pituitary gland. AVP causes free water retention in the kidneys, reducing plasma tonicity and increasing urinary tonicity (Fig. 1). The circumventricular osmoreceptors are exquisitely sensitive to changes in tonicity; perturbations as small as 1% cause changes in AVP release.

Regulation of plasma volume is controlled primarily by the renin-angiotensin-aldosterone axis. Baroreceptors and the juxtaglomerular apparatus control the secretion of angiotensin II and aldosterone, which causes retention of sodium and (following an osmotic gradient) water (volume). Effective intravascular volume is sensed by baroreceptors in the carotid sinus and aortic arch. As they are stretch receptors, they do not sense total plasma volume per se but rather the hemodynamically effective local intravascular volume. Regulation of volume is largely independent of water balance; the exception is states of significant effective intravascular hypovolemia, where the defense of tonicity is sacrificed for the defense of volume and free water is retained. This occurs due to direct AVP release in response to baroreceptor signaling and changes to the set point and gain of the AVP response curve due to plasma tonicity. Simply, hypovolemia leads to greater AVP secretion for a given tonicity and reduces the threshold for secretion to a lower tonicity (Fig. 2). Although a much greater perturbation in volume is required for AVP secretion compared to tonicity, the AVP response to volume depletion is exponential and stronger, although this diminishes with age.

Corticotropin-releasing hormone (CRH), adrenocorticotropic hormone (ACTH), and glucocorticoids have a complex relationship with AVP secretion and water balance (Fig. 3).

CRH and AVP have a synergistic relationship. AVP potentiates the effect of CRH on corticotrophs during stress, increasing ACTH secretion 30-fold compared to CRH simulation alone [1, 2], although this AVP action is subject to tachyphylaxis [3]. Indeed, in animal models, AVP, and not CRH, mediates acute stress-induced glucocorticoid release, while CRH provides basal stimulatory tone [4]. The relative importance of AVP in the acute stress response in humans requires confirmation. CRH is also an AVP secretagogue [5] so as to harness this synergistic effect during a stress response [6, 7]. CRH is stimulated by hypovolemia, and downstream glucocorticoid secretion is enhanced by concomitant AVP release in response to volume depletion [8]. CRH also changes the gain of the AVP response to hypertonicity, leading to greater secretion for a given osmotic stimulus [9, 10]. Despite the well-known effects of AVP on ACTH secretion, the reciprocal relationship is less clear. ACTH may be an AVP secretagogue [11], but human evidence is lacking.

Cortisol is the primary glucocorticoid receptor agonist; however, cortisol is also a mineralocorticoid receptor agonist, with potential implications for salt and water balance. Cortisol has similar potency on the mineralocorticoid receptor as aldosterone, but basal plasma concentrations are 2–3 orders of magnitude greater than those of aldosterone [12]. Inactivation of cortisol to cortisone by 11β-HSD2 allows aldosterone to primarily mediate the mineralocorticoid response in the principal cells of the nephron. Centrally, 11β-HSD2 expression is low in comparison to classic mineralocorticoid target sites [13], suggesting that hypothalamic control of salt and water balance may be affected by both the glucocorticoid and mineralocorticoid effects of cortisol [14].

Glucocorticoids inhibit AVP gene transcription [4] and AVP secretion [15, 16] and may change the osmostat set point of the hypothalamus [17] (Fig. 4). The vasopressin gene promotor region contains regulatory elements which bind glucocorticoid-receptor complexes [18, 19] and appear to mediate tonic glucocorticoid suppression of AVP transcription [17]. As the AVP is the primary mediator of glucocorticoid release during acute stress, this tonic glucocorticoid suppression may be a protective mechanism preventing an inappropriately high burden of hypothalamic-pituitary-adrenal (HPA) axis activation accumulating from minor stimuli that occur frequently in physiological conditions [16]. Supraphysiological glucocorticoid levels lead to complete suppression of tonicity-dependent AVP secretion [20], while glucocorticoid deficiency leads to persistent AVP release independent of tonicity and volume [21, 24].

Glucocorticoids also contribute to maintenance of the osmostat set point. Supraphysiological glucocorticoid levels increase the tonicity threshold for AVP release [25], and reduce the gain of response, meaning less AVP is secreted for a given increase in tonicity (Fig. 2). Conversely, glucocorticoid deficiency decreases the tonicity threshold, meaning AVP is secreted when tonicity is “normal,” and increases the gain, meaning the AVP response to hypertonicity and hypovolemia is exaggerated. When glucocorticoid deficiency is severe, physiological suppression of AVP in hypotonic states is abolished [21].

Peripheral actions of glucocorticoids also affect water balance. In vitro studies of collecting duct cells suggest that glucocorticoids enhance AVP-dependent transcription of aquaporin-2 (AQP2) [26], likely due to a glucocorticoid-responsive element in the AQP2 gene promotor region [27]. In contrast, in vivo studies have demonstrated that AQP2 expression is elevated in states of glucocorticoid deficiency [28, 29]. This apparent contradiction may reflect differences between in vitro and in vivo effects of glucocorticoids on the collecting duct or the effects of dysregulated AVP secretion, overwhelming the suppressive effects of glucocorticoid deficiency on AQP2 gene transcription.

In the setting of central diabetes insipidus (DI), glucocorticoid deficiency reduces free water clearance, evidence for a direct effect on renal water handling independent of AVP [30, 33]. Impaired aquaresis in glucocorticoid deficiency is thought to relate to a reduction in renal blood flow, glomerular filtration, and direct effects on the water permeability of the distal tubules [30, 34].

Glucocorticoids augment the sympathetic nervous system’s vasoconstrictive effects on the peripheral vasculature [35] and contribute to the mineralocorticoid mediated defense of plasma volume [36]. In states of glucocorticoid deficiency, this leads to a reduction in effective intravascular volume and subsequent secretion of AVP; however, this is not as severe as in cases of primary adrenal failure, where there is concomitant mineralocorticoid deficiency.

Glucocorticoids modulate peripheral baroreceptor responses [37], increasing tonic signaling but decreasing responsiveness to changes in blood pressure (i.e., a rightward shift with decreased gain). These changes are also reflected in renal sympathetic outflow [38], which contributes to control of plasma volume. Glucocorticoid deficiency is thus expected to reduce baroreceptor and renal sympathetic nerve basal tone but increase their responsiveness (gain) to changes in blood pressure. Together, these effects may alter the secretion of AVP in response to changes in effective intravascular volume in states of glucocorticoid deficiency.

Key to the management of hypotonicity (hyponatremia) is discerning the “appropriateness” of the AVP response [39]. AVP is secreted in response to elevated tonicity or decreased effective intravascular volume. The physiological response to hypotonicity is complete suppression of AVP secretion and a corrective aquaresis. In states of intravascular hypovolemia, defense of tonicity is sacrificed for defense of volume, and free water is retained; this AVP secretion is physiological (“appropriate”). An “inappropriate” response is ongoing AVP secretion inappropriate to serum tonicity and the intravascular volume state [40], manifesting as persistently hypertonic urine and free water retention.

Hyponatremia can thus be classified based on the “appropriateness” of the AVP response to tonicity and volume (Fig. 5). The amount of AVP secretion is estimated by urinary tonicity (twice the sum of the urinary sodium and potassium concentrations), with markedly diluted urine (e.g., urinary Na <20 mmol/L) characteristic of suppressed AVP. The intravascular volume state is notoriously difficult to measure [41], and thus in the absence of overt hypovolemia or edematous states, diagnosis is primarily guided by the renal response.

The persistent, dysregulated AVP secretion seen in hypocortisolemia is physiologically inappropriate to both tonicity and volume, and thus it is indistinguishable from the syndrome of inappropriate antidiuresis (SIADH) [21, 29]. A diagnosis of SIADH therefore requires exclusion of glucocorticoid insufficiency and obvious hypervolemia or hypovolemia. In patients with euvolemic hyponatremia admitted to an endocrine unit, all of whom underwent appropriate workup for hypocortisolism, 20% had evidence for hypopituitarism as the cause of their hyponatremia [42]. Conversely, up to 10% of patients with hypopituitarism have concomitant hyponatremia [43].

For patients with hyponatremia due to SIADH, management consists of correcting free water balance through reduction in free water intake with or without interventions to increase free water excretion. Treatment should be tailored to the degree of renal water retention, quantified by urinary tonicity (sum of urinary sodium and potassium concentrations) or, equivalently, renal effective free water clearance. The treatment of glucocorticoid deficiency-related hyponatremia relies on high clinical suspicion, early diagnosis, and institution of physiological glucocorticoid replacement.

DI is a polyuria-polydipsia syndrome most often caused by inadequate secretion of AVP in response to an osmotic stimulus. This manifests as hypotonic polyuria with or without hypernatremia. Central DI is often caused by infundibular injury, which can also impair the anterior pituitary and the HPA axis. Central DI can be masked by concurrent glucocorticoid deficiency due to abovementioned effects on renal blood flow, glomerular filtration rate, and water permeability, which impair renal free water clearance. Thus, in patients at risk of DI, all assessments must be performed when glucocorticoid replete to increase the sensitivity for diagnosis of DI.

Pituitary surgery is the first-line therapy for a range of pituitary pathologies. One-third of patients with pituitary adenomas have preoperative central glucocorticoid deficiency, and a further one-third of patients may develop acute central glucocorticoid deficiency after surgery [44]. The absolute or relative central hypocortisolism that may occur after pituitary surgery is an important consideration for postoperative hyponatremia.

The use of perioperative glucocorticoid replacement after pituitary surgery is variable between institutions. Some centers continue supraphysiological “stress-dose” glucocorticoid replacement in the perioperative period, while others tailor glucocorticoid replacement based on early postoperative morning serum cortisol. The diagnosis of secondary (central) adrenal insufficiency after pituitary surgery can be difficult. A low morning serum cortisol (<100 nmol/L) is diagnostic, and glucocorticoid replacement should be initiated. Conversely, a morning cortisol >400 nmol/L suggests sufficient systemic glucocorticoids, but these patients may still have inadequate reserve and an impaired stress response. Intermediate results require stimulatory testing; however, the regular short ACTH (tetracosactide, Synacthen^®) stimulation test is not reliable in acute secondary adrenal insufficiency as the adrenals have not had time to atrophy [45]. In these cases, an insulin-induced hypoglycemia tolerance test may be necessary to establish the diagnosis; however, this is rarely performed. More commonly, those with intermediate results are prophylactically “covered” with physiological glucocorticoid replacement until an ACTH stimulation test can appropriately be performed.

Postoperative hyponatremia complicates 15–20% of pituitary surgery, although symptomatic hyponatremia represents only one-third of these cases [46]. When symptomatic, patients may present with headache, lethargy, altered consciousness, and seizures. In patients with acute hyponatremia after pituitary surgery, the most common cause is SIADH. It is thought to be caused by an isolated second phase of the triple phase response and local irritation by inflammatory cytokines in the operative bed. During surgery, there may be infundibular injury insufficient to establish DI (which requires >80% of the stalk to be damaged [47]) but sufficient such that the injured neurons undergo degeneration and release their stored AVP days after injury (Fig. 6). This coincides with a pro-inflammatory cytokine peak [48, 49] that also induces non-osmotic AVP secretion. Together, this produces a delayed, transient SIADH state [50]. Up to 10% of postoperative hyponatremia may be attributable to hypocortisolism [51]; however, the associated hyponatremia tends to be less severe and establish itself earlier than that associated with SIADH.

In patients with postoperative hyponatremia after pituitary surgery, the mainstay of treatment is fluid restriction (Fig. 7). In those with highly concentrated urine with a negative effective free water clearance [urinary tonicity exceeds plasma tonicity, (Na)_u+(K)_u > (Na)_p], or if free water restriction monotherapy fails or is not tolerated, supplemental solute may be given to induce an aquaresis. This may be administered as salt tablets or hypertonic saline. Concurrently, the adequacy of the glucocorticoid axis should be established. A morning serum cortisol <400 nmol/L is suggestive of potential hypocortisolism, especially in the early postoperative period, and glucocorticoid replacement should be administered. With adequate replacement, tonicity normalizes within 3–5 days [42].

In addition to volume control, the HPA axis has substantial influence on water balance. This relates to the effects of CRH and cortisol on AVP secretion and the peripheral effects of cortisol on hemodynamics and renal water handling. The glucocorticoid axis should be interrogated in disorders of water balance, particularly after pituitary surgery.

Not applicable.

There are no conflicts of interest to disclose.

No funds, grants, or other support was received for this project.

Dr. Mendel Castle-Kirszbaum, Prof. Tony Goldschlager, Dr. Margaret Dao Yuan Shi, and Prof. Peter J Fuller contributed to the research, drafting, and editing and approved the final submission of the manuscript.

Not applicable.