Distribution of different types of replacement fluids

Figure 13.2 shows the relative effects on the compartments of the addition of identical volumes of water, saline and colloid solutions. Thus, 1 L of water given intravenously as 5% glucose is distributed equally into all compartments, whereas the same amount of 0.9% saline remains in the extracellular compartment. The latter is thus the correct treatment for extracellular water depletion – sodium keeping the water in this compartment. The addition of 1 L of colloid with its high oncotic pressure stays in the vascular compartment and is a treatment for hypovolaemia.

Figure 13.2 Relative effects of the addition of 1 L of: (a) water, (b) saline 0.9% and (c) a colloid solution.

Regulation of extracellular volume (Fig. 13.3)

The extracellular volume is determined by the sodium concentration. The regulation of extracellular volume is dependent upon a tight control of sodium balance, which is exerted by normal kidneys. Renal Na⁺ excretion varies directly with the effective circulating volume. In a 70 kg man:

Figure 13.3 Regulation of extracellular volume. (a) Sequence of events in which a decrease in cardiac output or peripheral arterial dilatation initiates renal sodium and water retention. (b) Mechanism of impaired escape from the actions of aldosterone and resistance to atrial natriuretic peptides (ANP).

(Modified from Schrier RW. Renal and Electrolyte Disorders, 7th edn. Philadelphia: Lippincott Williams and Wilkins; 2010, with permission.)

The unifying hypothesis of extracellular volume regulation in health and disease proposed by Schrier states that the fullness of the arterial vascular compartment – or the so-called effective arterial blood volume (EABV) – is the primary determinant of renal sodium and water excretion. Thus effective arterial blood volume constitutes effective circulatory volume for the purposes of body fluid homeostasis.

The fullness of the arterial compartment depends upon a normal ratio between cardiac output and peripheral arterial resistance. Thus, diminished EABV is initiated by a fall in cardiac output or a fall in peripheral arterial resistance (an increase in the holding capacity of the arterial vascular tree). When the EABV is expanded, the urinary Na⁺ excretion is increased and can exceed 100 mmol/L. By contrast, the urine can be rendered virtually free of Na⁺ in the presence of EABV depletion and normal renal function.

These changes in Na⁺ excretion can result from alterations both in the filtered load, determined primarily by the glomerular filtration rate (GFR), and in tubular reabsorption, which is affected by multiple factors. In general, it is changes in tubular reabsorption that constitute the main adaptive response to fluctuations in the effective circulating volume. How this occurs can be appreciated from Table 13.2 and Figure 13.4 and Figure 12.2 (see p. 563), which depicts the sites and determinants of segmental Na⁺ reabsorption. Although the loop of Henle and distal tubules make a major overall contribution to net Na⁺ handling, transport in these segments primarily varies with the amount of Na⁺ delivered; that is, reabsorption is flow-dependent. In comparison, the neurohumoral regulation of Na⁺ reabsorption according to body needs occurs primarily in the proximal tubules and collecting ducts.

Table 13.2 Mechanisms of sodium transport in the various nephron segments

Figure 13.4 The nephron – electrolyte and water exchange: 180 L of water and 26 000 mmol of sodium/day enter the nephrons via the afferent arterioles to the kidneys. Removal or addition of electrolytes results in the excretion of approximately 1 L of water and 60–180 mmol of sodium/day.

Neurohumoral regulation of extracellular volume

This is mediated by volume receptors that sense changes in the EABV rather than alterations in the sodium concentration. These receptors are distributed in both the renal and cardiovascular tissues.

High-pressure arterial receptors (carotid, aortic arch, juxtaglomerular apparatus) predominate over low-pressure volume receptors in volume control in mammals. The low-pressure volume receptors are distributed in thoracic tissues (cardiac atria, right ventricle, thoracic veins, pulmonary vessels) and their role in the volume regulatory system is marginal.

Aldosterone and possibly ANP are responsible for day-to-day variations in Na⁺ excretion, by their respective ability to augment and diminish Na⁺ reabsorption in the collecting ducts.

A salt load, for example, leads to an increase in the effective circulatory and extracellular volume, raising both renal perfusion pressure, and atrial and arterial filling pressure. The increase in the renal perfusion pressure reduces the secretion of renin, and subsequently that of angiotensin II and aldosterone (see Fig. 12.5), whereas the rise in atrial and arterial filling pressure increases the release of ANP. These factors combine to reduce Na⁺ reabsorption in the collecting duct, thereby promoting excretion of excess Na⁺.

By contrast, in patients on a low Na⁺ intake or in those who become volume-depleted as a result of vomiting and diarrhoea, the ensuing decrease in effective volume enhances the activity of the renin-angiotensin-aldosterone system and reduces the secretion of ANP. The net effect is enhanced Na⁺ reabsorption in the collecting ducts, leading to a fall in Na⁺ excretion. This increases the extracellular volume towards normal.

With more marked hypovolaemia, a decrease in GFR leads to an increase in proximal and thin ascending limb Na⁺ reabsorption which contributes to Na⁺ retention. This is brought about by enhanced sympathetic activity acting directly on the kidneys and indirectly by stimulating the secretion of renin/angiotensin II (see Fig. 13.3b) and non-osmotic release of antidiuretic hormone (ADH), also called vasopressin. The pressure natriuresis phenomenon may be the final defence against changes in the effective circulating volume. Marked persistent hypovolaemia leads to systemic hypotension and increased salt and water absorption in the proximal tubules and ascending limb of Henle. This process is partly mediated by changes in renal interstitial hydrostatic pressure and local prostaglandin and nitric oxide production.

Volume regulation in oedematous conditions

Sodium and water are retained despite increased extracellular volume in oedematous conditions such as cardiac failure, hepatic cirrhosis and hypoalbuminaemia. Here the principal mediator of salt and water retention is the concept of arterial underfilling due either to reduced cardiac output or diminished peripheral arterial resistance. Arterial underfilling in these settings leads to reduction of pressure or stretch (i.e. ‘unloading’ of arterial volume receptors), which results in activation of the sympathetic nervous system, activation of the renin-angiotensin-aldosterone system and non-osmotic release of ADH. These neurohumoral mediators promote salt and water retention in the face of increased extracellular volume. The common nature of the degree of arterial fullness and neurohumoral pathway in the regulation of extracellular volume in health and disease states forms the basis of Schrier’s unifying hypothesis of volume homeostasis (Fig. 13.3a).

Mechanism of impaired escape from actions of aldosterone and resistance to ANP

Not only is the activity of the renin-angiotensin-aldosterone system increased in oedematous conditions such as cardiac failure, hepatic cirrhosis and hypoalbuminaemia, but also the action of aldosterone is more persistent than in normal subjects and patients with Conn’s syndrome, who have increased aldosterone secretion (see p. 989).

In normal subjects, high doses of mineralocorticoids initially increase renal sodium retention so that the extracellular volume is increased by 1.5–2 L. However, renal sodium retention then ceases, sodium balance is re-established, and there is no detectable oedema. This escape from mineralocorticoid-mediated sodium retention explains why oedema is not a characteristic feature of primary hyperaldosteronism (Conn’s syndrome). The escape is dependent on an increase in delivery of sodium to the site of action of aldosterone in the collecting ducts. The increased distal sodium delivery is achieved by high extracellular volume-mediated arterial overfilling. This suppresses sympathetic activity and angiotensin II generation, and increases cardiac release of ANP with resultant increase in renal perfusion pressure and GFR. The net result of these events is reduced sodium absorption in the proximal tubules and increased distal sodium delivery which overwhelms the sodium-retaining actions of aldosterone.

In patients with the above oedematous conditions, e.g. heart failure, escape from the sodium-retaining actions of aldosterone does not occur and therefore they continue to retain sodium in response to aldosterone. Accordingly they have substantial natriuresis when given spironolactone, which blocks mineralocorticoid receptors. Alpha-adrenergic stimulation and elevated angiotensin II increase sodium transport in the proximal tubule, and reduced renal perfusion and GFR further increases sodium absorption from the proximal tubules by presenting less sodium and water in the tubular fluid. Sodium delivery to the distal portion of the nephron, and thus the collecting duct, is reduced. Similarly, increased cardiac ANP release in these conditions requires optimum sodium concentration at the site of its action in the collecting duct for its desired natriuretic effects. Decreased sodium delivery to the collecting duct is therefore the most likely explanation for the persistent aldosterone-mediated sodium retention, absence of escape phenomenon and resistance to natriuretic peptides in these patients (Fig. 13.3b).

Regulation of water excretion

Body water homeostasis is affected by thirst and the urine concentrating and diluting functions of the kidney. These in turn are controlled by intracellular osmoreceptors, principally in the hypothalamus, to some extent by volume receptors in capacitance vessels close to the heart, and via the renin-angiotensin system. Of these, the major and best-understood control is via osmoreceptors. Changes in the plasma Na⁺ concentration and osmolality are sensed by osmoreceptors that influence both thirst and the release of ADH (vasopressin) from the supraoptic and paraventricular nuclei of the anterior hypothalamus.

ADH plays a central role in urinary concentration by increasing the water permeability of the normally impermeable cortical and medullary collecting ducts. There are three major G-protein coupled receptors for vasopressin (ADH):

Activation of the V_1A receptors induces vasoconstriction while V_1B receptors appear to mediate the effect of ADH on the pituitary, facilitating the release of ACTH. The V₂ receptors mediate the antidiuretic response as well as other functions.

The ability of ADH to increase the urine osmolality is related indirectly to transport in the ascending limb of the loop of Henle, which reabsorbs NaCl without water. This process, which is the primary step in the countercurrent mechanism, has two effects: it makes the tubular fluid dilute and the medullary interstitium concentrated. In the absence of ADH, little water is reabsorbed in the collecting ducts, and a dilute urine is excreted. By contrast, the presence of ADH promotes water reabsorption in the collecting ducts down the favourable osmotic gradient between the tubular fluid and the more concentrated interstitium. As a result, there is an increase in urine osmolality and a decrease in urine volume.

The cortical collecting duct has two cell types (see also p. 597) with very different functions:

The ADH-induced increase in collecting duct water permeability occurs primarily in the principal cells. ADH acts on V₂ (vasopressin) receptors located on the basolateral surface of principal cells, resulting in the activation of adenyl cyclase. This leads to protein kinase activation and to preformed cytoplasmic vesicles that contain unique water channels (called aquaporins) moving to and then being inserted into the luminal membrane. Four renal aquaporins have been well characterized and are localized in different areas of the cells of the collecting duct. The water channels span the luminal membrane and permit water movement into the cells down a favourable osmotic gradient (Fig. 13.5). This water is then rapidly returned to the systemic circulation across the basolateral membrane. When the ADH effect has worn off, the water channels aggregate within clathrin-coated pits, from which they are removed from the luminal membrane by endocytosis and returned to the cytoplasm. A defect in any step in this pathway, such as in attachment of ADH to its receptor or the function of the water channel, can cause resistance to the action of ADH and an increase in urine output. This disorder is called nephrogenic diabetes insipidus.

Figure 13.5 Aquaporin-mediated water transport in the renal collecting duct. Stimulation of the vasopressin 2 receptor causes cAMP-mediated insertion of the aquaporin into the apical membrane, allowing water transport down the osmotic gradient.

(Adapted from Connolly DL, Shanahan CM, Weissberg PL. Water channels in health and disease. Lancet 1996; 347:211, with permission from Elsevier.)

Plasma osmolality

In addition to influencing the rate of water excretion, ADH plays a central role in osmoregulation because its releaseis directly affected by the plasma osmolality. At a plasma osmolality of <275 mosmol/kg, which usually represents a plasma Na⁺ concentration of <135–137 mmol/L, there is essentially no circulating ADH. As the plasma osmolality rises above this threshold, however, the secretion of ADH increases progressively.

Two simple examples will illustrate the basic mechanisms of osmoregulation, which is so efficient that the plasma Na⁺ concentration is normally maintained within 1–2% of its baseline value.

Osmoregulation versus volume regulation

A common misconception is that regulation of the plasma Na⁺ concentration is closely correlated with the regulation of Na⁺ excretion. However, it is related to volume regulation, which has different sensors and effectors (volume receptors) from those involved in water balance and osmoregulation (osmoreceptors).

The roles of these two pathways should be considered separately when evaluating patients.

In some cases, both volume and osmolality are altered and both pathways are activated. For example, if a person with normal renal function eats salted potato chips and peanuts without drinking any water, the excess Na⁺ will increase the plasma osmolality, leading to osmotic water movement out of the cells and increased extracellular volume. The rise in osmolality will stimulate both ADH release and thirst (the main reason why many restaurants and bars supply free salted foods), whereas the hypervolaemia will enhance the secretion of ANP and suppress that of aldosterone. The net effect is increased excretion of Na⁺ without water.

This principle of separate volume and osmoregulatory pathways is also evident in the syndrome of inappropriate ADH secretion (SIADH). Patients with SIADH (see p. 993) have impaired water excretion and hyponatraemia (dilutional) caused by the persistent presence of ADH. However, the release of ANP and aldosterone is not impaired and, thus, Na⁺ handling remains intact. These findings have implications for the correction of the hyponatraemia in this setting which initially requires restriction of water intake.

ADH is also secreted by non-osmotic stimuli such as stress (e.g. surgery, trauma), markedly reduced effective circulatory volume (e.g. cardiac failure, hepatic cirrhosis), psychiatric disturbance and nausea, irrespective of plasma osmolality. This is mediated by the effects of sympathetic overactivity on supraoptic and paraventricular nuclei. In addition to water retention, ADH release in these conditions promotes vasoconstriction owing to the activation of V_1A (vasopressin) receptors distributed in the vascular smooth muscle cells.

Regulation of cell volume

Most cells respond to swelling or shrinkage by activating specific metabolic or membrane-transport processes that return cell volume to its normal resting state. Within minutes after exposure to hypotonic solutions and resulting cell swelling, a common feature of many cells is the increase in plasma membrane potassium and chloride conductance. Although extrusion of intracellular potassium certainly contributes to a regulatory volume decrease, the role of chloride efflux itself is modest, given the relatively low intracellular chloride concentration. Other intracellular osmolytes, such as taurine and other amino acids, are transported out of the cell to achieve a regulatory volume decrease. By contrast, these regulatory mechanisms are operative in reverse to protect cell volume under hypertonic conditions, as is the case in the renal medulla. The tubular cells at the tip of renal papillae, which are constantly exposed to a hypertonic extracellular milieu, maintain their cell volume on a long-term basis by actively taking up smaller molecules, such as betaine, taurine and myoinositol, and by synthesizing more sorbitol and glycerophosphocholine.

Clinical use of diuretics

Loop diuretics

These potent diuretics are useful in the treatment of any cause of systemic extracellular volume overload. They stimulate excretion of both sodium chloride and water by blocking the sodium-potassium-2-chloride (NKCC2) channel in the thick ascending limb of Henle (Fig. 13.6) and are useful in stimulating water excretion in states of relative water overload. They also act by causing increased venous capacitance, resulting in rapid clinical improvement in patients with left ventricular failure, preceding the diuresis. Unwanted effects include:

Urate retention causing gout

Hypokalaemia

Hypomagnesaemia

Decreased glucose tolerance

Allergic tubulointerstitial nephritis and other allergic reactions

Myalgia – especially with high-dose bumetanide

Ototoxicity (due to an action on sodium pump activity in the inner ear) – particularly with furosemide

Interference with excretion of lithium, resulting in toxicity.

Figure 13.6 Transport mechanisms in the thick ascending limb of the loop of Henle. Sodium chloride is reabsorbed in the thick ascending limb by the bumetanide-sensitive sodium-potassium-2-chloride cotransporter (NKCC2). The electroneutral transporter is driven by the low intracellular sodium and chloride concentrations generated by the Na⁺/K⁺-ATPase and the kidney-specific basolateral chloride channel (ClC-Kb). The availability of luminal potassium is rate-limiting for NKCC2, and recycling of potassium through the ATP-regulated potassium channel (ROMK – renal outer medulla K⁺ channel) ensures the efficient functioning of the NKCC2 and generates a lumen-positive transepithelial potential. Genetic studies have identified putative loss of function mutations in the genes encoding NKCC2 1, ROMK 2, ClC-Kb 3, and barttin 4 in subgroups of patients with Bartter’s syndrome. In contrast to the normal condition, loss of function of NKCC2 impairs reabsorption of sodium and potassium. Inactivation of the basolateral ClC-Kb and barttin reduces transcellular reabsorption of chloride. Loss of function of any of these will reduce the transepithelial potential and thus decrease the driving force for the paracellular reabsorption of cations (K⁺, Mg²⁺, Ca²⁺ and Na⁺). Paracellin-1 is necessary for the paracellular transport of Ca²⁺ and Mg²⁺. In most patients with Bartter’s syndrome, urinary calcium excretion is increased. Hypercalcaemia or increased activation of calcium-sensing receptor inactivates ROMK and causes Bartter’s syndrome. Ka and Kb, kidney-specific basolateral chloride channel. ROMK, renal outer medullary potassium channel.

There is little to choose between the drugs in this class. Bumetanide has a better oral bioavailability than furosemide, particularly in patients with severe peripheral oedema, and has more beneficial effects than furosemide on venous capacitance in left ventricular failure.

FURTHER READING

Bagshaw SM, Bellomo R, Kellum JA. Oliguria, volume overload, and loop diuretics. Crit Care Med 2008; 34(4 Suppl): S172–S178.

Thiazide diuretics (see p. 719)

These are less potent than loop diuretics. They act by blocking a sodium chloride channel in the distal convoluted tubule (Fig. 13.7). They cause relatively more urate retention, glucose intolerance and hypokalaemia than loop diuretics. They interfere with water excretion and may cause hyponatraemia, particularly if combined with amiloride or triamterene. This effect is clinically useful in diabetes insipidus. Thiazides reduce peripheral vascular resistance by mechanisms that are not completely understood but do not appear to depend on their diuretic action, and are widely used in the treatment of essential hypertension. They are also used extensively in mild to moderate cardiac failure. Thiazides reduce calcium excretion. This effect is useful in patients with idiopathic hypercalciuria, but may cause hypercalcaemia. Numerous agents are available, with varying half-lives but little else to choose between them. Metolazone is not dependent for its action on glomerular filtration, and therefore retains its potency in renal impairment.

Figure 13.7 Transport mechanisms in the distal convoluted tubule. Under normal conditions, sodium chloride is reabsorbed by the apical thiazide-sensitive sodium-chloride cotransporter (NCCT) in the distal convoluted tubule. The electroneutral transporter is driven by the low intracellular sodium and chloride concentrations generated by the Na⁺/K⁺-ATPase and an, as yet, undefined basolateral chloride channel. In this nephron segment, there is an apical calcium channel and a basolateral sodium-coupled exchanger. Physiological evidence indicates that the mechanisms for the transport of magnesium are similar to those for calcium. In Gitelman’s syndrome, putative loss of function mutations in the sodium-chloride cotransporter (NCCT (X)) lead to decreased reabsorption of sodium chloride and increased reabsorption of calcium. Functional overactivity of NCCT leads to Gordon’s syndrome by a new mechanism (see text). TRPM6, a member of the transient receptor potential family.

Resistance to diuretics

Resistance may occur as a result of:

Management. Intravenous administration of diuretics may establish a diuresis. High doses of loop diuretics are required to achieve adequate concentrations in the tubule if GFR is depressed. However, the daily dose of furosemide must be limited to a maximum of 2 g for an adult, because of ototoxicity. Intravenous albumin solutions restore plasma oncotic pressure temporarily in the nephrotic syndrome and allow mobilization of oedema but do not increase the natriuretic effect of loop diuretics.

Combinations of various classes of diuretics are extremely helpful in patients with resistant oedema. A loop diuretic plus a thiazide inhibit two major sites of sodium reabsorption; this effect may be further potentiated by addition of a potassium-sparing agent. Metolazone in combination with a loop diuretic is particularly useful in refractory congestive cardiac failure, because its action is less dependent on glomerular filtration. However, this potent combination can cause severe electrolyte imbalance. Both aminophylline and dopamine increase renal blood flow and may be useful in refractory cardiogenic sodium retention. In addition, theophyllines, by inhibiting phosphodiesterase activity in the inner medullary collecting duct, prolong the action of cyclic GMP (a second messenger of ANP).

Effects on renal function

All diuretics may increase plasma urea concentrations by increasing urea reabsorption in the medulla. Thiazides may also promote protein breakdown. In certain situations diuretics also decrease GFR:

Avoiding ODS

A rapid rise in extracellular osmolality, particularly if there is an ‘overshoot’ to high serum sodium and osmolality, will result in the osmotic demyelination, syndrome (ODS), formally known as central pontine demyelination, which is a devastating neurologic complication. Plasma sodium concentration in patients with hyponatraemia should not rise by more than 8 mmol/L per day. The rate of rise of plasma sodium should be even lower in patients at higher risk for ODS, e.g. patients with alcohol excess, cirrhosis, malnutrition, or hypokalaemia. Other factors predisposing to demyelination are pre-existing hypoxaemia and CNS radiation (see above). ODS is diagnosed by the appearance of characteristic hypointense lesions on T₁-weighted images and hyperintense on T₂-weighted images on MRI; these take up to 2 weeks or longer to appear.

The pathophysiology of ODS is not fully understood. The most plausible explanation is that the brain loses organic osmolytes very quickly in order to adapt to hyponatraemia so that osmolarity is similar between the intracellular and extracellular compartments. However, neurones reclaim organic osmolytes slowly in the phase of rapid correction of hyponatraemia, resulting in an hypo-osmolar intracellular compartment and lead to shrinkage of cerebral vascular endothelial cells. Consequently the blood–brain barrier is functionally impaired, allowing lymphocytes, complement, and cytokines to enter the brain, damage oligodendrocytes, activate microglial cells and cause demyelination.

The most crucial issue in the treatment of hyponatraemia is to prevent rapid correction. A rapid rise in plasma sodium is almost always due to a water diuresis, which happens when vasopressin (ADH) action stops suddenly, for example with volume repletion in patients with intravascular volume depletion, cortisol replacement in patients with Addison disease, resolution of non-osmotic stimuli for vasopressin release such as nausea or pain. However, sometimes chronic hyponatraemia can develop in the absence of vasopressin excess. Even in these cases, water diuresis due to increased distal delivery of filtrate is the main cause of rapid rise in plasma sodium.

In the absence of vasopressin, it is generally assumed that the total urine volume is equal to the volume of filtrate delivered to the distal nephron, which is the GFR minus the volume reabsorbed in the proximal convoluted tubule (PCT). Approximately 80% of the GFR is reabsorbed in PCT under normal circumstance (increases even more in the presence of intravascular volume depletion). However, in real life water excretion will be less than the volume of distal delivery of filtrate, even in the absence of vasopressin, because a significant degree of water is reabsorbed in the inner medullary collecting duct through its residual water permeability, prompted by a very high osmotic force in the interstitium (see Fig. 12.2).

Even a modest water diuresis in the elderly with reduced muscle mass is large enough to cause a rapid rise in plasma sodium. Moreover, there is a higher risk for ODS if hypokalaemia is present. In such cases if plasma sodium rises too quickly due to anticipated water diuresis, administration of desmopressin to stop the water diuresis is beneficial. If plasma sodium rises regardless then lowering plasma sodium to the maximum limit of correction (<8 mmol/L per day) with the administration of 5% glucose solution is the best strategy.

Reversible hyponatraemia culminating in hypernatraemia

In many patients, the cause of water retention is reversible (e.g. hypovolaemia, thiazide diuretics). On correction of the cause, vasopressin levels fall and plasma sodium rises by up to 2 mmol/L per hour as a result of excretion of dilute urine. This excessive water diuresis should be anticipated and prevented by use of desmopressin.

Patients who are chronically hyponatraemic with concomitant hypokalaemia are especially susceptible to overcorrection. Plasma sodium is a function of the ratio of exchangeable body sodium plus potassium to total body water, so potassium administration increases sodium concentration. For example, a mildly symptomatic hyponatraemic patient with a plasma sodium of <120 mmol/L and potassium of <2 mmol/L can potentially develop ODS as a result of overcorrection of hyponatraemia simply as a direct result of replacing the large potassium deficit.

Antidiuretic hormone antagonists (vasopressin antagonists)

Vasopressin V₂ receptor antagonists (see p. 645), which produce a free water diuresis, are being used in clinical trials for the treatment of hyponatraemic encephalopathy. Three oral agents, lixivaptan, tolvaptan and satavaptan, are selective for the V₂ (antidiuretic) receptor, while conivaptan blocks both the V_1A and V₂ receptors.

These agents produce a selective water diuresis without affecting sodium and potassium excretion; they raise the plasma sodium concentration in patients with hyponatraemia caused by the SIADH, heart failure and cirrhosis.

The efficacy of oral tolvaptan in ambulatory patients has been demonstrated in patients with hyponatraemia (mean plasma sodium 129 mmol/L) caused by the SIADH, heart failure, or cirrhosis who had a sustained rise in plasma sodium to 136 mmol/L for 4 weeks. Tolvaptan is now approved for use in patients with euvolaemic hyponatraemia and those with SIADH. In addition, intravenous conivaptan is available and is also approved for the treatment of euvolaemic hyponatraemia (i.e. SIADH) in some countries. The approved dosing for conivaptan is a 20 mg bolus followed by continuous infusion of 20 mg over 1–4 days. The continuous infusion increases the risk of phlebitis, which requires the use of large veins and changing the infusion site every 24 hours.

Causes

The most common causes of chronic hypokalaemia are diuretic treatment (particularly thiazides) and hyperaldosteronism. Acute hypokalaemia is often caused by intravenous fluids without potassium and redistribution into cells. The common causes are shown in Table 13.12.

Table 13.12 Causes of hypokalaemia

Rare causes

These rare causes are discussed in detail because they show the mechanisms of how diuretics can affect the kidney.

Vitamin D-dependent rickets type I

Also known as pseudovitamin D-deficient rickets, this is caused by 1α-hydroxylase deficiency due to inactivating mutations in its gene. It can be corrected with high daily doses of vitamin D. This condition manifests clinically in the first year of life with severe hypocalcaemia often complicated by tetany, moderate hypophosphataemia and enamel hypoplasia. The characteristic biochemical findings are normal serum levels of 25-hydroxyvitamin D and low values of 1,25–dihydroxyvitamin D and usually relatively high PTH levels. The treatment of choice is replacement therapy with calcitriol.

Vitamin D-dependent rickets type II

This is a form of vitamin D resistance and is now known as hereditary vitamin D-resistant rickets. It is an autosomal recessive disorder and is usually caused by loss of function mutations in the gene encoding the vitamin D receptor. The clinical manifestations vary widely, depending upon the type of mutation within the vitamin D receptor and the amount of residual vitamin D receptor activity. Affected children usually develop rickets within the first 2 years of life with alopecia in two-thirds of cases which is due to lack of vitamin D receptor action within keratinocytes. The treatment involves a therapeutic trial of calcitriol and calcium supplementation. Long-term infusion of calcium into a central vein is a possible alternative for severely resistant patients. Oral calcium therapy may be sufficient once radiographic healing has been observed.

Decreased renal reabsorption of phosphate

Dent’s disease

Dent’s disease is the generally accepted name for a group of hereditary tubular disorders including X-linked recessive nephrolithiasis with renal failure, X-linked recessive hypophosphataemic rickets, and idiopathic low-molecular-weight proteinuria. It is characterized by low-molecular-weight proteinuria, hypercalciuria, hyperphosphaturia, nephrocalcinosis, kidney stones and eventual renal failure, with some patients developing rickets or osteomalacia.

Dent’s disease is caused by loss of function mutation of a proximal tubular endosomal chloride channel, ClC5. This chloride channel, along with the proton pump, is essential for acidification of proximal tubular endosomes. The process is linked with normal endocytosis, degradation and recycling of absorbed proteins, vitamins and hormones. Defective endosomal acidification (owing to the mutated ClC5 gene) results in impaired endosomal degradation and recycling of endocytosed hormones such as PTH with clinical phenotype of hyperparathyroidism. Moreover, the receptors (megalin and cubilin) for reabsorption of low-molecular-weight proteins and albumin in the proximal tubules are decreased in Dent’s disease. This explains low-molecular-weight proteinuria and excessive urinary leaks of cytokines, hormones and chemokines. This urinary profile is associated with progressive renal fibrosis and more rapid decline in renal function.

Renal reabsorption of bicarbonate

The plasma [HCO₃⁻] is normally maintained at approximately 25 mmol/L. In individuals with a normal glomerular filtration rate (120 mL/min), about 4500 mmol of bicarbonate is filtered each day. If this filtered bicarbonate were not reabsorbed, the plasma [HCO₃⁻] would fall, along with blood pH. Thus, maintenance of normal plasma [HCO₃⁻] requires that essentially all of the bicarbonate in the glomerular filtrate be reabsorbed (Fig. 13.10).

Figure 13.10 Resorption of sodium bicarbonate in the renal (mainly proximal) tubule. Bicarbonate is reclaimed by the secretion of H⁺ in exchange for Na⁺ into the tubule. This results in the formation of H₂CO₃, which is then broken down to CO₂. This is reabsorbed and converted back to H₂CO₃, which now dissociates into H⁺ and HCO₃⁻. The net result is reabsorption of Na⁺ and HCO₃⁻. This process is dependent on carbonic anhydrase within the cells and on the luminal surface of the tubular cell.

The proximal convoluted tubule reclaims 85–90% of filtered bicarbonate; by contrast, the distal nephron reclaims very little. This difference is caused by the greater quantity of luminal (brush border) carbonic anhydrase in the proximal tubule than in the distal nephron. As a result of these quantitative differences, bicarbonate that escapes reabsorption in the proximal tubule is excreted in the urine.

Proximal tubular bicarbonate reabsorption is catalysed by the Na⁺/K⁺-ATPase pump located in the basolateral cell membrane. By exchanging peritubular potassium ions for intracellular sodium ions, the pump keeps the intracellular sodium concentration low, allowing sodium ions to enter the cell by moving down the sodium concentration gradient from the tubule lumen to the cell interior. Hydrogen ions are transported in the opposite direction (at the Na⁺-H⁺ antiporter), thereby maintaining electroneutrality. Before bicarbonate enters the proximal tubule, it combines with secreted hydrogen ions, forming carbonic acid. In the presence of luminal carbonic anhydrase (CA-IV) carbonic acid rapidly dissociates into carbon dioxide and water, which can then rapidly enter the proximal tubular cell. In the cell, carbon dioxide is hydrated by cytosolic carbonic anhydrase (CA-II), ultimately forming bicarbonate, which is then transported down an electrical gradient from the cell interior, across the membrane into the peritubular fluid, and into the blood. In this process, each hydrogen ion secreted into the proximal tubule lumen is reabsorbed and can be resecreted; there is no net loss of hydrogen ions or net gain of bicarbonate ions.

Renal excretion of [H⁺] (Fig. 13.11)

More acid is secreted into the proximal tubule (up to 4500 nmol of hydrogen ions each day) than into any other nephron segment. However, the hydrogen ions secreted into the proximal tubule are almost completely reabsorbed with bicarbonate; consequently, proximal tubular hydrogen ion secretion does not contribute significantly to hydrogen ion elimination from the body. The excretion of the daily acid load requires hydrogen ion secretion in more distal nephron segments.

Figure 13.11 Renal excretion of H⁺. Secretion of H⁺ from the cortical collecting ducts is indirectly linked to Na⁺ reabsorption. Intracellular potassium is exchanged for sodium in the principal cell. Aldosterone stimulates H⁺ secretion by entering the principal cell, where it opens Na⁺ channels in the luminal membrane and increases Na⁺/K⁺-ATPase activity. The movement of cationic Na⁺ into the principal cells then creates a negative charge within the tubule lumen. K⁺ moves from the electrochemical gradient and into the lumen. Aldosterone apparently also stimulates the H⁺-ATPase directly in the intercalated cell, further enhancing H⁺ secretion. When the urinary pH falls to 4.0–4.5, further H⁺ secretion by the α-intercalated cells ceases. The filtration of titratable acids (e.g. phosphoric acid, H₂PO₄) raises the intraluminal pH and permits this process to continue. Secreted H⁺ binds to the conjugate anion of a titratable acid (HPO₄²⁻ in this case) and is excreted in the urine. The H⁺ to be secreted arises from the reassociation of H₂O and CO₂ in the presence of carbonic anhydrase; thus, a bicarbonate molecule is regenerated each time an H⁺ is eliminated in the urine.

Most dietary hydrogen ions come from sulphur-containing amino acids that are metabolized to sulphuric acid (H₂SO₄), which then reacts with sodium bicarbonate as follows:

Excess sulphate is excreted in the urine, whereas excess hydrogen ions are buffered by bicarbonate and lower the plasma [HCO₃⁻]. This fall in plasma [HCO₃⁻] leads to a slight decrease in the blood pH, although a smaller decrease in the blood pH than would have occurred if buffer were unavailable. The subsequent excretion of hydrogen ions takes place primarily in the collecting duct and results in the regeneration of 1 mmol of bicarbonate for every mmol of hydrogen ions excreted in the urine.

The collecting duct has three types of cells:

Secretion of hydrogen ions from the cortical collecting duct is indirectly linked to sodium reabsorption. Aldosterone has several facilitating effects on hydrogen ion secretion. Aldosterone opens sodium channels in the luminal membrane of the principal cell and increases Na⁺/K⁺-ATPase activity. The subsequent movement of cationic sodium into the principal cell creates a negative charge within the tubule lumen. Potassium ions from the principal cells and hydrogen ions from the α-intercalated cells move out from the cells down the electrochemical gradient and into the lumen. Aldosterone also stimulates directly the H⁺-ATPase in the α-intercalated cell, further enhancing hydrogen ion secretion.

When hydrogen ions are secreted into the lumen of the collecting tubule, a tiny, but physiologically critical, fraction of these excess hydrogen ions remains in solution. Here, they increase the urinary [H⁺] and lower urinary pH below 4.0. Nevertheless, below this urine pH, inhibition of proton-secreting pumps such as H⁺-ATPase severely restricts kidney secretion of more hydrogen ions. Consequently, secretion of hydrogen ions depends on the presence of buffers in the urine that maintain the urine pH at a level higher than 4.0.

In the presence of alkali excess, the homeostatic needs are reversed. Although the kidney can excrete excess alkaline load by reducing reabsorption of filtered bicarbonate in the proximal and distal tubule, the collecting ducts also contribute by secreting bicarbonate brought about by switching to β-intercalated cells. This switch enables kidneys to secrete bicarbonate and conserve H⁺ ions.

Buffer systems in acid excretion

Two buffer systems are involved in acid excretion: the titratable acids such as phosphate, and the ammonia system. Each system is responsible for excreting about half of the daily acid load of 50–100 mmol under physiological conditions (Fig. 13.11).

Ammonium (NH₄⁺)

In the setting of metabolic acidosis, titratable acids cannot increase significantly because the availability of titratable acid is fixed by the plasma concentration of the buffer and by the GFR. The ammonia buffer system, by contrast, can increase several hundred-fold when necessary. Consequently, impaired renal excretion of hydrogen ions is always associated with a defect in ammonium excretion (Fig. 13.12).

Figure 13.12 The ammonia buffering system in the kidney. All ammonia used to buffer H⁺ in the collecting duct is synthesized in the proximal convoluted tubule, and glutamine is the main source of this ammonia. As glutamine is metabolized, α-ketoglutarate (α-KG) is formed, which ultimately breaks down to bicarbonate that is then secreted into the peritubular fluid at an Na⁺-HCO₃⁻ cotransporter.

All ammonia used to buffer urinary hydrogen ions in the collecting tubule is synthesized in the proximal convoluted tubule. Glutamine is the primary source of ammonia. It undergoes deamination catalysed by glutaminase, resulting in α-ketoglutaric acid (Fig. 13.12) and ammonia. Once formed, ammonia can diffuse into the proximal tubule lumen and become acidified, forming ammonium. Once in the proximal tubule lumen, ammonium flows along the tubule to the thick ascending limb of Henle’s loop. Here, it is transported out of the tubule into the medullary interstitium. Ammonium then dissociates to ammonia, leading to a high interstitial ammonia concentration. The notion that ammonia diffuses down its concentration gradient into the lumen of the collecting tubule has recently been challenged by the discovery of rhesus (Rh) associated glycoproteins acting as ammonia transport proteins also called RhCG/Rhcg which are expressed in the basolateral and apical surfaces of DCT, inner medullary collecting duct and type A intercalated cells. These proteins play a fundamental role in renal ammonia excretion under both basal and acidosis states. Once secreted, NH₃ reacts with the hydrogen ions secreted by the collecting tubular cells to form ammonium. Because ammonium (NH₄) is not lipid-soluble, it is trapped in the lumen and excreted in the urine as ammonium chloride. Two conditions predominantly promote ammonia synthesis by the proximal tubular cell: systemic acidosis and hypokalaemia.

Glutamine metabolism and ureagenesis in the liver were thought to play a role in acid–base homeostasis. The liver was believed to contribute to regulation of acid–base balance by controlling the rate of ureagenesis and therefore bicarbonate consumption in response to changes in plasma acidity. Studies in human volunteers have concluded that ureagenesis is a maladaptive process for acid–base regulation and that ureagenesis has no discernible homeostatic effect on acid–base equilibrium in humans.

FURTHER READING

Weiner ID, Verlander JW. Molecular physiology of the Rh ammonia transport proteins. Curr Opin Nephrol Hypertens 2010; 19(5):471–477.

Respiratory acidosis

This is caused by retention of CO₂. The P_aCO₂ and [H⁺] rise. Renal retention of bicarbonate may partly compensate, returning the [H⁺] towards normal (see p. 876).

Respiratory alkalosis

Increased removal of CO₂ is caused by hyperventilation, so there is a fall in P_aCO₂ and [H⁺] (see p. 876).

The anion gap

The first step is to identify whether the acidosis is due to retention of H⁺Cl⁻ or to another acid. This is achieved by calculation of the anion gap.

Because there are more unmeasured anions than cations, the normal anion gap is 10–18 mmol/L, although calculations with more sensitive methods place this at 6–12 mmol/L. Albumin normally makes up the largest portion of these unmeasured anions. As a result, a fall in the plasma albumin concentration from the normal value of about 40 g/L to 20 g/L may reduce the anion gap by as much as 6 mmol/L, because each 1 g/L of albumin has a negative charge of 0.2–0.28 mmol/L.

Renal tubular acidosis (RTA)

This term refers to systemic acidosis caused by impairment of the ability of the renal tubules to maintain acid–base balance. This group of disorders is uncommon and only rarely a cause of significant clinical disease.

Urinary anion gap

Another useful tool in the evaluation of metabolic acidosis with a normal anion gap is the urinary anion gap:

This calculation can be used to distinguish the normal anion-gap acidosis caused by diarrhoea (or other gastrointestinal alkali loss) from that caused by distal renal tubular acidosis. In both disorders, the plasma [K⁺] is characteristically low. In patients with renal tubular acidosis, urinary pH is always greater than 5.3.

Although excretion of urinary hydrogen ions in the patient with diarrhoea should acidify the urine, hypokalaemia leads to enhanced ammonia synthesis by the proximal tubular cells. Despite acidaemia, the excess urinary buffer increases the urine pH to a value above 5.3 in some patients with diarrhoea.

Whenever urinary acid is excreted as ammonium chloride, the increase in urinary chloride excretion decreases the urinary anion gap. Thus, the urinary anion gap should be negative in the patient with diarrhoea, regardless of the urine pH. On the other hand, although hypokalaemia may result in enhanced proximal tubular ammonia synthesis in distal renal tubular acidosis, the inability to secrete hydrogen ions into the collecting duct in this condition limits ammonium chloride formation and excretion; thus, the urinary anion gap is positive in distal renal tubular acidosis.

Uraemic acidosis

Kidney disease causes acidosis in several ways. Reduction in the number of functioning nephrons decreases the capacity to excrete ammonia and H⁺ in the urine. In addition, tubular disease may cause bicarbonate wasting. Acidosis is a particular feature of those types of CKD in which the tubules are particularly affected, such as reflux nephropathy and chronic obstructive uropathy.

Chronic acidosis is most often caused by chronic kidney disease, where there is a failure to excrete fixed acid. Up to 40 mmol of hydrogen ions may accumulate daily. These are buffered by bone, in exchange for calcium. Chronic acidosis is therefore a major risk factor for renal osteodystrophy and hypercalciuria.

Chronic acidosis has also been shown to be a risk factor for muscle wasting in renal failure, and may also contribute to the inexorable progression of some types of renal disease.

Uraemic acidosis should be corrected because of these effects on growth, muscle turnover and bones. Oral sodium bicarbonate 2–3 mmol/kg daily is usually enough to maintain serum bicarbonate above 20 mmol/L, but may contribute to sodium overload. Calcium carbonate improves acidosis and also acts as a phosphate binder and calcium supplement, and is commonly used. Acidosis in end-stage kidney failure is usually fully corrected by adequate dialysis.

Lactic acidosis

Increased lactic acid production occurs when cellular respiration is abnormal, because of either a lack of oxygen in the tissues (‘type A’) or a metabolic abnormality, such as drug-induced (‘type B’) (Table 13.24). The most common cause in clinical practice is type A lactic acidosis, occurring in septic or cardiogenic shock. Significant acidosis can occur despite a normal blood pressure and P_aCO₂, owing to splanchnic and peripheral vasoconstriction. Acidosis worsens cardiac function and vasoconstriction further, contributing to a downward spiral and fulminant production of lactic acid.

Diabetic ketoacidosis (see p. 1017)

There is a high-anion-gap acidosis due to the accumulation of acetoacetic and hydroxybutyric acids, owing to increased production and some reduced peripheral utilization.

Classification and definitions

Metabolic alkalosis has been classified on the basis of underlying pathophysiology (Table 13.25).

Table 13.25 Causes of metabolic alkalosis

The most common group is due to chloride depletion which can be corrected without potassium repletion. The other major grouping is that due to potassium depletion, usually with mineralocorticoid excess. Metabolic alkalosis due to both potassium and chloride depletion also occurs.

Chloride may be lost from the gut, kidney or skin. The loss of gastric fluid rich in acid results in alkalosis because bicarbonate generated during the production of gastric acid returns to the circulation. In Zollinger–Ellison syndrome (see p. 370) or gastric outflow obstruction these losses can be massive. Although sodium and potassium loss in the gastric juice is variable, the obligate urinary loss of these cations is intensified by bicarbonaturia, which occurs during disequilibrium.

Chloruretic agents all directly produce loss of chloride, sodium and fluid in the urine. These losses in turn promote metabolic alkalosis by several mechanisms:

Urinary losses of chloride exceed those for sodium and are associated with alkalosis even when potassium depletion is prevented. The cessation of events that generate alkalosis is not necessarily accompanied by resolution of the alkalosis. A widely accepted hypothesis for the maintenance of alkalosis is chloride depletion rather than volume depletion. Although normal functioning of the proximal tubule is essential for bicarbonate absorption, the collecting duct appears to be the major nephron site for altered electrolyte and proton transport in both maintenance and recovery from metabolic alkalosis. During maintenance, the α-intercalated cells in the cortical collecting duct do not secrete bicarbonate because insufficient chloride is available for bicarbonate exchange. When chloride is administered and luminal or cellular chloride concentration increases, bicarbonate is promptly excreted and alkalosis is corrected.

Metabolic alkalosis in hypokalaemia is generated primarily by an increased intracellular shift of hydrogen ion causing intracellular acidosis. Potassium depletion is also associated with enhanced ammonia production with increased obligate net acid excretion. However, the role of intracellular acidosis is supported by the correction of the alkalosis by infusion of potassium without any suppression of renal net excretion. The correction is assumed to occur by the movement of potassium into and hydrogen ion out of the cell, which titrates extracellular fluid bicarbonate.

Milk–alkali syndrome in which both bicarbonate and calcium are ingested produces alkalosis by vomiting, calcium-induced bicarbonate absorption and reduced GFR. Cationic antibiotics in high doses can cause alkalosis by obligatory bicarbonate loss in the urine.