Section 15: Renal

INTRODUCTION

Acid-base disturbances occur frequently in the acutely ill hospitalized patient. Many physicians are intimidated by the complexity of acid-base analysis, as multiple, partially offsetting disorders can be challenging to diagnose. However, in the hospital setting, especially the critically ill, it is crucial that patients with acid-base disturbances be quickly identified and the abnormality be accurately interpreted. Swift intervention to treat the underlying causes is often necessary to avoid the often lethal consequences of severe acid-base disturbances.

NORMAL ACID-BASE HOMEOSTASIS

A typical Western diet generates about 15,000 mmol of volatile acids, in the form of CO₂, and ~70 mmol (1 mmol/kg) of fixed acid each day. CO₂ is excreted by normal respiration. Fixed acids are buffered by intra- and extracellular buffers. New buffers, predominantly HCO₃^–, must be continuously produced to replace buffers consumed in titrating fixed acids. Complex acid-base homeostasis mechanisms, which include chemical buffering in conjunction with the excretion of CO₂ by the respiratory system and new HCO₃^– production by the kidneys, normally maintain the blood pH between 7.35 and 7.45.

Acid-base homeostasis requires complementary functions of the respiratory and the renal systems. The kidneys play a central role by reabsorbing filtered bicarbonate, approximately 4000 mmol/d, and generating new bicarbonate. The central nervous system and the respiratory systems control the arterial CO₂ tension (PaCO₂). Respiratory CO₂ elimination is determined by alveolar ventilation. Kidneys generate new bicarbonate through the process of net acid excretion. Under normal conditions, net acid excretion balances fixed acid production. A disturbance of either respiratory CO₂ elimination or the balance between fixed acid generation and renal net acid excretion results in an acid-base disorder.

SIMPLE ACID-BASE DISORDERS

These disorders involve only a single acid-base disorder, and include metabolic acidosis, metabolic alkalosis, respiratory acidosis, and respiratory alkalosis. A key distinguishing feature is that the pH is always abnormal, as the compensation is never complete. Simple acid-base disorders can be classified as acute or chronic based on the degree of metabolic (renal) compensation (Table 238-1).

MIXED ACID-BASE DISORDERS

When multiple acid-base disorders exist simultaneously and are not compensatory responses, they are classified as mixed acid-base disorders. When these occur in the critically ill patient, they can lead to dangerous extremes of pH. However, the pH may also be normal or near normal when acidosis and alkalosis coexist in the same patient. A patient with diabetic ketoacidosis who also has metabolic acidosis resulting from chronic kidney disease is an example of a mixed acid-base disorder. Triple acid-base disorders are not uncommon, and are often not recognized unless the clinician specifically considers this possibility. For example, a patient with metabolic acidosis due to lactic acidosis complicating severe sepsis may develop metabolic alkalosis from intravenous bicarbonate administration, with superimposed respiratory alkalosis due to hyperventilation from mechanical ventilation. Another example is the patient with respiratory acidosis due to altered consciousness, metabolic acidosis due to diabetic ketoacidosis, and metabolic alkalosis from vomiting may have an identical arterial blood gas (ABG). It is critical to always interpret acid-base abnormalities in the clinical context of the patient.

APPROACH TO ACID-BASE DISORDERS

The first step in interpreting acid-base disorders is to perform a detailed history and physical examination. Next, one should simultaneously measure arterial blood gas (ABG) and plasma chemistries. A venous blood gas can be utilized under some circumstances, but is less accurate than an arterial blood gas. Blood gas analyzers directly measure the pH and PaCO₂; the bicarbonate value reported in an ABG analysis is calculated based from the pH and PaCO₂. Blood bicarbonate concentration is measured in the metabolic panel as total dissolved CO₂, which is ~95% bicarbonate. The samples can be validated by comparing the calculated HCO₃^– value reported on the arterial blood gas measurement with the measured HCO₃^– value on the chemistry panel. If the difference between the two values is greater than 2 mmol/L, the samples may not have been drawn simultaneously, or a laboratory error may be present. Excessive heparin in the syringe used to obtain the arterial blood sample can also cause confounding results. Repeating the laboratory studies may be helpful in determining the cause of this discrepancy.

Next, identify the primary acid-base disorder by looking at the arterial pH, HCO₃^–, and PaCO₂ (Table 238-2). If either respiratory acidosis or alkalosis, then determine whether the condition is acute or chronic from the change in the serum bicarbonate (Table 238-1). Finally, calculate the anion gap (AG) and ΔAG (see below). Further evaluation will be guided by the type of acid-base disorder.

THE ANION GAP

The anion gap is the difference between the amount of unmeasured cations and unmeasured anions in plasma. It normally represents unmeasured anions in plasma, such as anionic proteins, phosphate, sulfate, and organic anions. In modern laboratories, using ­ion-sensitive electrodes to measure electrolyte concentrations, the mean anion gap is 7 mmol/L and the normal range is 3 to 11 mmol/L. A metabolic acidosis associated with an elevated anion gap is termed an anion gap metabolic acidosis, and metabolic acidosis with a normal anion gap is normal is termed a nonanion gap metabolic acidosis. High AG acidosis occurs when excess acid anions, such as acetoacetate and lactate, accumulate in extracellular fluid.

The anion gap may decrease either due to a rise in unmeasured cations (such as paraproteins) or a fall in anionic albumin (Table 238-3). The most common cause of an unexpectedly low anion gap is hypoalbuminemia. The expected anion gap should be corrected for albumin levels. A fall in serum albumin by 1 g/dL from the normal value (4.0 g/dL) decreases the anion gap by 2.5 mmol/L. For example, a measured anion gap of 12 mmol/L is usually normal. However, in a patient with a serum albumin of 1 g/dL, the expected AG is 12 to 2.5 (4.0 – 1) = 4.5 mmol/L.

A normal HCO₃^– does not preclude an increased anion gap. For example, a patient who simultaneously has separate conditions causing an anion gap metabolic acidosis and metabolic alkalosis may have a normal, high or low, measured HCO₃^–. In this setting, recognizing that an elevated AG indicates the presence of AG metabolic acidosis may alert the clinician to the possibility of a mixed acid-base disorder.

ΔAG AND ΔHCO₃^– (ASSESSING THE DELTA GAP)

Interpreting mixed acid-base disorders can be tricky, especially when the combination of disturbances result in a normal pH, HCO₃^–, and PaCO₂. In such situations, comparing the change in HCO₃^– (ΔHCO₃^–) and the change in the AG (ΔAG) can be useful. The ΔHCO₃^– is calculated as ΔHCO₃^– = [HCO₃^–_actual] – 24, where is 24 is the normal serum bicarbonate concentration, and ΔAG is calculated as ΔAG = AG_actual – 7, where is 7 is the normal AG. The normal AG should be adjusted based on the serum albumin, as detailed above. If the ΔHCO₃^– and ΔAG differ from each other by more than 3, then multiple metabolic acid-base disorders may be present. If ΔAG > ΔHCO₃^-, then a simultaneous anion gap metabolic acidosis and metabolic alkalosis should be considered. If ΔHCO₃^– > ΔAG, then a combined anion gap metabolic acidosis and nonanion gap metabolic acidosis may be present.

PLASMA OSMOLAR GAP

The normal range for serum osmolality is 285 to 290 mOsm/L. The major molecules contributing to the serum osmolality include sodium, urea nitrogen, and glucose. Alcohol (ethanol) will contribute to osmolality if present at levels consistent with intoxication. The osmolar gap is obtained by subtracting the calculated osmolarity (see formulae below) from the measured osmolarity. A difference of more than 10 mOsm/L is abnormal (osmolar gap) suggests another solute, such as lactate, ethanol, ethylene glycol, methanol, contrast dye, and mannitol.

METABOLIC ACIDOSIS

The two major categories of clinical metabolic acidosis are anion gap metabolic acidosis and nonanion gap metabolic acidosis (Table 238-4). While the latter is sometimes termed hyperchloremic acidosis, we discourage use of this term. Abnormalities in the serum sodium concentration, either hyponatremia or hypernatremia, necessarily alter the serum chloride, and if this is not recognized can lead to incorrect categorization of the metabolic acidosis using only the serum chloride.

Metabolic acidosis may occur because of an increased production of endogenous acids, such as lactate and ketoacids, accumulation of endogenous acids, as in renal failure, or loss of bicarbonate, as in diarrhea. Metabolic acidosis can affect the respiratory, cardiac, and nervous systems, and in severe cases may lead to death. The normal respiratory response to metabolic acidosis is to increase alveolar ventilation and CO₂ elimination, partially correcting the arterial pH. Patients may exhibit the deep and labored respiration known as Kussmaul breathing.

We recommend the following evaluation pathway for metabolic acidosis (Figure 238-1). First, calculate the corrected anion gap. The uncorrected anion gap is often abnormal in chronically ill patients due to low serum albumin and poor nutritional status. If the corrected anion gap is greater than 11, then the patient has an anion gap metabolic acidosis. Anion gap metabolic acidosis may result from a number of underlying conditions, summarized in the mnemonic in the figure.

If the corrected anion gap is not elevated, then the differential diagnosis becomes limited to excessive GI bicarbonate losses versus inadequate renal net acid excretion. To differentiate these possibilities, calculate the urine anion gap. If it is significantly negative, that is, less than –20 mmol/L, then the kidneys are excreting significant net acid, and the nonanion gap metabolic acidosis is likely due to excessive GI bicarbonate losses. This is most commonly due to losses of small intestinal or colonic fluids, as in chronic diarrhea. If the urine anion gap is not significantly negative, then this indicates a failure of renal acid excretion. Hyperkalemia is a common cause of this, and if present indicates a type IV renal tubular acidosis (RTA). If hyperkalemia is not present, then the differential diagnosis is between uremic acidosis and RTA. These can be differentiated in most cases by assessing the estimated GFR. Because many patients with type IV RTA also have chronic kidney disease, the coexistence of type IV RTA and uremic acidosis is not unusual. This can be identified by persistence of the metabolic acidosis after correction of hyperkalemia.

TREATMENT PRINCIPLES

Metabolic acidosis indicates the presence of an important underlying condition. Treatment directed at the underlying condition should be instituted. Whether metabolic acidosis should be treated with alkali is less apparent, and depends in large part on whether acute or chronic metabolic acidosis is present.

For chronic metabolic acidosis, the benefit of alkali treatment is clear-cut, and includes improved bone and muscle integrity, improved growth in children, and preservation of renal function in patients with chronic kidney disease. There may also be improvements in thyroid hormone function and in regulation of serum glucose. Oral alkali therapy, typically in the form of NaHCO₃, should be instituted on an elective basis, and dosing adjusted every 3 to 4 days targeting a serum HCO₃^– of 24 mmol/L.

For acute metabolic acidosis, the benefit of alkali treatment is less certain. Clinical trials have shown no clinically significant effect of alkali treatment of acute metabolic acidosis on either mortality or on cardiovascular performance. However, treating acute metabolic acidosis with alkali might be considered if arterial pH is less than 7.1. If alkali treatment is used, a target pH of no more than 7.2 is reasonable, as aggressive alkali therapy may have deleterious side effects. Slow IV administration of 50 to 100 mEq of NaHCO₃, over 30 to 45  minutes, can be used. It is essential to monitor plasma electrolytes during the course of therapy, since the [K⁺] may decline as pH rises. Administration of large amounts of concentrated NaHCO₃ in the form of multiple ampules can cause acute hypernatremia because of the associated sodium load. Rapid intravenous alkali administration can acutely decrease the ionized calcium, which may impair cardiac function. Patients with simultaneous intravascular volume depletion and metabolic acidosis may benefit from volume resuscitation from intravenous fluids prepared from the addition of 150 mmol sodium bicarbonate (three ampules) per 1 L of D5W.

HIGH AG ACIDOSIS

The four most common causes of a high AG metabolic acidosis are lactic acidosis, ketoacidosis, ingested toxins, and renal failure. Therefore, the initial screening to differentiate the high AG acidoses should include relevant history for evidence of drug and toxin ingestion, determination of blood lactic acid, glucose, ethanol, beta-hydroxybutyrate, blood urea nitrogen (BUN), and creatinine, inspection of the urine for oxalate crystals, and identification of predisposing factors for lactic acidosis, such as hypotension, shock, cardiac failure, leukemia, cancer, unrecognized bowel ischemia, thiamine deficiency, and drug or toxin ingestion.

LACTIC ACIDOSIS

Lactic acidosis almost always indicates the presence of impaired tissue perfusion, most commonly due to hypotension, arterial disease, or sepsis. Rarely, an unsuspected neoplasm produces lactic acid and is the causative factor. Identifying and treating the underlying cause is central to therapy. Alkali therapy is usually reserved for severe acidemia (pH <7.1).

D-LACTIC ACIDOSIS

D-lactic acidosis is increasingly recognized. It occurs most often in short bowel syndrome, when undigested sugars passing into the colon are fermented to D-lactic acid by lactobacilli. The resulting low pH favors further overgrowth of lactobacilli, leading to a vicious cycle of worsening acidosis. D-lactic acid is also sometimes produced during diabetic ketoacidosis, and it is a metabolic product of propylene glycol, which is used as a stabilizing agent for lorazepam and diazepam infusions. D-lactic acidosis typically presents as encephalopathy, with delirium, ataxia, gait disturbance, slurred speech, and agitation. Special assays are required to detect ­D-lactate. Treatment includes correcting the acidosis, decreasing carbohydrate intake, and antibiotics to clear the overgrowth of lactobacilli.

DIABETIC KETOACIDOSIS

In diabetic ketoacidosis, treating with insulin prevents production of ketones and enables endogenous metabolism of accumulated ketones to bicarbonate, thus correcting the acidosis. Bicarbonate therapy is rarely needed except with extreme acidemia (pH <7.1), and then in limited amounts in patients with hemodynamic instability.

ALCOHOLIC KETOACIDOSIS

Mixed acid-base disorders are common in alcoholic ketoacidosis. The degree of ketosis and ketonuria can often be underestimated on initial presentation. Patients are often volume depleted, and respond to volume expansion. With restoration of circulation by resuscitation with isotonic saline, the preferential accumulation of beta-hydroxybutyrate is then shifted to acetoacetate. This paradoxically results in an increasingly positive nitroprusside reaction as the patient improves (the nitroprusside ketone reaction detects acetoacetic acid, but not beta-hydroxybutyrate). Patients should receive thiamine prior to administration of glucose-containing solutions.

ACIDOSIS FROM DRUGS AND TOXINS

Salicylates

Intoxication typically results in either respiratory alkalosis or a combined respiratory alkalosis and anion gap metabolic acidosis, depending on the stage and level of intoxication. ­Treatment includes gastric lavage to remove unabsorbed salicylate, and forced alkaline diuresis to increase renal salicylate excretion. ­Acetazolamide administration can be useful in patients with simultaneous alkalemia.

Alcohols

In the appropriate clinical setting, identification of an osmolar gap may be the first indication of a poison-associated anion gap acidosis. Ethylene glycol, methanol, and isopropyl alcohol all cause an elevated osmolar gap, but only the first two cause a high anion gap acidosis. Because ethylene glycol and methanol can cause CNS effects, at least initially, similar to that of ethanol, we recommend assessment of the serum osmolar gap in patients presenting in the outpatient setting with an acute anion gap metabolic acidosis, particularly if there is history, physical examination, or laboratory evidence of recent alcohol ingestion.

Ingestion of ethylene glycol, present in antifreeze, can be suspected by observing urine oxalate crystals, the presence of an osmolar gap in serum, and a high anion gap acidosis. Some antifreeze agents contain a fluorescent dye which may be detected by viewing the urine with the aid of a Wood’s lamp. Methanol metabolism results in the formation of formaldehyde and formic acid, which can cause severe optic nerve and central nervous system damage; methanol ingestion should be suspected in patients presenting in the outpatient setting with an anion gap metabolic acidosis and acute visual disturbances.

Treatment of methanol and ethylene glycol poisoning includes the prompt institution of a saline or osmotic diuresis, supplementation with thiamine, pyridoxine, and folic acid to maximize metabolism by nontoxic pathways, and inhibition of alcohol dehydrogenase with fomepizole. Fomepizole therapy has replaced therapy with ethanol, as the decline in ethylene glycol level is more predictable, and fomepizole does not cause the neurological obtundation seen during ethanol infusion. Correction of systemic acidosis with sodium bicarbonate may reduce the tissue penetration of toxic metabolites. Indications for emergent hemodialysis include the presence of ocular manifestations, severe (AG >30, pH <7.25) or worsening metabolic acidosis, blood methanol or ethylene glycol concentrations >50 mg/dL, renal failure, or deteriorating vital signs despite intensive supportive care.

Acetaminophen

Chronic acetaminophen administration, even at therapeutic doses, can cause an anion gap metabolic acidosis in susceptible individuals. Pyroglutamic acid is the fixed acid generated. The pathogenesis is thought due to chronic glutathione deficiency. Risk factors for its development include chronic illness, malnutrition, and female sex. This diagnosis should be differentiated from an acute acetaminophen overdose, which causes acute liver disease. This diagnosis should be considered in patients with anion gap metabolic acidosis of unclear etiology in the presence of chronic ingestion of acetaminophen.

Renal failure

Uremic acidosis is characterized by impaired excretion of ammonia and titratable acid. Metabolic acidosis with renal failure may be either anion gap or nonanion gap, and the determining factors appear to be retention of phosphates, sulfates, and other uremic ions. Simultaneous hyperkalemia may further inhibit renal ammonia metabolism and contribute to type IV RTA. Significant loss of bone mass may occur, as the retained acid is buffered by alkaline salts released from bone. Oral alkali replacement to maintain the [HCO₃^–] at a level of ~24 mmol/L is recommended. Treating the patient with chronic kidney disease and metabolic acidosis with sodium bicarbonate appears to slow the progression of chronic kidney disease and may decrease the risk of development of end-stage renal disease, without worsening either blood pressure control or proteinuria.

NORMAL AG METABOLIC ACIDOSIS

This disorder results from either failure of the kidneys to produce adequate net acid excretion to balance endogenous acid production, or from alkali loss from the intestinal tract. Intestinal tract fluids, except in the stomach, contain relatively high concentrations of bicarbonate, and excessive losses, such as with profound diarrhea, can lead to metabolic acidosis. Failure of the kidneys to generate adequate net acid excretion can result either from renal tubular acidosis (Table 238-5) or from advanced, typically stage IV or V, CKD. Because patients with type IV RTA typically have CKD in conjunction with their hyperkalemia, the differentiation of type IV RTA from CKD-induced acidosis is inexact. We recommend correcting the hyperkalemia, in which case the metabolic acidosis with type IV RTA in the presence of CKD should resolve. If the metabolic acidosis does not completely resolve, then a component of CKD-associated acidosis is present. Because some patients, particularly those with body image disorders and laxative abuse, may not be forthcoming about diarrhea, it is important to differentiate diarrhea-induced metabolic acidosis from RTA through the use of appropriate laboratory tests, such as the urinary anion gap. These tests are based on the fact that the renal response to metabolic acidosis resulting from intestinal bicarbonate losses is to increase net acid excretion in the form of ammonia, whereas patients with renal tubular acidosis cannot generate this response.

URINARY AG

The urinary anion gap (UAG) is used to indirectly assess urinary ammonia excretion. It is the sum of the urine sodium and urine potassium concentrations, minus the urine chloride concentration, all measured in mmol/L:

Urinary ammonium excretion is increased in patients with diarrhea, and reduced in patients with renal tubular acidosis. If the UAG is negative in a patient with metabolic acidosis, then urinary ammonia, present almost exclusively as urinary NH₄⁺, is appropriately increased. In metabolic acidosis from bicarbonate loss in diarrhea, the UAG is typically –20 to –50 mmol/L. The absence of a negative UAG indicates a failure of urinary ammonia excretion to increase, and suggests that RTA or advanced CKD is the cause of the acidosis. Clinicians must recognize that the formula used for the urine anion gap differs from that used for the plasma or serum anion gap. Not recognizing this difference can lead to serious errors and incorrect diagnoses.

TYPE I RTA (DISTAL RTA)

Type I RTA is characterized by hypokalemia, normal AG metabolic acidosis, low urinary NH₄⁺ excretion (as reflected by a positive UAG), and a high urine pH (pH ≥6.5) in the presence of untreated metabolic acidosis. These patients commonly come to medical attention either because of growth retardation and failure to thrive in children, or recurrent renal stone formation in adults. Urinary citrate, an important inhibitor of urinary calcium oxalate crystallization, is decreased in distal RTA, contributing to the nephrolithiasis. Type I RTA in children is frequently caused by genetic mutations in distal tubular ion transporters. In adults, type I RTA may result from acute interstitial nephritis, or autoimmune diseases, such as systemic lupus erythematosus and Sjögren syndrome.

Oral alkali supplementation, dosed at 1 mmol/kg/d, is the mainstay of therapy. Patients typically require 50% of the alkali as potassium citrate to treat the concomitant hypokalemia. Sodium alkali salts, such as sodium citrate or sodium bicarbonate, are used to correct the remainder of the metabolic acidosis, while avoiding excessive potassium supplementation and development of hyperkalemia. The dosage of sodium and potassium alkali salts should be adjusted to obtain normal serum bicarbonate and potassium levels.

TYPE II RTA (PROXIMAL RTA)

Type II RTA is characterized by hypokalemia and nonanion gap metabolic acidosis. In contrast to type I RTA, urine pH in untreated type II RTA is ≤ 6.0. Many patients with type II RTA exhibit the ­Fanconi syndrome (type II RTA, glycosuria, generalized aminoaciduria, and phosphaturia). Causes of type II RTA include inherited genetic disorders, such as cystinosis, hereditary fructose intolerance, and Wilson disease, and acquired disorders, such as multiple myeloma, renal transplant rejection, and certain medications, including acetazolamide, ifosfamide, and the antiretroviral drug tenofovir. ­Tenofovir is currently a common cause of acquired Fanconi syndrome, characterized by the recent onset of impaired renal function, with hypokalemia, nonanion gap metabolic acidosis, and glycosuria in a nondiabetic patient in an taking tenofovir for HIV or hepatitis B infection.

Identifying and correcting the underlying causes are central to treating acquired forms of proximal RTA. If the underlying cause cannot be reversed, oral alkali therapy is used. Large amounts of bicarbonate, typically 10 to 15 mmol/kg/d, are required. As the serum bicarbonate increases, renal bicarbonate losses increase, causing the need for very high doses. Similar to type I RTA, 50% of the alkali should be given initially as a potassium salt such as potassium citrate, and 50% as a sodium salt, either sodium citrate or sodium bicarbonate. Doses of these are adjusted as necessary to normalize serum potassium and bicarbonate.

TYPE IV RTA

Type IV RTA is characterized by hyperkalemia, normal anion gap metabolic acidosis, and disturbances in the renin-aldosterone axis leading to impaired release or action of aldosterone. Most patients have concomitant chronic kidney disease, most commonly diabetic nephropathy and interstitial nephritis. Other causes or contributing factors include HIV, sickle cell disease, urinary tract obstruction, lupus nephritis, amyloidosis, myeloma, adrenal insufficiency, kidney transplant rejection, and a variety of medications, such as cyclosporine, angiotensin converting enzyme (ACE) inhibitors, angiotensin receptor blockers (ARBs), NSAIDs, trimethoprim, heparin, and potassium-sparing diuretics. In patients with hyperkalemia and CKD, it may be difficult to differentiate type IV RTA from uremic metabolic acidosis. We recommend correcting the hyperkalemia. If the metabolic acidosis does not resolve after hyperkalemia has been corrected then a component of uremic metabolic acidosis is present.

The urine pH in most patients with type IV RTA is acidic, usually ≤ 6.0, but may be elevated. Acidosis is usually mild, and bicarbonate therapy may not be needed. In most cases, type IV RTA develops because hyperkalemia suppresses renal ammonia production and thus net acid excretion. Accordingly, decreasing dietary potassium is beneficial and often corrects the metabolic acidosis. Thiazide diuretics, which are more effective at increasing renal potassium excretion then loop diuretics, may also be helpful. The robust thiazide-type diuretic metolazone is often helpful in patients with advanced CKD.

Type IV RTA is rarely a result of primary adrenal insufficiency. These patients typically present with low blood pressure, hyperkalemia, and nonanion gap metabolic acidosis. The urinary pH will often be 6.5 or higher. The diagnosis of adrenal insufficiency can be confirmed with appropriate serologic tests. Treatment with glucocorticoids and mineralocorticoids is efficacious in correcting type IV RTA in this setting. In general, exogenous mineralocorticoids should be avoided in patients with chronic kidney disease, as they may contribute to progression of chronic kidney disease and an earlier necessity for initiation of renal replacement therapy.

We generally do not recommend discontinuing ACE inhibitors or ARBs in patients with type IV RTA, because these medications decrease cardiovascular events and slow the progression of chronic kidney disease.

METABOLIC ALKALOSIS

Patients with metabolic alkalosis may be asymptomatic or present with delirium, cardiac arrhythmias, and neuromuscular irritability when the pH exceeds 7.55. This diagnosis is most commonly prompted by the recognition of the elevated serum bicarbonate in routine serum chemistry analyses in the asymptomatic patient. Occasional patients come to medical attention because of compensatory hypoventilation, which may lead to hypoxia, respiratory failure, or pneumonia. Patients may have a history of diuretic use, vomiting, and poorly controlled hypertension. Laboratory values demonstrate an elevated arterial pH, increased serum HCO₃^–, and an increase in PaCO₂ secondary to compensatory alveolar hypoventilation.

Metabolic alkalosis involves both a generation and a maintenance phase. The generation phase involves either acid loss, from either loss of acidic gastric fluid, increased renal net acid excretion associated with either hypokalemia or hyperaldosteronism, or decreased plasma volume, leading to a decreased volume of distribution for bicarbonate. The maintenance phase either involves total body chloride depletion, thereby impairing renal chloride-dependent bicarbonate excretion, or ongoing acid losses, typically either gastric fluid or from aldosterone-stimulated renal acid excretion.

Assessment of the urine chloride may be helpful in differentiating the different causes of the maintenance phase (Table 238-6). A low urine chloride, typically <20 mmol/L, indicates total body chloride depletion. Chloride supplementation, typically in the form of sodium chloride, given either orally or intravenously, will often result in improvement in the metabolic alkalosis. A nonsuppressed urine chloride, typically >20 mmol/L, indicates chloride-unresponsive metabolic alkalosis. These patients will typically not be helped with chloride supplementation. Instead, treating their ongoing gastric acid or renal acid losses is critical to their therapy. Hypokalemia directly stimulates renal acid excretion. Treatment of hypokalemia, if present, may be helpful. Many of these patients have a degree of hyperaldosteronism, either primary or secondary, and benefit from treatment with either mineralocorticoid receptor blockers or inhibitors of the renin-angiotensin system if there is evidence of secondary hyperaldosteronism. Direct inhibition of renal acid excretion, such as the carbonic anhydrase inhibitor acetazolamide, may also be useful. Other causes of metabolic alkalosis include Bartter syndrome, Gitelman syndrome, Liddle syndrome, villous adenoma, and milk-alkali (calcium-alkali) syndrome.

TREATMENT

The treatment of metabolic alkalosis depends on whether patient has chloride-responsive or chloride-unresponsive metabolic alkalosis. Patients with chloride-responsive metabolic alkalosis improve with chloride administration. If they are hypokalemic, this can be provided as potassium chloride. If the serum potassium is normal, then the chloride is administered as sodium chloride, typically as intravenous saline solution. Chloride-unresponsive metabolic alkalosis requires treatment of the underlying condition, and generally does not respond well to either NaCl or KCl administration. Very rarely, acid therapy may be necessary. Intravenous acid preparations include either HCl or NH₄Cl, 100 mmol/L, and should be administered only through a central vein. NH₄Cl should be avoided in patients with liver disease because it may precipitate hepatic encephalopathy.

RESPIRATORY ACIDOSIS

Respiratory acidosis is diagnosed by an increase in PaCO₂ and decrease in pH. Patients may be minimally symptomatic, or acutely anxious, agitated, obtunded or comatose. Tremor, asterixis, and myoclonic jerking may be present on examination. Cerebral vasodilation may occur, with headaches and signs of raised intracranial pressure, such as papilledema, abnormal reflexes, and focal muscle weakness.

Differentiating acute from chronic respiratory acidosis is critical for accurate diagnosis and treatment. Acute respiratory acidosis increases HCO₃^– by 1 mmol/L for every 10 mm Hg increase in PaCO₂, whereas in chronic respiratory acidosis (> 3-5 days), renal adaptation increases the HCO₃^– by 4 mmol/L for every 10 mm Hg increase in PaCO₂. The serum HCO₃^– usually does not rise above 38 mmol/L. General anesthetics, sedatives, head trauma, alcohol, intracranial tumors, syndromes of sleep-disordered breathing, diseases of the motor neurons, neuromuscular junction and skeletal muscle, and improperly adjusted mechanical ventilation may result in respiratory acidosis. In a stable mechanically ventilated patient, respiratory acidosis may develop secondary to a sudden rise in CO₂ production, as from fever, agitation, sepsis, or overfeeding. However, it is more often due to a fall in alveolar ventilation from worsening pulmonary function or sudden occlusion of the upper airway.

TREATMENT

The goal of therapy is improved ventilation. Underlying respiratory conditions, such as acute exacerbations of COPD, should be aggressively treated. Sedatives and other central nervous system depressant drugs should be stopped, and the narcotic antagonist, naloxone, should be administered if narcotic overdose or excess is suspected. Patients may require noninvasive positive pressure ventilation, or endotracheal intubation and assisted ventilation. Severe hypercapnia should be cautiously corrected, as a rapid fall in pH may provoke cardiac arrhythmias, reduced cerebral perfusion, and seizures.

RESPIRATORY ALKALOSIS

The diagnosis of respiratory alkalosis depends on measurement of arterial pH and PaCO₂. Other laboratory findings may include a reduced serum K⁺ and elevated serum Cl^–. Acute respiratory alkalosis is not usually associated with increased renal HCO₃^– excretion, but within hours net acid excretion is reduced. Acutely, HCO₃^– concentration falls by 2.0 mmol/L for each 10 mm Hg decrease in PaCO₂. Chronic hypocapnia reduces the serum HCO₃^– by 4.0 mmol/L for each 10 mm Hg decrease in PaCO₂. However, it is unusual to observe a plasma HCO₃^– <12 mmol/L from respiratory alkalosis alone.

Chronic respiratory alkalosis is usually asymptomatic due to renal compensation. Acute respiratory alkalosis (hyperventilation) presents with lightheadedness, confusion, syncope, tachycardia, anxiety, and dyspnea. Alkalosis promotes the binding of calcium to albumin, leading to decreased serum ionized calcium, and may lead to circumoral and acral paresthesias, tetany, and seizures.

The differential diagnosis of respiratory alkalosis depends on whether it is acute or chronic. Chronic respiratory alkalosis is most typically associated with living at high altitude, pregnancy, and acute or chronic liver disease. The differential diagnosis of acute respiratory alkalosis is much more extensive. Many cardiopulmonary disorders associated with hypoxia, including pulmonary embolism, pneumonia, and pneumothorax, are associated with respiratory alkalosis in their early to intermediate stages. Respiratory alkalosis is common during mechanical ventilation. Liver failure, early ­Gram-negative septicemia, and central nervous system insults such as meningitis, encephalitis, brain tumor, or head injury may lead to respiratory alkalosis. Salicylates, theophylline, aminophylline, thyroxine, and catecholamines, and other drugs, may cause respiratory alkalosis. Hyperventilation from anxiety or panic attacks is a diagnosis of exclusion. It is important to consider other conditions such as pulmonary embolism, acute coronary syndrome, and hyperthyroidism prior to making the diagnosis of a panic attack.

TREATMENT

Management is limited to correction of the underlying disorder. Acid infusions, antidepressants, and sedatives are not recommended. Respiratory alkalosis in pregnant women does not require treatment.

ACKNOWLEDGMENTS

Preparation of this chapter supported by grant funds from the NIH (R01-DK045788) and the Department of Veterans Affairs (1I01BX000818).

SUGGESTED READINGS

INTRODUCTION

Acute kidney injury (AKI) is defined as a potentially reversible sudden deterioration in renal function due to prerenal, intrarenal, or postrenal causes. AKI is frequently accompanied by dysregulation of extracellular fluid volume and electrolytes, and a marked increase in the retention of nitrogenous and nonnitrogenous waste products over a period of hours to weeks. AKI may be oliguric (< 400 mL/d) or nonoliguric (> 400 mL/d). AKI can also be defined as an acute and sustained increase in serum creatinine of 0.5 mg/dL (44.2 μmol/L), if the baseline is less than 2.5 mg/dL (221 mmol/L), or an increase in serum creatinine more than 20% if the baseline is more than 2.5 mg/dL (221 mmol/L).

There have been several attempts to achieve consensus among intensivists and nephrologists on the definition of AKI. The Acute Dialysis Quality Initiative (ADQI) group published the RIFLE classification of AKI in 2004 based on three severity categories (risk, injury, and failure) and two clinical outcome categories (loss and end-stage renal disease) (Table 239-1). The parameters assessed in the RIFLE classification are changes in serum creatinine level, glomerular filtration rate (GFR), or urine output (UO). The baseline serum creatinine level and GFR may not be readily available. Hence the consensus committee recommends the use of the Modification of Diet in Renal Disease (MDRD) equation to estimate GFR/1.73 m². The proportional decrease in GFR is calculated from 75 mL/min per 1.73 m², the agreed-upon lower limit of normal. (The RIFLE criteria for adults have been adapted in children and are termed the pRIFLE criteria.)

A modified RIFLE criteria schema has been proposed by the Acute Kidney Injury Network (AKIN). The AKIN diagnostic criteria (Table 239-2) and the classification/staging system (Table 239-3) for AKI are an abrupt (within 48 hours) reduction in kidney function currently defined as an absolute increase in serum creatinine of more than or equal to 0.3 mg/dL (≥ 26.4 μmol/L), a percentage increase in serum creatinine of more than or equal to 50% (1.5-fold from baseline), or a reduction in urine output (documented oliguria of less than 0.5 mL/kg/h for more than 6 hours). The absolute increase in the serum creatinine levels (≥ 0.3 mg/dL) in this diagnostic criterion is based on epidemiologic data demonstrating that changes in serum creatinine levels of 0.3 to 0.5 mg/dL are associated with increased mortality risk. Also, the timeline of within 48 hours is deliberately included in the diagnostic criteria because of data demonstrating poorer outcomes within this period. The clinical utility and universal adoption of either RIFLE or AKIN criteria remain uncertain, and future studies are needed to demonstrate their validity.

EPIDEMIOLOGY

AKI IN THE HOSPITALIZED PATIENT

AKI is a complication of up to 18% of all hospital admissions in the United States. Major causes of hospital-acquired AKI include volume depletion resulting in decreased renal perfusion, major surgery, septic shock, congestive cardiac failure, contrast nephropathy, and aminoglycoside antibiotics. Acute tubular necrosis (ATN) is identified as the most frequent clinicopathologic entity in hospital-acquired AKI, followed by prerenal azotemia, acute-onset chronic renal failure, and urinary tract obstruction.

AKI IN THE CRITICALLY ILL PATIENT

AKI occurs in 25% or more of patients admitted to critical care units, with major causes including sepsis, multiorgan failure, hypotension, nephrotoxin administration, and prerenal factors. AKI affects mostly older patients who have chronic morbidities or are severely ill on admission to the hospital. The overall in-hospital mortality rate in intensive care unit (ICU)-associated AKI is approximately 60%.

COMMUNITY-ACQUIRED AKI

Community-acquired AKI accounts for 1% of the hospital admissions in the United States. Community-acquired AKI in developed nations mainly affects older patients, with causes including acute tubular necrosis, prerenal azotemia, acute-onset chronic renal failure, and obstructive uropathy. In developing countries, AKI is predominantly a disease of infants and children and is due to prerenal etiologies, such as dehydration from acute diarrheal illness. Falciparum malaria, HIV/AIDS, obstetrical mishaps, dengue fever, snake bites, insect stings, botanical and chemical nephrotoxins, acute glomerulonephritis, hemolytic uremic syndrome, and alternative medical therapies are important etiological factors of AKI in the tropics. Crush injury from natural disasters such as earthquake contributes to regional epidemics of AKI.

AKI AMONG CHILDREN

Recent studies indicate that the incidence of AKI in pediatric patients is increasing. This may be related to the high rates of AKI in hospitalized children in the setting of cardiac surgery and stem cell transplantation.

CAUSES OF AKI

The causes of AKI can be classified under three broad categories: prerenal, renal, and postrenal (Table 239-4). Prerenal azotemia and ischemic ATN account for 75% of AKI.

PRERENAL CAUSES

Prerenal causes are characterized by a reversible loss of kidney function and a drop in GFR due to decreased renal perfusion, while the integrity of renal structural components is maintained. Prerenal causes account for approximately 70% of community-acquired AKI and 40% of hospital-acquired cases. Prerenal azotemia is typically due to decreased effective circulating volume, true hypovolemia, or impaired renal perfusion.

Decreased effective circulating volume may be due to cardiac failure, aortic stenosis, nephrotic syndrome, cirrhosis, hepatorenal syndrome, acute pancreatitis, cardiac tamponade, sepsis, or systemic vasodilatation in conjunction with sepsis or anesthesia.

Intravascular volume depletion, a possible result of dehydration due to vomiting, diarrhea, poor fluid intake, fever, and diaphoresis, is the most common cause of prerenal azotemia in the outpatient setting. Excessive urination (polyuria) due to excess diuretics, diabetes insipidus, or poorly controlled diabetes mellitus can also cause AKI. Other causes of volume depletion include gastrointestinal bleeding and plasma loss due to burns, trauma, and anaphylaxis.

Decreased renal perfusion may result from renal artery stenosis or renal vein thrombosis, or more often from drugs such as nonsteroidal anti-inflammatory drugs (NSAIDs). NSAIDs interfere with glomerular autoregulation, especially in patients above the age of 60 with additional risk factors for prerenal azotemia, such as atherosclerotic vascular disease, preexisting CKD, hypotension, diuretic use, cirrhosis, nephrotic syndrome, and congestive cardiac failure. Immunosuppressive drugs, such as tacrolimus and cyclosporine, lead to acute kidney injury by inducing vasoconstriction of the afferent and efferent glomerular arterioles of the kidneys. Angiotensin-converting enzyme (ACE) inhibitors may cause AKI in patients with unilateral or bilateral renal artery stenosis.

INTRARENAL CAUSES

Intrarenal causes of AKI are characterized by the loss of kidney function due to structural damages to glomeruli, tubules, vessels, or interstitium. They are often categorized based on the primary site of renal injury.

Acute tubular necrosis (ATN) is the most common cause of intrinsic renal failure, and is also the most common cause of hospital-acquired acute kidney injury. It usually results from tubular ischemia and inflammation, as in sepsis and shock, or tubular toxins such as heme pigments (such as myoglobin from rhabdomyolysis or hemoglobin from intravascular hemolysis), cisplatin, ethylene glycol, and myeloma light chains. Ischemic tubular necrosis may occur in patients with sustained prerenal azotemia. Ifosfamide, a chemotherapeutic agent, is known to cause acute tubular dysfunction in the proximal tubule.

Hospital-acquired ATN is often multifactorial. For example, it may occur in a septic patient exposed to a potentially nephrotoxic drug, such as an aminoglycoside or amphotericin, or after the administration of a radiocontrast agent in a hypovolemic patient with preexisting renal dysfunction.

Patients with severe renal ischemia may develop cortical necrosis with injury to both tubules and glomeruli. Causes include severe sepsis, dehydration, snake bites, obstetrical catastrophes, thrombotic microangiopathies, and malaria. The prognosis for recovery is less favorable than for ATN.

Glomerular disease in the form of rapidly progressive glomerulonephritis (RPGN) and acute proliferative glomerulonephritis may cause AKI. RPGN may be primary or secondary in origin, the latter being associated with systemic diseases like systemic lupus erythematosus, Wegener granulomatosis, polyarteritis nodosa, Henoch-Schönlein purpura, and Goodpasture syndrome. RPGN can also progress to end-stage renal disease (ESRD) in days to weeks. Acute proliferative glomerulonephritis occurs in patients with bacterial endocarditis, poststreptococcal infection, and postpneumococcal infection.

Vascular disease (ie, microvascular and macrovascular renal arterial disease) can cause AKI. Thrombotic thrombocytopenic purpura, hemolytic uremic syndrome, HELLP syndrome (hemolytic anemia, elevated liver enzymes, low platelet count), glomerular capillary thrombosis, and atheroembolic disease may all cause AKI. Patients undergoing interventional or invasive procedures involving the major vessels and those with arrhythmias are at an increased risk for AKI from atheroemboli.

Interstitial disease, such as acute interstitial nephritis (AIN), often results from an allergic reaction to an offending agent (Table 239-5). Withdrawal of the offending agent frequently results in reversal of AKI. AIN may also occur in systemic inflammatory conditions such as sarcoidosis, systemic lupus erythematosus (SLE), Legionnaires' disease, and hantavirus infection.

POSTRENAL CAUSES

Postrenal AKI is characterized by loss of kidney function due to intrinsic or extrinsic masses obstructing the urinary collecting system, from the tip of the papillae to the urethral meatus. The most common causes are prostatic hypertrophy or carcinoma, cervical cancer, and other causes of bladder neck obstruction. Less common causes include bilateral renal stones, bladder carcinoma, retroperitoneal fibrosis, colorectal cancer, blood clots in the urinary tract, papillary necrosis, and neurogenic bladder.

PATHOPHYSIOLOGY

Figure 239-1 depicts the interaction of hemodynamic, immunological, and inflammatory factors in mediating AKI.

EVALUATION

HISTORY AND PHYSICAL EXAMINATION

A comprehensive history and physical examination is recommended in patients with AKI. Patients should be asked about nonspecific symptoms of azotemia, such as anorexia, nausea, vomiting, malaise, fatigue, pruritus, metallic taste in the mouth, and dyspnea. Bone pain may suggest multiple myeloma. The patient's daily fluid intake and output, daily weights, inpatient and outpatient medications, including NSAIDs and other over-the-counter medications, and outpatient laboratory data should be reviewed. Recent radiology studies should be assessed to determine if the patient has a history of recent contrast use in angiography or computed tomographic (CT) imaging.

On physical examination, patients should be assessed for signs of volume overload, such as peripheral edema, pulmonary rales, and elevated neck veins. Physical findings of advanced renal failure include asterixis, myoclonus, and pericardial or pleural rubs. A minority of patients with acute interstitial nephritis have fever and rash. Muscle pain, bluish discoloration of the toes, livedo reticularis, and loss of foot or ankle pulses suggest atheroembolic renal failure. Palpable purpura, mononeuritis multiplex, hemoptysis, or prominent upper respiratory tract symptoms, such as sinusitis or otitis, suggests glomerulonephritis associated with systemic vasculitis. Other clinical clues for the diagnosis of AKI are listed in Table 239-6.

URINE EVALUATION

A urine specimen should be collected for dipstick, microscopic sediment analysis, culture, and urine chemistries, including sodium, protein, creatinine, and osmolality. Heme-positive urine with few erythrocytes on microscopy suggests myoglobinuria or hemoglobinuria, as in rhabdomyolysis and transfusion reactions. Proteinuria and hematuria suggest glomerular injury. Microscopic evaluation of the urine sediment is crucial. The presence of dysmorphic red cells (Figure 239-2) and red cell casts (Figure 239-3) should raise the possibility of glomerulonephritis. The presence of fat either deposited in epithelial cells (oval fat bodies), in casts (fatty casts) or free fat suggests nephrotic syndrome. The presence of white cells and white cell casts suggests tubulointerstitial inflammation or a pyelonephritis. Hyaline casts and granular casts are suggestive of prerenal azotemia (Figures 239-4 and Figure 239-5). The presence of epithelial cells and pigmented granular casts (muddy brown casts) suggests acute tubular necrosis (ATN) (Figure 239-6). Eosinophiluria, once thought to be diagnostic of AIN, is neither sensitive nor specific for this condition, as it may be seen in pyelonephritis, prostatitis, glomerulonephritis, atheroembolism, and transplant rejection. In ethylene glycol ingestion, double-pyramid, ovoid, or dumbbell-shaped oxalate crystals are usually present in the urine. Amber-colored rhomboid uric acid crystals may be found in urine in tumor lysis syndrome.

Urine indices help to differentiate prerenal azotemia from ATN. The fractional excretion of sodium (FE_Na) is the percentage of sodium filtered by the glomeruli that is excreted in the urine. Patients that are volume depleted are sodium-avid and have a low FE_Na. It is calculated as follows:

In prerenal azotemia, urine osmolality is usually > 500 mOsm/kg, the urinary sodium concentration is < 20 mmol/L, and the fractional excretion of sodium (FE_Na) is < 1%, whereas in tubular necrosis and urinary obstruction the urine osmolality is < 350 mOsm/kg, the urine sodium concentration is > 40 mmol/L, and the FE_Na is > 1% (in ATN, the FE_Na is usually, but not always, > 2%). There are exceptions: the FE_Na may be < 1% in acute tubular injury due to radiocontrast exposure or heme pigment, and also in acute glomerulonephritis. In early urinary tract obstruction, the urinary sodium concentration and FE_Na can be low. The FE_Na is also not reliable in the setting of chronic kidney disease, where a high FE_Na may be seen in patients with prerenal azotemia, especially early in the clinical course. Typical urine findings in different causes of AKI are listed in Table 239-7.

BLOOD TESTS

Serum blood urea nitrogen (BUN), creatinine, electrolytes, calcium, phosphorous, and albumin should be obtained in all patients. Serum creatinine concentration typically rises by 1 to 1.5 mg/dL daily when there is a marked decrement in kidney function. Most hospital laboratories are now able to calculate glomerular filtration (estimated GFR or eGFR) using an equation. Other useful blood tests include uric acid, creatinine kinase, serum immunoelectrophoresis, serum osmolal gap, and complete blood count with differential. In malignancies, serum urate and calcium levels are often high. Eosinophilia suggests allergic interstitial nephritis. Elevated serum creatine kinase is present in rhabdomyolysis. Abnormal serum electrophoresis is present in multiple myeloma. In alcohol ingestions such as ethylene glycol or methanol, there is high osmolal gap.

EVALUATION FOR OBSTRUCTION

Urinary tract obstruction should be excluded in patients presenting with AKI. Renal ultrasonography is the preferred imaging technique, with a sensitivity of 80% to 85% and a specificity approaching 100% for the diagnosis of obstruction. Ultrasonography also helps to identify kidney stones and kidney size. To locate the exact site of obstruction and correct it, percutaneous anterograde urography (via percutaneous puncture of the renal pelvis under fluoroscopy) and retrograde urography (via cystoscopy and ureteral catheterization) may be necessary. In patients with retroperitoneal fibrosis, obstruction of the urinary tract may be difficult to visualize by sonography, and computed tomography or magnetic resonance imaging may be necessary. Bladder catheterization can be useful in ruling out urethral obstruction.

RENAL BIOPSY IN AKI

Percutaneous renal biopsy has a limited role in the immediate evaluation and treatment of AKI. However, it is valuable in diagnosing primary renal diseases, such as glomerulonephritis and AIN, after prerenal and postrenal causes of AKI have been excluded. Renal biopsy also plays a role in the evaluation of early allograft dysfunction in renal transplant patients, and in management decisions related to subsequent use of immunosuppressive therapy.

NOVEL BIOMARKERS OF AKI

The definition of AKI is still based on a rise in serum creatinine and fall in urine volume and GFR. However, serum creatinine is not an optimal marker of renal function as it is influenced by age, sex, race, diet, exercise, and lean muscle mass. Moreover, secretion of creatinine may account for 10% to 40% of its excretion, potentially masking a drop in GFR. The development of sensitive and specific new biomarkers of renal function and injury has been an area of intense research. Serum cystatin C, a cysteine protease inhibitor produced by all nucleated human cells and freely filtered into the glomerulus, appears to be a more accurate marker of renal function and a more robust predictor of all-cause mortality in elderly persons than serum creatinine. Potential urinary biomarkers of early AKI are under investigation, such as kidney injury molecule-1, shed by proximal tubular cells after ischemic or nephrotoxic damage; interleukin-18, generated in the proximal tubule in AKI; osteopontin, a cytokine involved in kidney inflammation and repair; and the cytoprotective protein clusterin, a fatty-acid-binding protein, and neutrophil-gelatinase-associated lipocalin, which are overexpressed by tubular cells undergoing injury and oxidative stress. However, their precise role in the diagnosis and management of AKI has yet to be defined.

MANAGEMENT OF AKI

The management of AKI is directed at identifying and treating the underlying cause and preventing and treating complications. This may involve reversal of hemodynamic instability, elimination of nephrotoxins and avoidance of additional insults, and correction of electrolyte disturbances, uremia, and acid-base disorders.

Maintenance of euvolemic status is central to the management of ischemic acute renal failure. Assessment and correction of volume status are based on careful physical examination, and at times by invasive monitoring. Composition of the fluid replaced should match that of the fluid lost. Minor to moderate hemorrhagic loss is treated with normal saline (0.9% NaCl), whereas severe blood loss is best managed by transfusing packed red blood cells. Repletion of fluid is usually done with normal saline in hypovolemic patients; however, most patients with AKI have volume overload.

Distinguishing true hypovolemic AKI from hepatorenal syndrome (HRS) is a challenge in patients with cirrhosis and ascites. A gentle fluid challenge should be performed in these patients, with slow administration of fluids titrated to jugular venous pressure or central venous/pulmonary artery pressure. In true hypovolemia, urine output increases and the serum creatinine falls, whereas in HRS, a fluid challenge is usually ineffective. In patients with AKI and volume overload, loop diuretics such as furosemide, dosed between 20 and 100 mg, should be administered as an intravenous bolus, initially every 6 hours. The dose can be increased if an adequate response is not achieved. If bolus dosing of furosemide is ineffective, a continuous infusion of furosemide may be initiated, although evidence for its efficacy compared to intermittent diuretic dosing is limited. When oliguria or anuria persists despite conservative management, dialysis is indicated.

Treatment of intrarenal AKI depends on the underlying pathology. Immunosuppressive agents, including steroids and cytotoxic drugs, may be beneficial in treating acute glomerulonephritis or vasculitis. Angiotensin-converting enzyme (ACE) inhibitors are helpful in treating AKI associated with hypertension and scleroderma renal crisis, but should generally be avoided in patients with hemodynamically related kidney injury. AKI due to malignant hypertension should be treated by vigorous control of blood pressure. Possible culprit drugs should be stopped in patients with possible AIN. While the role of corticosteroids in the management of AIN remains unclear since no randomized, controlled trials have been performed, and retrospective studies have shown conflicting results, most clinicians treat patients with a 1- to 2-week course of high dose steroids (prednisone 60 mg/d, rapidly tapered over a 1-2-week period).

Obstruction of the urethra or bladder neck can be relieved temporarily by insertion of a transurethral or suprapubic catheter. Obstructive lesions of the ureter are treated by stenting, either by anterograde percutaneous or retrograde cystoscopic placement. Patients often undergo a postobstructive diuresis for several days following relief of the obstruction, and care should be taken to avoid dehydration. Transient salt wasting syndrome occurs in approximately 5% of patients, and requires administration of intravenous normal saline to maintain blood pressure.

The most common electrolyte disturbance in AKI is hyperkalemia. Initial treatment should include calcium gluconate if the potassium level is > 6 mEq/L, or if electrocardiographic changes are present. Potassium may be transiently shifted into cells with intravenous insulin (10 units) and glucose (25 g) as a bolus infusion. As acidosis favors hyperkalemia by promoting extracellular movement of potassium, intravenous sodium bicarbonate (three ampules in 1 L of 5% dextrose) leads to translocation of potassium into cells and may also facilitate renal potassium excretion. Treatment with sodium bicarbonate is less effective acutely. Sodium polystyrene sulfonate (Kayexalate) given orally (25-50 g mixed with 100 mL of 20% sorbitol) or as enema (50 g in 50 mL of 70% sorbitol and 150 mL of tap water) is a more definitive therapy that leads to excretion of potassium from the body. Kayexalate enemas with 70% sorbitol are rarely associated with colonic necrosis, especially in critically ill patients, and should be avoided if other therapeutic options are available.

More recently sodium zirconium cyclosilicate (ZS-9) and patiromer sorbitex calcium have been shown to be effective in treating hyperkalemia. ZS-9 is an a nonabsorbable cation trap that selectively binds K⁺ in exchange for H⁺ and Na⁺; it binds K⁺ immediately upon ingestion. In contrast, patiromer is a nonabsorbable polymer that enhances K⁺ excretion by the exchange of Ca²⁺, predominantly in the distal colon. Dialysis remains an important option if hyperkalemia is severe or is refractory to treatment.

Acidosis may be corrected by administering sodium bicarbonate when the pH level is less than 7.2, or the serum bicarbonate level is less than 15 mEq/L. Bicarbonate may be administered orally (a 300 mg tablet contains 3.6 mEq of sodium bicarbonate) or intravenously. Ampules for intravenous administration are provided as 7.5% sodium bicarbonate (44.6 mEq/50 mL) or 8.4% sodium bicarbonate (50 mEq/50 mL). The required bicarbonate dose is measured by the following equation: bicarbonate deficit (mEq) = (lean body weight, LBW) (0.5) (desired HCO₃ – actual HCO₃). If acidosis cannot be corrected, dialysis is performed.

Diet plays an important role in patients with AKI who become nutritionally deficient. The daily caloric intake should be 30 to 45 kcal (126-189 kJ) per kg per day. In nondialysis patients, protein intake should be restricted to 0.6 g/kg/d. Patients receiving dialysis should have 1 to 1.5 kg/d.

RENAL REPLACEMENT THERAPY (RRT) IN AKI

Dialysis as a treatment modality in AKI is most useful in the elimination of metabolic wastes and uremic toxins, but it is also effective in the maintenance of fluid, electrolyte, and acid-base status. There is limited consensus among nephrologists regarding choice of dialysis, timing of initiation, frequency and intensity of dialysis, and impact on patient outcomes, due to lack of definitive evidence from studies conducted thus far in the setting of AKI (Table 239-8).

CONTINUOUS RENAL REPLACEMENT THERAPY VERSUS HEMODIALYSIS

In critically ill patients, continuous renal replacement therapy (CRRT) is now a popular and frequently preferred alternative to intermittent hemodialysis, especially in hemodynamically unstable patients.

CRRT can be either arteriovenous or venovenous, such as continuous arteriovenous hemodialysis (CAVHD) and continuous venovenous hemodialysis (CVVHD). CVVHD is preferred over CAVHD because of the high ultrafiltration rate, predictable blood flow rate, avoidance of arterial puncture, ease of vascular access (either femoral or internal jugular vein, via a double lumen catheter), and reduced risk of heparin-induced bleeding.

Compared to intermittent dialysis, CRRT has several advantages, including lower risk of hemodynamic instability, ease of administration of nutritional support, and more accurate control of fluid and metabolic status. Disadvantages include prolonged anticoagulation, nursing burden, cost, and limited mobility.

Sustained low-efficiency dialysis (SLED) or extended daily dialysis (EDD) is another increasingly popular dialytic modality of choice for critically ill patients and has the benefits of both continuous and intermittent dialysis methods. The advantages of SLED are good hemodynamic tolerability, high clearance levels of even small solutes, convenient treatment schedules, and cost effectiveness.

PERITONEAL DIALYSIS

Peritoneal dialysis (PD) is generally not a favored therapeutic modality for AKI. However, it is the preferred mode of renal replacement therapy in resource-limited settings and the developing world. Advantages include easy portability and access placement, better hemodynamic stability among patients, and the avoidance of systemic anticoagulation and its associated complications. Disadvantages include the risk of injury to the viscera during catheter placement, low levels of solute clearance, metabolic instability in critically ill patients, and the high risk of peritonitis.

DIALYSIS-RELATED COMPLICATIONS

Dialysis-induced complications include hypotension and complement cascade activation during the blood-dialyzer interaction. Hypotension worsens existing AKI because of impaired renal autoregulatory mechanisms. The upregulation of neutrophil adhesion molecules due to blood-dialyzer membrane interaction result in further damage to renal tissues. Unlike cuprophane, biocompatible membranes like polymethylmethacrylate or polyacrylonitrile do not activate complement to any great extent, and the use of biocompatible membranes is associated with improved recovery of renal function and reduced mortality among AKI patients requiring dialysis.

PROGNOSIS

The mortality rate of AKI ranges from 20% to 50%, and up to 70% in the ICU and surgical setting. Advanced age, severe underlying disease, preexisting poor nutritional status, and multisystem organ failure are associated with increased mortality. About 10% to 20% who suffer AKI will require maintenance dialysis. Infection, cardiorespiratory complications, and underlying disease states are the major causes of death in patients with AKI, rather than AKI itself. The availability of health care in a given country greatly influences the prognosis for AKI.

CONSULTATION

Nephrology consultation is appropriate in AKI when the cause is unclear after initial testing has been performed; if assistance is required with assessment and management of volume status; if intrinsic renal disease, such as glomerulonephritis or AIN, is suspected, and renal biopsy may be needed for diagnosis; and if an indication for dialysis is present, such as refractory fluid overload, severe metabolic acidosis, hyperkalemia, pericarditis, mental status changes, or overdose with a dialyzable toxin or drug.

DISCHARGE CHECKLIST

• Has close outpatient follow-up been arranged, with clinical assessment and measurement of serum creatinine and electrolytes?

• Has the patient been warned to avoid dehydration and drugs that may worsen renal function, such as NSAIDs?

• Have drug doses been adjusted for estimated GFR?

• If the patient seems likely to need dialysis in the near future, have they been told to avoid phlebotomy and intravenous placement in one arm to preserve it for graft or fistula creation?

SUGGESTED READINGS

INTRODUCTION

Abnormalities of calcium metabolism are common in hospital practice. Hypercalcemia has a prevalence of 0.1% in the general population and 1% among hospitalized patients. In the inpatient setting, hypercalcemia often portends serious illness, especially malignancy. Hypocalcemia is also common in the hospital, especially in patients with chronic renal failure or sepsis. Hypocalcemia may also be a manifestation of vitamin D deficiency, which has a prevalence of up to 80% on specialized geriatric inpatient units.

CALCIUM METABOLISM

Precise regulation of calcium homeostasis is essential because of the critical role of calcium in many physiological activities. It is the major mineral of bone. It also plays major roles in neuronal transmission, muscle contraction, and blood coagulation. Calcium is also required for the proper functioning of many enzymes, endocrine secretory processes, and biochemical signaling pathways.

NORMAL SERUM CALCIUM LEVELS

A typical laboratory range for serum total calcium concentration is between 8.4 and 10.2 mg/dL. Approximately half of this total amount is bound to albumin, with the remainder in free (ionized) form. The normal free calcium concentration range is 4.5 to 5.3 mg/dL. A small fraction (10%) of circulating calcium is complexed with anions, such as citrate and phosphate.

REGULATION OF CALCIUM HOMEOSTASIS

The three organ systems that together regulate serum calcium are the gastrointestinal tract, kidneys, and skeleton. The two principal regulatory hormones are parathyroid hormone (PTH) and 1,25-­dihydroxyvitamin D₃. PTH is a peptide secreted from the parathyroid glands in its active full-length configuration, known as PTH(1-84). Its plasma half-life is very short, on the order of 3 to 5 minutes. The major regulator of PTH secretion is the free calcium concentration in extracellular fluid. Elevated levels of free or ionized calcium promptly block secretion of PTH, while reduced serum ­calcium levels promptly increase secretion of PTH.

1,25-dihydroxyvitamin D₃ is produced by a sequence of activation steps (Figure 240-1), starting with the generation of cholecalciferol (vitamin D₃) through exposure of skin to ultraviolet light of a specified wavelength (90-315 nm). Cholecalciferol or its plant analogue, ergocalciferol (vitamin D₂), can also be obtained by dietary sources or in nutritional supplements. Cholecalciferol or ergocalciferol is converted in the liver to a hydroxylated form, 25-hydroxyvitamin D₃ or 25-hydroxyvitamin D₂. The 25-hydroxylated forms of vitamin D are converted to their active forms by a second hydroxylation step in the kidney leading to 1,25-dihydroxyvitamin D₂ or D₃. Both dihydroxylated forms of vitamin D are active in human subjects, although there is controversy over whether vitamin D₃ is more potent than vitamin D₂. PTH maintains serum calcium concentrations by conserving calcium that has been filtered at the kidney glomerulus and by mobilizing calcium from bone. 1,25-dihydroxyvitamin D maintains serum calcium by facilitating absorption of calcium from the gastrointestinal tract and, like PTH, mobilizing calcium from bone. Under normal conditions, the calcium absorbed by the gut (approximately 150-200 mg/d) is matched by the calcium eliminated by the kidney. At the dynamic skeletal interface, as much as 500 mg of calcium is turned over daily. This process is in a steady state, with net calcium neither gained nor lost. Thus, under normal circumstances, there are no significant fluctuations in body calcium stores, nor is there any major change in circulating serum calcium concentrations.

Changes in free or ionized calcium concentration are registered virtually instantly by parathyroid cells via the calcium-sensing receptor (CaSR). This receptor is located on the parathyroid cell surface, where its extracellular domain senses binding of calcium ions. If the circulating calcium concentration rises, the Ca²⁺-CaSR complex leads to a rise in intracellular calcium, inhibiting both PTH secretion and synthesis. If the serum calcium concentration falls, the Ca²⁺-CaSR complex sends a reduced signal to the cell, leading to an increase in PTH secretion and synthesis.

1,25-dihydroxyvitamin D decreases PTH production, although not as powerfully as does the ionized calcium signal. There is a stronger interaction between levels of 25-hydroxyvitamin D and PTH. They have an inverse relationship, with PTH levels rising when 25-hydroxyvitamin D levels fall below approximately 25 to 30 ng/mL. In turn, increased PTH stimulates the 1-alpha hydroxylase enzyme in the kidney that converts 25-hydroxyvitamin D to 1,25-­dihydroxyvitamin D (Figure 240-2). When PTH levels are elevated (ie, primary hyperparathyroidism), 1,25-dihydroxyvitamin D levels increase. When PTH levels are low (ie, hypoparathyroidism), 1,25-dihydroxyvitamin D levels are typically low. The three organ systems (bone, gastrointestinal tract, and kidneys) and the two calcium-regulating hormones (PTH and 1,25-dihydroxyvitamin D) work together to maintain normal calcium homeostasis. When they are not perturbed by disease or by the aging process, they are an exquisitely sensitive and effective servomechanism.

LABORATORY MEASUREMENT OF BLOOD CALCIUM

The measurement of serum calcium may be helpful when a disturbance of calcium metabolism is suspected. However, in many disorders of calcium metabolism, such as osteoporosis or Paget disease of bone, the serum calcium concentration is typically normal. Serum measurements may be performed by spectrophotometry or by atomic absorption spectrophotometry, with the latter yielding more accurate measurements. Spuriously high readings may occur if the tourniquet is in place too long before blood is drawn and hemoconcentration occurs. Under these circumstances, the measured serum calcium value can rise by as much as 0.4 mg/dL. On the other hand, the sample can read falsely low if the blood sample is obtained from a central, high-flow site via a central venous catheter. For most clinical situations, the total serum calcium is measured. This may need to be corrected for the circulating albumin concentration. For every 1 g/dL reduction in the serum albumin, the total calcium is adjusted upward by 0.8 mg/dL. This may be calculated as follows: Corrected total calcium = measured total calcium + 0.8 (4.0 – serum albumin)

In theory, free or ionized serum calcium is a more accurate physiological measurement than the adjusted total serum calcium concentration, but the sampling technique (the blood has to be free-flowing and not impeded by a tourniquet) and strict anaerobic collection conditions are problematic. Moreover, the measuring instrument has to be in regular use and properly calibrated. Samples have to be measured immediately. These technical issues somewhat limit the clinical utility of the ionized calcium measurement.

HYPERCALCEMIA

Signs and symptoms of hypercalcemia may be absent or subtle, except when calcium is significantly elevated or has increased rapidly. The diagnostic workup of hypercalcemia is usually straightforward (Figure 240-3) because two causes, primary hyperparathyroidism and malignancy-associated hypercalcemia, account for approximately 90% of cases. In addition, most individuals with primary hyperparathyroidism are asymptomatic and discovered on routine biochemical screening tests, while most individuals with malignancy-associated hypercalcemia have a known advanced malignancy at the time that hypercalcemia occurs. If the malignancy is not known, it is generally quickly apparent. When neither of these two etiologies is readily apparent, identification of the other potential etiologies requires a comprehensive history, physical examination, laboratory tests, and, occasionally, diagnostic imaging studies.

PRESENTING SYMPTOMS AND HISTORY

Many individuals with mild hypercalcemia (serum calcium level < 11 mg/dL) are asymptomatic, although some may report mild fatigue, vague changes in cognitive function, depression, or constipation. Symptomatic manifestations of hypercalcemia are more apparent when the serum calcium concentration is between 12 and 14 mg/dL. These symptoms include anorexia, nausea, weakness, and depressed mental status. As hypercalcemia may induce polyuria and nephrogenic diabetes insipidus, dehydration may occur should the compensatory polydipsia not keep up with urinary water losses. When serum calcium levels rise above 14 mg/dL, profound dehydration, renal dysfunction, and central nervous system changes, such as progressive lethargy, disorientation, and coma, may develop.

In addition to the absolute magnitude of the serum calcium elevation, the rate of increase in serum calcium also influences symptoms. Individuals who are chronically hypercalcemic may have relatively few symptoms, even with serum calcium values up to 15 to 16 mg/dL. In contrast, those whose calcium level has risen abruptly may have symptoms at much more modest calcium levels. Elderly or debilitated patients are more likely to be affected by hypercalcemia than younger individuals.

The medical record may contain clues to etiology. Prescription medications (Table 240-1), foods, and vitamin and nutritional supplements should be reviewed. A careful family history might uncover a familial endocrine condition. A history of family members with endocrine tumors of the pituitary or pancreas suggests multiple endocrine neoplasia type 1 syndrome (MEN-1). A family history of pheochromocytoma or medullary thyroid cancer is consistent with MEN-2 syndrome. Patients with sarcoidosis may have a history of unexplained fever, lymphadenopathy, skin rashes, or pulmonary symptoms. Bone pain suggests myeloma or other malignancies, although it may also be a nonspecific finding of hypercalcemia.

THE PHYSICAL EXAMINATION

The physical examination is directed at identifying signs of hypercalcemia. Evidence of dehydration such as orthostasis or dry mucous membranes may be present, although hypercalcemia must be marked and prolonged for these physical findings to be appreciated. The physical examination is often normal in patients with hypercalcemia, especially if calcium levels are only modestly elevated. Rarely, severe and prolonged hypercalcemia may produce a visible horizontal deposit of calcium salts on the cornea, a finding called band keratopathy.

Effort should be made to identify signs of common causes of hypercalcemia, such as malignancy and primary hyperparathyroidism. The physical examination in primary hyperparathyroidism, like most hypercalcemic states, is usually not noteworthy. A mass is virtually never found in the neck, because enlarged parathyroid glands are still too small to be felt. However, when the serum calcium is markedly elevated, a neck mass may signify a parathyroid carcinoma. Symptomatic kidney stones might be accompanied by costovertebral tenderness. Enlarged lymph nodes suggest sarcoid, lymphoma, or metastatic carcinoma.

DIAGNOSIS

Laboratory studies

The first step in evaluating hypercalcemia is adjustment for serum albumin. If the corrected serum calcium is elevated, it should be repeated. Renal function should also be assessed, because hypercalcemia may develop or worsen in the setting of acute renal failure. If hypercalcemia is confirmed, the next step is measurement of serum PTH. The PTH level is the most important test for distinguishing between the two most common causes of hypercalcemia, primary hyperparathyroidism and malignancy-associated hypercalcemia (Table 240-1). The so-called intact immunochemiluminometric assay for PTH assay primarily measures the intact molecule, PTH(1-84), as well as a large circulating fragment that is foreshortened at the amino terminus, PTH(7-84). A more specific assay that measures only PTH(1-84), the bio-intact assay, is also available, but it has not shown any clear advantages over the older assay, which has been in clinical use for over 20 years. When the creatinine clearance falls below 60 mL/min, these assays may begin to show elevations in PTH due to the accumulation of inactive fragments, and also perhaps due to increased secretory activity of the parathyroids (secondary hyperparathyroidism).

When the parathyroid glands are functioning normally, hypercalcemia should suppress PTH levels. Hypercalcemia is said to be PTH-mediated if serum calcium is elevated, and the PTH level is high or inappropriately normal. In this latter situation, one is usually dealing with primary hyperparathyroidism, although familial hypocalciuric hypercalcemia (FHH) and medication-induced hypercalcemia, as from thiazide diuretics or lithium, can also be associated with elevated PTH levels. When PTH levels are appropriately suppressed in hypercalcemia, the differential diagnosis includes malignancy, granulomatous disease, medications, milk-alkali syndrome, thyrotoxicosis, and adrenal insufficiency.

Other recommended tests in the evaluation of hypercalcemia include serum electrolytes and 25-hydroxyvitamin D. Levels of 25-hydroxyvitamin D typically exceed 150 ng/mL in vitamin D toxicity due to excess intake. Levels this high cannot be achieved by sun exposure alone. High 1,25-(OH)₂D levels may be seen in any granulomatous disease, particularly sarcoidosis or certain lymphomas. Inorganic phosphorus measurement may be helpful, as a low-normal serum phosphate is often seen in primary hyperparathyroidism, while high phosphate may be seen in vitamin D intoxication. An elevated serum creatinine may indicate dehydration or true renal dysfunction due to renal deposition of calcium salts or other causes. An elevated alkaline phosphatase level suggests elevated bone turnover. This may be confirmed by measuring bone-specific alkaline phosphatase or other indices of bone turnover, such as serum osteocalcin, serum C-terminal collagen peptide measurement, or urinary N-terminal collagen peptide. Most forms of hypercalcemia are accompanied by hypercalciuria (24-hour urine calcium excretion > 4 mg/kg/24 hours). However, in primary hyperparathyroidism, renal calcium excretion is lower than expected for the degree of hypercalcemia, because PTH conserves filtered calcium in the distal renal tubule.

Additional tests

The electrocardiogram may show a shorted QT_c interval, particularly if hypercalcemia has occurred over a short period of time. Bone mineral density (BMD) by dual energy x-ray absorptiometry (DXA) may be helpful. In primary hyperparathyroidism, there is a typical pattern of BMD with relative preservation of cancellous bone, as in the lumbar spine, and significant loss of cortical bone, as in the femoral neck and distal third of the radius. Abdominal imaging studies (CT or ultrasound) may identify renal stones or nephrocalcinosis. Serum and urine protein electrophoresis should be obtained if myeloma is suspected. Skeletal radiographs may reveal lytic lesions of multiple myeloma or other malignancies. In primary hyperparathyroidism, skeletal radiographs may show subperiosteal bone resorption or brown tumors of bone, but are rarely needed for diagnosis.

CAUSES OF PTH-MEDIATED HYPERCALCEMIA

Primary hyperparathyroidism

Elevation of both serum calcium and PTH concentrations, in the absence of lithium use or low urinary calcium excretion as seen in familial hypocalciuric hypercalcemia, supports a diagnosis of primary hyperparathyroidism. In this condition, PTH levels are usually within 1.5 to 2.0 times above the upper limit of normal. Extremely high levels of PTH raise the specter of parathyroid carcinoma. Typical primary hyperparathyroidism is associated with mild hypercalcemia, within 1 mg/dL above the upper limit of normal. The PTH level may be elevated, but may also fall in the upper portion of the normal range, which is inappropriate in hypercalcemia. Normocalcemic primary hyperparathyroidism is a new diagnostic entity applied to patients whose total and free serum calcium levels are normal, but in whom the PTH level is consistently elevated. In the absence of a secondary cause for elevated PTH levels, it is felt that these individuals have an early form of primary hyperparathyroidism.

Primary hyperparathyroidism is the most common cause of hypercalcemia in outpatients. The incidence is estimated to be approximately 21.6 per 100,000 person-years. The mean age at diagnosis is in the sixth decade of life, and there is a female-to-male ratio of 2:1. The clinical manifestations depend largely on the severity of the hypercalcemia. When primary hyperparathyroidism was first described more than 80 years ago, most patients presented with advanced disease with overt radiographic abnormalities of bone (osteitis fibrosa cystica) and kidneys (nephrolithasis or nephrocalcinosis). Since the introduction more than 40 years ago of automated multichannel autoanalyzers for measuring serum chemistry, primary hyperparathyroidism is most often diagnosed by routine blood testing, well before the development of other signs or any symptoms. It also may be uncovered during the evaluation of osteoporosis or during the workup of renal stone disease. The most common clinical presentation today is mild asymptomatic hypercalcemia. In 75% to 80% of cases, a solitary, benign parathyroid adenoma is present. Hyperplasia involving multiple parathyroid glands is found in 15% to 20% of cases, and parathyroid carcinoma is present in less than 0.5%. On occasion, double adenomas are found. Patients with MEN-1 or MEN-2 usually have parathyroid hyperplasia involving all parathyroid glands.

Parathyroid surgery is always indicated in symptomatic primary hyperparathyroidism, unless there are medical contraindications. The role of parathyroid surgery in asymptomatic primary hyperparathyroidism is more controversial. According to the guidelines of the Fourth International Workshop on the Management of Asymptomatic Primary Hyperparathyroidism, indications for surgery in asymptomatic patients include a serum calcium > 1 mg/dL above the upper limit of normal; creatinine clearance < 60 mL/min; 24-hour urine calcium > 400 mg/d and increased stone risk by biochemical stone risk analysis; presence of nephrolithiasis or nephrocalcinosis by x-ray, ultrasound or CT; T-score < –2.5 at lumbar spine, hip, or distal third of the radius; vertebral fracture by x-ray, CT, MRI or VFA; and age < 50. Patients who do not meet these guidelines can be followed expectantly. Thiazide diuretics and lithium should be avoided. Dietary calcium should not be restricted, because such restriction may promote further elevation of PTH, and possibly have adverse effects on bone mass. In patients who are vitamin D deficient, cautious replacement of vitamin D is advised. Patients should maintain hydration. Bisphosphonates increase lumbar spine BMD in primary hyperparathyroidism, without a major effect on the serum calcium concentration. The calcimimetic agent, cinacalcet, reduces serum calcium in primary hyperparathyroidism without having a major effect on BMD. Cinacalcet is indicated for use in patients with parathyroid cancer, as well as patients with primary hyperparathyroidism who are unable to undergo parathyroidectomy. Alendronate has not been approved by the Food and Drug Administration (FDA) for use in primary hyperparathyroidism.

Lithium can change the set point for the calcium-sensing receptor on the parathyroid gland, such that a higher serum calcium concentration is needed to inhibit PTH secretion. This can lead to mild biochemical abnormalities, such as high levels of calcium and high-normal to elevated PTH levels, that mimic primary hyperparathyroidism, but do not require medical intervention.

Thiazide-associated hypercalcemia also occurs. Many patients with hypercalcemia on thiazides probably have primary hyperparathyroidism. When thiazide therapy is discontinued, the hypercalcemia often persists, and the diagnosis of primary hyperparathyroidism is made.

Familial hypocalciuric hypercalcemia

Familial hypocalciuric hypercalcemia, also known as benign familial hypercalcemia, is a rare genetic condition caused by inactivating mutations in the CaSR. This results in lack of sensitivity of the parathyroid cell to ambient serum calcium, a higher set point for the extracellular ionized calcium concentration, and inappropriately normal to mildly elevated PTH levels. Patients with FHH have chronic asymptomatic hypercalcemia, with very low urinary calcium excretion. The relatively low urinary calcium excretion in FHH helps distinguish it from primary hyperparathyroidism, although low urinary calcium excretion may also occur in individuals with primary hyperparathyroidism. A family history of asymptomatic mild hypercalcemia, especially in individuals younger than 40 years, is suggestive of FHH. Other supportive evidence for FHH includes a very low urinary calcium to creatinine clearance ratio (< 0.01), and a history of family members who have undergone noncurative parathyroidectomy for presumed primary hyperparathyroidism. When FHH is suspected, further evaluation is necessary, such as screening of other family members for hypercalcemia. Genetic testing for FHH may be appropriate, as it may otherwise be exceedingly difficult to distinguish FHH from primary hyperparathyroidism.

Tertiary hyperparathyroidism

Conditions associated with low serum calcium are usually also associated with chronically elevated PTH levels, which is an appropriate physiological response. This is called secondary hyperparathyroidism. The rise in PTH may restore the serum calcium to normal, or calcium may remain low or in the low-normal range. Secondary hyperparathyroidism is not a hypercalcemic state. Common causes of secondary hyperparathyroidism include vitamin D deficiency, intestinal malabsorption of calcium or vitamin D, renal-based hypercalciuria, severe nutritional calcium deficiency, and especially chronic renal insufficiency. Correction of the underlying cause usually returns serum PTH concentrations to normal. Normalization of PTH may be relatively rapid, if the cause is of recent onset, or it may be protracted, if the associated condition has been longstanding. In patients with prolonged secondary hyperparathyroidism, the reactive state can become semiautonomous, leading to hypercalcemia. This condition, known as tertiary hyperparathyroidism, is most often seen in patients with poorly controlled chronic kidney disease. Tertiary hyperparathyroidism is usually associated with hyperplasia of multiple glands, but may also be caused by a parathyroid adenoma from a single clone of parathyroid cells.

Further investigations

Most patients with primary hyperparathyroidism have a serum calcium concentration below 11 mg/dL. Serum phosphate tends to be in the low-normal range (2.5-3.2 mg/dL). Rarely, a nonanion gap hyperchloremic acidosis from a PTH-induced defect in bicarbonate resorption may be seen. Urinary calcium excretion tends to be in the upper range of normal. However, hypercalciuria in primary hyperparathyroidism does not always predispose to renal stones, despite the fact that hypercalciuria is a risk factor for kidney stones in euparathyroid subjects. Bone turnover markers tend to be at the upper limit or normal, but occasionally can be frankly elevated.

Once the diagnosis of primary hyperparathyroidism is made, it should be determined whether or not the patient meets clinical criteria for parathyroidectomy (Table 240-2). If the clinical situation is appropriate, consideration should be given to the possibility of one of the MEN syndromes, particularly if the patient is young, or has a personal or family history of a related endocrinopathy. A diagnosis of MEN-1 or MEN-2 should prompt a search for multiple parathyroid gland disease.

CAUSES OF PTH-INDEPENDENT HYPERCALCEMIA

If the serum calcium concentration is elevated but the PTH level is appropriately suppressed, the patient has hypercalcemia due to causes other than hyperparathyroidism (PTH-independent hypercalcemia). Cancer is the most common cause. Other causes include thyrotoxicosis, vitamin D intoxication, sarcoidosis, immobilization, Addison disease, and various drugs and supplements.

In hypercalcemia of malignancy, calcium is usually moderately or severely elevated, and PTH is low or undetectable. Significant dehydration and generalized debility are usually evident, along with other cancer-related symptoms. Usually, the diagnosis of malignancy has already been established when patients become hypercalcemic. Hypercalcemia of malignancy has two forms: humoral hypercalcemia of malignancy (HHM) and local osteolytic hypercalcemia. HHM results from tumor production of a circulating factor with systemic effects on calcium metabolism, acting on skeletal calcium release, renal calcium handling, or intestinal calcium absorption. The usual cause of HHM is parathyroid hormone-related protein (PTHrP). Normally, PTHrP serves as a paracrine factor in tissues such as bone, skin, breast, uterus, placenta, and blood vessels, where it is involved in cellular calcium handling, smooth muscle contraction, and growth and development. The amino terminus of the PTHrP peptide is closely homologous with native PTH, and they share a common receptor. When PTHrP circulates at supraphysiologic concentrations, it produces effects similar to PTH, activating osteoclasts to resorb bone, decreasing renal calcium output, and increasing renal phosphate clearance.

Tumors that produce HHM by secreting PTHrP are typically squamous cell carcinomas of the lung, esophagus, head and neck, or cervix. Other tumors that may elaborate PTHrP include adenocarcinoma of the breast or ovary, renal carcinoma, transitional cell carcinoma of the bladder, islet cell tumors of the pancreas, T-cell lymphoma, and pheochromocytoma. As tumors that produce PTHrP do so in relatively small amounts, the syndrome typically develops in patients with a large tumor burden. It is therefore unusual for HHM to be the presenting feature of a cancer. The diagnosis may be confirmed by a commercially available radioimmunoassay for PTHrP. Care should be taken to ensure that blood for PTHrP levels is drawn and handled correctly to avoid spurious low results. Rarely, HHM is caused by the unregulated production of 1,25-dihyroxyvitamin D, usually by B-cell lymphomas, or other mediators that interfere with calcium homeostasis.

The other major mechanism of malignancy-associated hypercalcemia is the direct invasion of bone by tumor, with lytic destruction and calcium release. While this was formerly thought to be a mechanical process, it now appears to be driven by the local elaboration of cytokines leading to osteoclast-mediated bone resorption. In local osteolytic hypercalcemia, PTHrP and calcitriol are within normal limits. Bony metastases are usually obvious on imaging studies. The classic tumor associated with this syndrome is multiple myeloma, although breast cancer and certain lymphomas may also be responsible. Local osteolytic hypercalcemia may be perpetuated by a positive feedback loop. Factors produced by bone promote the growth and survival of metastases, and the tumor induces osteoclasts to produce factors promoting tumor growth, bone resorption, and hypercalcemia. Interruption of this positive feedback loop is the rationale for the use of bisphosphonates in the treatment of multiple myeloma.

PTH-independent hypercalcemia also occurs in sarcoidosis, tuberculosis, and other granulomatous diseases. Macrophages in the granuloma convert 25-hydroxyvitamin D to 1,25-dihydroxyvitamin D, via an unregulated 1-α hydroxylase enzyme. 25-hydroxyvitamin D levels are typically not elevated. When serum 25-hydroxyvitamin D levels are elevated, excessive vitamin D intake becomes the more likely etiology. Endocrine conditions that may occasionally lead to hypercalcemia include severe hyperthyroidism, which stimulates bone resorption, and Addison disease, where volume depletion reduces calcium clearance and control of calcium absorption is mitigated by glucocorticoid deficiency.

Immobilization stimulates bone resorption and may increase serum calcium levels, particularly in bedbound hospitalized patients. This is usually seen in persons with high bone turnover, such as adolescents and patients with unrecognized hyperparathyroidism or Paget disease of bone. Drugs and dietary supplements may lead to hypercalcemia. Vitamin D intoxication and excessive intake of vitamin A, which activates bone resorption, are occasional culprits. Thiazide diuretics may cause hypercalcemia due to enhanced renal retention of calcium. In many cases, this develops in individuals with underlying mild primary hyperparathyroidism.

In patients with an extensive negative workup, the rare possibility of occult malignancy should be considered, especially when PTHrP is elevated. Further imaging studies would then be needed for tumor localization, including a plain chest radiograph or a computed tomographic scan of the chest to rule out lung malignancy. If these are unrevealing, consideration should be given to otolaryngoscopic examination, esophagoscopy, or CT of the abdomen, followed by radiographic or endoscopic evaluation of the genitourinary tract if necessary.

TREATMENT OF HYPERCALCEMIA

Hypercalcemia that requires urgent management is usually due to malignancy, rather than primary hyperparathyroidism. Urgent management includes aggressive rehydration, bisphosphonate therapy to decrease bone resorption, and elimination of contributing factors, such as calcium or vitamin D supplements, thiazide diuretic therapy, and immobilization. Second-line therapies include calcitonin to increase renal calcium excretion, and glucocorticoids to diminish intestinal calcium absorption.

Saline hydration

Most patients with emergent hypercalcemia are dehydrated due to anorexia and polyuria. Intravascular volume should be aggressively restored with intravenous normal saline, with an initial bolus of 500 to 1000 mL, followed by maintenance fluids at a rate of 200 mL/h or more, depending on the patient’s renal function and cardiac reserve (Table 240-3). Typically, patients require 3 to 4 L for rehydration in the first 24 hours. Patients need careful monitoring of fluid intake and output to prevent fluid overload. Normal saline dilutes serum calcium, and facilitates calciuresis by increasing glomerular filtration rate and the amount of filtered calcium, and decreasing tubular calcium reabsorption. Administration of furosemide or other loop diuretics to further promote calcium excretion may be considered after intravascular volume is restored. However, the use of loop diuretics to treat hypercalcemia has not been studied in randomized controlled trials, and may not be superior to vigorous use of saline alone. Thiazide diuretics should be avoided, as they enhance calcium reabsorption.

Bisphosphonates

The major target of medical management in severe hypercalcemia is osteoclast-mediated bone resorption. First-line therapy is an intravenous bisphosphonate, such as pamidronate or zoledronic acid. Pamidronate is administered in a dosage of 30 to 90 mg intravenously over several hours. Serum calcium levels should decline in 24 to 48 hours, although the maximal effect may not be evident for several days. Zoledronic acid is given at a dosage of 4 mg intravenously, over no less than 15 minutes. It appears to have a greater potency and a longer duration of action than pamidronate. The need for repeat treatment with either pamidronate or zoledronic acid depends on the aggressiveness of the underlying malignancy. The first dose of intravenous bisphophonates may be associated with fever, headache, arthralgias, and myalgias. Intravenous bisphosphonates should be used with caution in renal dysfunction. Dose reduction of zoledronic acid is recommended for creatinine clearance below 60 mL/min, and use in patients with creatinine clearance below 30 mL/min is not recommended. Pamidronate may be used with caution in patients with renal insufficiency, but the dose should be infused slowly, over 4 to 6 hours. The newer bisphophonate ibandronate may be associated with a lower risk of nephrotoxicity than other intravenous agents.

Denosumab

Denosumab is a RANK ligand inhibitor that interferes with osteoclast development and maturation. For hypercalcemia of malignancy, 120 mg subcutaneously is administered every 4 weeks, with additional 120 mg doses on days 8 and 15 of the first month of therapy. Common side effects include nausea and dyspnea. Denosumab is associated with osteonecrosis of the jaw, so a dental exam should be performed prior to therapy, and invasive dental procedures should be avoided during therapy. Atypical femur fractures occur rarely with denosumab.

Other approaches to emergent hypercalcemia

Intravenous bisphosphonates do not act immediately. If serum calcium needs to be reduced quickly, combined subcutaneous calcitonin (4 IU/kg every 12 hours) and intravenous bisphosphonate has become popular. Although rather weak, calcitonin acts rapidly, probably by facilitating urinary calcium excretion. The combination of a short-acting and long-acting anticalcemic can be very effective. In severe or refractory cases, hemodialysis against a low-calcium bath may be employed. Plicamycin and gallium nitrate are treatments of largely historical interest, either because of toxicity (plicamycin) or ineffectiveness (gallium nitrate).

Glucocorticoids

In myeloma, vitamin D intoxication, or disorders associated with ectopic production of 1,25-dihydroxyvitamin D, such as sarcoidosis and lymphoma, glucocorticoids can be very effective. Glucocorticoids impair vitamin D action, inhibit intestinal calcium absorption, and may have a direct antitumor effect.

Addressing the underlying disorder

Successful management of acute hypercalcemia also requires treating the underlying etiology. When primary hyperparathyroidism is the cause, parathyroid surgery is indicated when the patient is stable enough to undergo the procedure. In malignancy-associated hypercalcemia, surgery, radiotherapy or chemotherapy may be appropriate. However, because hypercalcemia is often an end-stage complication of malignancy, such interventions may not be warranted.

DISCHARGE CHECKLIST: HYPERCALCEMIA

• Has outpatient follow-up been arranged, with short-term repeat measurements of calcium, creatinine, and other electrolytes?

• Is there a long-range plan to prevent recurrent hypercalcemia, such as repeat bisphosphonate dosing?

• Have patients been instructed to seek prompt care if recurrent symptoms of hypercalcemia develop, such as nausea, vomiting, malaise, and polyuria?

• For patients with a new diagnosis of hypercalcemia of malignancy, have they been educated as to their underlying condition? Has outpatient oncology follow-up been arranged?

• If hypercalcemia has arisen in the setting of advanced malignancy with poor prognosis, has hospice therapy been considered?

HYPOCALCEMIA

Hypocalcemia is a serum calcium level which is below normal after correction for the albumin concentration. As with hypercalcemia, a free (ionized) calcium determination on a correctly collected sample can be useful to confirm hypocalcemia.

PRESENTING SYMPTOMS AND HISTORY

Chronic hypocalcemia, unless severe, is usually asymptomatic. Signs and symptoms become more likely when albumin-adjusted serum calcium levels fall below 7.5 to 8 mg/dL. These include numbness, paresthesias, and muscle spasms, and in severe cases, seizures and carpal, pedal, or laryngeal spasm.

Important historical features include low dietary calcium and vitamin D intake, minimal sun exposure, gastrointestinal tract disease that may reduce vitamin D absorption, such as chronic pancreatitis, celiac disease, and inflammatory bowel disease, and alcohol, which decreases parathyroid hormone secretion both directly, and also indirectly, by causing magnesium depletion. A family history of hypocalcemia suggests a genetic cause of hypoparathyroidism or an inherited abnormality of vitamin D metabolism. Prior neck surgery or neck irradiation may lead to hypoparathyroidism. A history of adrenal insufficiency and mucocutaneous candidiasis suggests autoimmune polyendocrine syndrome type 1. Acute pancreatitis, rhabdomyolysis, and tumor lysis may lead to tissue precipitation of calcium, and massive blood transfusion may lead to intravascular precipitation of calcium with citrate.

PHYSICAL FINDINGS AND DIAGNOSTIC TESTING

The physical examination is not sensitive for hypocalcemia. There may be a positive Chvostek sign (ipsilateral contraction of the facial muscles, induced by tapping on the facial nerve at a point about 1 cm below the zygomatic arch and 2 cm anterior to the earlobe). Trousseau sign may be elicited by inflating a blood pressure cuff on the arm above systolic pressure for 3 minutes. It is considered positive if carpopedal spasm develops, with flexion of the wrist, metacarpophalangeal joints, and thumb, and hyperextension of the fingers (Figure 240-4). Patients with these signs of neuromuscular irritability are at risk of frank tetany or seizures. QT_c prolongation may be evident on the electrocardiogram.

The serum calcium must be adjusted for albumin, as above, and a low value confirmed with a measurement of serum ionized calcium. Levels of phosphate, magnesium, creatinine, PTH, and 25-­hydroxyvitamin D should be determined. 24-hour urinary calcium may be occasionally helpful. It is low in hypoparathyroidism and vitamin D deficiency, and high in patients with familial hypocalcemia with hypercalciuria, due to activating mutations in the calcium-sensing receptor. Genetic testing is available for some inherited disorders leading to hypocalcemia.

DIFFERENTIAL DIAGNOSIS

After confirming that calcium levels are truly low, exclusion of hypoparathyroidism by checking the level of PTH is central to the diagnostic workup of hypocalcemia (Figure 240-5). The most common causes of hypoparathyroidism are previous thyroid, parathyroid, or other neck surgery, and autoimmune destruction (Table 240-4). Autoimmune damage to the parathyroid glands may occur in isolated fashion, or in connection with failure of other endocrine glands, such as premature ovarian failure, hypothyroidism, and Addison disease, and mucocutaneous candidiasis. Infiltration of the parathyroid glands, as may occur in hemochromatosis, Wilson disease, and metastatic cancer, can lead to hypoparathyroidism. Congenital absence of the parathyroid glands may be seen in DiGeorge syndrome. Functional hypoparathyroidism may result from severe hypomagnesemia, because magnesium is necessary for both PTH release and PTH action. This is commonly seen in hospitalized alcoholic patients who are often markedly hypomagnesemic.

The other major category of hypocalcemia includes conditions in which the parathyroid glands respond appropriately to hypocalcemia, and PTH is elevated. In vitamin D deficiency, serum calcium concentrations may be low or in the low-normal range, because of compensatory increases in PTH, with secondary mobilization of skeletal calcium and reduction in renal calcium excretion. Clinical manifestations of vitamin D deficiency include osteomalacia, pathologic fractures, falls, and muscle weakness. Vitamin D deficiency has also been linked to autoimmune disease, cancer, and cardiovascular disease. Dietary vitamin D deficiency in the elderly is common, but often overlooked. Other adults at risk for vitamin D deficiency include darker complexion and low sun exposure. Recent reports suggest that vitamin D deficiency may be widely prevalent.

Other causes of hypocalcemia include pseudohypoparathyroidism, a genetic disorder of PTH resistance associated with elevated parathyroid hormone levels, moon facies, short stature, and short fourth metacarpals. In acute pancreatitis, fatty acids released through the action of pancreatic enzymes complex with calcium. Hypocalcemia due to the formation of calcium phosphate complexes occurs in severe hyperphosphatemic states, such as renal failure, rhabdomyolysis, and tumor lysis. Hypocalcemia may also be seen in patients given multiple red blood cell transfusions containing calcium chelators to prevent clotting. In patients with critical illness, hypocalcemia is probably multifactorial, arising as a consequence of the release of procalcitonin and other acute phase reactants, hypoalbuminemia, hypomagnesemia, and blunted PTH secretion.

TREATMENT OF HYPOCALCEMIA

When hypocalcemia is severe and symptomatic, calcium gluconate should be administered by slow intravenous infusion. A typical ­calcium infusion is prepared with 10 ampules (100 mL) of 10% calcium gluconate (93 mg elemental calcium/ampule) in 1 L of D₅W, administered at 50 mL/h. For an average-sized person, this is equivalent to 15 mg calcium/kg of body weight. Serum calcium should be tested frequently, and the rate of infusion adjusted to maintain calcium levels in the low-normal range. Deficiencies in magnesium or vitamin D should also be corrected. In severe cases, hypocalcemia may recur quickly after discontinuation of the calcium infusion, so oral calcium should be administered concurrent with tapering the infusion.

In mild or moderate hypocalcemia, patients may be given oral calcium carbonate or calcium citrate, starting at 1000 to 1500 mg of elemental calcium daily in divided doses with meals. Patients should be instructed to take calcium with a protein meal, particularly patients with hypochlorhydria or achlorhydria, or who are on proton pump inhibitors. The protein meal supplies the acid that may be missing in these individuals, and which is required for the absorption of calcium carbonate. Calcium citrate does not require acid for absorption. If appropriate, vitamin D, in the form of cholecalciferol (vitamin D₃) or ergocalciferol (vitamin D₂), should be provided. Recommended daily intake for vitamin D is currently being revised upward. The official recommendation for adults 50 years and older, 400 to 600 IU/d, is acknowledged by most experts to be inadequate. A popular approach to normalizing vitamin D levels in deficient individuals is to provide a weekly capsule of 50,000 IU for 8 to 12 weeks. This approach requires a prescription for ergocalciferol, because cholecalciferol is currently unavailable by prescription in the United States in this form.

Chronic hypoparathyroidism may require long-term administration of vitamin D and 1,25-dihydroxyvitamin D as needed to maintain normal levels. The goal of therapy is to maintain the serum calcium at a level at which the patient is asymptomatic. To avoid hypercalciuria, serum calcium levels are often maintained in the lower range of normal. Periodic monitoring for hypercalciuria and nephrocalcinosis in these patients may be appropriate. Hypoparathyroidism is the last classic endocrine deficiency disease for which the missing hormone is not available as an approved therapy. However, recent studies have shown promise in the use of recombinant parathyroid hormone (teriparatide) for hypoparathyroidism.

DISCHARGE CHECKLIST: HYPOCALCEMIA

• Have repeat measurements of calcium, creatinine, and other electrolytes been arranged within 1 to 2 weeks of hospital discharge?

• Has outpatient endocrinology follow-up been arranged for patients with hypoparathyroidism?

HYPOPHOSPHATEMIA

Hypophosphatemia is frequent in hospitalized patients. It is most often caused by alcoholism, malnutrition, eating disorders, diabetic ketoacidosis, and refeeding of malnourished patients. It is also seen in burns, sepsis, trauma, in severe respiratory alkalosis, and as a complication of treatment with diuretics, some bisphosphonates, the antiretroviral drug tenofovir, sucralfate, and aluminum hydroxide-containing antacids. In primary hyperparathyroidism, the serum phosphate has a tendency to be in the low-normal range, but frank hypophosphatemia is uncommon.

Symptoms of hypophosphatemia are unusual if the serum phosphate is > 2.0 mg/dL. Subjects with mild to moderate hypophosphatemia (serum phosphate between 1.5-2.0 mg/dL) may display muscle weakness, nausea, and vomiting, and anorexia. Those with severe hypophosphosphatemia (serum phosphate <1.5 mg/dL) may be at risk for rhabdomyolysis, hemolytic anemia, impaired leukocyte and platelet function, impaired oxygenation of tissues, confusion, seizures, and coma.

Oral phosphate repletion is sufficient for most patients, although amounts exceeding 500 mg/d may be associated with diarrhea. Hypomagnesemia and hypokalemia should be corrected, if present. Patients with severe hypophosphatemia may require intravenous repletion, but this should be done only under by experts and according to institutional protocols.

HYPERPHOSPHATEMIA

Most patients with hyperphosphatemia have diminished renal excretion of phosphate, usually due to acute or chronic kidney disease. Hypoparathyroidism and pseudohypoparathyroidism are also classically associated with hyperphosphatemia. It can also be seen in acromegaly. Less often, hyperphosphatemia results from transcellular phosphate shifts, as in diabetic ketoacidosis (even despite total body phosphate depletion), or cellular injury, such as rhabdomyolysis, trauma, and tumor lysis syndrome. Patients may be asymptomatic, although symptoms of concomitant hypocalcemia and other electrolyte disturbances are often present. Vascular and soft tissue calcification is common in hyperphosphatemic patients with chronic kidney disease, particularly if the calcium × phosphate product exceeds 55. Hyperphosphatemia is treated with phosphate binders and dietary phosphate restriction, as discussed in greater detail in Chapter 245.

SUGGESTED READINGS

INTRODUCTION

Potassium and magnesium disorders are common in the hospital setting. About 12% of hospitalized patients have hypokalemia, 3% have hyperkalemia, and 11% have hypomagnesemia. Potassium disorders are a particular challenge for hospitalists, as the first clinical manifestation of a severe potassium abnormality may be a cardiac arrhythmia.

Potassium (K⁺) is the most abundant intracellular cation. In a 70 kg adult, the total body K⁺ is approximately 3500 mmol (50 mmol/kg). About 98% is located in the intracellular compartment, with the majority localized in muscle (70%). Because only 2% of the total body K⁺ is in the extracellular compartment, laboratory results correlate poorly with total body K⁺. Normal levels of extracellular potassium are between 3.5 and 5.0 mmol/L, while intracellular K⁺ levels range between 140 and 150 mmol/L.

Similarly, plasma magnesium (Mg²⁺) accounts for only 1% of the total body magnesium stores, and plasma Mg²⁺ correlates poorly with total body magnesium. Plasma magnesium is normally 1.7 to 2.3 mg/dL (0.70-0.95 mmol/L). Magnesium abnormalities can also have major clinical consequences, such as torsades de pointes. Hypokalemia is often accompanied by hypomagnesemia, which renders potassium repletion less effective.

POTASSIUM BALANCE

Dietary K⁺ is excreted mostly in the urine (90%), with a minor component of fecal excretion (10%). In severe chronic kidney disease, stool losses may exceed 25% to 50% of dietary intake. Under normal conditions, the kidney can adjust potassium excretion over a wide range depending on K⁺ intake. The recommended daily K⁺ intake for patients with normal renal function is 4700 mg (120 mmol). If K⁺ control mechanisms are intact, an increased daily intake of 400 mmol can be tolerated with < 1 mmol/L increase in plasma K⁺ values. Daily K⁺ intake for patients on hemodialysis or with severe chronic kidney disease is typically 2000 mg (51 mmol) per day. Conversely, many patients on peritoneal dialysis can tolerate a normal daily intake of K⁺ due to losses of potassium in the urine and peritoneal fluid.

HYPERKALEMIA

ETIOLOGY

Hyperkalemia is defined as a plasma K⁺ > 5.0 mmol/L. Hyperkalemia can result from redistribution, reduced K⁺ excretion, or increased K⁺ intake. Medications may lead to hyperkalemia by affecting either potassium redistribution or excretion (Table 241-1).

Pseudohyperkalemia is a laboratory artifact due to potassium release from cells prior to laboratory measurement. It should be suspected when hyperkalemia is reported in the setting of red blood cell hemolysis, leukocytosis (> 70,000/mm³) or thrombocytosis (> 500,000/mm³), and no ECG changes are present. It has become less common with the increased use of plasma for potassium measurement, rather than serum. Plasma is the supernatant collected from heparinized whole blood, whereas serum is the supernatant remaining after centrifugation of a clotted whole blood sample. Serum K⁺ is normally 0.3 mmol/L above the plasma K⁺, but may be higher if K⁺ is released from the clot formed in the tube. If pseudohyperkalemia is suspected, the plasma potassium should be measured instead of the serum potassium. ­Pseudohyperkalemia can also occur due to potassium release from muscle cells following prolonged constriction with a tourniquet or limb exercise while a tourniquet is in place.

Redistribution of potassium

Redistribution of potassium from the intracellular to the extracellular space can occur in severe acidosis due to nonorganic acid metabolic acidosis, hyperosmolar states, tissue breakdown, and hyperkalemic periodic paralysis. Organic acids such as lactic acid and ketoacids are less likely to cause hyperkalemia than non-organic acids. These organic acids have greater transmembrane mobility, allowing movement into cells with H⁺, rather than movement of K⁺ out of cells in exchange for H⁺. Hyperkalemia in diabetic ketoacidosis usually results from insulin deficiency and hyperosmolality, rather than acidosis. Hyperglycemia increases extracellular osmolality, drawing water from cells down the osmotic gradient. Potassium follows the water movement (solvent drag), and hyperkalemia results.

Tissue breakdown may liberate large amounts of intracellular potassium, resulting in rapid, life-threatening increases in extracellular potassium. Rhabdomyolysis, tissue necrosis, tumor lysis with chemotherapy, and large hematomas are common causes. In rhabdomyolysis, hypokalemia may precede hyperkalemia, and contribute to muscle breakdown by causing vasoconstriction and decreased blood flow to the involved muscle.

Hyperkalemic periodic paralysis is an autosomal dominant disorder involving the muscle cell sodium channel. During these episodes, potassium moves from the intracellular to extracellular space, accompanied by movement of sodium and water into the cell. The hyperkalemia is accompanied by transient weakness or paralysis.

Decreased potassium excretion

Potassium excretion in the distal nephron depends upon adequate urine flow, aldosterone, and activity of the basolateral Na⁺/K⁺-ATPase. The reabsorption of luminal Na⁺ through aldosterone-sensitive epithelial sodium channels (ENaCs) creates an electrochemical gradient for K⁺ excretion in the urine. Hyperkalemia can occur in aldosterone-resistant or deficient states by attenuating the electrochemical gradient for K⁺ secretion. Acquired mineralocorticoid resistance from diabetes mellitus, obstructive uropathy, chronic tubulointerstitial disease, sickle cell anemia, lupus nephritis, and medications, as well as decreased mineralocorticoid production (including medication induced), are commonly associated with type IV renal tubular acidosis (RTA). Type IV RTA is characterized by hyperkalemia, normal anion gap hyperchloremic acidosis, with serum HCO₃ values ranging between 16 and 22 mmol/L, and decreased urinary NH₄⁺ production, causing a positive urine anion gap [Urine (Na⁺) + (K⁺) – (Cl^–)]. In acute and chronic tubular injury, a decreased responsiveness to aldosterone can contribute to hyperkalemia. For acute kidney injury associated with prerenal azotemia or hypovolemic states, decreased urinary flow to the distal tubule and a reduction in ­sodium-potassium exchange can contribute to hyperkalemia.

Excessive potassium intake

Excessive intake of K⁺ rarely causes hyperkalemia in the setting of normal renal function. Renal secretion of potassium is typically adequate at glomerular filtration rates (GFR) above 20 to 30 mL/min/1.73 m². Patients with end-stage renal disease (ESRD) on hemodialysis usually tolerate a daily K⁺ intake of 2000 mg (51 mEq). Upregulated gut potassium excretion and shifts in transcellular K⁺ prevent hyperkalemia between dialysis sessions. The loss of these mechanisms may result in the rapid development of hyperkalemia, especially with large exogenous loads of potassium, such as massive blood transfusions. Irradiation of blood and increased age of the blood increase the amount of free potassium that is released during the blood transfusion. Seven-day-old blood has approximately 23 mmol/L of K⁺, while 42-day-old blood has approximately 50 mmol/L of K⁺.

SIGNS AND SYMPTOMS

Clinical effects of hyperkalemia relate to altered membrane excitability due to changes in the transcellular potassium gradient. Severe hyperkalemia leads to cardiac arrhythmias and conduction abnormalities. It may also cause weakness of the lower extremities, progressing superiorly to cause flaccid paralysis and respiratory failure. This presentation may mimic Guillain-Barré syndrome, but is easily differentiated by the response to potassium correction. Hyperkalemia may also contribute to metabolic acidosis by interfering with renal ammonium excretion.

EVALUATION OF HYPERKALEMIA IN THE HOSPITALIZED PATIENT

In the hospital setting, the initial evaluation of hyperkalemia includes monitoring for life-threatening arrhythmias, checking for pseudohyperkalemia, eliminating exogenous sources of potassium, evaluating renal function, and evaluating for rapid transcellular shifts of potassium (Figure 241-1).

ECG changes

Even mildly elevated K⁺ levels may be associated with ECG changes, especially in the setting of rapid rises in plasma K⁺ values. If ECG changes from hyperkalemia are found or if the plasma K⁺ value is ≥ 7.0, continuous telemetry is indicated and should continue until the plasma K⁺ value is ≤ 5.8 and there are no hyperkalemic ECG changes. Frequent chemistry checks are indicated for plasma K⁺ greater than 5.8 or for hyperkalemia associated with ECG changes, severe hyperglycemia, or any clinical condition which predisposes the patient to rapid changes in plasma K⁺.

Plasma potassium levels do not correlate with specific ECG changes. A symmetric increase in T wave height may be seen initially. Hyperkalemia-induced peaked T waves may be difficult to distinguish from the hyperacute T waves of myocardial injury. As potassium levels rise further, flattening of the P waves, prolongation of the PR interval, and prolonged QRS duration occur. Eventually, atrial standstill and a sine wave ECG pattern may be seen (Figure 241-2).

Urine potassium, transtubular potassium gradient, and urine potassium/creatinine

A 24-hour urine K⁺ measurement can help differentiate between renal and nonrenal causes of hyperkalemia. With hyperkalemia, urinary excretion of K⁺ should exceed 40 mmol/d. Alternatively, the transtubular potassium gradient (TTKG), a measurement of potassium secretion by the distal nephron corrected for urine osmolality, has been used:

K_U and K_S are the concentrations of K⁺ in the urine and serum, and S_Osm and U_Osm are the osmolalities of the serum and urine, respectively. Plasma potassium and osmolality can be used instead of serum values to estimate TTKG with minimal effect on clinical interpretation. The accuracy of the TTKG has been called into question by recent studies of urea and potassium handling in the renal tubule. Previous studies using TTKG have suggested that patients with normal renal function and normal potassium intake have a TTKG of 8 to 9. In hyperkalemia, a low TTKG (< 5-7) suggests an inappropriately low secretion of potassium. A high TTKG with hyperkalemia suggests normal aldosterone action and an extrarenal cause of hyperkalemia, except in cases of volume depletion where aldosterone secretion is enhanced with a TTKG > 7, but total renal potassium excretion is limited by low urine flow.

When the TTKG is inappropriately low in the setting of hyperkalemia, an increase in the TTKG to >10 hours after the administration of 0.05 mg of fludrocortisone suggests hypoaldosteronism. If fludrocortisone has no effect on the TTKG, drug-induced or intrinsic renal resistance to aldosterone are likely.

If further studies cast doubt on the usefulness of TTKG, the urine potassium/creatinine ratio will likely be used more frequently when a 24-hour urine potassium is not available. It has been suggested that a patient with hyperkalemia and a normal renal response should have a spot urine ratio of >200 mmol K⁺/g creatinine (=22.6 mmol K⁺/mmol creatinine). For a typical patient with a daily creatinine excretion of over 1 g/d, this is significantly more than the 24-hour urinary K⁺ excretion of >40 mmol used by some clinicians as a cutoff for adequate renal potassium clearance in the setting of hyperkalemia. This further underscores the point that it is difficult to define an exact cutoff for expected renal potassium excretion or urine K⁺/Cr in the face of hyperkalemia without additional investigation.

TREATMENT OF HYPERKALEMIA

Treatments exist for hyperkalemia which cause net excretion or removal of potassium, such as gastrointestinal resins and laxatives or hemodialysis. As these therapies may not act immediately or may involve logistical difficulties, they are often used in conjunction with short-acting temporizing measures, such as cardiac membrane stabilization with calcium, and agents such as insulin that cause transcellular potassium shifts (Table 241-2).

Cardiac membrane stabilization

Intravenous calcium stabilizes the cardiac membrane by inhibiting membrane depolarization. It is the most important initial treatment for severe hyperkalemia. Two forms of calcium are commonly available: calcium gluconate and calcium chloride. Calcium gluconate is preferred because it can be administered through a peripheral intravenous line, whereas calcium chloride requires a central venous line to prevent tissue necrosis. Tissue necrosis can occur if calcium chloride leaks from the venous access into the surrounding tissue. A 10 mL ampule of calcium gluconate contains 90 mg (2.3 mmol) of elemental calcium, and a 10 mL ampule of calcium chloride contains 272 mg (7.0 mmol) of elemental calcium.

The initial dose of calcium gluconate is 10 mL of 10% solution infused over 2 to 3 minutes. An equivalent amount of elemental calcium is contained in 3.3 mL of 10% calcium chloride. The onset of action is 1 to 3 minutes, and duration of action is 30 to 60 minutes. Calcium cannot be mixed with bicarbonate solutions, because precipitation of CaCO₃ occurs.

Administration of intravenous calcium to patients taking digoxin requires extreme caution, as calcium has been shown to potentiate the effects of digoxin toxicity in animal models, especially at very rapid infusion rates. The risk of calcium administration in this setting can be reduced by infusing the calcium gluconate over 20 to 30 minutes. Another option is to administer digoxin immune fab to neutralize the effect of the digoxin.

Potassium redistribution

Insulin is the most reliable means of inducing transcellular potassium shifts. The usual dose is 10 units of regular insulin intravenously, followed immediately by 50 mL of 50% dextrose (25 g of dextrose). For a blood glucose > 250, insulin can be administered alone with close glucose monitoring. The effect begins in 10 to 20 minutes, peaks at 30 to 60 minutes, and lasts 4 to 6 hours. Potassium levels typically drop by 0.5 to 1.2 mmol/L. An infusion of 10 units of regular insulin can also be administered over 1 hour in 10% dextrose.

Beta-2 agonists have an additive effect with insulin in transiently reducing plasma potassium by redistribution. High doses of nebulized albuterol are used, typically 10 to 20 mg of nebulized albuterol in 4 mL of normal saline over 10 minutes. Plasma potassium levels usually fall by 0.5 to 1 mmol/L. The effect begins in 30 minutes, peaks at 90 minutes, and lasts 2 to 4 hours. Intravenous albuterol has also been used, but it is not available in the United States. As some patients, including those with renal failure, have a reduced response to albuterol, it should not be the only agent used. Caution should be exercised for individuals at risk for side effects such as cardiac ischemia from the resulting increase in heart rate.

Sodium bicarbonate (NaHCO₃) does not reliably lead to the redistribution of potassium, and it should not be considered as first-line therapy for hyperkalemia. This is especially true in high anion gap acidosis, where hyperkalemia is usually not a direct consequence of the presence of organic acids. If intravenous or oral NaHCO₃ is used to treat metabolic acidosis caused by nonorganic anions, which is usually associated with a normal anion gap, plasma potassium may fall, but it is not preferred treatment even in this setting. Oral NaHCO₃ is useful for chronic treatment of type IV renal tubular acidosis.

Potassium removal

Sodium polystyrene sulfonate (Kayexalate) exchanges Na⁺ for K⁺ in the gastrointestinal tract. When taken orally, sorbitol has been typically added to the resin to speed passage through the gastrointestinal tract. However, an FDA warning has been issued about the risk of colonic necrosis when kayexalate is used with sorbitol, and thus the powdered form of kayexalate not premixed with sorbitol is preferred. If Kayexalate with sorbitol is the only form available, diluting with water is appropriate. Each gram of resin binds 0.5 to 1.2 mmol of K⁺. Oral doses of 15 to 60 g are typical. Sodium polystyrene sulfonate can also be administered rectally as a retention enema, at a dose of 30 to 60 g in 250 mL of water every 6 hours. The solution should be introduced by gravity, flushed with an additional 50 to 100 mL of nonsodium containing fluid, retained 30 to 60 minutes or longer, and cleansed with 250 to 1000 mL of nonsodium containing solution at body temperature. Sorbitol should not be used rectally due to risk of colonic necrosis. Oral or rectal Kayexalate should not be used in postoperative patients for the same reason.

Favorable data have been published for two novel potassium reduction agents. Sodium zirconium cyclosilicate (ZS-9) is a highly selective potassium trap, whereas patiromer is a nonabsorbed polymer which exchanges calcium for potassium, primarily in the distal colon. Patiromer is FDA approved, ZS-9 is under FDA review, and further studies are needed to evaluate their suitability for acute potassium reduction. Initial studies have focused on their use for more chronic potassium reduction, which may allow broader use of potassium-sparing medications.

Dialysis is required for treatment of refractory hyperkalemia. Intermittent hemodialysis removes potassium most rapidly. Continuous renal replacement therapy is an option for patients who have ongoing causes of severe hyperkalemia, such as tissue necrosis or rhabdomyolysis. Peritoneal dialysis provides gradual removal of potassium. Although peritoneal dialysis is rarely used in developed countries when the other modalities are available, peritoneal dialysis is an established therapy for acute kidney injury. Since electricity is not required for its use and access is placed in the peritoneal cavity rather than in large veins, it is an option for hyperkalemia treatment in a variety of situations such as difficulty with placing vascular access for hemodialysis, disasters associated with overwhelming caseloads, or power failure

Diuretics are unreliable for the acute treatment of hyperkalemia in the setting of compromised renal function. In patients with adequate renal function, the combination of a loop and thiazide diuretic is more effective for potassium removal than either alone. Of the loop diuretics, torsemide and bumetanide have higher bioavailability than furosemide.

Potassium intake reduction

The typical daily potassium intake for a patient with end-stage renal disease is 2000 mg (51 mmol) of potassium per day, but patients with acute kidney injury may require even more stringent potassium restriction.

HYPOKALEMIA

ETIOLOGY

Hypokalemia (K⁺ < 3.5 mmol/L) can result from redistribution, increased potassium excretion from renal and nonrenal sources, or decreased potassium intake. Medications, especially loop and thiazide diuretics, frequently cause hypokalemia. Hypokalemia is common with tubular toxins such as amphotericin B and cisplatin. High doses of penicillin or semisynthetic penicillins such as ticarcillin and carbenicillin occasionally cause renal potassium wasting in the distal nephron due to a transient anion effect. Inhalation of toluene (glue sniffing) may cause distal renal tubular acidosis with associated hypokalemia.

Pseudohypokalemia

Falsely low serum or plasma potassium values may occur in conditions such as acute leukemia due to time-dependent uptake of potassium by the increased abnormal white cell mass. Rapid analysis of the sample, or storing the sample at 4°C prior to analysis, can prevent the potassium uptake and confirm the diagnosis of pseudohypokalemia.

Redistribution

Insulin directly stimulates potassium entry into cells by increasing the activity of the Na⁺/K⁺-ATPase pump. This mechanism is independent of stimulation of cellular glucose entry. ­Beta-2-adrenergic activators stimulate Na⁺/K⁺-ATPase mediated cellular potassium uptake by a slightly different cellular mechanism. Thus, insulin and ­beta-2-adrenergic activation may act synergistically to cause hypokalemia. Aldosterone stimulates direct cellular uptake and redistribution due to increased Na⁺/K⁺-ATPase activity. It also acts in the distal renal tubule to enhance potassium excretion.

Thyrotoxic periodic paralysis causes severe hypokalemia by redistribution, in conjunction with extremity and limb girdle weakness, hypophosphatemia, and hypomagnesemia. It is more common in patients of Asian and Hispanic origin. Attacks often occur during rest after vigorous physical activity, and can also be precipitated by carbohydrate-rich meals. Other rare genetic forms of hypokalemic periodic paralysis also exist.

Nonrenal potassium losses

Common nonrenal causes of hypokalemia include intestinal loss of potassium from diarrhea, celiac disease, ileostomy, and chronic laxative abuse. Potassium loss from vomiting and nasogastric suctioning can result in hypokalemia, although renal potassium losses from aldosterone activation may be more important in this setting. ­Potassium losses through the skin are usually low, except in the setting of extreme physical exertion. Severe burns may lead to hypokalemia by multiple mechanisms.

Renal potassium loss

Aldosterone-producing adenomas (Conn syndrome) cause hypertension and hypokalemia by stimulation of aldosterone receptors in the distal renal tubule. Cortisol also activates the aldosterone receptor. Normally, cortisol is converted to cortisone by ­11-beta-hydroxysteroid dehydrogenase-2 (11βHSD-2) before it can reach the aldosterone receptor. Very high cortisol levels, as seen in Cushing syndrome, overwhelm the ability of 11βHSD-2 to degrade cortisol to cortisone, and the nondegraded cortisol activates the aldosterone receptor in the distal tubule and precipitates hypokalemia. Similarly, glycyrrhizinic acid, a component of black licorice that is sometimes added to chewing tobacco, inhibits 11βHSD-2, preventing cortisol degradation and causing hypertension and hypokalemia. With vomiting and nasogastric suctioning, potassium wasting in the urine occurs through aldosterone secretion in the setting of volume depletion, as noted above.

Hypokalemia can be associated with an RTA, due to disease of either the proximal (type II RTA) or distal (type I RTA) renal tubules. Either can result from autoimmune, genetic, endocrine, medication-induced, toxin-induced, or idiopathic causes. Rare causes of hypokalemia associated with metabolic alkalosis involve defects in the thick ascending limb of Henle transport proteins (Bartter syndrome), mutations in the thiazide-sensitive Na⁺/Cl^– cotransporter in the distal convoluted tubule (Gitelman syndrome), or increased activation of the epithelial sodium channel (ENaC) in the distal tubule (Liddle syndrome).

Decreased potassium intake

Decreased potassium intake is rarely a cause of hypokalemia, except in severe malnutrition. For patients with normal renal function who consume the recommended 4700 mg (120 mmol) of potassium per day, 90% (108 mmol) is excreted in the urine. When intake is sharply reduced, urinary potassium loss decreases to < 15 to 20 mmol/d in order to conserve potassium.

SIGNS AND SYMPTOMS

Individuals with serum potassium levels between 3.0 and 3.5 mEq/L are often asymptomatic. However, there is a risk of cardiac arrhythmias for individuals with predisposing conditions such as coronary artery disease. At serum potassium levels between 2.5 and 3.0 mEq/L, patients report generalized weakness and constipation. When serum potassium levels drop below 2.5 mEq/L, there is an increased risk of muscle necrosis and rhabdomyolysis. At serum potassium levels less than 2.0 mEq/L, ascending paralysis and respiratory failure can occur.

EVALUATION OF HYPOKALEMIA

An algorithm for the diagnosis of hypokalemia is presented in Figure 241-3. Pseudohypokalemia and transcellular shifts should be excluded. Magnesium levels should be measured early in the workup of hypokalemia. Many disorders, such as diarrhea and excess diuresis, deplete both potassium and magnesium. Moreover, hypomagnesemia may lead to renal potassium wasting via potassium channels in the distal tubule, and make hypokalemia more refractory to treatment.

The next step is determining whether hypokalemia arose from a renal or nonrenal source. An obvious cause may be apparent, such as profuse diarrhea or escalating doses of a loop or thiazide diuretic. When the cause of hypokalemia is unclear, a 24-hour urine potassium measurement may differentiate between renal and nonrenal losses. In the setting of hypokalemia, urinary K⁺ losses should fall to less than 15 to 20 mmol/d. Potassium excretion above this level suggests a renal contribution to hypokalemia.

When a 24-hour urine K⁺ is not available, a spot urine ratio of 22 mmol K⁺/g creatinine (=2.5 mmol K⁺/mmol creatinine) marks the cutoff between hypokalemia secondary to intracellular shifts (if < 22 mmol K⁺/g Cr) and renal loss (if >22 mmol K⁺/g creatinine).

The TTKG (see above) has been used to help establish the cause of hypokalemia, but the use of TTKG has been cast into doubt by recent studies of urea and K⁺ handling in the renal tubule. An inappropriately high TTKG (> 4) in hypokalemia has been interpreted as suggesting an increased distal potassium secretion and renal potassium loss. A low TTKG can occur with nonrenal potassium wasting, with urinary potassium losses from osmotic diuresis, with hypokalemia secondary to diuretics which were discontinued at the time of TTKG measurement, or with hypokalemia associated with K⁺ shifts.

Patients with renal potassium wasting should be further classified by acid-base status. Patients with acidosis may have RTA, diabetic ketoacidosis, or tubular dysfunction from drugs such as amphotericin B or acetazolamide. Patients with alkalosis and hypertension may have mineralocorticoid excess or Liddle syndrome. Hypokalemia with alkalosis and normal or low blood pressure may be caused by vomiting, diuretics, and Bartter or Gitelman syndrome. The spot urine chloride is a useful diagnostic tool in evaluating the etiology of hypokalemia in the setting of metabolic alkalosis, with a spot low urine chloride (< 10 mmol/L) suggesting volume depletion and a chloride-responsive state.

ECG changes

ECG changes in hypokalemia are shown in Figure 241-2. U waves appear following the T waves, and become progressively more prominent in comparison to the T waves as potassium levels decrease. Ultimately, the U wave merges with the T wave, and the QT interval appears prolonged.

TREATMENT OF HYPOKALEMIA

Potassium repletion is the cornerstone of therapy for hypokalemia. As extracellular potassium comprises a fraction of the total body potassium store, relatively large amounts of potassium are required to correct the total body potassium deficit. A plasma K⁺ 1 mmol/L below normal corresponds to a total body potassium deficit of approximately 200 to 400 mmol, and a drop in plasma K⁺ to 2 mmol/L below normal requires 400 to 800 mmol for repletion. Typically, daily repletion is significantly less than the total body deficit as the time required for redistribution is prolonged. Underlying disorders such as metabolic alkalosis that are causing or perpetuating hypokalemia must also be addressed. ­Continuous telemetry and frequent electrolyte checks should be performed in severe hypokalemia, or in conditions in which the serum potassium may decline rapidly, such as diabetic ketoacidosis (DKA) and hyperosmolar hyperglycemic state (HHS).

Oral repletion

If there is no immediate threat to life, oral potassium can be used to treat hypokalemia. Generally, potassium chloride (KCl) is indicated for hypokalemia associated with diuretic use or volume depletion. A typical initial dose in a patient with normal renal function is 40 to 100 mmol (40-100 mEq) per day, in two to three divided doses. ­Liquid, wax matrix, and microencapsulated forms exist. Compliance is poor with the liquid form due to the strong taste. Although the wax matrix form is easier to swallow, it has been associated with erosions of the gastrointestinal tract. The microencapsulated formulation is associated with the fewest complications.

Other potassium preparations are available for different indications. Oral potassium phosphate is found in many foods, and is indicated for combined potassium and phosphorus depletion. Terminology may be misleading. For example, Neutra-Phos actually has more potassium than K-Phos. Potassium bicarbonate is useful for the treatment of both metabolic acidosis and hypokalemia, while potassium citrate may prevent renal stones.

Intravenous repletion

KCl is preferred for intravenous repletion. Potassium phosphate may be used for dual phosphorus and potassium depletion. Potassium can be infused through a peripheral intravenous line at a maximum rate of 10 mmol/h. Higher rates require a central venous line and continuous ECG monitoring. Infusion rates of 20 to 40 mmol/h are reserved for cases of life-threatening hypokalemia requiring emergent correction. One liter bags of IV fluids typically have a maximum of 60 mEq of K⁺ added in order to avoid infusing an excess amount of K⁺.

Intravenous potassium is a common cause of iatrogenic hyperkalemia. A typical intravenous dose with normal renal function is 20 to 40 mmol (20-40 mEq). Although 20 mmol of intravenous KCl might increase plasma K⁺ by 0.25 mmol/L, transcellular shifts make it difficult to predict the effect of therapy. Renal potassium clearance generally decreases significantly at a GFR below 20 to 30  mL/min/1.73 m², and requires reduction in potassium dosing and additional monitoring.

Hypokalemia in DKA and HHS

The use of insulin in DKA and HHS drives potassium into the intracellular space, and also decreases the hyperglycemia-induced osmolar driving force for movement of potassium from the intracellular to the extracellular space. Rapid and sometimes life-threatening potassium shifts may result. In 2009, the American Diabetes Association recommended that insulin should not be started until the serum potassium is known. If the serum potassium is < 3.3 mmol/L, insulin therapy is held until the potassium is repleted to 3.3 mmol/L or above. If the serum potassium is ≥ 3.3 mmol/L, insulin therapy can be initiated. For serum potassium ≥ 3.3 and < 5.2 mmol/L, potassium supplementation as 20 to 30 mEq K⁺ in each liter of IV fluid is given during insulin therapy. The goal is to achieve serum potassium values between 4 to 5 mmol/L. Potassium supplementation is initially held for serum potassium values > 5.2 mmol/L, but is often subsequently required, as total body stores of potassium are usually depleted and insulin plus intravenous fluid therapy eventually unmasks the total body potassium deficit.

Thyrotoxic hypokalemic periodic paralysis

Oral propranolol (3 mg/kg) is first-line treatment for thyrotoxic periodic paralysis because it rapidly reverses hypokalemia, hypophosphatemia, and hypomagnesemia, and is not associated with rebound hyperkalemia. Propranolol 1 mg IV pushed slowly every 10 minutes, up to a total of 3 mg IV, is an alternative regimen. Aggressive potassium repletion for this disorder has been associated with a 25% or greater incidence of hyperkalemia, so if oral or IV potassium is given, subsequent close serial monitoring of plasma potassium is warranted. Treatment to establish a euthyroid state is the long term priority to prevent future attacks.

MAGNESIUM BALANCE

A typical American diet contains 300 to 400 mg/d of elemental magnesium. Approximately 30% to 40% of dietary magnesium is absorbed in the gut. Additionally, 40 mg/d of magnesium is secreted in the small intestine, of which 20 mg/d is reabsorbed in the colon and rectum. Approximately 100 mg appears in the urine each day, which is 5% of the filtered load. Specific cutoffs for hypomagnesemia and hypermagnesemia are difficult to establish because of the poor correlation between extracellular concentration and total body stores. Plasma magnesium levels of 1.7 to 2.3 mg/dL (0.70-0.95 mmol/L) are considered normal, but a normal serum level may be present despite total body magnesium depletion.

HYPOMAGNESEMIA

ETIOLOGY OF HYPOMAGNESEMIA

Hypomagnesemia results from low dietary intake, gastrointestinal losses, and renal losses. Poor intake of magnesium is common in alcoholics and hospitalized patients receiving inadequate magnesium supplementation in parenteral nutrition or intravenous fluids. The causes of impaired gastrointestinal absorption include diarrhea, inflammatory bowel disease, laxative abuse, proton pump inhibitors, and small bowel resection.

Osmotic and loop diuretics provoke urinary losses of magnesium. Acutely, thiazide diuretics increase magnesium absorption in the distal convoluted tubule, but long-term use can reduce magnesium reabsorption and cause hypomagnesemia. Urinary magnesium wasting is also seen in alcoholics. Many nephrotoxic drugs, such as amphotericin B, aminoglycosides, cisplatin, foscarnet, and cyclosporine, interfere with magnesium reabsorption in the thick ascending limb or distal convoluted tubule and cause magnesium wasting. Rare familial disorders, such as Gitelman syndrome, are also associated with urinary magnesium losses.

Miscellaneous causes of hypomagnesemia include acute pancreatitis, in which magnesium and calcium are saponified in necrotic fat, hungry bone syndrome, where magnesium, calcium, and phosphate are absorbed by bone after parathyroidectomy for hyperparathyroidism, and diabetic ketoacidosis, where magnesium levels fall due to osmotic diuresis and insulin-related transmembrane shifts.

HYPOMAGNESEMIA AND ASSOCIATED ELECTROLYTE ABNORMALITIES

Hypokalemia

Hypomagnesemia and hypokalemia often coexist due to similar common underlying etiologies, such as excess gastrointestinal losses and diuretics. Hypokalemia is often difficult to treat without magnesium repletion.

Hypercalcemia

Elevated ionized serum calcium levels induce renal Mg²⁺ wasting. Hypomagnesemia is common in hypercalcemia of malignancy. However, in hypercalcemia secondary to hyperparathyroidism, magnesium deficiency is rare due to the parathyroid hormone (PTH)-induced stimulation of renal Mg²⁺ reabsorption.

Hypocalcemia

Hypomagnesemia may cause hypocalcemia due to inhibition of PTH secretion and by induction of skeletal resistance to PTH. ­Hypocalcemia may be present in up to half of patients with hypomagnesemia.

CLINICAL MANIFESTATIONS

Mild hypomagnesemia may be asymptomatic. Severe hypomagnesemia leads to neuromuscular, neurologic, and cardiovascular symptoms. Neuromuscular abnormalities include hyperreflexia, carpopedal spasm, delirium, seizures, tetany, and paralysis. Chvostek and Trousseau signs may be present. ECG manifestations include torsades de pointes, premature ventricular contractions, ventricular tachycardias, and ventricular fibrillation. There is also an increased risk of digitalis cardiac toxicity. As serum magnesium levels do not always correlate with total body magnesium stores, normomagnesemic magnesium depletion may be considered in patients with unexplained hypocalcemia and hypokalemia and clinical risk factors for magnesium deficiency.

EVALUATION OF HYPOMAGNESEMIA

Urine studies

Urine studies are useful to evaluate renal vs nonrenal causes of hypomagnesemia. The fractional excretion of magnesium (Fe_Mg) is given by:

The 0.7 in the denominator is a correction factor for the 30% of plasma magnesium bound to plasma proteins. A Fe_Mg of > 3% in a patient with normal GFR indicates renal magnesium loss. A 24-hour magnesium collection can also be obtained and is normally 3 to 5 mmol (75-125 mg)/24 hours. In the presence of hypomagnesemia, normal kidneys should be able to reduce the 24-hour urinary excretion of magnesium even further, to 1 mmol or less.

TREATMENT OF HYPOMAGNESEMIA

Oral repletion

The most popular formulation for oral replacement is magnesium oxide (242 mg = 20 mEq Mg²⁺ per 400 mg tablet), with a typical dose of 400 mg two to three times per day. Magnesium chloride, magnesium gluconate, magnesium lactate, and magnesium L-aspartate are other options. Diarrhea is a common side effect. It may be reduced with the use of a sustained release formulation, such as magnesium chloride (64 mg per 535 mg tablet). The potassium-sparing diuretics triamterene and amiloride, which block ENaC in the distal renal tubule, can assist in treatment of hypomagnesemia refractory to oral supplementation.

Intravenous repletion

Symptomatic or severe hypomagnesemia should be treated with intravenous magnesium. For active seizures or cardiac arrhythmias, an initial dose of 8 to 16 mEq of Mg²⁺ (1-2 g of MgSO₄7H₂O) is administered over 2 minutes. For nonemergency repletion, 64 mEq of Mg²⁺ (8 g of MgSO₄7H₂O) can be given over the first 24 hours, followed by 32 mEq of Mg²⁺ daily for six additional days. Since magnesium is renally cleared, the dose should be reduced by 25% to 50% and the plasma magnesium level monitored after each dose for GFR < 20 to 30 mL/min/1.73 m².

HYPERMAGNESEMIA

ETIOLOGY

Hypermagnesemia usually results from iatrogenic causes, such as magnesium treatment for preeclampsia or eclampsia, or inadvertent administration of excessive doses of magnesium-containing supplements, laxatives, Epsom salts, enemas, or antacids. The risk of hypermagnesemia is particularly high for patients with severely impaired renal function.

CLINICAL MANIFESTATIONS

Clinical manifestations are unusual with plasma magnesium levels < 4.5 to 5 mg/dL. Above this range, nausea, vomiting, cutaneous flushing, hyporeflexia, and mild hypotension can be seen. For plasma magnesium levels > 7 to 10 mg/dL, there may be loss of tendon reflexes, muscle weakness, and hypotension. Respiratory muscle paralysis occurs when magnesium levels exceed 12 to 15 mg/dL. ECG changes with plasma magnesium values > 5 mg/dL include prolonged PR interval, an increased QRS interval, prolonged QT interval, and bradycardia. Complete heart block is seen for plasma Mg²⁺ > 10 to 15 mg/dL, and cardiac arrest for levels > 15 mg/dL.

TREATMENT

In mild cases, stopping magnesium administration may be sufficient. Dialysis can be performed for extreme cases. Intravenous calcium (100-200 mg of elemental calcium given over 5-10 minutes) can be used to temporarily antagonize the effects of magnesium until dialysis can be performed. Details regarding intravenous calcium administration can be found in the section on treatment of hyperkalemia. Intravenous volume infusion may be helpful in promoting magnesium excretion in patients who are not volume overloaded and who have adequate renal function.

DISCHARGE CHECKLIST

• For patients with potassium disorders at risk for cardiac arrhythmias, is the potassium level within the normal range prior to discharge?

• For patients with potassium disorders at low risk of cardiac arrhythmias, is the potassium level between 3 and 5.6? Are clinical symptoms or ECG changes associated with potassium disorders absent? If the patient is discharged with mild hypokalemia or mild hyperkalemia, is there a clinical plan to achieve a normal plasma potassium (3.5-5) which can be re-evaluated at follow-up?

• Have patients with hypokalemia been counselled regarding potential dietary sources of potassium? (These include dark leafy greens, avocadoes, peaches, prunes, raisins, potatoes, squash, beans, and fish, as well as the commonly cited bananas and orange juice.)

• Has follow-up testing been arranged for potassium, magnesium, and creatinine, if appropriate?

SUGGESTED READINGS

INTRODUCTION

Sodium and water disturbances are among the most commonly encountered disorders in hospitalized and critically ill patients. Sodium and water balance are independently regulated by mechanisms that are designed to maintain circulatory integrity and plasma osmolality, respectively. Sodium balance is regulated by changes in sodium intake and excretion, whereas plasma osmolality is regulated by changes in water intake and water excretion.

Water is the predominant constituent of the human body. In healthy individuals, it makes up 60% of male body weight and 50% of female body weight. Body water is distributed between two compartments: the intracellular fluid compartment, containing 55% to 65%, and the extracellular fluid compartment, containing the remaining 35% to 45%. The extracellular fluid compartment is further subdivided into the interstitial space and the intravascular space. The interstitial space comprises approximately 75% of the extracellular fluid compartment, whereas the intravascular space contains 25%. Total body water diffuses freely between the intracellular space and the extracellular space in response to solute concentration gradients. Therefore, the amount of water in each compartment depends entirely on the quantity of solute in that compartment. The major extracellular solute is sodium, while potassium is the major intracellular solute. This solute distribution is maintained by active transport, via the Na⁺/K⁺-ATP-dependent pumps found on cell membranes.

PATHOPHYSIOLOGY OF SODIUM BALANCE

The extracellular fluid compartment depends on the total body sodium content as well as the integrity of the mechanisms responsible for its maintenance. Sodium content is normally tightly regulated by modulating renal retention and excretion when there is deficiency or excess of extracellular fluid. The operative homeostatic mechanisms include an afferent sensing limb and an efferent effector limb. Disorders of either sensing or effector mechanisms can lead to failure to adjust renal sodium handling, resulting in hypertension and edema with positive sodium balance, or hypotension and hypovolemia with negative sodium balance.

EFFECTIVE ARTERIAL BLOOD VOLUME

Effective arterial blood volume refers to the blood volume detected by baroreceptors in the arterial circulation. The effective arterial blood volume can change independently of the total extracellular fluid volume, leading to sodium and water retention in different clinical situations.

Afferent limb sensing sites include low-pressure cardiopulmonary receptors (atrial, ventricular, and pulmonary stretch receptors), high-pressure arterial baroreceptors (carotid, aortic arch, and renal sensors), and central nervous system and hepatic receptors. Activation of these afferent sites engages several effector mechanisms. Activation of the sympathetic nerves stimulates proximal tubular sodium reabsorption. Renin release from the renal juxtaglomerular apparatus stimulates the formation of angiotensin II, a potent vasoconstrictor. Angiotensin II also stimulates the adrenal gland to secrete aldosterone, which increases tubular sodium reabsorption. Other effector mechanisms include prostaglandins that modulate renal blood flow and sodium handling, natriuretic peptides that enhance renal sodium excretion in response to atrial stretching, and pituitary secretion of arginine vasopressin (AVP), a modest vasoconstrictor that also enhances renal water retention.

REGULATION OF WATER BALANCE

Osmoregulation is the regulation of water intake and excretion to maintain constant body osmolality (tonicity). Disorders of water homeostasis result in hypo- or hypernatremia, whereas disorders of total body sodium content result in extracellular volume depletion or excess. Since osmolality in all body fluids is essentially equal, it can be estimated by measuring the plasma osmolality. The normal plasma osmolality is 275 to 290 mOsm/kg. It is kept within this range by hypothalamic osmoreceptors, which are capable of sensing a 1% to 2% change in tonicity. Osmoreceptors provoke thirst and release of the antidiuretic hormone AVP. The combined effect of AVP stimulation and thirst results in water intake and retention, lowering plasma osmolality and sodium concentration by dilution, whereas suppression of thirst and AVP secretion leads to the contrary.

DISORDERS OF SODIUM BALANCE

EXTRACELLULAR FLUID VOLUME CONTRACTION

Extracellular fluid (ECF) volume contraction is caused by loss of sodium and water in excess of intake. This may result from renal losses, extrarenal losses, or sequestration in potential spaces, such as the peritoneal cavity, retroperitoneum, pleural spaces, or muscles, which are not in hemodynamic equilibrium with the ECF (“third-spacing”).

Extrarenal causes

The gastrointestinal tract is the most common extrarenal source of fluid loss. Vomiting or nasogastric suction may cause volume loss with metabolic alkalosis. Diarrhea may result in volume depletion with metabolic acidosis. Third-space sequestration, excessive sweat production, and loss of skin barrier from burns or exudative skin lesions may lead to significant ECF volume depletion. Hemorrhage, as from gastrointestinal bleeding or trauma, may lead to significant volume loss.

Renal losses

Diuretics may cause renal sodium wasting, volume contraction, and acid-base disturbances if abused or inappropriately prescribed. Genetic and acquired tubular disorders can result in renal sodium wasting and volume contraction. Mineralocorticoid deficiency or resistance may lead to renal sodium wasting. This may occur in the setting of primary adrenal insufficiency (Addison disease), hyporeninemic hypoaldosteronism secondary to diabetes mellitus, or other chronic interstitial renal diseases. Severe hyperglycemia or high levels of blood urea after relief of urinary tract obstruction can lead to obligatory renal sodium and water loss secondary to glucosuria or urea diuresis, respectively.

Clinical manifestations

The symptoms of hypovolemia are nonspecific. They range from mild postural symptoms, thirst, muscle cramps, and weakness, to drowsiness and disturbed mentation with profound volume loss. Physical examination may reveal tachycardia, cold clammy skin, postural or recumbent hypotension, and reduced urine output depending on the degree of volume loss (Table 242-1). Rarely, diminished skin turgor may be seen. Reduced jugular venous pressure (JVP) noted at the base of the neck is a useful parameter of volume depletion and may roughly estimate the central venous pressure (CVP). However, an elevated CVP does not exclude hypovolemia in patients with underlying cardiac failure or pulmonary hypertension. The absence of symptoms or obvious physical findings does not preclude volume depletion in an appropriate clinical setting, and hemodynamic monitoring or administration of a fluid challenge may sometimes be necessary.

Laboratory findings consistent with volume depletion include hemoconcentration and increased serum albumin concentration. However, anemia or hypoalbuminemia may confound interpretation of these laboratory values. In healthy individuals, the ratio of blood urea nitrogen (BUN) (mg/dL) to serum creatinine (mg/dL) approximately equals 10. In volume depletion, this ratio increases because of a differential increase in the tubular reabsorption of urea. Several clinical conditions confound this ratio. Upper gastrointestinal hemorrhage and administration of corticosteroids increase urea production and the ratio of BUN to creatinine. Malnutrition and underlying liver disease diminish urea production, and thus the BUN-to-creatinine ratio is less useful in these clinical settings.

In the absence of renal salt wasting or diuretics, elevated urine osmolality and specific gravity may indicate hypovolemia. Hypovolemia normally triggers avid renal sodium reabsorption, resulting in low urine sodium concentration and low fractional excretion of sodium. The fractional excretion of sodium (FE_Na) is calculated using the following formula: FE_Na = [U_Na × P_Cr/ U_Cr × P_Na] × 100, where U_Na and U_Cr are urinary sodium and creatinine concentrations, and P_Na and P_Cr are plasma sodium and creatinine concentrations, respectively. Decreased FE_Na (< 1%) is consistent with ECF volume depletion. However, arterial underfilling due to low cardiac output or systemic arterial vasodilatation, as in cirrhosis, is also associated with a FE_Na less than 1%, in spite of increased ECF and edema.

Treatment

Volume depletion is treated by replacing fluid deficits and ongoing losses with a replacement fluid closely resembling the lost fluid. Clinical parameters help to distinguish mild to moderate from severe volume loss (Table 242-1), but invasive monitoring may be needed in some patients. Mild volume contraction can usually be corrected via the oral route. In hypovolemic shock with evidence of life-threatening circulatory collapse or organ dysfunction, intravenous fluid must be given as rapidly as possible until clinical parameters improve. A slower, more judicious approach is warranted in the elderly and in those with underlying cardiac conditions to avoid overcorrection with subsequent pulmonary edema.

Crystalloid solutions are effective, as they distribute primarily in the ECF. Isotonic saline is usually the initial replacement fluid of choice in volume-depleted patients. Colloid-containing solutions, such as 5% and 25% albumin and hetastarch (6% hydroxyethyl starch), are second-line agents for the treatment of hypovolemia. They remain within the vascular compartment by virtue of their large molecular sizes. Hyperoncotic starch solutions should be avoided as they increase the risk of acute kidney injury, need for renal replacement therapy, and mortality.

Hypovolemic shock may be accompanied by lactic acidosis due to tissue hypoperfusion. Fluid resuscitation restores tissue oxygenation and decreases the production of lactate. Correction of acidosis with sodium bicarbonate has several potential adverse effects, including increasing tonicity, worsening intracellular acidosis, and reducing ionized calcium concentration. Therefore, its use to manage lactic acidosis in the setting of volume depletion is not recommended, unless the arterial pH is below 7.1.

EXTRACELLULAR FLUID VOLUME EXPANSION

Excess extracellular fluid accumulation usually results from sodium and water retention by the kidneys. Renal sodium retention may be primary or a compensatory response to reduced effective arterial blood volume. It manifests clinically with edema and weight gain, commonly in response to congestive heart failure (CHF), cirrhosis with ascites, or nephrotic syndrome.

Renal sodium and water retention in CHF involves activation of the renin-angiotensin-aldosterone system (RAAS), as well as nonosmotic AVP stimulation, which drives the hyponatremia associated with CHF. Arterial underfilling is also responsible for water and sodium retention in cirrhosis. Unlike CHF and cirrhosis, in which the kidneys are structurally normal, the nephrotic syndrome is characterized by kidneys that are functionally impaired. Rarer causes of edema include drug-induced edema, which may be seen with ingestion of systemic vasodilators, such as minoxidil and diazoxide, dihydropyridine calcium channel blockers, and thiazolidinediones. Nonsteroidal anti-inflammatory drugs (NSAIDs) can exacerbate volume expansion in CHF and cirrhotic patients by decreasing vasoregulatory prostaglandins in the kidneys.

Clinical manifestations

Patients with left heart failure may present with exertional dyspnea, orthopnea, and paroxysmal nocturnal dyspnea, whereas patients with predominant right heart failure or biventricular failure may exhibit weight gain and lower-limb swelling. Physical examination may reveal JVP elevation, pulmonary crackles, a third heart sound, or peripheral edema in the ankles or sacrum. Nephrotic patients classically present with periorbital edema because of their ability to lie flat during sleep. Patients with severe nephrotic syndrome may exhibit marked generalized edema with anasarca. Cirrhotic patients present with ascites and lower-limb edema consequent to portal hypertension and renal sodium and water retention. Physical examination may reveal stigmata of chronic liver disease, such as jaundice, spider angiomata, palmar erythema, Dupuytren contractures, and splenomegaly.

Therapy

Management of ECF volume expansion consists of treating the underlying cause and achieving negative sodium balance with dietary sodium restriction and diuretics. Moderate dietary sodium restriction (2-3 g/d; 86-130 mmol/d) should be instituted. Restriction of total fluid intake is usually only necessary for hyponatremic patients. Medications that promote sodium retention, such as NSAIDs, corticosteroids, estrogens, and androgens, should be eliminated, if possible. Diuretics are the cornerstone of therapy to remove excess volume. Other measures can be used when the response to diuretics is inadequate. Extracorporeal fluid removal by ultrafiltration can be utilized in patients with acute decompensated heart failure with renal insufficiency or diuretic resistance. In cirrhosis, large-volume paracentesis with albumin infusion (6-8 g per liter of ascitic fluid removed) can be employed. Interventions to shift ascitic fluid to a central vein, such as a transjugular intrahepatic portosystemic shunt, may reduce refractory ascites and improve GFR and sodium excretion.

Diuretics are the mainstay of therapy for edematous states. As ECF volume expansion in CHF and cirrhosis may be a compensatory mechanism for arterial underfilling, a judicious approach is necessary to avoid a precipitous fall in cardiac output and tissue perfusion. Diuretics can be classified into five classes, based on their predominant sites of actions along the nephron (Figure 242-1). They inhibit the reabsorption of sodium and an accompanying anion, usually chloride. The resultant natriuresis decreases the ECF. However, the time course of natriuresis is limited because renal mechanisms attenuate the sodium excretion.

The most serious adverse effects of diuretics are electrolyte and acid-base disturbances. Loop diuretics increase the excretion of potassium, magnesium, and calcium. Thiazide diuretics have similar effects on potassium and magnesium, but unlike loop diuretics, they decrease urinary calcium losses. Thiazide diuretics interfere with urine-diluting mechanisms, and may therefore pose a risk of hyponatremia. Patients on diuretic therapy require electrolyte monitoring, and may need oral supplementation of potassium and magnesium. Addition of a potassium-sparing diuretic may need to be considered in patients with severe or recurrent hypokalemia. The volume depletion caused by thiazides and loop diuretics leads to increased tubular urate reabsorption, promoting hyperuricemia and gout. Ototoxicity may occur with large intravenous doses of loop diuretics, particularly when an aminoglycoside is coadministered. Gynecomastia may develop with spironolactone therapy.

Long-term loop diuretic tolerance occurs as a consequence of hypertrophy of the distal nephron segment and enhanced sodium reabsorption from increased exposure to solutes not absorbed proximally. This can be addressed by combining loop and thiazide diuretics, as the latter blocks reabsorption distal to the loop of Henle.

Diuretic resistance refers to edema that is or has become refractory to a given diuretic. Diuretic resistance has several causes. Chronic kidney disease is associated with decreased secretion and tubular delivery of diuretics, reducing their concentration at the active site in the tubular lumen. Arterial underfilling in cirrhosis and CHF is associated with diminished nephron responsiveness to diuretics because of increased proximal tubular sodium reabsorption, leading to decreased delivery of sodium to the distal sites of diuretic action. Nonsteroidal anti-inflammatory drugs block prostaglandin-mediated increases in renal blood flow, leading to sodium retention.

Treatment of diuretic resistance includes salt restriction, increasing the dose of loop diuretics, administering more frequent doses, and using combination therapy to sequentially block more than one site in the nephron, as that may result in a synergistic interaction between diuretics (Figure 242-2). In patients who respond poorly to intermittent doses of a loop diuretic, a continuous intravenous infusion can be tried. This has been associated with a better safety profile but similar efficacy in clinical trials. Ultrafiltration may be necessary in highly resistant edematous patients. Table 242-2 lists effective diuretic doses in patients with CHF, nephrotic syndrome, advanced cirrhosis, and renal failure.

DISORDERS OF WATER BALANCE

HYPONATREMIA

Hyponatremia (plasma sodium concentration below 135 mmol/L) is the most common electrolyte abnormality in hospitalized patients, occurring in 15% to 30% of medical inpatients, with over half of cases being acquired during hospitalization. Most patients with hyponatremia have hypo-osmolar or hypotonic hyponatremia, with excess water in relation to solute. However, hyponatremia does not always signify a hypotonic state. Rarely, patients have hypertonic hyponatremia, or translocational hyponatremia, due to osmotically active solutes that do not readily penetrate into cells, such as mannitol, contrast media, and sorbitol. To equalize osmolality across cell membranes, water shifts out of cells to the extracellular space, leading to a drop in plasma sodium concentration. The most common cause is excess extracellular glucose, as in poorly controlled diabetes mellitus. Plasma sodium concentration falls by 1.6 mmol/L for every 100 mg/dL rise in the plasma glucose concentration. In pseudohyponatremia (isotonic hyponatremia), hyponatremia occurs when the solid phase of plasma, primarily lipids and proteins, is greatly expanded, as in hypertriglyceridemia and myeloma and other paraproteinemic disorders, leading to an artifactual decrease in plasma sodium concentration.

Etiology and diagnosis

Hypotonic hyponatremia is the most common plasma sodium disorder in the hospital setting, indicating an excess of water relative to sodium in plasma. This can occur with a decrease in total body sodium, or hypovolemic hyponatremia; a near-normal total body sodium, or euvolemic hyponatremia; or an excess of total body sodium, or hypervolemic hyponatremia (Figure 242-3). Since sodium is the primary cation in the ECF compartment, its concentration determines ECF volume. A comprehensive history and a physical examination focused on the evaluation of the ECF volume allows for the classification of the hyponatremia into one of three categories: hypovolemic, euvolemic, or hypervolemic.

In hypovolemic hyponatremia, deficits of both total body sodium and water lead to ECF volume depletion, with consequent AVP secretion and decreased solute-free water excretion.

Renal causes of solute-rich fluid loss include diuretics, mineralocorticoid deficiency, salt-losing nephropathy, such as polycystic kidney disease and interstitial nephritis, and partial urinary tract obstruction. Fluid losses may also be extrarenal, especially gastrointestinal, such as gastroenteritis, peritonitis, pancreatitis, and ileus, as well as hemorrhage or excessive sweating. Historical features may include vomiting, diarrhea, diuretic use, thirstiness, or hyperglycemia with glucosuria. Physical examination may reveal tachycardia, orthostatic hypotension, and flat neck veins. Helpful urinary studies include urine sodium and the Fe_Na, which are typically less than 10 mmol/L and 1%, respectively with extrarenal fluid losses. Urine sodium and Fe_Na greater than 10 mmol/L and 1%, respectively, usually indicate renal losses, such as secondary to diuretic use and osmotic diuresis.

In euvolemic hyponatremia, total body sodium content is generally normal, but there is hypo-osmolality from a relative gain of water. Physical examination is notable for the absence of both features of volume overload, such as peripheral edema, ascites, pulmonary edema, and features of volume depletion, such as tachycardia, low blood pressure, or flat neck veins. Concentrations of BUN and uric acid are normal or low, and the urine sodium is usually greater than 20 mEq/L. In this setting, hyponatremia is often due to the syndrome of inappropriate antidiuretic hormone secretion (SIADH), attributable to release of AVP from the pituitary or, rarely, an ectopic source such as lung carcinoma. This leads to impaired water excretion, while the regulation of sodium balance is virtually unaffected. Ingestion or administration of water is also required, since a high level of AVP alone is not usually sufficient to produce hyponatremia. Causes of SIADH include infectious, inflammatory, neoplastic, and vascular disorders of the central nervous and respiratory systems, pain, narcotics, anticonvulsants, antidepressants, neuroleptics, and other drugs which act on the central nervous system (Table 242-3). However, in many hospitalized elderly patients with SIADH, the cause is obscure. Several entities should be considered prior to diagnosing SIADH. These include profound hypothyroidism, adrenal insufficiency, renal impairment, and primary polydipsia. Hyponatremia may also be seen in excessive beer consumption (beer potomania), as beer is rich in carbohydrates and water, but poor in solutes and electrolytes.

To make the diagnosis of SIADH (Table 242-4), patients must be euvolemic, and the urinary osmolality must exceed 100 mOsm/kg of water despite a low plasma osmolality (< 275 mOsm/kg of water). Because of the low urine volume in SIADH, the urinary sodium concentration is high, usually greater than 40 mmol/L, in the steady state. However, if the patient is volume depleted or on a salt-restricted diet, the urine concentration may be lower. Expanding the ECF volume with normal saline or normal salt intake increases urinary sodium, but does not correct the hyponatremia. Supportive laboratory tests for SIADH include hypouricemia (serum uric acid < 4 mg/dL) and low BUN (< 10 mg/dL). When the diagnosis is still uncertain, ECF volume depletion can be ruled out by infusing 2 L of normal saline over 24 to 48 hours. Correction of hyponatremia, along with a drop in urinary osmolality, suggests volume depletion rather than SIADH. Caution must be exercised when administering saline to hyponatremic patients, because in a subset of patients this can induce brisk water diuresis and a rapid rise in serum sodium. Osmotic demyelination, with severe neurologic morbidity and even death, may occur with rapid correction of plasma sodium concentration (greater than 9 mEq/L rise in 24 hours) in patients with chronic hyponatremia.

Frequent monitoring of serum and urinary electrolytes is therefore warranted. If rapid overcorrection occurs, administration of hypotonic fluid or even AVP may be needed to halt the progressive rise in serum sodium.

In hypervolemic hyponatremia, both total body sodium and water are increased, but water is comparatively more so. Underlying causes include cardiac failure and liver cirrhosis, in which the kidneys are intrinsically normal, but respond to decreased effective arterial perfusion with increased AVP levels and activation of RAAS. Acute and chronic kidney disease may be associated with hyponatremia if water intake exceeds the ability to excrete equivalent volumes. Hyponatremia is a risk factor for diminished survival in cardiac failure and cirrhosis. In the absence of diuretics, Fe_Na should be less than 1% in both conditions. However, a Fe_Na greater than 1% is seen in the hypervolemic hyponatremic patient with acute or chronic kidney disease with tubular dysfunction leading to suboptimal sodium and water absorption.

The clinical manifestations of hyponatremia are mainly neurologic and caused by intracellular fluid shifts, leading to brain swelling and edema. The severity of symptoms depends on the rate of decline of serum sodium concentration, because the adaptive mechanisms to protect brain cell volume require a longer time to be fully functional. These mechanisms involve a gradual loss of potassium and organic solutes to restore cell volume to normal. Symptoms include nausea, vomiting, headache, lethargy, seizures, and a progressively decreased level of consciousness, ending in coma and death. Neurologic signs of severe acute hyponatremia may include depressed deep tendon reflexes, hypothermia, pseudobulbar palsy, and Cheyne-Stokes respirations.

Psychiatric patients are at risk for hyponatremia for many reasons, including:

• Thirst-producing anticholinergic side effects of antidepressants and antipsychotics

• SIADH from typical and atypical antipyschotics, tricyclic antidepressants, and selective serotonin reuptake inhibitors

• Obsessive-compulsive behavior

• Lack of satiety from water consumption in patients hospitalized for alcoholism or anorexia

Schizophrenic patients are particularly at risk: up to 14% have hyponatremia, and up to 25% have polydipsia. This predisposition may be multifactorial, relating to a decreased osmotic threshold for AVP release, enhanced renal sensitivity to AVP, and a defect in the osmoregulation of thirst. Clozapine may have a lesser tendency to cause polydipsia and hyponatremia than other antipsychotics. Unfortunately, it has uncommon but severe adverse effects, including agranulocytosis, seizures, myocarditis, and orthostatic hypotension.

Therapy

Hypovolemic hyponatremia is treated with volume repletion with isotonic saline to expand ECF volume and interrupt AVP release. Volume expansion should be continued until blood pressure is restored and the patient demonstrates clinical euvolemia. When the initial volume estimate is equivocal, a fluid challenge with 0.5 to 1 L of isotonic saline can be both diagnostic and therapeutic.

Hypervolemic hyponatremia

The treatment of hyponatremia in the context of volume overload requires management of the underlying cardiac failure or liver cirrhosis. Sodium and water restriction as well as diuretics should be employed as necessary. When hyponatremia occurs, fluid restriction to amounts less than insensible losses plus urine output is necessary to cause a negative solute-free water balance, but is often difficult to achieve. Diuretic-resistant cases of CHF and liver cirrhosis may be treated with ultrafiltration and large-volume paracentesis, respectively.

Euvolemic hyponatremia

The treatment of hyponatremia in euvolemic patients depends on the degree of symptoms and on whether it has developed acutely (within hours) or chronically (over more than 48 hours). Acute hyponatremia with onset in the hospital is most often related to hypotonic solutions. The rapid development of hyponatremia with very low sodium values (< 110-115 mmol/L) may cause fluid to shift into brain cells, possibly leading to cerebral edema and death, as the adaptive changes cannot keep pace with intracellular fluid shifts leading to brain swelling within the confined skull space.

In acute severe hyponatremia with neurologic symptoms, the treatment of choice is 3% hypertonic saline, infused initially at 1 to 2 mL/kg/h (Table 242-5). For each 100 mL of 3% hypertonic saline, the serum sodium concentration will increase approximately by 2 mmol/L. This should be continued until a less severe level of hyponatremia has been attained (ie, 125-130 mmol/L), or symptoms resolve. If seizures, obtundation, or coma are present, hypertonic saline may be infused at 4 to 6 mL/kg/h for a short period of time. An alternative approach that has been successfully used in marathon runners is infusing 100 mL bolus of 3% hypertonic saline, enough to increase the serum Na approximately by 2 mmol/L. A small, quick increase in the serum Na (2-4 mmol/L) is effective in treating acute hyponatremia because reducing brain swelling even slightly will substantially decrease intracerebral pressure. If severe neurologic symptoms persist or worsen, or if the serum sodium is not improving, a 100-mL bolus of hypertonic saline can be repeated one or two more times at 10-minute intervals.

When hyponatremia develops over several days, brain cells extrude organic solutes from their cytoplasm, allowing intracellular osmolality and plasma osmolality to equalize without a large increase in cell water. Because of this rapid normalization of chronic hyponatremia may also lead to cerebral edema. Hyponatremia should be corrected very gradually when it has been present for more than 48 hours, or the duration is unknown. Serum sodium initially be increased by 4 to 6 mEq/L during the first 24 hours and by less than 9 mEq/L over any given 24-hour period. It should be corrected even more slowly in high-risk patients with severe malnutrition, alcoholism, or advanced liver disease, who have an impaired ability to protect their intracellular volume by generating osmotically active molecules such as glycine, taurine, and myoinositol.

Rapid correction of chronic hyponatremia, outpacing the brain’s ability to recapture lost organic osmolytes, may result in osmotic demyelination syndrome, which especially affects glial cells of the brainstem. Patients improve initially as hyponatremia resolves, but within days develop new, progressive, and sometimes permanent neurologic deficits, including fluctuating level of consciousness, quadriparesis, pseudobulbar palsy, ataxia, dysarthria, and locked-in syndrome. Magnetic resonance imaging (MRI) may show T2-weighted hyperintensities in the brainstem, especially the pons, consistent with demyelination.

In patients with chronic asymptomatic hyponatremia, hypothyroidism and secondary adrenal insufficiency should be sought and treated appropriately if present. Potentially responsible medications should be discontinued. Ideally, SIADH is treated with therapy directed against its underlying cause. However, if the cause cannot be identified or expeditiously treated, hyponatremia should be treated conservatively with fluid restriction. To achieve negative water balance, daily fluid intake must be significantly limited to less than the 24-hour urine output plus insensible losses. The maximum tolerated fluid intake is proportional to the oral osmotic load, so adequate intake of dietary protein and salt should be encouraged. Many patients will find the recommended degree of fluid restriction intolerable and difficult to comply with. Oral medications that may be helpful include urea (30 g/d), demeclocycline (300-600 mg twice daily), or lithium, but these therapies have poor tolerability, inconsistent responses, and significant toxicities.

Recently, vasopressin receptor antagonists (vaptans) have been introduced for the treatment of hyponatremia. These agents act by increasing electrolyte-free water excretion (aquaresis) and hence raising serum sodium concentration. Conivaptan, a combined V1 and V2 receptor antagonist, has been approved by the United States Food and Drug Administration (FDA) for 4-day intravenous use in the hospital setting to treat euvolemic and hypervolemic hyponatremia. However, there are insufficient data to use conivaptan in the treatment of acute symptomatic hyponatremia at this time. Conivaptan might be considered particularly for those with moderate to severe hyponatremia and symptoms, but not seizures, delirium, or coma, which would warrant the use of hypertonic saline. The selective V2 receptor antagonist, tolvaptan, is orally active and has been approved by the FDA for euvolemic and hypervolemic hyponatremia. However, the FDA warns that tolvaptan should not be used in any patient for longer than 30 days, and should not be given to patients with liver disease, including cirrhosis. Increased thirst is seen in patients treated with vaptans and may limit the rise in serum sodium. Moreover, overly rapid correction of the hyponatremia may occur, which can lead to irreversible neurologic injury.

Hyponatremia: Discharge Checklist

• Have patient and family been educated about signs and symptoms of hyponatremia?

• Has the patient and family been educated about any restrictions on fluid intake, if necessary?

If indicated, specify total liquid intake, not only water. Aim for a fluid restriction that is 500 mL/d below the 24-hour urine volume.

• Do not restrict sodium or protein intake unless indicated.

• If the patient is being discharged on a vaptan, ensure that treatment is not taken beyond 30 days.

• Have blood draws for electrolyte testing and outpatient follow-up been arranged for 1 to 2 weeks after discharge?

HYPERNATREMIA

Although not as common as hyponatremia, hypernatremia is not infrequent in hospitalized patients, particularly those in intensive care units. Hypernatremia is always associated with hyperosmolality, as sodium salts are the major extracellular solutes. Hyperosmolality leads to an exodus of water from cells to the extracellular compartment to maintain osmotic equilibrium, with a loss of intracellular volume and shrinkage of brain cells. The major defense against sodium elevation is the stimulation of thirst and AVP, leading to increased water intake and retention. While AVP is important, thirst provides greater protection against hypernatremia. Thus, hypernatremia due to water loss occurs mainly in vulnerable patients with altered mental status and those who are unable to obtain water, such as infants, the elderly, and the severely ill.

Causes of hypernatremia are grouped into three categories, according to the total body sodium content as estimated by extracellular volume (Figure 242-4). Thus, the diagnostic approach to the hypernatremic patient hinges on the assessment of extracellular volume. The history should include a review of recent and current medications, vomiting, diarrhea, increased urinary losses (polyuria), recent fluid intake, and the presence or absence of thirst. Physical examination must assess volume and neurologic status.

In hypovolemic hypernatremia, there is a predominant loss of solute-free water leading to elevation in the serum sodium concentration. This can occur either from renal or extrarenal sources. Extrarenal losses include gastrointestinal causes, with diarrhea being more likely than vomiting to cause hypernatremia. Osmotic diarrhea, as induced by lactulose, malabsorption, and some infectious diarrheas, may result in water being lost in excess of sodium and potassium, leading to hypernatremia. Secretory diarrhea, in contrast, produces fluid with electrolyte content similar to plasma, leading to loss of volume and potassium, but not hypernatremia. Insensible salt and water losses through evaporation from the respiratory tract and sweat, brought on by fever or strenuous exercise in hot temperatures, may lead to hypernatremia, because of the hypotonic nature of the lost fluid.

The most common renal causes of sodium and water loss are loop diuretics and osmotic diuresis. Loop diuretics may lead to hypernatremia because they provoke loss of isotonic or hypotonic fluid. Osmotic diuresis from nonelectrolyte solutes, such as urea and glucose, causes hypernatremia by impairing tubular reabsorption of sodium and water. Most hypotonic fluid losses do not cause hypernatremia, unless inadequate free water is ingested or infused to replace the ongoing losses. Hence, these disorders usually also involve some component of inadequate fluid intake.

In hypervolemic hypernatremia, there is excess sodium in the extracellular compartment with extracellular volume expansion. This is usually seen in hospitalized patients who receive excess saline or sodium bicarbonate to treat metabolic acidosis, or those receiving enteral or parenteral feedings without adequate free water administration.

Euvolemic hypernatremia

Loss of water without sodium increases fluid tonicity, but usually does not result in clinically evident volume depletion because water is mainly distributed in the intracellular compartment. Euvolemic hypernatremia occurs when there is loss of fluid that is low in sodium and potassium salts. The classic example is diabetes insipidus (DI) from either defects in AVP production or release (central DI), a failure of renal response to AVP (nephrogenic DI), or, rarely, from rapid degradation of AVP by vasopressinase during pregnancy. Most patients with DI present with polyuria and polydipsia; hypernatremia is usually not present unless fluid intake is inadequate.

Central DI can be partial or complete. Causes include head trauma, brain surgery, and inflammatory, neoplastic, and infiltrative disorders of the hypothamalus and pituitary. However, no cause is identified in about half of the cases. Nephrogenic DI is usually a congenital or acquired defect of vasopressin V2 receptors or the vasopressin-dependent water channel (aquaporin 2). Acquired causes include medullary or interstitial renal disease, such as interstitial nephritis, polycystic kidney disease, partial urinary tract obstruction, and advanced chronic kidney disease. Hypercalcemia and hypokalemia may also cause nephrogenic DI, as well as various drugs that impair response to AVP, such as lithium. The various forms of DI must be differentiated from primary polydipsia in patients who present with polyuria.

Urine sodium, volume, and osmolality should be measured to determine the integrity of the AVP-renal axis. The normal response to hypernatremia is increased AVP release, resulting in urine osmolality that can reach a maximum of 1000 mOsm/kg in normal young individuals, and above 500 mOsm/kg in elderly patients who usually have reduced responsiveness to AVP. Reduced urine osmolality indicates either impaired AVP release (central DI) or action (nephrogenic DI).

The various forms of DI may be differentiated from primary polydipsia in patients with polyuria by performing the water deprivation test, along with administration of AVP. An increase in urine osmolality with exogenous AVP is consistent with central DI, whereas lack of response suggests nephrogenic DI. With fluid restriction leading to a 3% weight loss, patients with primary polydipsia should spontaneously increase their urine osmolality and will not increase their urine osmolality more with exogenous vasopressin. With partial central DI, fluid restriction will not maximally stimulate AVP, thus a further increase in urinary osmolality (>10%) will occur with exogenous vasopressin.

Clinical manifestations

Patients without perturbed neurologic status and with normal thirst mechanisms should complain of thirst. If allowed to progress, hypernatremia becomes associated with confusion, restlessness, irritability, and other neurologic manifestations of cellular dehydration and brain cell shrinkage. Clinical signs may include muscular twitching, hyperreflexia and ataxia. Severe elevations of serum sodium may lead to focal and grand mal seizures, intracerebral hemorrhage, and death. More severe symptoms and signs occur in patients with rapidly rising serum osmolality, extremely elevated serum osmolality, especially above 325 mOsm/kg, and at extremes of age, with the very young and the very old being most vulnerable. Hypernatremia developing over a longer period of time enables brain cells to undergo osmotic adaptation by generating osmolytes (idiogenic osmoles) to guard against cell shrinkage. While protective, they may nonetheless predispose the brain to edema if hypernatremia is corrected rapidly by favoring intracellular fluid shifts.

Therapy

In hypernatremia with extracellular volume depletion, restoration of ECF volume is the primary therapeutic target. Isotonic saline is the fluid of choice, and the volume and rate of administration should be guided by clinical parameters including pulse rate, orthostatic blood pressure measurements, and urine output. Once euvolemia is established, further fluid therapy should be delivered to gradually correct tonicity in the form of hypotonic (0.45%) saline.

In rare patients who develop hypernatremia with extracellular volume expansion, loop diuretics may be employed. Ultrafiltration may sometimes be needed to treat hypernatremia in patients with advanced renal impairment.

In patients with hypernatremia without extracellular volume depletion or expansion, the total body sodium is unchanged, but there is a deficit of total body water. Water can be replaced either orally or parenterally with 5% dextrose in water (D5W) or 0.45% NaCl (half-normal saline). The estimated fluid deficit can be estimated by a formula that uses total body water (TBW), because sodium concentration reflects tonicity in all body compartments. For example, a 75 kg man with a serum sodium concentration of 157 mmol/L would need water as estimated by the following:

Then, desired body water is calculated from the following formula:

The estimated water deficit is the difference between desired body water and actual body water. In this example, the administration of 5.5 L (50.5-45 L) should correct the water deficit and normalize the serum sodium concentration. In addition to estimated water deficit, estimated ongoing water losses during the time of repletion should be added. This includes 500-1000 mL/d of insensible losses, with larger amounts in febrile or mechanically ventilated patients.

Water replacement should be carried out gradually over hours to days, unless there is evidence that hypernatremia has evolved quickly. This is because water replacement needs to keep pace with the osmotic adaptation process by which the brain cells keep from shrinking. Replacement also has to be tailored to the presence of neurologic symptoms and signs, which call for more rapid serum sodium correction. Acute symptomatic hypernatremia that developed over hours may be rapidly corrected, as the risk of cerebral edema is minimal. In patients with acute symptomatic hypernatremia, serum sodium should be lowered by 1 mmol/L/h. Once the patient is no longer symptomatic, the correction rate can be reduced to 0.5 mmol/L/h. Hypernatremia that has developed over more than 24 hours, or is of unknown duration, should be corrected with a maximal rate of 0.5 mmol/L/h. Half of the water deficit should be replaced over the first 24 hours, with the remainder over the following 24 hours or longer. The cornerstone of central DI treatment is desmopressin (dDAVP), administered either nasally, intravenously, or subcutaneously. Patients with nephrogenic DI do not respond to exogenous desmopressin. Salt restriction and thiazide diuretics produce mild volume contraction, which helps to decrease urinary output. Nonsteroidal anti-inflammatory drugs also help reduce urine output. Patients with lithium-associated nephrogenic DI may benefit from amiloride.

Hypernatremia: Discharge Checklist

• Have patient and family been educated about signs and symptoms of hypernatremia?

• Are patients being discharged on gastric tube feedings also receiving free water supplementation via the tube several times daily to prevent hypernatremia?

• If the patient has diabetes insipidus, is the treatment appropriate for the form.

Central DI is treated with desmopressin (dDAVP) whereas salt restriction, thiazide diuretics and NSAIDS are useful in the treatment of nephrogenic DI.

• Have blood draws for electrolyte testing and outpatient follow-up been arranged for 1 to 2 weeks after discharge?

SUGGESTED READINGS

EPIDEMIOLOGY

Kidney stones inflict recurrent episodes of excruciating pain and significant morbidity on a substantial portion of the population, including many young and otherwise healthy individuals. In the United States, the current lifetime incidence of nephrolithiasis is 13% for men and 7% for women. Without treatment, the 5-year recurrence rate following an initial episode is up to 50%. A recent report from the National Health and Nutrition Examination Survey for the period of 1994 to 2010 describes a 70% increase in prevalence of self-reported kidney stones, extending across men and women of all age groups.

Regional variations in the frequency and nature of kidney stone disease exist within the United States, with an increased prevalence in the southeastern region of the country. This variation may be related to differences in climate and sunlight exposure, as well as dietary habits and beverage consumption. Kidney stones develop more frequently among Caucasians than African Americans. Stones in the upper urinary tract are frequently seen in industrialized countries and are associated with a more affluent lifestyle, including high animal protein consumption, gout, and components of the metabolic syndrome, including hypertension, impaired glucose tolerance, increased waist circumference, high triglycerides, and low-high-density lipoprotein cholesterol. Bladder stones are more commonly seen in developing countries and more frequently affect individuals with a poor socioeconomic status.

Patients with kidney stones typically present with renal colic, characterized by severe pain and autonomic symptoms such as lightheadedness, diaphoresis, nausea, and vomiting. The severity of symptoms often results in a visit to a hospital emergency room, often requiring hospitalization and absenteeism from work. In the United States, kidney stones account for more than 2 million outpatient visits, over 600,000 emergency room visits, and approximately 0.4% of hospital admissions. Complications may arise, such as urinary tract obstruction and pyelonephritis, or the need for stone removal by instrumentation, surgery, or extracorporeal shock wave lithotripsy (ESWL). Patients with recurrent stone disease also have a heightened risk of chronic kidney disease. In the year 2000, the annual cost of kidney stones in the United States, including hospitalizations, professional charges, and lost productivity, was estimated at $5.3 billion.

PATHOPHYSIOLOGY AND RISK FACTORS

Kidney stones can form from several substances excreted in the urine, and frequently consist of two or more different substances (Table 243-1). Calcareous (calcium oxalate, phosphate, or mixed) stones are by far the most common, accounting for over 80% of kidney stones. Metabolic defects leading to stone formation include hypercalciuria in over 65% of cases, and less frequently hyperuricosuria, hyperoxaluria, hypocitraturia, or some combination thereof.

For a kidney stone to form, the concentration of a dissolved salt must exceed its solubility in urine, a condition known as supersaturation. Supersaturation is favored by increased urinary excretion of stone-forming salts, optimal urinary pH (Table 243-2), and decreased urinary volume, which leads to increased urinary concentration. The presence of crystallization facilitators in the urine, such as uric acid, or the absence of crystallization inhibitors, such as citrate, also contributes to stone formation.

Risk factors for kidney stone formation can be divided into diet, stone-provoking conditions, stone-provoking drugs, and anatomic abnormalities. Dietary factors promoting nephrolithiasis include low fluid intake, which promotes urinary supersaturation, high sodium and animal protein intake, which promote hypercalciuria, and high oxalate intake, which promotes hyperoxaluria. In addition, a family history of kidney stones and a personal history of gout also increase the risk of kidney stone disease.

Table 243-3 displays a selected list of stone-provoking conditions, stone-provoking drugs, and anatomic urologic abnormalities associated with stone disease. Of note, roux-en-Y gastric bypass surgery, a common bariatric surgical procedure, is associated with increased intestinal absorption of oxalate and hyperoxaluria, resulting in a long-term risk of kidney stone formation. Major stone-provoking drugs include uricosuric agents, such as probenecid, losartan, and fenofibrate, carbonic anhydrase inhibitors, such as acetazolamide and topiramate, and the antiretroviral agent indinavir, which crystallizes in the urine.

Oxalobacter formigenes and Lactobacillus spp. are intestinal bacteria that produce an oxalate-degrading enzyme. The resulting increased intestinal degradation of oxalate may protect against stone formation. Lack of intestinal colonization with O. formigenes is prevalent in patients with recurrent calcium oxalate stones due to hyperoxaluria.

CLINICAL PRESENTATION

Patients with kidney stones most often present to the hospital with renal colic. Others present with a urinary tract infection, impaired renal function from obstructive uropathy, or chronic kidney damage. Increasingly, incidental kidney stones are diagnosed on imaging studies in patients admitted to the hospital for other reasons.

Typical renal colic begins with the acute onset of severe pain that awakens the patient from sleep. Colic typically lasts for 1 to 4 hours, followed by gradual improvement. The patient may move ceaselessly, looking for a comfortable position, and may also suffer from nausea, vomiting, and sweating.

As shown in Figure 243-1, the anatomic site of the stone in the urinary tract influences the radiation of the pain to different locations. Several conditions of the genitourinary, ­musculoskeletal, gastrointestinal, and cardiovascular systems may produce hematuria, back pain, or abdominal pain, mimicking kidney stones (Table 243-4).

INITIAL DIAGNOSTIC EVALUATION

The initial workup of a patient with suspected renal colic includes a urinalysis, urine culture, limited blood work, and an imaging study. The urinalysis may reveal an acid pH (<5.5), suggesting uric acid or cystine stones, or an alkaline pH (>7.5), suggesting calcium phosphate stones in the setting of renal tubular acidosis, or struvite stones, associated with infection with a urea-splitting bacteria such as Proteus or Klebsiella. Microscopic examination of the urinary sediment may reveal specific crystals (Figure 243-2).

Urate crystals can be rhomboid, needle shaped, or amorphous in shape (Figures 243-2A to C). Oxalate crystals can be dumbbell or cigar shaped (Figure 243-2D, calcium oxalate monohydrate) or envelope shaped (Figure 243-2E, calcium oxalate dihydrate). ­Phosphate crystals can be in a coffin-lid structure (Figure 243-2F, triple phosphate) or a granular precipitate (Figure 243-2G, amorphous phosphate). Cystine crystals are hexaedric and flat (Figure 243-2H), and pathognomonic for cystinuria. A cyanide-nitroprusside test resulting in purple-red discoloration of the urine confirms the presence of cystine crystals.

A limited blood work panel for the evaluation of kidney stone disease includes serum electrolytes to evaluate for renal tubular acidosis; blood urea nitrogen and creatinine to assess kidney function; calcium, albumin, and phosphorus to evaluate for primary hyperparathyroidism; and uric acid to evaluate for gout. An elevated serum creatinine may reflect acute kidney injury from bilateral obstructive uropathy, unilateral obstruction in the setting of a single functioning kidney, preexisting chronic kidney disease, extracellular fluid volume depletion, or infection. A urine culture should be obtained to rule out concomitant urinary tract infection, particularly with the urea-splitting bacteria such as Proteus that lead to struvite stones. Fever should trigger a more thorough investigation for pyelonephritis and sepsis.

The imaging modality of choice is a nonenhanced helical computed tomography (CT) scan, which has 98% sensitivity and 97% specificity for the diagnosis of kidney stones as small as 1 mm in diameter. Of note, indinavir stones are radiolucent and undetectable by CT scan. Ultrasonography can only visualize the kidney and proximal ureter, detecting stones equal to or greater than 3 mm in diameter. Consequently, it has a sensitivity and specificity of 61% and 97%, respectively. However, due to the lack of radiation exposure, ultrasonography may have a role for pregnant women and children. An abdominal plain x-ray (ie, kidneys, ureters, and bladder x-ray) visualizes the entire urinary tract proximal to the urethra, detecting 90% of stones greater than 2 mm in diameter, but cannot detect radiolucent uric acid stones.

TREATMENT OF AN ACUTE EPISODE

INITIAL EMERGENCY ROOM MANAGEMENT