Disorders of Acid-Base and Potassium Balance Part 2

Causes of Metabolic Acidosis with a Normal Anion Gap

A normal anion gap in metabolic acidosis can arise from several causes, including administration of inorganic acids such as hydrochloric acid, loss of bicarbonate from the gastrointestinal tract (e.g., in diarrhea) or the kidneys (e.g., type 2 renal tubular acidosis [RTA]), and impaired renal excretion of hydrogen ions (e.g., renal insufficiency and type 1 RTA) [see Table 2]. In the resultant conditions, electroneutrality is maintained by a decrease in plasma bicarbonate and replacement by chloride. Conse-quently, these disorders are sometimes collectively referred to as the hyperchloremic acidoses.

Table 1 Causes of High-Anion-Gap Metabolic Acidosis
Renal failure*	Ingested agents and toxins
Lactic acidosis	Salicylates
Ketoacidosis*	Ethylene glycol
Rhabdomyolysis	Methanol Toluene*

Acid and chloride administration The infusion of amino acid solutions during parenteral nutrition supplies abundant hydrochloric acid. In certain settings, administration of a sodium chloride solution also may lower the plasma [HCO3-] by dilution (a condition termed expansion acidosis), leading to a fall in blood pH.

Bicarbonate or other alkali losses Bicarbonate can be lost from the body via the gastrointestinal tract or the kidneys. Compared with blood, bowel contents are alkaline because pancreatic and biliary secretions add bicarbonate, which is later exchanged for chloride in the ileum and colon to maintain a normal acid-base status. Gastrointestinal losses of bicarbonate (or bicarbonate precursors such as lactate and acetate) are most commonly observed in patients with severe diarrhea. The diagnosis of diarrhea as the cause of a normal-anion-gap acidosis is usually apparent from the patient’s history [see Diagnosis, below]. Hypokalemia resulting from stool potassium losses also supports the diagnosis.

Less commonly, metabolic acidosis resulting from gastrointestinal alkali losses is caused by pancreatic fistulas, biliary drainage, or urinary diversions to the colon or small bowel.8 It also occurs after pancreatic transplantation in patients who lose bicarbonate via a pancreas-bladder anastomosis.

Table 2—Causes of Metabolic Acidosis with a Normal Anion Gap

Acid and chloride administration

Hyperalimentation*

NH4Cl, HCl for treatment of severe metabolic acidosis*

Bicarbonate (or other alkali) losses

Gastrointestinal alkali loss+

Diarrhea, pancreatic-intestinal-biliary fistulas, pancreatic transplantation with drainage into the urinary bladder

Ureteral diversions^

Ureterosigmoidostomy, ileal bladder (if obstructed)

Type 2 RTA+

Recovery from ketoacidosis^

Posthypocapnia+

Reduced renal and hydrogen excretion

Renal insufficiency*

Type 1 RTA [see Table 3]

Hypokalemic forms*

Hyperkalemic forms+

Type 4 RTA* [see Table 4]

* Normal or increased plasma K+.

^ Normal or decreased plasma K+.

RTA—renal tubular acidosis

Renal bicarbonate losses cause the acidemia observed in type 2, or proximal, RTA and in patients who are posthypocapnic. In type 2 RTA, the normal threshold for bicarbonate reabsorption is reduced. Therefore, bicarbonate can no longer be reabsorbed at a rate adequate to maintain the normal plasma level of about 25 mEq/L. During this phase, bicarbonaturia occurs, leading to a urine pH of greater than 5.3. Bicarbonate wasting ceases, however, once the fall in plasma [HCO3-] has stabilized; at this time, the urine pH may be less than 5.

Proximal RTA can be observed in a number of circumstances, including the administration of carbonic anhydrase inhibitors and in some patients with multiple myeloma. In multiple myeloma, toxic damage to the proximal tubular cells by the filtered myeloma light chains often leads to a generalized reduction in proximal reabsorption (including glucose and phosphate) in addition to bicarbonate. This group of maladies, caused by proximal tubular dysfunction, is called the Fanconi syndrome.

Hypocapnia leads to a fall in the proximal tubular reabsorp-tion of bicarbonate. After 1 to 3 days, the plasma level of bicarbonate decreases. Because this renal adaptive process requires the same time to cease, a sudden increase in the PaCO2 does not immediately alter proximal bicarbonate reabsorption. This posthypocapnic metabolic acidosis resolves spontaneously within 24 to 72 hours.

Reduced renal hydrogen ion excretion Reduced renal acid excretion is observed in three conditions: renal insufficiency, type 1 (distal) RTA, and type 4 RTA (hypoaldosteronism) [see Tables 3 and 4]. The acidosis of renal failure is primarily caused by a reduction in the number of nephrons. In contrast, type 1 RTA is characterized by a reduction in renal acid excretion by each nephron. Because the total quantity of ammonia that can be synthesized is also reduced in renal failure, the urine pH is lower than 5.3 in most patients.

Type 1 RTA, which may be acquired in association with a number of disorders, such as Sjogren syndrome, occurs most commonly when hydrogen ions cannot be pumped out of the a-intercalated cells into the collecting tubular lumen or when they leak back into the cell from the tubule lumen [see Figure 3]. The mutations underlying congenital type 1 RTA may be caused by mechanisms similar to those found in the acquired form or by translocation of the Cl-/HCO3- transporter, from the basolateral to the apical (luminal) membrane of the a-intercalated cells.11 As a result, the urine cannot be maximally acidified, and the urine pH is always higher than 5.3. Furthermore, hypokalemia is characteristically present, in part because of enhanced distal nephron Na+-K+ exchange, a process that is necessary to maintain sodium balance because hydrogen ions cannot be secreted in response to sodium reabsorption.

A hyperkalemic form of distal RTA also has been described; it occurs most often in patients with urinary tract obstruction. This disorder is characterized by extensive tubular injury involving aldosterone-sensitive hydrogen secretion as well as aldosterone-independent hydrogen secretion. The urine pH in patients with this disorder is above 5.

The most important clinical complication of hypokalemic type 1 RTA is the formation and deposition of calcium phosphate salts that can cause calculi throughout the kidney (nephrocalcinosis). A major factor contributing to crystal formation is hypocitraturia caused by an increase in proximal tubular citrate reabsorption. Because calcium citrate is significantly more soluble than calcium phosphate, a decrease in urinary citrate facilitates the precipitation of calcium phosphate crystals in the collecting tubular lumen.

Table 3 Causes of Type 1 RTA

Hypokalemic Forms	Hyperkalemic Forms
Primary	Urinary tract obstruction Sickle cell anemia Systemic lupus erythematosus Renal transplant rejection
Idiopathic
Genetic
Familial
Marfan syndrome
Ehlers-Danlos syndrome
Nephrocalcinosis
Chronic hypercalcemia
Medullary sponge kidney
Hypergammaglobulinemic states
Amyloidosis*
Cryoglobulinemia
Cirrhosis Drugs and toxins
Amphotericin B
Lithium carbonate
Toluene+
Autoimmune diseases
Sjogren syndrome*
Thyroiditis
Chronic active hepatitis
Primary biliary cirrhosis

*May also cause type 2 RTA [see Table 5].

+May also cause metabolic acidosis with an elevated anion gap [see Table 1].

Type 4 RTA may be caused by a number of medications, including nonsteroidal anti-inflammatory drugs (NSAIDs), an-giotensin-converting enzyme (ACE) inhibitors, angiotensin II blockers, cyclosporine, and heparin. However, it is most often observed in patients with diabetes mellitus [see Plasma Potassium Disorders, Hyperkalemia, below]. The most common electrolyte disturbance in type 4 RTA is hyperkalemia (plasma [K+] > 5.0 mEq/L), which is caused by impaired luminal reabsorption of sodium ions and, thus, reduced potassium ion secretion by the principal cells of the collecting tubule [see Figure 3]. This defect is caused by a reduction in aldosterone production or action. In contrast to hypokalemia, high plasma potassium levels impair renal ammonia production. Insufficient urinary buffer in the form of ammonia limits hydrogen ion secretion in the collecting tubule. Despite this reduced ability to excrete hydrogen ions, the urine is usually acidified (pH < 5.3). This apparently paradoxical finding occurs because the limited urinary buffer permits the free urine [H+] to exceed the secretory capacity of the a-intercalated cells, thus limiting further hydrogen ion transport from the cells and into the collecting tubular lumen. In contrast to patients with type 1 RTA, patients with type 4 RTA can excrete some acid, albeit in inadequate quantities; consequently, the urine pH is typically lower than 5.3 in patients with this disorder.

Potassium Imbalance in Metabolic Acidosis

Hyperkalemia may be observed in patients with normal an-ion gap metabolic acidosis; in respiratory acidosis, the shift of potassium from cells to plasma is smaller. A fall in the plasma [HCO3-] leads to an exchange of extracellular hydrogen ions for intracellular potassium ions. This defense mechanism permits intracellular buffering of hydrogen ions; the H+-K+ exchange maintains electroneutrality. Therefore, the patient who has diarrhea with a normal-anion-gap metabolic acidosis and a low plasma [K+] has significantly greater potassium depletion than the patient with the same plasma [K+] and a normal blood pH.

In contrast to what occurs in patients with a normal-anion-gap metabolic acidosis, potassium shifts caused by acidemia are less pronounced in patients with endogenous organic acidoses (e.g., lactic acidosis and ketoacidosis).12 When hyperkalemia develops in this setting, it is typically caused by renal failure, cellular catabolism that permits leakage of potassium from cells (lactic acidosis), or insulin deficiency and hyperglycemia, both of which promote potassium exit from cells.

Diagnosis

Clinical manifestations Kussmaul respirations suggest the presence of metabolic acidosis. The increase in tidal volume, rather than respiratory rate, that characterizes these ventilatory changes results from stimulation of the brain stem respiratory system by the low pH.

Secondary hypotension also may be observed in severely aci-demic persons. In this setting, the reduced blood pressure results from depressed myocardial contractility and arterial vasodilata-tion, which are induced by the decreased blood pH. Initially, elevated levels of circulating catecholamines counter the cardiovascular effects of acidemia; however, at a blood pH below 7.15 to 7.20, the effects of acidemia may predominate.13 Reentrant arrhythmias and a reduction in the threshold for ventricular fibrillation can occur, whereas the defibrillation threshold remains unaltered.

The symptoms of metabolic acidosis caused by renal failure depend on the cause, the rapidity with which the renal failure develops, and concomitant conditions that may be present (e.g.,congestive heart failure). There may be no abnormal clinical symptoms, despite the presence of azotemia and acidemia.

Table 4 Causes of Type 4 RTA and Aldosterone Resistance

Disorder	Cause
	Reduced activity of the renin-angiotensin system
	Hyporeninemic type 4 RTA (diabetes most common)
	Nonsteroidal anti-inflammatory drugs (with the possible exception of sulindac)
	Angiotensin-converting enzyme inhibitors
	Angiotensin II receptor blockers
	Cyclosporine
	AIDS*
Type 4 RTA	Reduced aldosterone synthesis
	Low cortisol levels
	Primary adrenal insufficiency
	Enzymatic deficiencies (primarily adrenal hyperplasia)
	Normal cortisol levels
	Heparin
	Immediately after removal of adrenal adenoma in primary aldosteronism
	Enzymatic deficiencies
Aldosterone resistance (normal or increased aldosterone levels)	Potassium-sparing diuretics, trimethoprim
	Pseudohypoaldosteronism (hereditary or acquired)
	Hyperkalemic type 1 RTA+

*Adrenalitis causing type 4 RTA may also occur in persons who are seropositive for HIV.

+In this setting, the urine is alkaline (pH > 5.3).

Ketoacidosis is often associated with increased thirst and polyuria. It is frequently precipitated by an unrelated insult, such as an infection, that may dominate the clinical presentation.

The distinctive characteristics of rhabdomyolysis are myalgia and myoglobinuria, the latter of which leads to reddish urine that has no significant red cells. In many patients, however, neither finding is present. Despite the lack of findings, serum crea-tine kinase levels are uniformly elevated.

The symptoms and signs of lactic acidosis are characteristically those of the underlying disturbance causing the disorder— namely, hypotension or sepsis. Most commonly, patients exhibit evidence of hypoperfusion, such as low blood pressure and cool or mottled extremities. Less commonly, a medication (e.g., met-formin) may be the cause. Patients with D-lactic acidosis may present with encephalopathy. L-Lactate dehydrogenase is the standard assay for the diagnosis of lactic acidosis; because this test does not measure D-lactic acid, a specific enzymatic assay for D-lactate (i.e., one that uses D-lactic dehydrogenase) is necessary to confirm the diagnosis of D-lactic acidosis.

In addition to acid-base changes, clinical findings that may accompany severe salicylate intoxication include tinnitus, hyper-pyrexia, vasodilatation leading to shock, and peripheral or pulmonary edema. Symptoms of acidosis caused by ingestion of methanol or ethylene glycol [see Ingested Agents and Toxins, above] may develop 12 to 36 hours after ingestion. In addition to acid-base changes, the initial symptoms that occur after ingestion of methanol include weakness, nausea, headache, and decreased vision, which can progress to blindness, coma, and death. Fundo-scopic examination may reveal a retinal sheen caused by retinal edema. After ethylene glycol is ingested, the earliest findings are neurologic abnormalities that range from drunkenness to coma. If the patient is not treated, these changes may be followed first by cardiopulmonary symptoms (tachypnea and pulmonary edema) and then by flank pain and renal failure caused by calcium oxalate crystal deposition, which may be seen in the urine sediment [see Figure2 in 10:XIINephrolithiasis].

Laboratory tests The diagnosis of metabolic acidosis is made relatively easily in the presence of a low blood pH and a low plasma [HCO3-]. The anion gap can then be used to identify a specific cause. The finding of concomitant hypokalemia or hy-perkalemia may also be useful.

Once the presence or absence of a high serum anion gap is determined, the respiratory defense against acidemia can be evaluated. The respiratory response begins immediately, although it may not be maximal for 12 to 24 hours. The appropriate respiratory compensation can be calculated by using the Winter14 formula:

The respiratory compensation is inadequate when the measured PaCO2 is higher than the expected value and is excessive when the PaCO2 is lower than the expected value. When these inadequate or excessive responses occur, a superimposed respiratory acidosis or respiratory alkalosis, respectively, is present [see Respiratory Acidosis and Alkalosis, below]. For example, if the serum [HCO3-] were 16 mEq/L in a patient with a low blood pH (metabolic acidosis), the expected Pco2 would be approximately 32[(1.5 x 16) + 8] mm Hg. A Pco2 below this value indicates a superimposed respiratory alkalosis.

Another laboratory calculation, the urinary anion gap, may be useful in defining the cause of metabolic acidosis when the serum anion gap is normal. For example, although the diagnosis of diarrhea is usually apparent from the patient’s history and the presence of hypokalemia, a profile of plasma electrolytes similar to that in persons with diarrhea may be observed in patients with type 1 RTA. These disorders can usually be distinguished by the urine pH; the urine tends to be acidic (pH < 5.3) in patients with diarrhea and tends to be alkaline (pH > 5.3) in patients with type 1 RTA. In some patients with diarrhea, however, the urine may be alkaline, presumably because ammonia production (induced by hypokalemia) increases to such an extent that urinary buffer is produced in excess of hydrogen ion secretion.

Calculation of the urinary anion gap, as shown in the following equation, may be quite useful in providing an estimate of urinary ammonia secretion:

Whenever secreted hydrogen ions are excreted as ammonium chloride (NH4+Cl-), an increase in urinary chloride excretion results. The increase in urinary chloride excretion decreases the urinary anion gap, leading to a negative value in most patients with diarrhea. By comparison, in type 1 RTA, the urinary anion gap is positive. As an example, a patient with a normal anion gap metabolic acidosis (e.g., [HCO3-] = 10 mEq/L), a low potassium level (hypokalemia), and an alkaline urine pH (6.0) could have diarrhea or type 1 RTA. If the urine [Na+] were 50 mEq/L, the urine [K+] were 28 mEq/L, and the urine [Cl-] were 55 mEq/L, the urinary anion gap would be +23, supporting a diagnosis of RTA.

Despite the potential value of the urinary anion gap in estimating urinary NH4+Cl- excretion, the presence of certain anions (e.g., p-hydroxybutyrate and acetoacetic acid) can contribute to a positive urinary gap. This may erroneously suggest that metabolic acidosis is caused by RTA rather than diarrhea. The urinary osmolal gap, which largely represents ammonium salts, can be used to confirm that the positive urinary anion gap is the result of type 1 RTA. The urinary osmolal gap is the difference between the urinary osmolality (Uosm) measured by the laboratory and that calculated by the following formula:

The multiple of 2 accounts for the anions accompanying sodium and potassium, whereas the divisors 2.8 and 18 reflect adjustments required to convert the blood urea nitrogen and glucose findings from the routinely used units of mg/dl to mmol/L or mOsm/kg. The normal urinary osmolol gap is 80 to 100 mmol/L, reflecting an NH4+ excretion rate of about half this value—namely, 40 to 50 mEq/L—as a result of the accompanying anion chloride. In the presence of metabolic acidosis, the value should be significantly increased; a normal or lower value in an individual with metabolic acidemia strongly supports impaired NH4+Cl- excretion—and, ultimately, type 1 RTA—as the underlying cause for the positive anion gap.