Distal Renal Tubular Acidosis

In distal (classic or type 1) RTA, the urine cannot be maximally acidified because of impaired hydrogen ion secretion in the collecting ducts, and urine pH typically is above 6.0, despite moderately to markedly decreased plasma HCO₃⁻ concentration. Increased urine pH (>6.0) in the presence of acidosis is the hallmark of distal RTA. Urinary tract infection by a urease-positive organism (e.g., Proteus sp., Staphylococcus aureus) must be ruled out before considering distal RTA. Urinary net acid excretion is decreased, but bicarbonaturia usually is mild because urinary HCO₃⁻ concentration is only 1 to 3 mEq/L in the pH range of 6.0 to 6.5. Nephrolithiasis (usually calcium phosphate stones), nephrocalcinosis (resulting from alkaline urine pH and decreased urinary citrate concentration), bone demineralization (resulting from loss of bone buffer stores during chronic acidosis), and urinary potassium wasting with hypokalemia are features of distal RTA in human patients. Mutations in cytosolic carbonic anhydrase, the basolateral Cl⁻/HCO₃⁻ anion exchanger, and luminal H⁺-ATPase that affect function of the α-intercalated cells have been associated with inherited forms of distal renal tubular acidosis in humans.¹⁷³ Urinary fractional excretion of HCO₃⁻ is normal (<5%) in distal RTA when plasma HCO₃⁻ concentration is increased to normal by alkali administration.

A diagnosis of distal RTA may be confirmed by an ammonium chloride tolerance test during which urine pH is monitored (using a pH meter) before and at hourly intervals for 5 hours after oral administration of 0.2 g/kg NH₄Cl. Under such conditions, the urine pH of normal dogs decreased to a minimum value of 5.16 at 4 hours after administration of ammonium chloride.²¹⁴ Dogs in this study also developed systemic acidosis (pH approximately 7.22 and HCO₃⁻ approximately 14 mEq/L at 2 hours after ammonium chloride administration). The amount of alkali required to correct the acidosis in human patients with distal RTA is variable but typically less than that required in proximal RTA. The required dosage of alkali in distal RTA may be as little as 1 mEq/kg/day (i.e., that required to offset daily endogenous acid production) or more than 2 to 4 mEq/kg/day. A combination of potassium and sodium citrate (depending on potassium balance) may be the preferred source of alkali.¹⁹⁶

Proximal Renal Tubular Acidosis

In proximal (type 2) RTA, renal reabsorption of HCO₃⁻ is markedly reduced and urinary fractional excretion of HCO₃⁻ is increased (>15%) when plasma HCO₃⁻ concentration is increased to normal. Bicarbonaturia is absent and urine pH is appropriately low when metabolic acidosis is present and plasma HCO₃⁻ concentration is decreased because distal acidifying ability is intact. When plasma HCO₃⁻ concentration is decreased, the filtered load of HCO₃⁻ is reduced, and almost all of the filtered HCO₃⁻ is reabsorbed in the distal tubules, despite the presence of the proximal tubular defect. Thus, proximal RTA can be viewed as a “self-limited” disorder in which plasma HCO₃⁻ stabilizes at a lower than normal concentration after the filtered load falls sufficiently enough that distal HCO₃⁻ reabsorption can maintain plasma HCO₃⁻ at a new but lower steady-state concentration. Mutations in renal tubular transport proteins, such as the electrogenic basolateral Na⁺/ 3HCO₃⁻ cotransporter⁷³ and one of the five forms of the luminal Na⁺/H⁺ antiporter, have been implicated in the pathogenesis of inherited forms of proximal renal tubular acidosis in humans.¹¹⁸

Other abnormalities of proximal tubular function typically accompany impaired HCO₃⁻ reabsorption in proximal RTA, and these include defects in glucose, phosphate, sodium, potassium, uric acid, and amino acid reabsorption. This combination of proximal tubular defects is known as Fanconi syndrome. Serum potassium concentration usually is normal in affected human patients at the time of diagnosis, but alkali therapy may precipitate hypokalemia and aggravate urinary potassium wasting, presumably by increasing distal delivery of sodium and HCO₃⁻.

The diagnosis of proximal RTA is made by finding an acid urine pH (<5.5 to 6.0) in the presence of hyperchloremic metabolic acidosis and a normal GFR but an increased urine pH (>6.0) and increased urinary fractional excretion of HCO₃⁻ (>15%) after plasma HCO₃⁻ concentration has been increased to normal by alkali administration. If present, the detection of other defects in proximal tubular function (e.g., glucosuria with normal blood glucose concentration) establishes the diagnosis. Correction of metabolic acidosis by alkali therapy is more difficult in proximal RTA than in distal RTA because of the marked bicarbonaturia that occurs when plasma HCO₃⁻ concentration is increased to normal. Alkali dosages in excess of 10 mEq/kg/day may be required to correct the plasma HCO₃⁻ concentration, and such therapy may result in frank hypokalemia. Thus, potassium citrate may be the preferred source of alkali.

Multiple renal tubular reabsorptive defects resembling Fanconi syndrome have been reported in young basenji dogs.^26-28,80 Clinical findings included polyuria, polydipsia, weight loss, dehydration, and weakness. Affected dogs had abnormal fractional reabsorption of glucose, bicarbonate, phosphate, sodium, potassium, and urate, and they had isolated cystinuria or generalized aminoaciduria. The renal tubular disorder in affected basenji dogs is thought to be the result of a metabolic or membrane defect affecting sodium movement or increased back leak or cell-to-lumen flux of amino acids. In one study, brush border membranes isolated from basenji dogs with Fanconi syndrome had decreased sodium-dependent glucose transport but no abnormality of cystine uptake despite the observed reabsorptive defect for cystine.¹⁵⁷ Defective urinary concentrating ability leads to isosthenuria or hyposthenuria, and the GFR may be normal initially but decreased later in the course of the disease. Hypokalemia has also been observed late in the course of the disease.⁸⁰ Death usually results from acute renal failure and papillary necrosis or acute pyelonephritis. A distinctive renal lesion is hyperchromatic karyomegaly of renal tubular cells.

Fanconi syndrome has been observed sporadically in other breeds^{81,143,156,182,213} and has been reported in association with administration of some drugs.^16,28,160 In one case, Fanconi syndrome developed in association with primary hypoparathyroidism and resolved after treatment with calcium and calcitriol.⁸⁸ Rickets in growing children and osteomalacia in adults are features of Fanconi syndrome in human patients that usually are not observed in affected dogs. However, congenital Fanconi syndrome and renal dysplasia were associated with histologic features of rickets in two Border terriers.⁶⁴ The skeletal abnormalities in one of the affected dogs resolved after treatment with calcitriol and potassium phosphate. Transient Fanconi syndrome and proximal renal tubular acidosis also have been reported in a dog with high liver enzyme activities, and toxin exposure was considered as a possible explanation.¹¹⁵ Idiopathic transient renal tubular dysfunction also has been reported in a Labrador retriever¹²⁶ and greyhound.¹ Fanconi-like syndrome occurred in Australian dogs that had been fed dried chicken treats from China in 2007 and another product (not containing chicken and not from China) in 2009.⁸⁷ Affected dogs had polyuria, polydipsia, glucosuria, acidosis, hypokalemia, hypophosphatemia, and azotemia. Most of them survived with conservative medical management. Finally, Fanconi syndrome has been reported in several dogs with copper storage hepatopathy, and tubular dysfunction resolved after copper chelation therapy.^10,111

In one report, an 8-year-old female German shepherd had hyperchloremic metabolic acidosis, polyuria, polydipsia, isosthenuria, glucosuria with normal blood glucose concentration, and alkaline urine pH (7.46) after oral administration of NH₄Cl.⁷⁰ The metabolic acidosis was unresponsive to NaHCO₃ administration at dosages up to 4 mEq/kg/day. This dog appeared to have distal (type 1) RTA and renal glucosuria. In another case of apparent distal RTA, a 5-year-old mixed breed dog was presented for evaluation of anorexia and was determined to have alkaline urine pH with hyperchloremic metabolic acidosis.¹⁹¹ In another report, an 8-year-old female German shepherd was presented for polyuria, polydipsia, weight loss, and lethargy.²⁴ It had a normal GFR, metabolic acidosis, hyposthenuria, and intermittent glucosuria. Fractional reabsorption of sodium, glucose, and HCO₃⁻ was decreased, but reabsorption of chloride, phosphate, potassium, urate, and amino acids was normal. The dog gained weight, and its clinical signs were reversed after treatment with NaHCO₃ at approximately 10 mEq/kg/ day. This dog appeared to have proximal (type 2) RTA.

Distal RTA has been reported in two cats with pyelonephritis caused by Escherichia coli.^77,236 Clinical signs included polyuria, polydipsia, anorexia, lethargy, enlarged kidneys, and isosthenuria. In one cat, urine pH was 5.0 at the time pyelonephritis was first diagnosed, but distal RTA was documented at a later time by the presence of hyperchloremic metabolic acidosis, alkaline urine pH, and failure to lower urine pH after oral administration of NH₄Cl.⁷⁷ Findings were similar for the other cat, but hyperphosphaturia and persistent hypokalemia also were detected.²³⁶ Distal RTA and hepatic lipidosis were reported in another cat without urinary tract infection²⁹ and in a cat with concurrent hyperaldosteronism and severe hypokalemia.²²⁸ Distal renal tubular acidosis also has been reported in association with immune-mediated hemolytic anemia in three dogs.²¹⁵ Distal renal tubular acidosis is associated with some immune-mediated diseases in human patients, but not specifically immune-mediated hemolytic anemia. The clinical features of proximal (type 2) and distal (type 1) RTA are summarized in Table 10-2.

Table 10-2 Clinical Features of Proximal and Distal Renal Tubular Acidosis

Fe, fractional excretion.

* Decreased fractional reabsorption of sodium, potassium, phosphate, urate, glucose, and amino acids.

Hyporeninemic hypoaldosteronism, characterized by hyperkalemia with decreased plasma renin and aldosterone concentrations, occurs in some human patients, notably those with diabetes mellitus who also have mild to moderate renal insufficiency.⁶⁵ The hyperchloremic metabolic acidosis observed in these patients has been called Type 4 RTA. This syndrome has not been characterized in veterinary medicine but should be considered in dogs and cats with hyperkalemia and mild to moderate hyperchloremic metabolic acidosis after hypoadrenocorticism has been ruled out by an adrenocorticotropic hormone (ACTH) response test. The diagnosis may be established by finding an inappropriately decreased plasma aldosterone concentration in the presence of hyperkalemia.

Pathophysiology

EG is first metabolized in the liver to glycoaldehyde by alcohol dehydrogenase. Glycoaldehyde uncouples oxidative phosphorylation and may contribute to neurologic signs observed early in the course of intoxication. Subsequent steps in metabolism produce glycolic and glyoxylic acids. Glycolic acid is primarily responsible for the severe metabolic acidosis that occurs in animals poisoned by EG.⁵⁰ Renal tubular injury results from glycoaldehyde, glycolic acid, and glyoxylic acids, and calcium oxalate crystals are deposited within renal tubules. The observation of these birefringent crystals in the presence of acute tubular nephrosis confirms the diagnosis of EG intoxication.

Vomiting, polydipsia, and polyuria may occur soon after ingestion of EG, but the owners of poisoned animals often do not detect these signs. Within 12 hours of ingestion, neurologic signs (e.g., lethargy, ataxia, stupor, seizures, coma) may develop. Cardiac and pulmonary manifestations (e.g., tachypnea, tachycardia) occur 12 to 24 hours after ingestion but rarely are detected in clinical cases. Oxalate crystals may be detected in the urine as early as 3 to 6 hours after ingestion of EG.^68,69 Renal failure occurs in dogs as early as 24 to 48 hours after ingestion and is manifested by anorexia, lethargy, vomiting, and oliguria or anuria.⁹⁷ In cats, azotemia may develop within 12 to 24 hours after ingestion of EG.⁶⁸ Unfortunately, most dogs and cats with EG poisoning are presented for veterinary attention after renal failure has already developed.

A severe normochloremic (i.e., high anion gap) metabolic acidosis occurs within 3 hours of EG ingestion and persists for at least 24 hours.^68,69,97,227 Serum hyperosmolality and osmolal gap peak 1 to 6 hours after ingestion and persist for 12 to 24 hours,^68,69,97 but the osmolal gap may be normal in animals presented later in the course of the disease.²²⁷ Activated charcoal preparations containing propylene glycol and glycerol can increase osmolality and osmolal gap, and potentially complicate the diagnosis of EG ingestion. Measured serum osmolality peaked at 4 hours (353 mOsm/kg), osmolal gap at 6 hours (52 mOsm/kg), and serum lactate concentration at 4 hours (4.5 mmol/L) after administration of 4 g/kg of an activated charcoal preparation containing propylene glycol and glycerol.³³ Results returned to baseline 24 hours after administration of the activated charcoal preparation.

Calcium oxalate dihydrate crystals (“Maltese cross” or “envelope” forms) may be observed in the urine, but calcium oxalate monohydrate crystals (“picket fence” or “dumbbell” forms) are observed more commonly. Calcium oxalate dihydrate crystals occasionally are found in the urine of normal dogs and cats, whereas calcium oxalate monohydrate crystals rarely are seen except in animals that have ingested EG (Fig. 10-5).^68,227 Crystals previously referred to as hippurates actually are calcium oxalate monohydrate crystals.^134,226 Other laboratory findings include azotemia, isosthenuria, hypocalcemia, hyperphosphatemia, and hyperglycemia.²²⁷ Hyperphosphatemia observed very early in the course of EG intoxication (3 to 12 hours after ingestion) probably is the result of the high phosphorus content of rust-retardant antifreeze preparations.^57,69 Hyperechogenicity of the renal cortex is observed on renal ultrasonography as early as 5 hours after ingestion of EG.²

Figure 10-5 Photomicrographs of (A) calcium oxalate monohydrate and (B) dihydrate crystals in urine sediment.

(From Chew DJ, DiBartola SP. Diagnosis and pathophysiology of renal disease. In: Ettinger SJ, editor. Textbook of veterinary internal medicine. Philadelphia: WB Saunders, 1989: 1907.)

Treatment

The response to treatment depends on the amount of EG ingested and the amount of time that elapses before treatment. In early studies, dogs that ingested less than 10 mL/kg EG were saved if treated within 2 to 4 hours of ingestion,^17,175,205 and cats survived up to 6 mL/kg EG if treated within 4 hours.¹⁸⁷ Treatment consists of inducing vomiting with apomorphine or performing gastric lavage with activated charcoal if ingestion has been recent (<8 hours before presentation). Severe hypocalcemia is corrected with calcium gluconate, and NaHCO₃ is administered to combat metabolic acidosis. A NaHCO₃ dosage of 1 to 2 mEq/kg may be used empirically. Calcium gluconate and NaHCO₃ must not be given simultaneously because calcium carbonate crystals form, and the solution becomes turbid. Attempts to stimulate urine production with furosemide (2 to 4 mg/kg) or mannitol (1 g/kg) usually are futile.

Alcohol dehydrogenase has greater affinity for ethanol than EG. For this reason, 20% ethanol has been administered intravenously to affected dogs at a dosage of 5.5 mL/kg every 4 hours for five treatments and then every 6 hours for four additional treatments.⁹⁶ Cats are treated with 20% ethanol at a dosage of 5 mL/kg every 6 hours for five treatments and then every 8 hours for four additional treatments. This treatment is unlikely to be of benefit if more than 12 to 24 hours have elapsed since ingestion of EG. Fomepizole (4-methylpyrazole) is a pharmacologic inhibitor of alcohol dehydrogenase that can be used to treat dogs with EG toxicosis.^67,69 In dogs, it is superior to ethanol because it does not cause central nervous system (CNS) depression, but it must be administered within 8 hours of EG ingestion. The dosage of fomepizole used in dogs with EG intoxication is 20 mg/kg intravenously, followed by 15 mg/kg intravenously at 12 and 24 hours and 5 mg/kg intravenously at 36 hours.^57,67,69 Unfortunately, fomepizole was not efficacious in EG-intoxicated cats unless administered at the same time as the EG was consumed.⁶⁸ A study to investigate the difference in efficacy of fomepizole between dogs and cats found that the percentage inhibition of canine and feline alcohol dehydrogenase was similar when the concentration of fomepizole applied to feline liver homogenates was 6 times higher than that applied to canine liver homogenates.⁵⁸ When cats that received lethal doses of EG were treated within 3 hours of ingestion using 125 mg/kg fomepizole followed by 31 mg/kg at 12, 24, and 36 hours, 5 of 6 survived.⁵⁹ One cat developed acute renal failure but recovered. Cats treated with this high dosage of fomepizole developed mild sedation, but no biochemical evidence of toxicity was identified.

Thiamine promotes conversion of glyoxylate to glycine, and pyridoxine promotes conversion of glyoxylate to α-hydroxy-β-ketoadipate (see Fig. 10-4). These vitamins may be administered to promote alternative pathways of glyoxylate metabolism, but efficacy has not been demonstrated for such treatment. In one study, all nonazotemic dogs treated with fomepizole within 2 to 8.5 hours after EG ingestion survived, whereas only 1 of 21 azotemic dogs treated 8.5 to 38 hours after ingestion survived.⁵⁷

Peritoneal dialysis or hemodialysis is necessary if the animal has anuric or oliguric renal failure at the time of presentation. Early dialysis may also be helpful to remove toxic intermediate metabolites. Despite dialysis, affected dogs may progress to end-stage renal disease and become dependent on dialysis. The prognosis for survival in adult dogs and cats with anuric or oliguric acute renal failure caused by EG intoxication is unfortunately very poor.^57,227

Pathophysiology

Overproduction of acetoacetic acid (pK’_a = 3.58) and β-hydroxybutyrate (pK’_a = 4.70) by the liver occurs in diabetes mellitus because of a deficiency of insulin and relative excess of glucagon. An increase in glucagon and a decrease in insulin shift the liver from its normal role in esterification of fatty acids into triglycerides to β-oxidation of fatty acids into ketoacids. At the normal pH of ECF (7.40), these organic acids are completely dissociated, and the hydrogen ions that are released titrate HCO₃⁻ and other body buffers. Acetone is formed by the nonenzymatic decarboxylation of acetoacetate and does not contribute additional fixed acid. The pathophysiology and treatment of diabetic ketoacidosis are discussed in detail in Chapter 20.

Metabolic acidosis is common in dogs and cats with diabetic ketoacidosis. In one series, mean plasma HCO₃⁻ concentration in 72 dogs with diabetic ketoacidosis was approximately 11 mEq/L at the time of diagnosis with a range of 4 to 20 mEq/L, whereas the mean HCO₃⁻ concentration in 20 affected cats was 13 mEq/L with a range of 8 to 22 mEq/L.⁸³ In an early study of dogs with diabetes mellitus, mean plasma HCO₃⁻ concentration was 13.7 mEq/L in eight survivors (range, 9.3 to 21.0 mEq/L) and 18.1 mEq/L in five nonsurvivors (range, 13.4 to 30.2 mEq/L).¹³⁸ In another study of dogs with diabetic ketoacidosis, mean arterial pH and HCO₃⁻ concentration were 7.201 (range, 6.986 to 7.395) and 11.1 mEq/L (range, 4.1 to 19.7 mEq/L) before treatment and 7.407 ± 0.053 and 18.2 ± 0.7 mEq/L 24 hours after treatment.¹⁴² Only three dogs (those with pH <7.1) received sodium bicarbonate treatment. Metabolic acidosis with median pH of 7.14 (range, 7.04 to 7.24) and HCO₃⁻ concentration of 10 mEq/L (range, 6 to 15 mEq/L) was found in 25 of 33 cats evaluated by venous blood gas analysis in a survey of cats with diabetic ketoacidosis.³¹ Cats with HCO₃⁻ concentrations below 14 mEq/L received bicarbonate supplementation of their fluids. In another series of diabetic cats, median total CO₂ was 13 mEq/L in ketoacidotic cats and 15 mEq/L in nonketoacidotic cats.⁶³ In a study of 116 dogs with diabetes mellitus, 43 (37%) had diabetic ketoacidosis with median venous blood pH of 7.228 (range, 6.979 to 7.374) and median bicarbonate concentration of 10.1 mEq/L (range, 4.0 to 19.3 mEq/L).⁷⁸ In a study of 127 dogs with ketoacidosis, acid-base status at presentation had substantial impact on outcome.¹¹⁷ Nonsurvivors had lower venous pH and larger base deficits, and for each unit improvement in base deficit there was a 9% increase in likelihood of discharge from the hospital.

The nitroprusside reagent (e.g., Acetest, Bayer, Tarrytown, N.Y.) detects only ketone (−C=O) groups (e.g., acetoacetate, acetone). The concentration of β-hydroxybutyrate typically exceeds that of acetoacetate in uncontrolled diabetic ketoacidosis, and the dipstick reaction underestimates the degree of ketonuria. This problem can be overcome by adding a few drops of hydrogen peroxide to urine, which nonenzymatically converts β-hydroxybutyrate to acetoacetate.¹⁷¹ When insulin is administered and metabolism of ketones proceeds, there is a shift toward acetoacetate, and the dipstick reaction transiently becomes more strongly positive. This possibility should be recognized by the clinician and should not cause concern. In a study of 116 diabetic dogs (of which 88 had not previously received insulin), all ketotic and ketoacidotic dogs and 21 of 32 (66%) “nonketotic” dogs (i.e., negative urine dipstick test for ketones) had abnormally high serum β-hydroxybutyrate concentrations (>0.15 mmol/L) at presentation.⁷⁸ Although not as readily available, measurement of plasma ß-hydroxybutyrate concentrations is more valuable than use of dipstick tests in the characterization of ketonemia in diabetic dogs and cats.^74,242,243 The increase in unmeasured anions (as reflected in the anion gap) gives a rough estimate of the concentration of ketoanions in serum. However, this estimate is inaccurate if lactic acidosis develops because lactate also is an unmeasured anion. In one study of diabetic dogs, however, acidosis was correlated primarily with serum ketone concentration, and not with serum lactate concentration.⁷⁹

To some extent, the anions of these ketoacids are excreted in urine along with sodium and potassium for electroneutrality. These organic anions are lost from the body and cannot be metabolized to HCO₃⁻ after correction of diabetic ketoacidosis with insulin therapy. Their loss thus contributes to depletion of body buffer and cation stores. Osmotic diuresis is induced by hyperglycemia and also contributes to the whole-body cation deficit. The extent of impairment in renal function may determine whether patients with diabetic ketoacidosis have an increased anion gap metabolic acidosis or hyperchloremic metabolic acidosis at the time of presentation. Patients with severe volume depletion have an increased anion gap because of retention of ketoanions, whereas those without volume depletion have hyperchloremia as a result of increased urinary excretion of the sodium and potassium salts of ketoanions and retention of chloride.^5,9

Treatment

The best treatment for the acidosis of uncontrolled diabetes mellitus is fluid therapy and insulin. Insulin administration allows glucose use by skeletal muscle and adipose tissue, decreases hepatic glucose production, prevents lipolysis and ketogenesis, and permits peripheral metabolism of ketoacids. Several regimens for administration of insulin to ketoacidotic dogs and cats have been described.⁸⁴ The particular protocol of insulin administration is probably less crucial to the ultimate outcome than the individualized care provided by the veterinarian during management of the diabetic animal.

Several factors may contribute to a delay in the repair of the HCO₃⁻ deficit in patients with diabetic ketoacidosis.¹⁰⁰ Ketoacid anions that have been excreted in the urine are lost to the body and cannot be metabolized to HCO₃⁻. After treatment with fluids and insulin, recovery may be faster in patients with a high anion gap because the retained ketoanions are metabolized, yielding HCO₃⁻.^7,9 Thus, withholding alkali may be more rational for diabetic patients with high anion gap metabolic acidosis than for those with hyperchloremic metabolic acidosis. Dilutional acidosis may occur if ECF volume (ECFV) is expanded with alkali-free solutions such as 0.9% saline. If hyperventilation persists, it may impair renal reabsorption of HCO₃⁻, and renal acid excretion may require several days to become fully augmented.

The use of NaHCO₃ to treat diabetic ketoacidosis is highly controversial, and clear benefits of its use have not been demonstrated in human patients. For example, there was no difference in recovery (based on rate of decrease of blood glucose and ketone concentrations and rate of increase of blood or cerebrospinal fluid [CSF] pH or HCO₃⁻ concentration) when NaHCO₃ was or was not administered to human patients with diabetic ketoacidosis who presented with blood pH values in the range of 6.90 to 7.14.¹⁶⁶ In another study, treatment with NaHCO₃ delayed resolution of ketosis in diabetic ketoacidosis.¹⁷⁸

There are several theoretical arguments against the use of NaHCO₃ in diabetic ketoacidosis. Acidosis in the CNS may develop after NaHCO₃ administration. The blood-brain barrier is permeable to CO₂ but less permeable to the charged HCO₃⁻ ion. If NaHCO₃ is administered, pH increases in ECF as the HCO₃⁻/ P_CO2 ratio increases, and compensatory hyperventilation decreases somewhat. As a result, P_CO2 increases and CO₂ diffuses into the CNS. However, bicarbonate diffusion into CNS lags behind that of CO₂. During this time, the HCO₃⁻ P_CO2 ratio and pH in the CNS may decrease. This has been referred to as paradoxical CNS acidosis.¹⁹² The frequency of occurrence of this complication and its clinical significance are uncertain.¹³⁵

The pathophysiology of diabetic ketoacidosis also affects oxygen delivery to tissues. Chronic acidosis shifts the oxygen-hemoglobin dissociation curve to the right, thus enhancing delivery of oxygen to the tissues. Conversely, phosphorus deficiency in diabetes decreases red cell 2,3-diphosphoglycerate concentration and causes a shift of the oxygen-hemoglobin dissociation curve back to the left. Correction of acidosis with NaHCO₃ shifts the curve farther to the left and potentially decreases oxygen delivery to tissues. However, administration of insulin and fluid therapy also lead to correction of the acidosis and should have a similar effect on the oxygen-hemoglobin dissociation curve.

Overzealous therapy with NaHCO₃ may contribute to late development of metabolic alkalosis because insulin promotes metabolism of retained ketoacid anions to HCO₃⁻. This excess HCO₃⁻ should be readily excreted in the urine if renal function is adequate. Other potentially detrimental effects of NaHCO₃ therapy include aggravation of hyperosmolality as a consequence of the obligatory sodium load, tetany resulting from a sudden decrease in ionized serum calcium concentration, and precipitation of severe hypokalemia as extracellular potassium ions move into cells during administration of insulin and correction of acidosis. For all these reasons, NaHCO₃ is not used unless severe acidosis (pH <7.1 to 7.2) is present and then only in small amounts (see the Treatment of Metabolic Acidosis section).

Pathophysiology

The metabolic acidosis of chronic renal failure is usually mild to moderate in severity (plasma HCO₃⁻ concentration, 12 to 15 mEq/L) and may be hyperchloremic early in the course of the disease process.²³⁹ Later in the course of the disease, the anion gap increases because of retention of phosphates, sulfates, and organic anions. Acid-base status is usually well preserved in chronic renal failure until GFR decreases to 10% to 20% of normal. In retrospective studies of small animal patients with chronic renal failure, plasma HCO₃⁻ concentrations were less than 16 mEq/L in 40% of dogs with chronic renal failure caused by amyloidosis⁷² and less than 15 mEq/L in 63% of cats with chronic renal failure of various causes.⁷¹ A high anion gap was observed in 43% of affected dogs (>25 mEq/L) and in 19% of affected cats (>35 mEq/L) in these studies. In acute renal failure, there has been insufficient time for the kidneys to adapt to the disease state, and the metabolic acidosis of acute renal failure is usually more severe than that observed in chronic renal failure. Complications such as sepsis and marked tissue catabolism may contribute to the severity of metabolic acidosis in acute renal failure.

Delivery of HCO₃⁻ from the proximal tubules to the distal nephron is increased in chronic renal failure.²³⁵ In dogs with experimentally induced unilateral renal disease, renal HCO₃⁻ reabsorption was not different in the diseased and control kidneys, but bicarbonaturia developed when the normal kidney was removed, and the contralateral diseased kidney was forced to function in a uremic environment.¹⁶⁵ The osmotic diuresis characteristic of uremia may thus contribute to the increased delivery of HCO₃⁻ to the distal tubules. Increased parathyroid hormone concentration as a result of renal secondary hyperparathyroidism does not seem to have important adverse effects on HCO₃⁻ reabsorption in experimentally induced renal disease in dogs.^12,206,207 The ability to lower urine pH maximally is preserved in chronic renal failure.

The main method by which the diseased kidney responds to chronic retention of fixed acid is by enhanced renal ammoniagenesis. Total ammonium excretion decreases during progressive chronic renal disease, but ammonium excretion is observed to be markedly increased when expressed per 100 mL GFR or per remnant nephron.^75,202 On a per nephron basis, the diseased kidney can increase its ammonium excretion threefold to fivefold.^219,235,238 This adaptive mechanism seems to be fully expended when the GFR decreases to less than 20% of normal. At this point, the diseased kidneys can no longer effectively cope with the daily fixed acid load, and a new steady state is established at a lower than normal plasma HCO₃⁻ concentration. The relatively mild decrease in plasma HCO₃⁻ concentration that is observed in chronic renal failure has been attributed to the contribution of the large reservoir of buffer (e.g., calcium carbonate) in bone. However, the capacity of the skeleton to buffer the amount of acid that accumulates in long-standing chronic renal failure has been questioned.¹⁷⁷ The decrease in total ammonium excretion that occurs in chronic renal failure may be counterbalanced by decreased urinary excretion of organic anions (e.g., citrate, lactate, pyruvate, ketoanions).⁵⁶ Metabolism of these retained organic anions would result in a net gain of HCO₃⁻ that would offset the decreased excretion of H⁺ in the form of NH₄⁺.

The amount of phosphate buffer available in urine in chronic renal failure is relatively fixed and likely to be at its maximum because of hyperphosphatemia and the effects of increased plasma parathyroid hormone concentration.^202,219 Furthermore, phosphorus binders and dietary phosphorus restriction are commonly used to treat chronic renal failure and may limit the amount of phosphate that can contribute to titratable acidity. When expressed on a per nephron basis, however, titratable acidity is increased in chronic renal failure.¹⁶⁴

Treatment

Whether to treat well-compensated mild to moderate metabolic acidosis in adult patients with chronic renal failure is controversial. The potential benefits of such treatment include minimizing potential depletion of bone buffers, preventing the catabolic effects of uremic acidosis on muscle protein, preventing tubulointerstitial damage resulting from complement activation by ammonia, and improving the patient’s ability to combat a superimposed acidotic crisis (e.g., acute diarrhea).¹⁹⁷ Thus, treatment with oral NaHCO₃ at a dosage of 0.5 to 1.0 mEq/kg/day or an amount sufficient to maintain plasma HCO₃⁻ concentration at 15 mEq/L or above is reasonable if the patient can tolerate the associated sodium load. One teaspoon of baking soda contains 5 g NaHCO₃ (1.3 g of which is sodium). An advantage of using calcium carbonate (e.g., Tums [GlaxoSmithKline, Brentford, UK], Os-Cal [GlaxoSmithKline]) as a phosphorus binder in chronic renal failure is that this compound can serve as both a source of alkali and a source of calcium, if small amounts of calcitriol (2 to 3 ng/kg/day) are also provided. The patient should be monitored for development of hypercalcemia when calcium carbonate and calcitriol are administered concurrently. Potassium and sodium citrate should not be used for alkali therapy in chronic renal failure patients that also are being treated with aluminum-containing phosphorus binders (e.g., aluminum hydroxide, aluminum carbonate) because citrate can increase aluminum absorption from the gastrointestinal tract in this clinical setting.¹⁶²

Normal Physiology

Lactate is a metabolic end product. Its production allows regeneration of cytosolic nicotinamide adenine dinucleotide (NAD⁺) during anaerobic metabolism, and its ultimate fate is reoxidation back to pyruvate:

The equilibrium of this reaction is far to the right, and the normal ratio of lactate to pyruvate is 10:1. The main determinants of cytosolic lactate concentration are the concentration of pyruvate and the NADH/NAD⁺ ratio, both of which are affected by mitochondrial oxidative function.

Pyruvate is produced in the cytosol by anaerobic glycolysis (Embden-Meyerhof pathway). Under aerobic conditions, NADH is oxidized to NAD⁺ in the mitochondria and pyruvate enters the mitochondria for conversion to acetylcoenzyme A (CoA) and use in the tricarboxylic acid (Krebs) cycle, or it is converted to oxaloacetate and used for gluconeogenesis in the liver and renal cortex. Under anaerobic conditions (e.g., tissue hypoxia), oxidative pathways in the mitochondria are disrupted, and NAD⁺ must be replenished by reduction of pyruvate to lactate in the cytosol. Thus, lactate accumulation is the price to be paid for maintaining energy production under anaerobic conditions.

At rest, skin, red cells, brain, skeletal muscle, and gut all produce lactate. During tissue hypoxia, skeletal muscle and gut become the major producers of lactate. The liver and kidneys are the main consumers of lactate, using it for gluconeogenesis (primarily in the liver) or oxidizing it to CO₂ and water. Protons are consumed when lactate is metabolized:

Gluconeogenesis

Oxidative metabolism

Both of these reactions require normal mitochondrial oxidative function. The protons are consumed when adenosine triphosphate (ATP) is synthesized from adenosine diphosphate (ADP) and when NADH is oxidized to NAD⁺ in the mitochondria.^136,144 Protons are released by hydrolysis of ATP to ADP and by reduction of NAD⁺ to NADH, reactions that occur mainly in the cytosol. The protons do not arise from dissociation of lactic acid because the anion lactate is the predominant metabolite at normal hepatocyte pH_i (pH_i = 7.00 to 7.20). Thus, lactic acidosis reflects the imbalance between ATP hydrolysis and synthesis and between reduction and oxidation of NAD⁺. The protons produced during anaerobic glycolysis are buffered by bicarbonate and nonbicarbonate buffers. Protons are consumed and the buffers replenished when lactate is metabolized to glucose or oxidized to CO₂ and water.

Pathophysiology

Lactic acidosis occurs when production of lactate by muscle and gut exceeds its use by liver and kidneys. Both pathways of lactate use depend on intact mitochondrial oxidative function, and clinical settings characterized by tissue hypoxia are the most common causes of lactic acidosis (see Box 10-2). Hepatic uptake of lactate is decreased when arterial PO₂ decreases to approximately 30 mm Hg.²²⁵ Severe acidosis further impairs hepatic uptake of lactate, and the liver eventually becomes a producer rather than a consumer of lactate.¹⁴⁰

In an experimental model of hypoxic lactic acidosis (type A) induced by ventilating dogs with 8% O₂, lactate concentration was more than 5 mEq/L, pH was less than 7.2, HCO₃⁻ concentration was less than 12 mEq/L, PO₂ was less than 30 mm Hg, and hepatocyte pH_i was less than 7.00.¹¹ When a similar degree of acidosis was created by infusing lactic acid into dogs with normal PO₂, hepatocyte pH_i remained greater than 7.00, and hepatic extraction of lactate (as a percentage of the delivered load) was approximately three times higher than that observed in the hypoxic animals. Hypoxemia reduces hepatic O₂ uptake, and hepatocyte pH_i decreases, presumably as a result of CO₂ accumulation within cells. This study demonstrated that impaired hepatic extraction of lactate is related to decreased hepatic O₂ uptake and pH_i but not to arterial pH. During severe hypoxia, increased lactate production by gut and muscle and decreased hepatic extraction of lactate lead to progressive lactic acidosis. Impaired hepatic extraction of lactate and increased splanchnic production also contribute to the lactic acidosis of sepsis in dogs.⁴⁸

Clinical Features

Lactic acidosis may occur in several clinical settings, especially those associated with poor perfusion and tissue hypoxia (e.g., cardiac arrest and cardiopulmonary resuscitation, shock, left ventricular failure). The clinician should strongly consider the possibility of lactic acidosis in such settings (see Box 10-2). Usually, lactic acidosis results from accumulation of the L isomer of lactate. D-Lactic acidosis, characterized by the accumulation of the D isomer, is rare but has been reported in human patients with “short-bowel syndrome” in whom gut bacteria metabolize glucose to D-lactate. Increased concentrations of D-lactate have been observed in cats fed propylene glycol^45,46 and in cats with diabetic ketoacidosis, possibly as a result of hepatic ketone metabolism.⁴⁷ Severe D-lactic acidosis has been documented in a cat with pancreatic insufficiency, likely as a consequence of intestinal bacterial overgrowth.¹⁸⁴

Lactic acidosis should be suspected whenever there is an unexplained increase in unmeasured anions (i.e., an unexplained increase in the anion gap). Confirmation requires measurement of plasma lactate concentration, but this has not been performed commonly in small animal practice. Care should be taken to prevent vascular stasis when collecting venous blood for lactate determinations, and blood samples should be centrifuged immediately after collection to prevent a spurious increase in lactate concentration related to anaerobic glycolysis by red cells. Lactate concentrations in dogs have been reported in many experimental studies.* From results of these studies, normal plasma lactate concentrations in dogs are expected to be less than 2 mEq/L. Control plasma lactate concentrations in cats were 1.46 mEq/L in one study.¹³ In an experimental model of hemorrhagic shock in dogs, plasma lactate concentration increased from 1.5 to 5.5 mEq/L but did not completely account for the observed increases in anion gap and strong ion gap.³⁰ Other organic anions (especially acetate and citrate) also contributed to the changes in the anion gap and strong ion gap.

Racing caused venous lactate concentrations in greyhounds to increase from 0.57 to 28.93 mEq/L, but lactate concentrations returned to 0.53 mEq/L 3 hours after exercise.¹¹⁹ Arterial pH decreased from 7.365 to 6.997 and returned to 7.372 3 hours after exercise, and HCO₃⁻ concentration decreased from 21.1 to 3.1 mEq/L and returned to 20.5 mEq/L 3 hours after exercise. Plasma potassium concentration does not increase in response to organic acidosis as it does in acute mineral acidosis.⁶ In the racing greyhounds, there was no change in plasma potassium concentration despite severe lactic acidosis.

Cardiac Arrest and Cardiopulmonary Resuscitation

Oxygen delivery to, and CO₂ removal from, tissues are dependent on adequate tissue perfusion. Cardiac arrest is an extreme example of impaired tissue perfusion. During cardiopulmonary resuscitation (CPR), reduced tissue perfusion and reduced O₂ delivery cause anaerobic metabolism and lactic acidosis. In dogs, lactate concentrations increased linearly during the time between cardiac arrest and the onset of CPR.³⁸ Lactate concentrations increased progressively during closed-chest CPR in dogs³⁹ and remained stable but did not decrease during 30 minutes of open-chest CPR.³⁸ In this model, closed-chest CPR did not provide adequate tissue perfusion and O₂ delivery to halt anaerobic metabolism.

During CPR, arterial blood gases reflect alveolar-arterial gas exchange, whereas mixed venous blood gases reflect tissue acid-base status and oxygenation.¹⁵⁴ Respiratory alkalosis develops in arterial blood as a result of mechanical ventilation, whereas respiratory acidosis develops in venous blood because of poor tissue perfusion and impaired transport of accumulated CO₂ to the lungs. In one study of human patients undergoing CPR, average arterial pH was 7.41, whereas average mixed venous pH was 7.15.²³⁷ Arterial P_CO2 averaged 32 mm Hg and mixed venous P_CO2 was 74 mm Hg, whereas arterial and venous HCO₃⁻ concentrations were similar.

Closed-chest CPR, initiated after 6 minutes of cardiac arrest, was studied in dogs.²⁰⁴ Sodium bicarbonate (2 mEq/kg) was administered after 20 minutes of cardiac arrest. Administration of NaHCO₃ increased both arterial and venous pH. Before NaHCO₃, arterial P_CO2 was approximately 40 mm Hg, and with CPR it decreased to 20 mm Hg as a result of mechanical ventilation. After NaHCO₃, arterial P_CO2 increased to 30 mm Hg. Venous P_CO2 was nearly 50 mm Hg, and it slowly increased during 30 minutes of cardiac arrest to 60 mm Hg in untreated dogs. Bicarbonate treatment caused venous P_CO2 to increase transiently to 100 mm Hg, and it decreased to 70 mm Hg 10 minutes after NaHCO₃ administration. The pH of CSF was not changed by NaHCO₃ administration.

The normal arteriovenous pH gradient in dogs is 0.01 to 0.04.^8,20,152 Reduced cardiac output increases arteriovenous pH and P_CO2 gradients as a result of arterial hypocapnia and venous hypercapnia.^8,20,154,237 The ventilation-to-perfusion ratio is increased because of decreased pulmonary blood flow, accounting for the observed arterial hypocapnia. Venous hypercapnia results from anaerobic metabolism and a greater than normal addition of CO₂ to venous blood from hypoperfused tissues and diminished CO₂ excretion in the lungs because of pulmonary hypoperfusion. These increases in arteriovenous pH and P_CO2 gradients occur only if pulmonary ventilation continues. Respiratory arrest abolishes arteriovenous pH and P_CO2 gradients.⁸ In summary, arterial P_CO2 is not an accurate reflection of CO₂ removal from tissues during CPR, and analysis of mixed venous P_CO2 is recommended.^{8,20,152,154,237}

During CPR and ventilation with 100% O₂, arterial PO₂ may be normal, but tissue perfusion is low (20% to 25% of normal).¹¹² After NaHCO₃ administration, additional CO₂ is produced, and venous hypercapnia persists if ventilation is inadequate. Improving tissue perfusion is much more important during CPR than is NaHCO₃ administration. Effective cardiac compression and adequate perfusion allow delivery of O₂ to and removal of CO₂ from tissues. Conversely, tissue acidosis is aggravated and pH_i is decreased by NaHCO₃ administration if the CO₂ generated cannot be removed from the tissues by the lungs. The increase in tissue CO₂ decreases pH_i because CO₂ diffuses more rapidly into cells than does the charged HCO₃⁻, thereby lowering the intracellular HCO₃⁻/P_CO2 ratio. Intracellular acidosis of the myocardium leads to impaired cardiac contractility, decreased cardiac output, and aggravation of lactic acidosis. Thus, the main goals of CPR are to provide adequate tissue perfusion by effective cardiac compression and to ventilate the patient with 100% O₂. In one study of short (5 minutes) and prolonged (15 minutes) cardiac arrest in dogs, NaHCO₃ administration improved acidosis without a significant increase in P_CO2.²³⁴ The authors concluded that NaHCO₃ might be useful to reverse the acidosis of cardiac arrest if ventilation is adequate and NaHCO₃ is administered in a reasonable therapeutic window.

Lymphosarcoma in Dogs

Dogs with lymphosarcoma had higher lactate concentrations than control animals, and their lactate concentrations increased significantly 30 minutes after administration of 500 mg/kg dextrose.²³¹ Blood lactate concentrations were higher before and 1 hour after infusion of lactated Ringer’s solution in dogs with lymphosarcoma as compared with control animals.²³⁰ Blood lactate concentration returned to baseline during the second hour of the 6-hour infusion. The authors concluded that dogs with stage III or IV lymphosarcoma might have abnormal carbohydrate metabolism and a transient inability to handle lactate loads. Tumors may produce increased amounts of lactate as a result of excessive anaerobic metabolism and possibly as a result of less than normal hepatic extraction of lactate. Induction of remission with doxorubicin chemotherapy did not improve hyperlactatemia in dogs with lymphosarcoma.¹⁷⁶

Treatment

The outcome of lactic acidosis depends on the severity and reversibility of the underlying disease process responsible for the acid-base disturbance. If treatment of lactic acidosis is to be successful, prompt diagnosis and correction of the underlying disease state are crucial. Tissue perfusion and oxygen delivery should be improved by aggressive fluid therapy to expand ECFV. Ventilation with O₂ should be considered if the patient’s spontaneous ventilation is inadequate. Infections should be treated with appropriate antimicrobial agents, and cardiac output should be improved, if necessary, by administration of inotropic agents. If the underlying disease cannot be corrected, the prognosis for patients with lactic acidosis is very poor. If the underlying disease can be corrected, the accumulated lactate is metabolized, yielding an equivalent amount of HCO₃⁻, and the acidosis is reversed.

When the pH of the patient’s blood decreases to below 7.1 to 7.2, administration of alkali is justified to prevent the detrimental effects of severe acidosis on the cardiovascular system (e.g., impaired myocardial contractility, impaired cardiovascular responsiveness to catecholamines, increased susceptibility to ventricular arrhythmias). Small doses of NaHCO₃ should be administered to increase the patient’s pH to 7.2.^4,112,144

Approximately 10% to 15% of administered NaHCO₃ is converted immediately to CO₂.¹¹² It is essential that ventilation increase to allow removal of accumulated CO₂ from the body. It is probably safe to administer NaHCO₃ if the patient can reasonably be expected to increase ventilation spontaneously. If not, administration of NaHCO₃ may be detrimental. In any case, NaHCO₃ should be administered slowly to minimize the increase in mixed venous P_CO2.

The volume of distribution (V_d) of administered HCO₃⁻ is variable, depending on the severity of the acidosis.³ Thus, there is no simple way to calculate the dosage of NaHCO₃ required to increase the pH to 7.2. Volumes of distribution of 0.21 and 0.5 have been recommended for calculation of the bicarbonate space.^4,112 Sodium bicarbonate should be used cautiously and only in amounts necessary to increase the pH to 7.2. It should be administered slowly over several minutes to a few hours, and at least 30 minutes should be allowed to elapse after the infusion before judging its effect.⁴

The use of NaHCO₃ in lactic acidosis is controversial.^170,220 Using the canine model of hypoxic lactic acidosis described above,¹¹ affected dogs were left untreated, treated with 2.5 mEq/kg NaHCO₃, or treated with 2.5 mEq/kg 1 M NaCl.^92,93 Animals treated with bicarbonate showed a greater decrease in pH and HCO₃⁻ concentration and higher lactate concentration than the other groups. Gut lactate production was greater in dogs that received NaHCO₃ than in dogs that received NaCl, and portal vein P_CO2 was higher in the group that received NaHCO₃. Arterial blood pressure and cardiac output declined in the untreated group and the group that received NaHCO₃ but were higher in the group that received NaCl. Increased portal vein P_CO2 and hepatic accumulation of lactate presumably caused hepatocyte pH_i to decrease. The ability of the liver to extract lactate depends on adequate hepatic blood flow and normal hepatocyte pH_i, both of which are decreased in this model. During hypoxia (PO₂ <30 mm Hg), the liver is unable to increase its lactate extraction, despite an increased load delivered from the ischemic gut. The investigators concluded that use of NaHCO₃ during lactic acidosis might not be effective and might even be detrimental.

Dichloroacetate (DCA) stimulates the enzyme pyruvate dehydrogenase, which converts pyruvate to acetyl CoA.⁶² In the canine model of hypoxic lactic acidosis described before,¹¹ DCA was compared with NaCl.⁹¹ DCA increased pH and HCO₃⁻ concentration and maintained a constant lactate concentration, whereas NaCl treatment was associated with a decrease in pH and HCO₃⁻ concentration and an increase in lactate concentration. Hepatic lactate extraction increased with DCA, whereas liver and muscle accumulation of lactate decreased. Muscle pH_i increased with DCA, but neither treatment changed arterial blood pressure or cardiac output. DCA was also studied in a cardiac arrest model in dogs.²¹⁶ This study compared DCA, DCA and NaHCO₃, NaHCO₃, and no treatment. Bicarbonate treatment increased arterial pH, but DCA did not. DCA did not decrease lactate concentration or increase pH in either the peripheral circulation or CNS. In a canine model of hemorrhagic shock, DCA administration decreased arterial lactate concentrations but was associated with decreased cardiac stroke volume, decreased myocardial efficiency, and reduced myocardial lactate consumption.¹⁵ Thus, there are conflicting results regarding the usefulness of DCA in canine models of lactic acidosis.

Carbicarb is an equimolar mixture of Na₂CO₃ and NaHCO₃ that limits the generation of CO₂ during the buffering process:

However, some of the HCO₃⁻ generated from this reaction can buffer H⁺ released from nonbicarbonate buffers and generate CO₂ in the presence of carbonic anhydrase:

In the canine model of hypoxic lactic acidosis described earlier,¹¹ 2.5 mEq/kg Carbicarb was compared with 2.5 mEq/kg NaHCO₃.²¹ Arterial pH increased after administration of Carbicarb but decreased after NaHCO₃. Mixed venous P_CO2 was unchanged after Carbicarb administration but increased after NaHCO₃. Arterial lactate concentration increased after administration of NaHCO₃ but stabilized after Carbicarb, whereas lactate use by the gut, muscle, and liver improved with Carbicarb but decreased after NaHCO₃. Hepatocyte pH_i increased after Carbicarb and decreased after NaHCO₃. Arterial blood pressure decreased to a lesser extent and cardiac output stabilized with Carbicarb, whereas cardiac output decreased with NaHCO₃. It was concluded that Carbicarb had a beneficial effect on myocardial contractility. Myocardial contractility may decrease after NaHCO₃ administration as a result of increased venous P_CO2 and decreased myocardial pH_i. Decreased cardiac output follows and leads to decreased blood flow and decreased O₂ delivery to gut, muscle, and liver, resulting in decreased lactate use and increased production. Carbicarb improved arterial pH without impairing myocardial contractility, presumably because it did not increase venous P_CO2. This study suggests that Carbicarb is superior to NaHCO₃ in the treatment of lactic acidosis in dogs.

In another study, Carbicarb was compared with sodium bicarbonate and hypertonic saline in a canine model of hemorrhagic shock.¹⁹ All dogs received identical sodium loads. Groups that received Carbicarb and sodium bicarbonate experienced similar increases in serum bicarbonate, but arterial P_CO2 increased more in bicarbonate-treated dogs than in those treated with Carbicarb. Hemodynamics, oxygen delivery, and oxygen consumption improved in all three groups, and these effects were attributed to the sodium load. Carbicarb, NaHCO₃, and NaCl were compared in a model of hypoxic lactic acidosis in anesthetized, mechanically ventilated dogs.¹⁹³ Carbicarb increased arterial pH, base excess, and cardiac index without an increase in lactate. Bicarbonate increased P_CO2, but no adverse effects of NaHCO₃ on hemodynamics or pH_i were detected.

A sodium-free 0.3 N solution of tromethamine (THAM) is another CO₂-consuming alkalinizing agent that is capable of buffering both nonvolatile (H⁺) and volatile (H₂CO₃ derived from CO₂) acid. THAM and sodium bicarbonate had similar buffering ability when evaluated in dogs with experimentally induced metabolic acidosis.¹⁶³ Dogs treated with THAM did not experience the transient hypernatremia and hypercapnia that were observed in bicarbonate-treated dogs.

Administration of Alkali

Acute administration of 4 mEq/kg NaHCO₃ to normal unanesthetized cats resulted in mild increases in venous blood pH and HCO₃⁻ concentration lasting 180 minutes.⁴² A slight decrease in serum chloride concentration persisted for 30 minutes, whereas a mild increase in P_CO2 persisted for 60 minutes. A solution of NaHCO₃ (6.6 mEq/L) infused over 30 minutes into anesthetized dogs caused transient increases in arterial P_CO2, pH, base excess, and standard bicarbonate concentration.¹⁰⁶ Prompt renal excretion of administered NaHCO₃ presumably prevented any persistent change in acid-base values in these acute studies. Renal acid excretion decreases, urine pH increases, and administered NaHCO₃ is excreted within hours. There is an acute increase in carbonic acid and P_CO2 as body buffers release H⁺ to combine with the administered HCO₃⁻. The excess NaHCO₃ is excreted in the urine, increased ventilation occurs in response to increased P_CO2, and acid-base balance is restored to normal.

When alkali is administered chronically, plasma HCO₃⁻ concentration becomes a function of the daily dosage administered but returns to normal within a few days after alkali administration is discontinued. If alkali is given to subjects rendered sodium avid by previous dietary salt restriction, smaller dosages of alkali result in greater increases in plasma HCO₃⁻ concentration than are observed when higher alkali dosages are used in subjects receiving normal amounts of dietary salt.

Sources of alkali other than NaHCO₃ may also contribute to metabolic alkalosis. Such organic anions include lactate that has accumulated during lactic acidosis, ketoacids in uncontrolled diabetes mellitus, and citrate in banked blood or that administered in an attempt to prevent recurrence of calcium oxalate urolithiasis. These organic anions yield HCO₃⁻ when metabolized:

This reaction often serves to replace the HCO₃⁻ titrated during development of the acidosis (e.g., lactic acidosis, diabetic ketoacidosis). If NaHCO₃ has been administered during treatment, however, metabolism of the organic anion after correction of the acidosis can result in metabolic alkalosis. If renal function is normal and volume depletion is not present, the kidneys promptly excrete the excess HCO₃⁻ and restore normal acid-base balance.

Administration of nonabsorbable alkali (e.g., aluminum hydroxide used as a phosphorus binder in patients with renal failure) usually does not cause metabolic alkalosis. Neutralization of H⁺ by Al(OH)₃ in the stomach results in the net addition of HCO₃⁻ to ECF. Combination of Al³⁺ with HCO₃⁻ secreted by the pancreas produces insoluble Al₂(CO₃)₃ in the duodenum, and there is no net increase in HCO₃⁻ ions in ECF. If, however, Al(OH)₃ is administered concurrently with a cationic exchange resin (e.g., polystyrene sulfonate), the resin can bind Al³⁺, leaving HCO₃⁻ secreted by the pancreas to be reabsorbed in the small intestine, thus resulting in alkalinization of ECF. When renal failure is present, the kidneys have reduced capacity to excrete retained HCO₃⁻, and metabolic alkalosis could result. This sequence of events is most likely to occur in an animal with oliguric renal failure that is treated concurrently with Al(OH)₃ for hyperphosphatemia and with polystyrene sulfonate for hyperkalemia.

Gastric Fluid Loss

The H⁺ and Na⁺ concentrations of gastric fluid are inversely related to one another, whereas the K⁺ concentration is relatively stable (approximately 10 mEq/L). The Cl⁻ concentration is very high (approximately 150 mEq/L) and remains remarkably constant even when hypochloremia develops. Subtracting the sum of the Na⁺ and K⁺ concentrations of gastric fluid from the Cl⁻ concentration yields an approximation of the H⁺ concentration. The composition of gastric fluid is compared with that of other body fluids in Figure 10-10. When a dog or cat vomits stomach contents, water is lost along with large amounts of HCl and small amounts of potassium and sodium.

Figure 10-10 Comparison of electrolyte composition of gastric juice to other body fluids using Gamblegrams. Strong ions are crosshatched.

(From Jones NL. Blood gases and acid-base physiology, 2nd ed. New York: Thieme Medical, 1987: 133.)

The H⁺ produced during gastric acid secretion originates from the dissociation of carbonic acid; thus an equal number of HCO₃⁻ ions are generated in ECF. In the normal animal, gastric acid secretion does not disturb acid-base balance because the increase in ECF HCO₃⁻ concentration that accompanies parietal cell H⁺ secretion is balanced by pancreatic HCO₃⁻ secretion in the duodenum, and Cl⁻ secreted into the stomach is recaptured lower in the gastrointestinal tract. When stomach contents are lost, H⁺ and Cl⁻ are removed from this system, and HCO₃⁻ secreted into the duodenum by the pancreas is no longer titrated by gastric H⁺ but is reabsorbed farther down in the gastrointestinal tract in place of Cl⁻. The normal relationship between gastric and pancreatic secretions in the gastrointestinal tract is shown in Figure 10-11. Continued loss of gastric fluid can result in marked increases in plasma HCO₃⁻ concentration, and chronic vomiting of stomach contents is the most common cause of metabolic alkalosis in small animal practice.

Figure 10-11 Normal relationship between gastric and pancreatic secretions in the gastrointestinal tract.

(Modified from Guyton AC. Textbook of medical physiology, 7th ed. Philadelphia: WB Saunders, 1986: 775–779.)

In studies of gastric alkalosis, experimental subjects are rendered sodium avid by feeding a low-salt diet. Gastric fluid is then continuously removed by nasogastric suction, and fluid and electrolyte losses other than HCl are quantitatively replaced.^131,132,172 The effects of repeated gastric drainage over 3 days on plasma HCO₃⁻ and chloride concentrations and on potassium, sodium, and chloride balance in experimental dogs are shown in Figure 10-12. Note that the resulting metabolic alkalosis is corrected by provision of NaCl despite a progressively negative potassium balance. In the clinical setting, persistent vomiting of stomach contents leads to fluid and electrolyte losses (H⁺ and Cl⁻ > Na⁺ and K⁺), and anorexia prevents adequate dietary intake of electrolytes. In patients with pyloric obstruction, gastrin secretion is enhanced, and gastric acid secretion is stimulated further. Overproduction of gastrin by a gastrin-secreting tumor may also stimulate gastric acid secretion. In one dog with gastrinoma, severe metabolic alkalosis and hypokalemia were associated with a history of chronic vomiting.²²¹

Figure 10-12 Plasma composition and electrolyte balance in a representative study of selective HCl depletion.

(From Needle MA, Kaloyanides GJ, Schwartz WB. The effects of selective depletion of hydrochloric acid on acid-base and electrolyte equilibrium. J Clin Invest 1964;43:1839, with copyright permission of the American Society for Clinical Investigation.)

Renal avidity for sodium and defense of the ECFV occur because of ongoing fluid and electrolyte losses in the vomiting animal or intake of a low-salt diet and nasogastric suction in the experimental setting. To maintain ECFV, the kidneys must reabsorb sodium by all available mechanisms. Because of ongoing loss of gastric HCl and insufficient dietary intake of salt, there is a chloride deficit; consequently, the kidneys must reabsorb less sodium with chloride and more sodium in exchange for hydrogen and potassium ions. The latter two mechanisms contribute to perpetuation of the metabolic alkalosis and development of potassium depletion as shown in Figure 10-13. The low urine pH during the maintenance phase reflects increased distal Na⁺-H⁺ exchange in the sodium-avid state. This observation has led to the term “paradoxical aciduria” to describe the finding of low urine pH in patients with metabolic alkalosis. However, consideration of the relevant pathophysiology shows that this reduction in urine pH is the appropriate renal response under the circumstances. The extent of potassium depletion that develops is related to the severity and chronicity of the metabolic alkalosis.

Figure 10-13 Effects of chloride and potassium depletion on acid-base balance. See text for explanation.

(Drawing by Tim Vojt.)

Provision of chloride as the sodium or potassium salt allows correction of the alkalosis because the kidneys may now preferentially reabsorb sodium with chloride and rely less on Na⁺-K⁺ and Na⁺-H⁺ exchange.^{14,130,131,172} This allows retained HCO₃⁻ to be excreted in the urine. Urine pH increases as HCO₃⁻ is excreted, indicating a favorable response to therapy. Chloride once again appears in the urine when the alkalosis is resolved. The critical factor in resolution of this form of alkalosis is the provision of chloride as a resorbable anion. Alkalosis can be corrected without provision of sodium or potassium as long as chloride is provided. Clinically, however, alkalosis is corrected by administering some combination of NaCl and KCl.

Diuretic Administration

Diuretics cause approximately equal losses of sodium and chloride in the urine, but the concentration of chloride in ECF is less than that of sodium by approximately 35 mEq/L. Thus, these drugs may cause chloride-responsive metabolic alkalosis by a disproportionate loss of chloride in urine and creation of a relative chloride deficit in ECF. Increased renal sodium avidity is also an important factor in development of the metabolic alkalosis and potassium depletion that may occur during diuretic administration.

Loop diuretics inhibit NaCl reabsorption in the thick ascending limb of Henle’s loop by competing with chloride for the Na⁺-K⁺-2Cl⁻ luminal carrier. This causes increased delivery of sodium to the distal nephron, where accelerated Na⁺-H⁺ and Na⁺-K⁺ exchange occurs as the kidneys attempt to retain more sodium. Increased reliance of the kidneys on these mechanisms for sodium reabsorption contributes to metabolic alkalosis and potassium depletion. These complications are less likely when thiazide diuretics are used. Thiazide diuretics inhibit NaCl transport in the distal tubule and connecting segment. They are less potent than the loop diuretics because their main effect occurs at sites in the nephron distal to those responsible for the majority of sodium reabsorption.

In response to hypokalemia, transcellular shifts of H⁺ from ECF into renal tubular cells may occur in exchange for K⁺. The resultant increase in intracellular H⁺ concentration facilitates renal Na⁺-H⁺ exchange and aggravates metabolic alkalosis. Stimulation of the renin-angiotensin-aldosterone system by decreased effective circulating volume also favors increased Na⁺-H⁺ and Na⁺-K⁺ exchange in the distal nephron. These latter effects are probably important in most forms of chloride-responsive metabolic alkalosis.

Many animals treated with diuretics have congestive heart failure as their primary disease process. If the treatment plan for the animal includes a low-sodium diet, renal sodium avidity is guaranteed and increases the tendency toward metabolic alkalosis and potassium depletion. Complications from diuretic therapy are unlikely if the animal is drinking water and eating a diet with adequate amounts of chloride. However, complications can develop if the animal becomes anorexic.

Posthypercapnia

Blood pH increases rapidly when P_CO2 is suddenly reduced in patients with chronic hypercapnia. This has been called posthypercapnic metabolic alkalosis. In such patients, plasma HCO₃⁻ concentration has previously been increased by adaptive changes in renal HCO₃⁻ reabsorption. In response to the lowered P_CO2, it takes several hours for the kidneys to decrease Na⁺-H⁺ exchange and begin to excrete the previously retained HCO₃⁻. It may take several days for the kidneys to excrete all of the excess HCO₃⁻, and sufficient chloride must be available during this time for reabsorption with sodium. Chloride deficiency during recovery from chronic hypercapnia plays a role in sustaining posthypercapnic metabolic alkalosis. Provision of chloride allows the alkalosis to be corrected.²⁰⁹ Posthypercapnic metabolic alkalosis occurs most commonly in human patients with chronic pulmonary disease who are treated by mechanical ventilation. It is important that P_CO2 is decreased slowly and that adequate chloride intake is provided to prevent this complication.

Primary Hyperaldosteronism

In primary hyperaldosteronism, increased secretion of aldosterone, usually by an adrenocortical tumor, results in sodium retention, volume expansion, hypernatremia, mild to moderate hypertension, potassium deficiency, hypokalemia, and metabolic alkalosis resistant to chloride administration. Plasma renin activity is low, but plasma aldosterone concentration is high. Affected human patients are in salt balance at an expanded ECFV and excrete ingested NaCl in the urine. Stimulation of distal nephron Na⁺-H⁺ and Na⁺-K⁺ exchange by excess mineralocorticoids is probably the most important pathophysiologic feature of primary hyperaldosteronism.

Several dogs and cats with primary hyperaldosteronism caused by aldosterone-producing adenomas or adenocarcinomas of the adrenal gland have been reported in the veterinary literature. Clinical features in affected animals included polyuria, polydipsia, weakness, hypertension, hypokalemia, hypernatremia, mild metabolic alkalosis, dilute urine, and extremely high serum aldosterone concentrations (for additional information and references see Chapter 5).

Hyperadrenocorticism

Metabolic alkalosis occurs in approximately one third of human patients with Cushing’s syndrome.¹⁰⁴ It is more common in patients with adrenocortical carcinomas and in those with ectopic production of ACTH by nonadrenal malignancies than in those with pituitary-dependent hyperadrenocorticism. The frequency of metabolic alkalosis and serum electrolyte disturbances in dogs with hyperadrenocorticism is uncertain. Serum sodium and potassium concentrations often are normal in dogs with hyperadrenocorticism. This may reflect the fact that 80% to 85% of dogs with hyperadrenocorticism have pituitary-dependent disease. In a large group of dogs with hyperadrenocorticism, 21 of 52 (40%) dogs had increased serum sodium concentrations and 25 of 52 (48%) had decreased serum potassium concentrations.¹³⁹ The relative frequency of pituitary- and adrenal-dependent disease was not reported in this study. In another study, mild hypernatremia and hypokalemia were observed occasionally in dogs with hyperadrenocorticism, and total CO₂ content was increased in 33% of affected dogs.¹⁸⁸ In another report, hypokalemia was found in only 5% of dogs with pituitary-dependent hyperadrenocorticism but in 45% of those with adrenocortical neoplasia.¹⁵⁸ A high rate of secretion of cortisol and other corticosteroids, such as desoxycorticosterone and corticosterone in patients with adrenocortical malignancies, could be responsible for hypernatremia, hypokalemia, and metabolic alkalosis in adrenal-dependent hyperadrenocorticism.