Hydrogen ion (H+

) combines with HCO3

− and shifts the equilibrium of Eq. 26-1 to

the right. In the proximal renal tubule lumen, carbonic anhydrase catalyzes the

dehydration of H2CO3

to CO2 and H2O, which are absorbed into the tubule cell, as

illustrated in Eq. 26-2 and in Figure 26-1. Within the tubule cell, H2O dissociates

into H+ and OH−

. The H+

is then secreted into the lumen by a Na

+–H+ exchanger.

Carbonic anhydrase then catalyzes the combination of OH− and CO2

to HCO3

−

, which

is carried into the circulation by a Na

+HCO3

− cotransporter.

4

To maintain acid–base balance, the kidney must reclaim and regenerate all the

filtered HCO3

−

. The daily amount that must be reabsorbed can be calculated by the

product of the glomerular filtration rate (GFR) and the HCO3

− concentration in ECF

(180 L/day GFR × 24 mEq/L HCO3

− = 4,320 mEq/day).

1 The proximal tubule

reabsorbs about 85% of the filtered HCO3

−

. The loop of Henle and the distal tubule

reabsorb about 10%.

5 Acid salts, such as HPO4

−

(pKα of 6.8), that have a pKα greater

than the pH of the urine (titratable acids) can accept a proton and be excreted as the

acid, thus regenerating an HCO3

− anion.

5 Sulfuric acid and other acids with a pKα

less than 4.5 are not titratable. Protons from these acids must be combined with

another buffer to be secreted. Glutamine deamination in proximal tubular cells forms

NH3

, which accepts these protons. In the collecting tubule, the NH4

+ produced is

lipid insoluble, trapping it in the lumen and causing its excretion, eliminating the

proton, and allowing for regeneration of HCO3

−

.

4–6 Figure 26-2 is a simplified

illustration of the buffering of these acids.

The daily metabolism of carbohydrates and fats generates about 15,000 mmol of

CO2

. Although CO2

is not an acid, it reversibly combines with H2O to form carbonic

acid (i.e., H2CO3

). Respiration prevents the accumulation of volatile acid through the

exhalation of CO2

. Metabolism of proteins and fats results in several fixed acids and

bases. Amino acids such as lysine and arginine have a net positive charge and serve

as acids. Compounds such as glutamate, aspartate, and citrate have a negative charge.

In general, animal proteins contain more sulfur and phosphates, producing an acidic

diet. Vegetarian diets consist of more organic anions, resulting in a more alkaline

diet.

7 Normally, fatty acids are metabolized to HCO3

−

; however, during starvation or

diabetic ketoacidosis, they may be incompletely oxidized to acetoacetate and βhydroxybutyric acid.

6 The typical diet generates a net nonvolatile acid load of about

70 to 100 mEq of H+

(1.0–1.5 mEq/kg)/day.

1,8 Renal excretion of 70 mEq in 2 L of

urine each day would require a pH of 1.5. Because the kidney cannot produce a pH

less than 4.5, most of this fixed acid load must be buffered. The primary buffers for

renal net acid excretion are NH3

−

/NH4

+ and titratable buffers, such as

HPO4

−

/H2PO4

2−

, as mentioned earlier.

7 The correct assessment of acid–base

disorders begins with an evaluation of appropriate laboratory data and an

understanding of the physiologic mechanisms responsible for maintaining a normal

pH.

Figure 26-1 Renal tubular bicarbonate reabsorption.

Figure 26-2 Renal tubular hydrogen ion excretion.

p. 557

p. 558

Laboratory Assessment

Laboratory data used to evaluate acid–base status are arterial pH, arterial carbon

dioxide tension (Paco2

), and serum bicarbonate (HCO3

−

).

9–11 These values are

obtained routinely with an arterial blood gas (ABG) determination. Acid–base

abnormalities occur when the concentration of Paco2

(an acid) or HCO3

−

(a base) is

altered. ABG measurements also include the arterial oxygen tension (Pao2

); however,

this value does not directly influence decisions regarding acid–base abnormalities.

Normal ABG values are listed in Table 26-1. When arterial pH is less than 7.35, the

patient is considered acidemic, and the process that caused acid–base imbalance is

called acidosis. Conversely, when the arterial pH is greater than 7.45, the patient is

considered alkalemic, and the causative process is alkalosis. The process is further

defined as respiratory in cases of an inappropriate elevation or depression of Paco2

or metabolic with an inappropriate rise or fall in serum HCO3

−

.

Acid–base balance is normally maintained by the primary extracellular buffer

system of HCO3

−

/CO2

. Components of this buffer system are measured routinely to

assess acid–base status. Other extracellular buffers (e.g., serum proteins, inorganic

phosphates) and intracellular buffers (e.g., hemoglobin, proteins, phosphates),

however, also contribute significant buffering activity.

1,7–10 Serum electrolytes are

obtained to calculate the anion gap, an estimate of the unmeasured cations and anions

in serum. The anion gap helps determine the probable cause of a metabolic

acidosis.

6,10,12–21 Urine pH, electrolytes, and osmolality help to further differentiate

among the possible causes of metabolic acidosis.

10,22–26

Acid–Base Balance, Carbon Dioxide Tension, and

Respiratory Regulation

In aqueous solution, carbonic acid (i.e., H2CO3

formed through the reaction

described in Eq. 26-1) reversibly dehydrates to form carbon dioxide (CO2

) and

water (H2O) as shown in Eq. 26-2.

The enzyme carbonic anhydrase (CA), present in red blood cells, renal tubular

cells, and other tissues, catalyzes the interconversion of carbonic acid and carbon

dioxide. Some of the carbon dioxide produced by dehydration of carbonic acid

remains dissolved in plasma, but most exists as a volatile gas:

In Eq. 26-3, k is a solubility constant that has a value of approximately 0.03 in

plasma at body temperature.

2,26 Virtually, all the carbonic acid in body fluids is in the

form of carbon dioxide. The Paco2

, a measure of carbon dioxide gas, is therefore

directly proportional to the amount of carbonic acid in the HCO3

−

/H2CO3 buffer

system. The normal range for Paco2

is 35 to 45 mm Hg.

Table 26-1

Normal Arterial Blood Gas Values

ABGs Normal Range

pH 7.35–7.45

Pao2

80–105 mm Hg

PaCO2

35–45 mm Hg

HCO3− 22–26 mEq/L

ABG, arterial blood gas.

The lungs can rapidly exhale large quantities of carbon dioxide and thereby

contribute significantly to the maintenance of a normal pH. Carbon dioxide formed

through the reaction described in Eq. 26-3 diffuses easily from tissues to capillary

blood and from pulmonary capillary blood into the alveoli where it is exhaled from

the body.

3 Pulmonary ventilation is regulated by peripheral chemoreceptors (located

in the carotid arteries and the aorta) and central chemoreceptors (located in the

medulla). The peripheral chemoreceptors are activated by arterial acidosis,

hypercarbia (elevated Paco2

), and hypoxemia (decreased Pao2

). Central

chemoreceptors are activated by cerebrospinal fluid (CSF) acidosis and by elevated

carbon dioxide tension in the CSF.

3 Activation of these chemoreceptors stimulates

the respiratory control center in the medulla to increase the rate and depth of

ventilation, which results in increased exhalation of carbon dioxide.

In clinical practice, the serum bicarbonate concentration usually is estimated from

the total carbon dioxide content when the serum concentration of electrolytes are

ordered on an electrolyte panel or calculated from the pH and Paco2 on an ABG

determination. These estimations of the serum bicarbonate concentration are more

convenient than directly measuring serum bicarbonate. The total carbon dioxide

content that is reported on serum electrolyte panels is determined by acidifying serum

to convert all the bicarbonate to carbon dioxide and measuring the partial pressure of

CO2 gas. Approximately 95% of the total carbon dioxide content is bicarbonate. The

serum bicarbonate concentration reported on ABG results is calculated from the

patient’s pH and Paco2 using the Henderson–Hasselbalch equation (Eq. 26-4). This

calculated bicarbonate concentration should be within 2 mEq/L of the measured total

carbon dioxide. The normal range of serum bicarbonate using these methods is 22 to

26 mEq/L.

10

2.

3.

4.

5.

7.

1.

6.

a.

b.

EVALUATION OF ACID–BASE DISORDERS

Acid–base disorders should be evaluated using a stepwise approach.

24,25

Obtain a detailed patient history and clinical assessment.

Check the arterial blood gas, sodium, chloride, and HCO3

−

. Identify all

abnormalities in pH, Paco2

, and HCO3

−

.

Determine which abnormalities are primary and which are compensatory based on

pH (Table 26-2).

If the pH is less than 7.40, then a respiratory or metabolic acidosis is primary.

If the pH is greater than 7.40, then a respiratory or metabolic alkalosis is primary.

If the pH is normal (7.40) and there are abnormalities in Paco2 and HCO3

−

, a mixed

disorder is probably present because metabolic and respiratory compensations rarely

return the pH to normal.

Always calculate the anion gap. If it is equal to or greater than 20, a clinically

important metabolic acidosis is usually present even if the pH is within a normal

range.

27

If the anion gap is increased, calculate the excess anion gap (anion gap – 10). Add

this value to the HCO3

−

to obtain corrected value.

28

If the corrected value is greater than 26, a metabolic alkalosis is also present.

If the corrected value is less than 22, a nonanion gap metabolic acidosis is also

present.

p. 558

p. 559

Table 26-2

Laboratory Values in Simple Acid–Base Disorders

Disorder Arterial pH Primary Change Compensatory Change

Metabolic acidosis ↓ ↓HCO3− ↓PaCO2

Respiratory acidosis ↓ ↑PaCO2− ↑HCO3−

Metabolic alkalosis ↑ ↑HCO3− ↑PaCO2

Respiratory alkalosis ↑ ↓PaCO2− ↓HCO3−

Consider other laboratory tests to further differentiate the cause of the disorder.

If the anion gap is normal, consider calculating the urine anion gap.

If the anion gap is high and a toxic ingestion is expected, calculate an osmolal gap.

If the anion gap is high, measure serum ketones and lactate.

Compare the identified disorders to the patient history and begin patient-specific

therapy.

METABOLIC ACIDOSIS

Metabolic acidosis is characterized by loss of bicarbonate from the body, decreased

acid excretion by the kidney, or increased endogenous acid production. Two

categories of simple metabolic acidosis (i.e., normal anion gap and increased anion

gap) are listed in Table 26-3. The anion gap (AG) represents the concentration of

unmeasured negatively charged substances (anions) in excess of the concentration of

unmeasured positively charged substances (cations) in the extracellular fluid. The

concentrations of total anions and cations in the body are equal because the body

must remain electrically neutral. Most clinical laboratories, however, measure only a

portion of these ions (i.e., sodium, chloride [Cl

−

], and bicarbonate). The

concentrations of other negatively and positively charged substances, such as

potassium (K+

), magnesium (Mg

+

), calcium (Ca

2+

), phosphates, and albumin, are

measured less often. The concentration of unmeasured anions normally exceeds the

concentration of unmeasured cations by 6 to 12 mEq/L, and the anion gap can be

calculated as follows:

Table 26-3

Common Causes of Metabolic Acidosis

Normal AG Elevated AG

Hypokalemic

Diarrhea

Fistulous disease

Ureteral diversions

Type 1 RTA

Type 2 RTA

Carbonic anhydrase inhibitors

Hyperkalemic

Hypoaldosteronism

Hydrochloric acid or precursor

Type 4 RTA

Potassium-sparing diuretics

Amiloride

Spironolactone

Triamterene

Renal Failure

Lactic Acidosis

(see Table 26-4)

Ketoacidosis

Starvation

Ethanol

Diabetes mellitus

Drug Intoxications

Ethylene glycol

Methanol

Salicylates

AG, anion gap; RTA, renal tubular acidosis.

Of the unmeasured anions, albumin is perhaps the most important. In critically ill

patients with hypoalbuminemia, the calculated AG should be adjusted using the

following formula: adjusted AG = AG + 2.5 × (normal albumin − measured albumin

in g/dL), where a normal albumin concentration is assumed to be 4.4 g/dL.

16–19 For

example, a hypoalbuminemic patient (serum albumin, 2.4 g/dL) with early sepsis and

lactic acidosis might have a calculated AG of 11 mEq/L; however, after the

calculation is corrected for the effect of the abnormal serum albumin concentration,

the presence of elevated AG acidosis is more prominent (the calculated AG is

adjusted: AG(adjusted) = 11 mEq/L + 2.5 × [normal albumin – measured albumin] = 16

mEq/L).

Metabolic acidosis with a normal AG (e.g., hyperchloremic metabolic acidosis)

usually is caused by loss of bicarbonate and can be further characterized as

hypokalemic or hyperkalemic.

5,23,26,29–36 Diarrhea can result in severe bicarbonate

loss and a hyperchloremic metabolic acidosis. Elevated AG metabolic acidosis

usually is associated with overproduction of organic acids or with decreased renal

elimination of nonvolatile acids.

26,37–39

Increased production of organic acids (e.g.,

formic, lactic acids) is buffered by extracellular bicarbonate with resultant

consumption of bicarbonate and appearance of an unmeasured anion (e.g., formate,

lactate).

24,37,38 The decrement in serum bicarbonate approximates the increment in the

AG, the latter being a good estimate of the circulating anion level. Prolonged hypoxia

results in lactic acidosis. Uncontrolled diabetes mellitus or excessive alcohol intake

with starvation can cause ketoacidosis. In the case of renal failure, the capacity for

H+ secretion diminishes, resulting in metabolic acidosis.

29 The accompanying

increased AG results from decreased excretion of unmeasured anions such as sulfate

and phosphate.

20

Normal Anion Gap (Hyperchloremic) Metabolic

Acidosis

EVALUATION

CASE 26-1

QUESTION 1: J.D., a 21-year-old, 75-kg woman, is hospitalized for evaluation of weakness. She has a history

of bipolar affective disorder, pica, and reports recent ingestion of paint from the walls of her house. J.D.’s only

current medication is lithium carbonate 300 mg 3 times a day (TID). On admission, she appears weak and

apathetic and complains of anorexia. Laboratory tests reveal the following:

Serum Na, 143 mEq/L

K, 3.0 mEq/L

Cl, 121 mEq/L

p. 559

p. 560

Albumin, 4.4 g/dL

pH, 7.28

Paco2

, 26 mm Hg

HCO3

−

, 12 mEq/L

Urine pH, 5.5

J.D.’s urine pH after an ammonium chloride (NH4Cl) 0.1 g/kg IV load is less than 5.1. A bicarbonate load of

1 mEq/kg infused intravenously (IV) for 1 hour induces bicarbonaturia (urinary pH, 7.0) and lowers the serum

potassium to 2.0 mEq/L. Her blood pH only increased to 7.31. What type of acid–base disorder is present?

Using a stepwise approach, we see that J.D.’s history gives a clue to the cause for

her acidosis. The low pH is consistent with a metabolic acidosis because her CO2

and HCO3

− are both reduced (Table 26-3). Alterations in pH resulting from a

primary change in serum bicarbonate are metabolic acid–base disorders.

Specifically, metabolic acidosis is associated with a decrease in serum HCO3

− and

decreased pH, whereas metabolic alkalosis is associated with an increase in serum

HCO3

− and increased pH. In respiratory disorders, the primary change occurs in the

Paco2

. If J.D. had a decrease in pH and increase in Paco2

, a respiratory acidosis

would be present. Because J.D. has a low Paco2 and decreased serum HCO3

−

, she

has a metabolic acidosis. In most cases of metabolic acidosis or alkalosis, the lungs

compensate for the primary change in serum HCO3

− concentration by increasing or

decreasing ventilation. Most stepwise approaches would next suggest the evaluation

of whether the decrease in Paco2 of 14 mm Hg for J.D. is consistent with respiratory

compensation (Table 26-4). A primary decrease in the serum bicarbonate to a level

of 12 mEq/L should result in a compensatory decrease in the Paco2 concentration by

12 to 14 mm Hg (Table 26-4). J.D.’s Paco2 has fallen by 14 mm Hg (normal, 40 mm

Hg; current, 26 mm Hg), confirming that normal respiratory compensation has

occurred. When values for Paco2 or serum HCO3

−

fall outside of normal

compensatory ranges, either a mixed acid–base disorder, inadequate extent of

compensation, or inadequate time for compensation should be suspected.

Nomograms, especially ones that are different for acute and chronic disorders, are

inherently difficult to memorize, however, and are often not available to the clinician

at the point of care. Following the stepwise approach advocated herein will enable

clinicians to identify most clinically important disorders without needing to depend

on tables or formulas.

Table 26-4

Normal Compensation in Simple Acid–Base Disorders

Disorder Compensation

a

Metabolic acidosis ↓PaCO2

(mm Hg) = 1.0–1.2 × HCO3− (mEq/L)

Metabolic alkalosis ↑PaCO2

(mm Hg) = 0.5–0.7 × ↑HCO3− (mEq/L)

Respiratory acidosis

Acute ↑HCO3− (mEq/L) = 0.1 × ↑PaCO2

(mm Hg)

Chronic

↑HCO3− (mEq/L) = 0.4 × ↑PaCO2

(mm Hg)

Respiratory alkalosis

Acute ↓HCO3− (mEq/L) = 0.2 × ↓PaCO2

(mm Hg)

Chronic ↓HCO3− (mEq/L) = 0.4–0.5 × ↓PaCO2

(mm Hg)

aBased on change from normal HCO3− = 24 mEq/L and PaCO2 = 40 mm Hg.

CAUSES

CASE 26-1, QUESTION 2: What are potential causes of metabolic acidosis in J.D.?

Steps 4 to 7 of the stepwise approach in the evaluation of acid–base disorders are

used to further determine the cause of the disorder. In patients with metabolic

acidosis, calculation of the AG serves as a first step in classifying the metabolic

acidosis and provides additional information about conditions that might be

responsible. J.D.’s calculated AG is 10 mEq/L ( Eq. 26-5). Thus, J.D. has

hyperchloremic metabolic acidosis with a normal AG.

The common causes of metabolic acidosis are presented in Table 26-3.

5,10,37

Normal AG metabolic acidosis usually is caused by gastrointestinal loss of

bicarbonate (diarrhea, fistulous disease, ureteral diversions); exogenous sources of

chloride (normal saline infusions); or altered excretion of hydrogen ions (renal

tubular acidosis). J.D. reports a history of both, pica resulting in paint ingestion

(perhaps lead-based paint) and chronic use of lithium. Both lead and lithium have

been associated with the development of renal tubular acidosis.

23,40

Renal Tubular Acidosis

CASE 26-1, QUESTION 3: How do the results of NH4Cl and sodium bicarbonate (NaHCO3

) loading help

identify the type of renal tubular acidosis in J.D.?

Renal tubular acidosis (RTA) is characterized by defective secretion of hydrogen

ion in the renal tubule with essentially normal GFR. Many medical conditions and

chemical substances have been associated with RTA.

23,26 The recognized forms are

type 1 (distal), type 2 (proximal), and type 4 (distal, hypoaldosterone). Type 1 RTA

is caused by a defect in the distal tubule’s ability to acidify the urine. The most

common causes in adults are autoimmune disorders, toluene sniffing in recreational

drug users, and marked volume depletion.

41 Type 2 RTA is caused by altered urinary

bicarbonate reabsorption in the proximal tubule as can occur with the use of

acetazolamide. Type 4 is characterized by hypoaldosteronism and impaired

ammoniagenesis.

23,34

Evaluation of bicarbonate reabsorption during bicarbonate loading and of

response to acid loading by infusion of ammonium chloride is useful in distinguishing

among the various types of RTA. In healthy subjects, approximately 10% to 15% of

the filtered bicarbonate escapes reabsorption in the proximal tubule, but it is

reabsorbed in more distal segments of the nephron. Urine bicarbonate excretion is

therefore negligibly small, and urine pH is maintained between 5.5 and 6.5.

Type 2 RTA is associated with a decrease in proximal tubular bicarbonate

reabsorption. The distal tubular cells partially compensate for this defect by

increasing bicarbonate reabsorption, but urinary bicarbonate excretion still is

increased. As occurred with A.B., serum HCO3

− concentration in patients with type 2

RTA may acutely fall below a threshold of 15 but then stabilize around 15 mEq/L.

10,23

At this point, distal bicarbonate delivery no longer is excessive, allowing the distal

nephron to acidify the urine appropriately and excrete acid in the form of titratable

ammonia and phosphate.

In type 1 RTA, a defect in net hydrogen ion secretion results from a back-diffusion

of H+

from the tubule lumen to the tubule cell. Patients with type 1 RTA cannot

reduce their urine pH below 5.5 even when systemic acidosis is severe.

34

p. 560

p. 561

J.D.’s response to the acid (NH4Cl) load demonstrates an ability to acidify the

urine (i.e., pH < 5.1), which helps rule out type 1 RTA. During bicarbonate loading

in patients with type 2 RTA, serum bicarbonate concentration is increased, and

abnormally large amounts of bicarbonate are again delivered to the distal tubule. Its

hydrogen secretory processes are overwhelmed, resulting in bicarbonaturia.

Administration of bicarbonate to J.D. produced bicarbonaturia and an elevation in

urine pH (7.0), with low blood pH (7.31). These findings indicate that the

reabsorption of bicarbonate in the proximal tubule is impaired, which is

characteristic of type 2 RTA. Type 4 RTA is unlikely given her initial serum

potassium of 3.0 mEq/L.

Lead-Induced

CASE 26-1, QUESTION 4: What is the cause of J.D.’s proximal RTA?

The most likely cause of A.B.’s proximal RTA is her exposure to presumably

lead-based paint. The pathogenesis of lead-induced type 2 RTA is unclear. Some

studies suggest that carbonic anhydrase deficiency in the proximal tubule is the major

factor, but these data are inconclusive.

CASE 26-1, QUESTION 5: Why is J.D. hypokalemic?

Bicarbonate wasting in proximal RTA is associated with sodium loss,

extracellular fluid reduction, and activation of the renin–angiotensin–aldosterone

axis. Aldosterone increases distal tubular sodium reabsorption and greatly augments

potassium and hydrogen ion secretion. This results in potassium wasting, which

explains J.D.’s hypokalemia.

42 When plasma bicarbonate achieves steady state, less

bicarbonate reaches the distal tubule, and the stimulus for aldosterone release is

removed. Therefore, J.D. experiences only a mild depletion of potassium body

stores. When J.D. is exposed to bicarbonate loading, the renin–angiotensin–

aldosterone axis is reactivated, and hypokalemia worsens. In addition, raising the

concentration of bicarbonate in the blood drives potassium intercellularly and

contributes to her hypokalemia.

TREATMENT

CASE 26-1, QUESTION 6: What treatment is indicated for J.D.?

Although it is rare for patients with type 2 RTA to develop severe acidosis and

potassium depletion chronically, it is not uncommon in an acute situation such as this.

J.D. has a bicarbonate deficit; thus, she should be treated with alkali replacement,

and the offending agent, if confirmed to be lead, should be removed concurrently. Her

serum potassium is also dangerously low, and bicarbonate correction could further

decrease it. J.D. needs potassium supplementation. The clinician should obtain

hourly blood samples for electrolytes until her potassium is greater than 3.5 mEq/L.

In adults such as J.D., chronic treatment often is not needed because acidosis is selflimited. J.D., however, should be treated with sodium bicarbonate until proximal

RTA resolves. Very large doses of bicarbonate (6–10 mEq/kg/day) would be

required to increase serum bicarbonate to the normal range.

10

In adults with proximal

RTA, however, the goal is to increase serum bicarbonate to no more than 18

mEq/L.

23 Bicarbonate can be provided as sodium bicarbonate tablets (8 mEq/600-mg

tablet) or Shohl’s solution. Shohl’s solution, USP, contains 334 mg citric acid and

500 mg sodium citrate per 5 mL. Sodium citrate is metabolized to sodium

bicarbonate in the liver. Shohl’s solution provides 1 mEq of sodium and 1 mEq of

bicarbonate per milliliter of solution. Therapy for J.D. should be initiated with 1

mEq/kg/day. The clinician should monitor A.B.’s lithium levels while she is

receiving alkali therapy. Sodium ingestion might increase renal lithium excretion and

exacerbate her bipolar disorder. Because of severe hypokalemia resulting from alkali

administration, supplemental potassium as chloride, bicarbonate, acetate, or citrate

salts also should be administered.

Metabolic Acidosis with Elevated Anion Gap

EVALUATION AND OSMOLAL GAP

CASE 26-2

QUESTION 1: G.D., a 64-year-old, 60-kg man, is brought to the emergency department (ED) by his family in

a semicomatose state. He was found lying on the floor of his garage near a partially empty bottle of windshield

wiper fluid 30 minutes ago. G.D. has a long history of alcohol abuse and recently diagnosed dementia. In the

ED, supine blood pressure (BP) is 120/60 mm Hg, pulse is 100 beats/minute, and respiratory rate is 40

breaths/minute. G.D.’s pupils are reactive, and mild papilledema is noted. Laboratory tests reveal the following:

Serum Na, 139 mEq/L

K, 5.8 mEq/L

Cl, 103 mEq/L

Blood urea nitrogen (BUN), 25 mg/dL

Creatinine, 1.4 mg/dL

Fasting glucose, 150 mg/dL

ABG include pH, 7.16; Paco2

, 23 mm Hg; and HCO3

−

, 8 mEq/L. His toxicology screen is negative for

alcohol, and his serum osmolality is 332 mOsm/kg. What acid–base disturbance is present in G.D., and what are

possible causes of the disorder?

G.D. has an acidosis (pH, 7.16; HCO3

−

, 8 mEq/L) with a large AG (28 mEq/L).

Subtracting 10 from the anion gap of 28 and adding this value to his serum

bicarbonate concentration (see Step 5 in the section Evaluation of Acid–Base

Disorders) yields a value of 26, suggesting no other metabolic abnormality is

present.

An elevated AG metabolic acidosis often indicates lactic acidosis resulting from

intoxications (e.g., salicylates, acetaminophen, methanol, ethylene glycol,

paraldehyde, metformin) or ketoacidosis induced by diabetes mellitus, starvation, or

alcohol.

14,21,25,38,43–48 Step 6 in the stepwise approach leads to the consideration of

additional laboratory tests that may be helpful in the differential diagnosis of an

elevated AG. These include serum ketones, glucose, lactate, BUN, creatinine, and

plasma osmolal gap.

25 Osmolal gap is defined as the difference between measured

serum osmolality (SO) and calculated SO using Eq. 26-6.

When the difference between measured and calculated SO is greater than 10

mOsm/kg, the presence of an unmeasured osmotically active substance, such as

ethanol, methanol, or ethylene glycol, should be considered.

25,49 G.D.’s calculated SO

is 295 mOsm/kg, compared with the measured value of 332; therefore, his osmolal

gap is 37 mOsm/kg. An increase in the anion gap and osmolal gap, without diabetic

ketoacidosis or chronic renal failure, suggests the possibility of metabolic acidosis

resulting from a toxic ingestion.

25 On the basis of G.D.’s presentation (papilledema,

history of alcohol abuse, increased osmolal gap, increased AG metabolic acidosis),

history of dementia, and partially empty bottle of windshield wiper fluid found at the

scene, methanol intoxication should be considered.

p. 561

p. 562

CAUSES

Methanol-Induced

CASE 26-2, QUESTION 2: How would G.D.’s methanol intake induce metabolic acidosis with an elevated

anion AG?

Methanol intoxication results in the formation of two organic acids, formic and

lactic acids, which consume bicarbonate with production of an AG metabolic

acidosis. Alcohol dehydrogenase in the liver metabolizes methanol to formaldehyde

and then to formic acid. The formic acid contributes to the metabolic acidosis and

also is responsible for the retinal edema and blindness associated with methanol

intoxication.

25,26,48

Serum lactic acid concentrations also are increased in patients with methanol

intoxication.

25 Lactic acidosis classically has been divided into type A, which is

associated with inadequate delivery of oxygen to the tissue, and type B, which is

associated with defective oxygen utilization at the mitochondrial level (Table 26-5).

Although these distinctions often are not clear, the lactic acidosis caused by methanol

intoxication is most consistent with the type B variety.

50

TREATMENT

CASE 26-2, QUESTION 3: How should G.D.’s methanol intoxication be managed acutely?

Antidotes

Because G.D.’s mental status is impaired and his respiratory rate is 40

breaths/minute, his airway was secured via endotracheal intubation and he was

placed on mechanical ventilatory support. Even though both ethanol and fomepizole

compete with methanol for alcohol dehydrogenase binding sites and could be used to

treat G.D., fomepizole is chosen because it is easier to dose and does not need

serum-level monitoring to ensure efficacy like ethanol.

26,48–52 Because ethanol and

fomepizole have much greater affinity for alcohol dehydrogenase than methanol, these

agents may reduce the conversion of methanol to its toxic metabolite, formic acid.

The unmetabolized methanol is then excreted by the lungs and kidneys. Fomepizole

can be given IV as a 15 mg/kg loading dose for 30 minutes, followed by bolus doses

of 10 mg/kg every 12 hours. Because of induction of metabolism of fomepizole,

doses should be increased to 15 mg/kg every 12 hours if therapy is required beyond 2

days.

48 Fomepizole is usually continued until the serum methanol concentration is less

than 20 mg/dL (6.2 mmol/L). Adverse effects of fomepizole are relatively mild; G.D.

should be monitored for headache, nausea, dizziness, agitation, metallic taste,

abnormal smell, and rash. Cofactor therapy with folinic acid or folic acid at a dosage

of 50 mg IV every 6 hours should be given to enhance the elimination of formate

along with IV thiamine because of G.D.’s history of chronic alcoholism.

Table 26-5

Common Causes of Lactic Acidosis

Type A Type B

Anemia Diabetes mellitus

Carbon monoxide poisoning Liver failure

Congestive heart failure Renal failure

Shock Seizure disorder

Sepsis Leukemia

Drugs

Didanosine

Ethanol

Isoniazid

Metformin

Methanol

Salicylates

Zidovudine

Because of its high cost and infrequent use, some hospitals might not have

fomepizole readily available. In such cases, ethanol is an alternative. Administration

of IV ethanol as an antidote can be technically difficult and may produce central

nervous system (CNS) depression.

48,51 For G.D., an IV-loading dose of 0.6 g/kg

ethanol solution should be administered over the course of 30 minutes, followed by a

continuous infusion of about 150 mg/kg/hour if the patient has been drinking, or 70

mg/kg/hour for nondrinkers if the patient was not drinking. Serum ethanol

concentration should be maintained at more than 100 mg/dL.

26,50 Charcoal may be

considered to bind other agents that may be coingested.

26,53

When other low-molecular-weight toxins, such as ethanol or ethylene glycol, are

not present, the serum methanol level can be estimated by multiplying the patient’s

osmolal gap by a standardized conversion factor of 2.6. G.D.’s osmolal gap of 37

mOsm/L, therefore, may reflect a methanol level of approximately 96 mg/dL (37

mOsm/L × 2.6). When methanol blood levels are higher than 50 mg/dL, hemodialysis

is indicated to rapidly reduce concentrations of methanol and its toxic metabolite.

The dosage of fomepizole or ethanol should be increased in patients receiving

hemodialysis to account for the increased elimination of these antidotes.

26,50 Ethylene

glycol poisoning can also be treated by using fomepizole or ethanol.

Bicarbonate

In general, a severe acidosis causes reduced myocardial contractility, impaired

response to catecholamines, and impaired oxygen delivery to tissues as a result of

2,3-diphosphoglycerate depletion. For this reason, some clinicians have judiciously

administered IV sodium bicarbonate to patients with metabolic acidosis in an attempt

to raise the arterial pH to about 7.20.

15,54,55

In G.D.’s case, bicarbonate therapy is

indicated in an attempt to raise the pH to 7.3 which converts formic acid (unionized

metabolite of methanol) to formate (ionized form) to decrease tissue penetration. If

IV sodium bicarbonate is given, the amount required to correct serum HCO3

− and

arterial pH can be estimated using Eq. 26-7 as follows:

Bicarbonate distributes to approximately 50% of total body weight (thus, the factor

of 0.5 L/kg in Eq. 26-7). To prevent overtreating, bicarbonate doses should only

attempt to increase the bicarbonate concentration by 4 to 8 mEq/L (see Case 26-2,

Question 4).

54 For G.D., the dose required to raise serum bicarbonate from 8 to 12

mEq/L amounts to 120 mEq of bicarbonate (0.5 L/kg × 60 kg × 4 mEq/L; Eq. 26-7).

Clinical assessment of the effect of bicarbonate can be determined about 30 minutes

after administration.

54 Arterial pH and serum bicarbonate concentrations should be

obtained before any additional therapy.

RISKS OF BICARBONATE THERAPY

CASE 26-2, QUESTION 4: What are the risks of G.D. bicarbonate therapy?

p. 562

p. 563

Concerns about the risks of bicarbonate administration and the failure of studies to

demonstrate significant short-term benefits have raised questions about the

appropriateness of bicarbonate therapy in metabolic acidosis, particularly in

ketoacidosis and lactic acidosis caused by cardiac arrest or other hypoxic events.

55–61

Bicarbonate administration can result in overalkalinization and a paradoxical

transient intracellular acidosis. Whereas arterial pH can increase rapidly after

bicarbonate administration, intracellular pH increases more slowly because of slow

penetration of the negatively charged bicarbonate ion across cell membranes. The

bicarbonate in plasma, however, is converted rapidly to carbonic acid, and the

carbon dioxide tension increases as a result (Eq. 26-2). Because CO2 diffuses into

cells more rapidly than HCO3

−

, the intracellular HCO3

−

/CO2

ratio decreases,

resulting in a decrease in intracellular pH. This intracellular acidosis will persist as

long as bicarbonate administration exceeds the CO2 excretion; therefore, adequate

tissue perfusion and ventilation must be provided in patients with diminished CO2

excretion (e.g., cardiac or respiratory failure).

56

Overalkalinization also will cause a shift to the left in the oxygen–hemoglobin

dissociation curve. This shift increases hemoglobin affinity for oxygen, decreases

oxygen delivery to tissues, and potentially increases lactic acid production and

accumulation.

26 Sodium bicarbonate administration also can cause hypernatremia,

hyperosmolality, and volume overload; however, the excessive sodium and water

retention usually can be avoided by the administration of loop diuretics.

26,49

Hypokalemia is another potential adverse effect of bicarbonate therapy. Acidosis

stimulates movement of potassium from intracellular to extracellular fluid in

exchange for hydrogen ions. When acidosis is corrected, potassium ions move

intracellularly, and hypokalemia can occur. This translocation of potassium tends to

reduce serum potassium levels by about 0.4 to 0.6 mEq/L for each 0.1 unit increase

in pH, although wide interpatient variability in this relationship exists.

5,8

In G.D. and

other patients with organic acid intoxications, raising extracellular pH helps to

provide a gradient to shift the toxin from the CNS and “trap” it into the blood and

urine, thus enhancing elimination. To prevent the risks of bicarbonate therapy, G.D.’s

mental status, serum sodium and potassium levels, and ABG should be monitored.

METABOLIC ALKALOSIS

Metabolic alkalosis is associated with an increase in serum bicarbonate

concentration and a compensatory increase in Paco2

(caused by hypoventilation). The

two general classifications of metabolic alkalosis, saline-responsive and salineresistant (Table 26-6), are usually distinguishable based on an assessment of the

patient’s volume status, BP, and urinary chloride concentration.

Table 26-6

Classification of Metabolic Alkalosis

Saline-Responsive Saline-Resistant

Diuretic therapy Normotensive

Extracellular volume contraction Potassium depletion

Gastric acid loss Hypercalcemia

Vomiting Hypertensive

Nasogastric suction Mineralocorticoids

Exogenous alkali administration Hyperaldosteronism

Blood transfusions Hyperreninism

Licorice

Saline-responsive metabolic alkalosis is associated with disorders that result in

the loss of chloride-rich, bicarbonate-poor fluid from the body (e.g., vomiting,

nasogastric suction, diuretic therapy, cystic fibrosis). Physical examination may

reveal volume depletion (e.g., orthostatic hypotension, tachycardia, poor skin turgor),

and the urinary chloride concentration often will be less than 10 to 20 mEq/L

(although urine chloride levels may be >20 mEq/L in patients with recent diuretic

use).

10,27,62

Severe hypokalemia or excessive mineralocorticoid activity can result in a salineresistant metabolic alkalosis, but this disorder is rare in comparison with salineresponsive metabolic alkalosis. Saline-resistant metabolic alkalosis should be

suspected in alkalemic patients with evidence of increased ECF volume,

hypertension, or high urinary chloride values (>20 mEq/L) without recent diuretic

use.