Metabolic acidosis may result from bicarbonate loss per se, administration of acid, or endogenous production and accumulation of acid. Loss of bicarbonate occurs in externalization of intestinal contents (e.g., vomiting, enterocutaneous fistulae) and renal wasting of bicarbonate (e.g., renal tubular acidosis, carbonic anhydrase inhibitor therapy). Administration of acid occurs primarily in total parenteral nutrition, where patients receive hydrochloric salts of basic amino acids. Finally, endogenous acids accumulate in renal tubular acidosis, ketoacidosis, and lactic acidosis.
Metabolic acidosis results in a decreased serum bicarbonate concentration, and its development causes respiratory compensation through increases in alveolar ventilation, resulting in a reduction in P co2. The steady-state relationship between the Pco2 and the bicarbonate concentration, determined from a study of 60 patients who had only metabolic acidosis of more than 24-h duration is shown in Eq (6)12i
When the bicarbonate concentration is greater than about 8 meq/dL, the relationship between P co2 and bicarbonate is simpler: Pco2 falls by 1 mmHg for every 1 meq/dL fall in bicarbonate. Use of these relationships allows the clinician to calculate the expected P co2 from the measured bicarbonate concentration. If the expected Pco2 value differs from the measured value in uncomplicated steady-state metabolic acidosis, a respiratory disorder also exists. For example, if the [HCO 3-] is 15, the expected Pco2 is about 30 mmHg. If it is higher than this value (say 35 mmHg), then by definition there is also concomitant primary respiratory acidosis (see Fig
21-2A). If the value is lower than expected (say 25 mmHg), then there is a concomitant respiratory alkalosis. This latter case is not an "overcompensation" but rather a second primary disturbance occurring simultaneously. These are important concepts. The body cannot tolerate both a metabolic and respiratory mechanism for acidosis simultaneously, as one cannot buffer or compensate for the other.
The ED patient's illness can, unfortunately, rarely be assumed to be in steady state. Pierce and colleagues, in physiologic studies of otherwise healthy persons with acute metabolic acidosis caused by diarrhea, discovered that the completeness of the respiratory response to metabolic acidosis depends upon the duration of the acidosis, the time course of its development, and its severity.13 If acidosis develops quickly, the Pco2 is often higher than that observed in steady state; the more rapid and severe the acidosis, the larger the difference between the observed P co2 and the predicted steady-state Pco2. When bicarbonate concentration is then held constant, steady-state Pco2 is reached in 11 to 24 h. When acidosis develops or is corrected more slowly, there is no lag in respiratory compensation.
There are limits to the adequacy of respiratory compensation during metabolic acidosis. A study early in the twentieth century of diabetic ketoacidosis by Kety and colleagues found that respiratory minute volume actually declined when pH fell below 7.10.14 This finding led both Albert et al.12 and Pierce et al.13 to initiate bicarbonate therapy if their subjects' pH fell below 7.10. The nature of respiratory compensation in untreated severe metabolic acidosis is therefore unknown. Furthermore, these studies appear to have ensconced 7.10 as the definition of "severe" metabolic acidosis. It is particularly important to appreciate any contribution to the acidosis from inadequate respiratory response. Simply initiating bicarbonate therapy when a pH of less than 7.1 is encountered may miss the respiratory insufficiency, which, if addressed, may obviate the need to use solutions containing HCO 3-. Further, administration of HCO3- in the face of inadequate ventilatory response actually exacerbates the respiratory acidosis, as the HCO 3- is converted to CO2 and H2O. The development of metabolic acidosis in which the pH is below 7.10 is probably associated with a very high risk of ventilatory insufficiency. There is another limit to respiratory compensation. The lowest level that P co2 can attain is about 12 mmHg. The restriction of air movement and the CO2 generated by the exertion required for rapid ventilation limits the attainable nadir for P co2. The superimposition of respiratory acidosis upon a patient in such straits will result in a rapid decline of pH to levels at which organ function and pharmacotherapy will fail. Mechanical ventilation should usually be instituted in such situations.
The serum potassium level is affected by metabolic acidosis. The movement of hydrogen ion into cells is associated with extrusion of potassium. Changes in potassium concentration are more substantial in inorganic acidosis, though elevated serum potassium values are typically seen in diabetic ketoacidosis. Generally, for each 0.10 change in the pH, serum [K+] will increase by approximately 0.5 meq/L. Whatever the mechanism of the acidosis, it is important to remember that low-normal or low serum potassium values probably reflect severe intracellular potassium depletion. The reversal of the acidosis in such circumstances may result in severe hypokalemia, with attendant cardiovascular effects.
Metabolic acidosis results in a depression of serum bicarbonate, whose negative charge must be replaced. The negatively charged species are either unmeasured anions (elevated-AG acidosis) or chloride (normal-AG acidosis, also referred to as hyperchloremic metabolic acidosis).
The causes of elevated-AG metabolic acidosis are listed in Table21:2. We reemphasize that the anion gap may be within the normal range even when a metabolic acidosis associated with increased concentrations of unmeasured anions is present. A comparison with the patient's steady-state AG is warranted or measurement of specific anions in such cases is indicated. However, caution is necessary when serum ketone testing is performed, as the chemical reaction used to measure serum ketones has an important limitation: the nitroprusside reaction for ketones is positive only for species whose carbonyl moiety has an a-methyl group. The major ketone present in the serum of patients with diabetic ketoacidosis may be b-hydroxybutyrate, which has no a-methyl group and is not detected by the nitroprusside reaction. The result may be a paradox: initial serum ketone assays in a patient with clinically severe diabetic ketoacidosis are only weakly positive yet rise in spite of clear clinical improvement. This occurs because appropriate resuscitation alters the hepatic NAD/NADH 2 ratio and the restoration of NAD concentrations allows oxidation of b-hydroxybutyrate to acetoacetate. The addition of several drops of hydrogen peroxide to serum will also oxidize b-hydroxybutyrate to acetoacetate in cases where this distinction is clinically important. See Chap.iii203 for a detailed discussion.
Lactic acidosis occurs whenever lactate production exceeds lactate utilization or metabolism and is classically of two types. The first, in which tissue hypoxia is present and lactate production is elevated, is referred to as type A. Normal tissue oxygenation and impairment of lactate utilization define the second, called type B. Type B lactic acidosis is further subdivided. B ! lactic acidosis is associated with systemic disorders such as diabetes, renal insufficiency, sepsis, and leukemia; type B2 is associated with various substances, especially biguanides (phenformin, metformin), salicylates, methanol, iron and isoniazid; and type B 3 is associated with hereditary metabolic diseases.
The pyruvate produced by glycolysis may be transported across mitochondrial membranes and metabolized in the Krebs cycle under aerobic conditions. However, under anaerobic conditions, it is oxidized to lactate by lactate dehydrogenase. This reaction is reversible, but the conversion of lactate to pyruvate in the liver requires NAD. Thus, it is not surprising that many patients with type B lactic acidosis have underlying liver disease. For example, an alcoholic may develop lactic acidosis after giving up heavy drinking because impaired gluconeogenesis prevents pyruvate fixation into glucose, while the metabolism of ethanol has left little NAD available to convert lactate to pyruvate. The distinction between type A and type B lactic acidosis is useful in conceptualizing the therapeutic approach. However, there is some impairment of lactate utilization in both types, usually because of impaired hepatic oxygenation or perfusion in the case of type A and because of underlying liver disease in type B.
Note that an ABG is completely immaterial in determining whether a wide-AG metabolic acidosis exists. The determination is made with simple venous electrolytes. The differential diagnosis falls into four broad categories: renal failure (uremia), lactic acidosis, ketoacidosis (DKA, AKA, SKA), and ingestions (methanol, ethylene glycol, salicylates).
Renal failure should be evident from the serum chemistries. Positive serum ketones point to one of the ketoacidoses. In known insulin-dependent diabetes mellitus (IDDM), DKA is likely, although there is usually a small component of lactic acidosis also. In alcoholics who have recently stopped binge drinking, AKA should be considered. Starvation ketosis will be found in patients with inadequate recent oral intake (fasting, dieting, or protracted vomiting).
Determination of the osmolal gap will help differentiate two of the ingestants from other etiologies. Elevated gaps are seen in methanol and ethylene glycol poisoning. Although methanol is measured in most hospital laboratories, determination of ethylene glycol levels is still performed off-site in many institutions. A widened osmolal gap without evidence of methanol ingestion may determine the diagnosis long before confirmatory laboratory evidence is available. Calculation adjustments to the osmolal gap may need to be made if ethanol is a coingestant (see Chap,.23 for detailed discussion).
When the diagnosis remains in doubt or poor tissue perfusion is a diagnostic possibility, lactate levels should be specifically sent. It should be noted that several poisonings may result in lactic acidosis, including isoniazid (INH), iron, carbon monoxide (CO), methemoglobin, and cyanide. This is but one reason why the authors shun the overtaught mnemonic of MUDPILES. For example, this mnemonic does not clearly reflect the fact that INH and iron exert their effects on the AG through lactic acidosis. Also, ethanol is frequently taught as a cause of wide-AG acidosis. Ethanol should never be considered the etiologic source of any significant metabolic acidosis. While ethanol alcohol metabolism may lead to very mild lactic acidosis, usually in association with AKA, its effect is very mild and all but clinically immaterial.
Severe acidosis that is resistant to treatment is seen in various B1 lactic acidoses and ingestions. AKA and SKA tend to be mild. Acidosis seen in initial stages of renal failure may be severe but is stable ([HCO3-] about 15 meq/L) in chronic renal failure. Concomitant acid-base disturbance may further assist in determining the etiology. The triple acid-base disturbance of wide-AG metabolic acidosis, metabolic alkalosis, and respiratory alkalosis is seen with sepsis (lactic acidosis), and salicylate poisoning. The latter may be associated with a mild temperature elevation also.
Finally, the relationship of [HCO3-] to the anion gap and expected Pco2 compensation must be examined in every patient with wide anion gap acidosis to determine if other (respiratory) acid-base disturbances exist (see Fig 2.1-.2..A.).
Differential Diagnosis of Unchanged (Normal) AG Acidosis
The non-AG type of acidosis is often referred to as "normal" anion gap acidosis. As discussed above, issues related to the AG are relative, so the term unchanged or non-AG type is preferred. Some texts refer to this as hyperchloremic metabolic acidosis.
Non-AG acidosis results from loss of bicarbonate, failure to excrete hydrogen ion, or administration of hydrogen ion. Bicarbonate may be lost from either the urine or gastrointestinal tract and is usually accompanied by potassium. However, potassium-sparing diuretics, hypoaldosteronism, urinary tract obstruction, and type IV renal tubular acidosis all result in loss of bicarbonate with retention of potassium ( Table. . . 21,-3).
uytdiq IWJL fifty unra Krira E-fty ibthutfibi «qp^TK Rmü a^jii jùlui-rjj, HjpOjlÉM ■. Il ' A'J-Jlk* Î1K1K)
KilUKkl -lsU *îi JM !'■ BÎUI lihiliruiAiiiL Kpf 3|
Antr dufTtu ktwj «f
■:Ir±rpj Ititt&rM. ûT H ' JAd Cl Mri tiki +1' HCCV MA f
TABLE 21-3 Causes of Normal Anion Gap Metabolic Acidosis
One should be wary of the traditional classification based on K +, as serum [K+] itself is dependent on the actual pH. Thus, in severe acidosis, a normal range [K +] may be deceiving unless the clinician corrects for the degree of acidosis.
Since all diuretics have a tendency to result in mild contraction alkalosis, the metabolic acidosis that occurs simultaneously with potassium-sparing diuretics may not be evident, as the two distinct physiologic processes may simply cancel each other out (see Fig 2.1.-2..C). Since the AG is unchanged, there is no clue that two opposed processes may be occurring.
Acetazolamide exerts its effect through carbonic anhydrase inhibition, inducing a functional RTA. Treatment of Acidosis
Acidemia has numerous negative physiologic consequences, impairing the function of many different organs through mechanisms not yet well understood. Cardiac contractile function is reduced, probably due to impaired oxidative phosphorylation as well as intracellular acidosis. The threshold for ventricular fibrillation falls while the defibrillation threshold rises. Hepatic and renal perfusion decline, along with systemic blood pressure, while pulmonary vascular resistance increases. The physiologic effects of catecholamines are attenuated, and when acidosis is sufficiently severe, vascular collapse may result. A catabolic state develops, including a generalized increase in metabolism, resistance to insulin, and inhibition of anaerobic glycolysis. Finally, the effect of hypoxia on all organs is aggravated. 15
The treatment of acidosis reflects that of the underlying disorder but particularly emphasizes restoration of normal tissue perfusion and oxygenation. As noted above, the most important step is to determine whether there is a respiratory component to the acidosis (i.e., a primary respiratory acidosis), because the treatment approach differs. If there is inadequate respiratory compensation, the most appropriate treatment will be to first correct the respiratory problem.
The adverse effects of acidemia make the concept of buffer therapy teleologically appealing, but the role of buffer therapy in both cardiac arrest and severe metabolic acidosis is uncertain. The traditional therapeutic buffer, sodium bicarbonate, may have negative effects in the treatment of acidosis. Bicarbonate therapy results in the generation of significant quantities of CO2, which diffuses readily into cells—particularly those of the central nervous system—and may therefore cause paradoxical worsening of intracellular acidosis. An abrupt CO 2 load may also exceed the ventilatory capacity of a maximally ventilating patient, producing respiratory failure. After successful treatment with bicarbonate, "overshoot" alkalosis may result; this problem occurs most commonly in organic acidosis, when reversal of the acidosis results in reformation of bicarbonate. Finally, bicarbonate therapy imposes an osmotic and sodium load (1000 meq/L of typical 1N solution). These concerns suggest that bicarbonate therapy should not be used in the ED treatment of mild to moderate metabolic acidosis. Bicarbonate therapy is reasonable, however, in situations where the effects of acidosis are so severe that they preclude or jeopardize therapy for the underlying disease ( Iablle.21:4).
TABLE 21-4 Indications for Bicarbonate Therapy in Metabolic Acidosis
When bicarbonate is used, Adrogue and Madias15 recommend administering 0.5 meq/kg bicarbonate for each meq/dL desired rise in [HCO2-]. The goal is to restore adequate buffer capacity ([HCO3-] > 8 meq/dL) or achieve clinical improvement in shock or dysrhythmias. Bicarbonate should be given as slowly as the clinical situation permits; 1.5 ampules of sodium bicarbonate in 500 mL D5W produces a nearly isotonic solution for infusion. Adequate time should be allowed for the desired effect to be achieved, and close monitoring of acid-base balance, especially in patients with organic acidosis, is critical.
Newer buffers appear to show promise in the treatment of metabolic acidosis. Carbicarb, an equimolar solution of sodium bicarbonate and sodium carbonate (CO 32-), produces significantly less CO2 than an equimolar dose of bicarbonate. The carbonate ion, a strong base, combines avidly with protons to form bicarbonate, resulting in increased pH, an increased bicarbonate concentration, and limited CO 2 production. Clinical studies of Carbicarb use in lactic acidosis models have shown improvements in pH with little or no change in Pco2J i7 However, large-scale studies in ED patients do not yet exist, and Carbicarb, while promising, remains experimental.
Was this article helpful?