Vous êtes sur la page 1sur 4

JASN Express. Published on March 5, 2008 as doi: 10.1681/ASN.



Bicarbonate Therapy in Severe Metabolic Acidosis

Sandra Sabatini and Neil A. Kurtzman

Department of Internal Medicine, Texas Tech University Health Sciences Center, Lubbock, Texas

ABSTRACT monly termed potential bicarbonate.

The utility of bicarbonate administration to patients with severe metabolic acidosis Giving bicarbonate to a patient with a
remains controversial. Chronic bicarbonate replacement is obviously indicated for true bicarbonate deficit is not controver-
patients who continue to lose bicarbonate in the ambulatory setting, particularly sial. Controversy arises when the de-
patients with renal tubular acidosis syndromes or diarrhea. In patients with acute crease in bicarbonate concentration is
lactic acidosis and ketoacidosis, lactate and ketone bodies can be converted back the result of its conversion to another
to bicarbonate if the clinical situation improves. For these patients, therapy must base, which, given time, can be converted
be individualized. In general, bicarbonate should be given at an arterial blood pH back to bicarbonate. If one knew that the
of 7.0. The amount given should be what is calculated to bring the pH up to 7.2. timely and efficient conversion of aceto-
The urge to give bicarbonate to a patient with severe acidemia is apt to be all but acetate and -hydroxybutyrate or lactate
irresistible. Intervention should be restrained, however, unless the clinical situation back to bicarbonate would occur with-
clearly suggests benefit. Here we discuss the pros and cons of bicarbonate therapy out morbidity or mortality, then there
for patients with severe metabolic acidosis. would be no reason even to contemplate
giving bicarbonate.
J Am Soc Nephrol : , . doi: 10.1681/ASN.2007121329
In considering acute bicarbonate re-
placement, four questions should be
considered: (1) What are the deleterious
Metabolic acidosis is an acid-base disor- (P 0.01). Also, 80% of nephrologists
effects of acidemia, and when are they
der characterized by a primary con- would consider the PCO2 in making their
manifest? (2) When is acidemia severe
sumption of body buffers including a fall decision to treat, whereas only 59% of enough to warrant therapy? (3) How
in blood bicarbonate concentration. intensivists would (P 0.02). For pa- much bicarbonate should be given, and
There are many causes (Table 1), and tients with lactic acidosis, 86% of neph- how is that amount calculated? (4) What
there are multiple mechanisms that min- rologists would treat with bicarbonate, are the deleterious effects of bicarbonate
imize the fall in arterial pH. A patient whereas two thirds of intensivists would therapy?
with metabolic acidosis may have a nor- give bicarbonate (P 0.05). A wider Severe acidemia causes a decrease in
mal or even high pH if there is another variance was noted in the therapy of dia- myocardial contractility, a fall in cardiac
primary, contravening event that raises betic ketoacidosis. Sixty percent of neph- output, and a fall in BP. Acidemia also
the bicarbonate concentration (vomit- rologists would treat with bicarbonate decreases the binding of norepinephrine
ing) or lowers the arterial PCO2 (respira- versus 28% of intensivists (P 0.01). to its receptors. It also shifts the oxyhe-
tory alkalosis). Metabolic acidosis differs Both groups would administer bicarbon- moglobin curve to the right, allowing
from acidemia in that the latter refers ate by constant infusion, targeting an ar- more O2 to be releasedthe Bohr Effect.
solely to a fall in blood pH and not the terial pH of 7.2 as a goal. Seventy-five Protons bind to intracellular proteins as
process. percent of nephrologists would calculate well as extracellular proteins, especially
A recent online survey by Kraut and the amount of bicarbonate required,
Kurtz1 highlighted the uncertainty over whereas only one third of intensivists
when to give bicarbonate to patients with would do this. Published online ahead of print. Publication date
metabolic acidosis. They reported that Metabolic acidosis results from a loss available at www.jasn.org.

nephrologists will prescribe therapy at a of bicarbonate from the body (e.g., diar- Correspondence: Dr. Neil A. Kurtzman, Depart-
higher pH compared with critical care rhea) or from its titration to an anionic ment of Internal Medicine, 3601 4th MS 9410, Texas
Tech University Health Sciences Center, Lubbock,
physicians. Forty percent of the intensiv- base that often can be converted back to TX 79430. Phone: 806-743-3181; Fax: 806-743-
ists would not give bicarbonate unless bicarbonate, such as seen in diabetic ke- 1092; E-mail: neil.kurtzman@ttuhsc.edu
the pH were 7.0; only 6% of nephrolo- toacidosis or lactic acidosis (Table 1). Copyright 2008 by the American Society of
gists would wait until pH gets this low This nonbicarbonate base anion is com- Nephrology

J Am Soc Nephrol : , 2008 ISSN : 1046-6673/00- 1


Table 1. Causes of severe metabolic acidosis is given (1.5 mmol/kg over 5 min).8 By
General Mechanism Specific Clinical Examples contrast, infusion with THAM (Hos-
True HCO3 deficit pira Inc., Lake Forest, IL) or CarbiCarb
kidney Renal tubular acidosis (International Medication System,
gastrointestinal Diarrhea South El Monte, CA) does not affect ar-
H gain terial PCO2.8,9 These observations have
exogenous acid NH4Cl administration, toxins led some investigators to recommend ei-
abnormal lipid metabolism Diabetic ketoacidosisa ther of these compounds as preferred
abnormal carbohydrate metabolism Lactic acid therapy.2
normal protein metabolism Uremic acidosis Bicarbonate therapy is also associated
Also a component of true HCO3 deficit because ketone bodies (potential HCO3) are lost in the urine
both before and after admission.
with an increase in mortality. This has
been noted in humans and experimental
albumin and hemoglobin. Thus, aci- The volume of distribution of bicar- animals under a variety of acidemic con-
demia may adversely affect cell functions bonate is approximately that of total ditions.10 12 The increase in mortality is
such as enzymatic reactions, ATP gener- body water. In patients with metabolic blamed on a fall in BP and cardiac out-
ation, fatty acid biosynthesis, and bone acidosis, it is said to vary from 50% to put. There are also shifts in ionized cal-
formation/resorption.2 4 100%, depending on the severity of the cium; in strong acid acidosis, potassium
All drugs that are the salts of weak acidemia.6 This distribution will obvi- also shifts out of the cell sensitizing the
bases or acids have a dissociation con- ously affect the calculated bicarbonate heart to abnormal electrical activity and
stant. Those that are the salts of a weak deficit. Any calculated amount will be subsequent arrhythmias. Moreover, a
acid will have more of the drug in the only approximate, of course. The patient paradoxic intracellular acidosis may
nonionized form in an acid environment should be carefully monitored and bicar- occur when giving bicarbonate therapy
and, thus, may manifest increased toxic- bonate administration altered, when because CO2 generated from its titration
ity. One such example of enhanced tox- given, to suit the course. Fernandez et al.7 freely diffuses across the cell membrane.
icity during acidemia is that of acetylsal- derived a more precise formula for calcu- In addition, both volume expansion and
icylic acid. Other salts of weak acids are lating the bicarbonate space: (0.4 2.6/ hypernatremia can occur; in patients
tolbutamide, methotrexate, phenobarbi- PHCO3) (body weight). A graphic repre- with compromised cardiac output, ful-
tal, and phenytoin. The degree of ioniza- sentation of the formula (Figure 5 in minate congestive heart failure with flash
tion is only one of the factors that deter- reference7) shows that the apparent bi- pulmonary edema may result.
mine a drugs solubility across cell carbonate space increases quite mark- Many in vitro studies show that intra-
membranes, but it is crucial in those edly with acidemia but decreases very lit- cellular alkalinization hastens cell death
body compartments in which pH may tle with alkalemia. Although they used after anoxia13; if cell water is maintained
change.5 actual data from several human studies, at pH 6.8, for example, more tissue re-
The optimum extracellular pH for all it is not clear that renal response to aci- mains viable.14,15 Bicarbonate adminis-
physiologic mechanisms and organ demia was accounted for as they ana- tration may stimulate superoxide forma-
functions is 7.4. By contrast, intracellular lyzed acute acid-base disorders. This for- tion, increase proinflammatory cytokine
pH is approximately 7.1 in virtually every mula, like any other guide to bicarbonate release, or enhance apoptosis. Whether
tissue studied. Many diverse mecha- treatment, should just be a starting point, these observations relate to human dis-
nisms are in place to maintain both ex- which is modified as events unfold. orders with acidemia is unknown. Re-
tracellular and intracellular pH within In some patients, only a small amount bound alkalemia may also occur after
this very narrow range. Deviations from of bicarbonate may be required. For ex- base administration, especially when the
normal pH will obviously decrease the ample, if a patient has a PCO2 of 13 PCO2 is low. Giving bicarbonate to both
efficiency of all reactions, although the mmHg and bicarbonate of 4 mEq/L, animals and humans increases blood lac-
degree will vary depending on the spe- then his arterial pH is 7.1. If the bicar- tate and ketone bodies.6,16 18 This po-
cific event. For example, whereas aci- bonate is doubled (raised to only 8 mEq/ tential bicarbonate will be converted
demia protects the central nervous sys- L), then the blood pH will increase to 7.4. back to actual bicarbonate unless it lost
tem against seizures, it sensitizes the This is true only if the PCO2 does not in the urine.
myocardium to arrhythmias. Because we change. In this example, given a static
do not measure intracellular pH, we PCO2, if the bicarbonate concentration
must use extracellular pH (arterial or ve- rises only 1 mEq/L, then the pH would be DIABETIC KETOACIDOSIS
nous) as a surrogate. Most authorities in above 7.2. Arterial PCO2 typically, how-
acid-base physiology would give bicar- ever, does not remain the same after In ketoacidosis, substantial amounts of
bonate to a patient with an arterial pH NaHCO3 infusion. In patients with se- acetoacetate and -hydroxybutyrate are
7.1, but, as we discuss, this is not a hard vere acidosis, it rises 6.7 1.8 mmHg lost in the urine before the patient arrives
and fast rule. when an infusion of sodium bicarbonate at the hospital. Thus, not only has the

2 Journal of the American Society of Nephrology J Am Soc Nephrol : , 2008


patient converted bicarbonate to poten- ninth time in diabetic ketoacidosis. He sponsible for the acidosis, base therapy
tial bicarbonate, he is truly bicarbonate was poorly responsive and had Kussmaul will be futile.
deficient. More urinary loss of ketone respirations. Before any therapy, he had a If bicarbonate is given, then its
bodies occurs after fluid administration plasma Na of 140 mEq/L, K of 4 mEq/L, amount should be calculated as the de-
and volume repletion. Hence, the ubiq- Cl of 109 mEq/L, CO2 of 3 mEq/L, and sired minus the observed bicarbonate
uitous hyperchloremic metabolic acido- creatinine of 1 mg/dl. The arterial pH concentration using a volume of distri-
sis we see the day after insulin therapy is was 6.95, PCO2 was 14 mmHg, and the bution of total body water. It should also
initiated. In ketoacidosis, it is almost calculated HCO3 was 3 mEq/L. Urine be assumed that the arterial PCO2 will not
never necessary to give bicarbonate even and blood ketones were strongly posi- change. The desired bicarbonate concen-
though the patient is bicarbonate defi- tive. He was treated with insulin and ap- tration at this unchanged PCO2 is that
cient unless renal function is perma- propriate fluid and electrolyte replace- which will give an arterial pH of 7.2. This
nently impaired. Therapy with fluids and ment. He was not given bicarbonate. The calculation is only an approximation. At
electrolytes restores extracellular volume next day he was fully oriented. His the end of 2 h, an arterial blood gas and
and renal blood flow, thus enhancing the plasma Na was 142, K was 4, Cl was 114, chemistries should be remeasured and a
renal excretion of acid and regenerating and CO2 was 18 mEq/L. The remainder new plan for the next 2 h made. Note that
bicarbonate. Okuda et al.18 demon- of his clinical course was unremarkable. patient 1 got no bicarbonate. He was oth-
strated in humans with diabetic ketoaci- erwise healthy with a normal cardiovas-
dosis, as well as in the in situ acidemic Patient 2 cular system, whereas patient 2 received
perfused rat liver (pH of 7.15), that bi- An 80-yr-old man was admitted with se- bicarbonate because he had a severely
carbonate therapy markedly increased vere congestive heart failure. He was hy- compromised cardiovascular system.
blood acetoacetate and -hydroxybu- potensive and oliguric. He had both pul- Thus, it is impossible to be dogmatic
tyrate levels. Infusion also increased monary and peripheral edema. His about the treatment of acidemia. No
blood lactate levels approximately three- baseline creatinine was known to be 1.6 hard and fast rule works for every pa-
fold. Others have reported similar find- mg/dl. On arrival at the emergency de- tient.
ings.19 Indeed, bicarbonate therapy actu- partment, his plasma Na was 135 mEq/L,
ally delays the removal of ketone bodies K was 4 mEq/L, Cl was 97 mEq/L, CO2
from the blood. was 7 mEq/L, and creatinine was 2.5 mg/ DISCLOSURES
dl. His arterial pH was 7.1, PCO2 was 20 None.
mmHg, and the calculated HCO3 was 6
LACTIC ACIDOSIS mEq/L. The blood lactate level was 20
mmol/L. The patient was intubated and REFERENCES
Lactic acidosis is an ominous event and placed on a respirator, keeping his PCO2
generally signifies severe tissue hypoxia. at 20 mmHg. Continuous venovenous 1. Kraut JA, I Kurtz: Use of base in the treat-
It may be secondary to an exogenous hemodialysis was begun with a bath con- ment of severe organic acidosis and critical
toxin such as cyanide or metformin or taining 14 mEq/L of bicarbonate. He was care physicians: Results of an online survey.
the result of severe tissue underperfusion given an infusion of 300 mEq of bicar- Clin Exp Nephrol 10: 111117, 2006
2. Luft F: Lactic acidosis update for critical care
from cardiogenic or hemorrhagic shock. bonate over 2 h; with a total body water
clinicians. J Am Soc Nephrol 12[Suppl 17]:
The mortality of lactic acidosis ap- of 43 L, one would aim for an HCO3 of 14 S15S19, 2001
proaches 80% or more. This is often be- mEq/L: (7 mEq/L 43 L 301 mEq). At 3. Schoolwerth A, Kaneko TM, Sediacek T,
cause of the inability to correct ade- the end of that time, his pH was 7.2 and Block CA, Remillard BD: Acid-base distur-
quately the underlying disorder(s). A the HCO3 was 13 mEq/L. Five days later, bances in the intensive care unit: Metabolic
acidosis. Semin Dial 19: 492 495, 2006
number of studies show even if blood he was transferred out of the intensive
4. Androgue HJ: Metabolic acidosis: Patho-
lactate level is lowered with drug therapy, care unit, his lactic acidosis resolved. physiology, diagnosis and management. J
mortality is unchanged.17,20,21 Nephrol 19: S62S69, 2006
5. Brunten LL, Lazo JS, Parker K, Buxton I, Blu-
FINAL THOUGHTS menthal D, eds.: Goodman and Gilmans
Pharmacological Basis of Therapeutics Dig-
ital Edition, 11th Ed., New York, McGraw
Bicarbonate therapy for metabolic aci- Hill, 2006
The following two cases demonstrate dosis is recommended at an arterial pH 6. Garella S, Dana CL, Chazan JA: Severity of
that therapy for acidemia requires flexi- varying from as low as 6.9 to as high as metabolic acidosis as a determinant of bi-
bility. 7.2. We suggest that bicarbonate therapy carbonate requirements. N Engl J Med 289:
121126, 1973
be given at pH 7.0 but that this target pH
7. Fernandez PC, Cohen RM, Feldman GM:
Patient 1 be a guide that is variable depending on The concept of bicarbonate distribution
A 20-yr-old man with a 5-yr history of clinical setting. Unless efforts are focused space: The crucial role of body buffers. Kid-
type 1 diabetes was admitted for the on reversing the underlying defects re- ney Int 36: 747752, 1989

J Am Soc Nephrol : , 2008 Bicarbonate and Metabolic Acidosis 3


8. Levraut L, Garcia P, Giunti C, Ichai C, 13. Curren RT, Gores GJ, Thurman RG, Lemas- 18. Okuda Y, Androgue HJ, Field JB, Nohara H,
Boureaba M, Ciebiera JP, Payan P, Grimaud ters JJ: Protection by acidic pH against an- Yamshita K: Counterproductive effects of
D: The increase in CO2 production induced oxic cell killing in the perfused rat liver: Ev- sodium bicarbonate in diabetic ketoacido-
by NaHCO3 depends on blood albumin and idence for a pH paradox. FASEB J 5: 207 sis. J Clin Endocrinol Metab 81: 314 319,
hemoglobin concentrations. Intensive Care 210, 1991 2000
Med 26: 558 564, 2000 14. Bing OH, Brooks WW, Messer JV: Heart 19. Makiasalo JH, Soini HO, Nordin HJ: Effects
9. Holmdahl MH, Wiklund L, Wetterberg T, muscle viability following hypoxia: Protec- of bicarbonate therapy on tissue oxygen-
Streat S, Wahlander S, Sutin K, Nahas G: The tive effect of acidosis. Science 180: 1297 ation during resuscitation of hemorrhagic
place of THAM in the management of aci- 1298, 1973 shock. Crit Care Med 17: 1170 1174,
demia in clinical practice. Acta Anaesthesiol 15. Zager RA, Schimpf BA, Gmur DJ: Physio- 1989
Scand 44: 524 527, 2000 logic pH: Effects on posthypoxic proximal 20. Rhee KH, Toro LO, McDonald GG, Nunnally
10. Sing RF, Branas CA, Sing RF: Bicarbonate tubular injury. Circ Res 72: 837 846, RL, Levin DL: Carbicarb, sodium bicarbon-
therapy in the treatment of lactic acidosis: 1993 ate, and sodium chloride in hypoxic lactic
Medicine or toxin? J Am Osteopath Assoc 16. Graf H, Leach W, Arieff AI: Evidence for a acidosis: Effect on arterial blood gases, lac-
95: 5257, 1995 detrimental effect of bicarbonate therapy in tate concentration, hemodynamic variables
11. Stacpoole PW: Lactic acidosis: The case hypoxic lactic acidosis. Science 227: 754 and intracellular pH. Chest 104: 913918,
against bicarbonate therapy. Ann Intern 756, 1985 1993
Med 105: 276 279, 1986 17. Bersin RM, Chatterjee, Arieff AI: Metabolic 21. Cooper DJ, Walley KR, Wiccs BR, Russell JA:
12. Gehlbach BK, Schmidt GA: Bench-to-bed- and hemodynamic consequences of so- Bicarbonate does not improve the dynamics
side review: Treating acid-base abnormali- dium bicarbonate administration in pa- in critically ill patients who lactic acidosis: A
ties in the intensive care unitThe role of tients with heart disease. Am J Med 87: prospective, controlled clinical study. Ann
buffers. Crit Care 8: 259 265, 2004 714, 1989 Intern Med 112: 492 498, 1990

4 Journal of the American Society of Nephrology J Am Soc Nephrol : , 2008