Lactic Acidosis

The new england journal of medicine

Review Article

Disorders of Fluids and Electrolytes Julie R. Ingelfinger, M.D., Editor Lactic Acidosis

Jeffrey A. Kraut, M.D., and Nicolaos E. Madias, M.D.

actic acidosis results from the accumulation of lactate and From Medical and Research Services, protons in the body fluids and is often associated with poor clinical out- Membrane Biology Laboratory, and the Division of Nephrology, David Geffen comes. The effect of lactic acidosis is governed by its severity and the clinical School of Medicine, University of Califor- L nia, Los Angeles, and the Veterans Af- context. Mortality is increased by a factor of nearly three when lactic acidosis ac- companies low-flow states or sepsis,1 and the higher the lactate level, the worse fairs Greater Los Angeles Healthcare Sys- 2 tem — both in Los Angeles (J.A.K.); and the outcome. Although hyperlactatemia is often attributed to tissue hypoxia, it can the Department of Medicine, Division of result from other mechanisms. Control of the triggering conditions is the only Nephrology, St. Elizabeth’s Medical Cen- effective means of treatment. However, advances in understanding its pathophysio- ter, and the Department of Medicine, Tufts University School of Medicine — logical features and the factors causing cellular dysfunction in the condition could both in Boston (N.E.M.). Address reprint lead to new therapies. This overview of lactic acidosis emphasizes its pathophysi- requests to Dr. Kraut at the Division of ological aspects, as well as diagnosis and management. We confine our discussion Nephrology, VHAGLA Healthcare Sys- tem, 11301 Wilshire Blvd., Los Angeles, to disorders associated with accumulation of the l optical isomer of lactate, which CA 90073; or at jkraut@ ucla.edu; or to represent the vast majority of cases of lactic acidosis encountered clinically. Dr. Madias at the Department of Medi- cine, St. Elizabeth’s Medical Center, 736 Cambridge St., Boston, MA 02135, or at Pathophysiological Features nicolaos.madias@steward.org. N Engl J Med 2014;371:2309-19. Normal Lactate Metabolism DOI: 10.1056/NEJMra1309483 The reaction integral to the generation or consumption of lactate is shown below: Copyright © 2014 Massachusetts Medical Society.

pyruvate + NADH + H+ ←→ lactate + NAD+.

Pyruvate is generated largely by anaerobic glycolysis (Embden–Meyerhof path- way). The redox-coupled interconversion of pyruvate and lactate occurs in the cytosol and is catalyzed by lactate dehydrogenase (LDH), a tetramer with five isoforms, each made up of different combinations of two subunits, LDHA and LDHB.3 The LDHA subunit has a higher affinity for pyruvate and its reduction than does LDHB; thus, the nature of the LDH isoforms in tissues affects lactate metabolism. The blood lactate:pyruvate ratio is normally 10:1, but it rises with an increased ratio of NADH concentration ([NADH]) to NAD+ concentration ([NAD+]) (redox state).4 Approximately 20 mmol of lactate per kilogram of body weight is produced in the human body daily, primarily by highly glycolytic tissues containing LDHA-rich LDH, such as skeletal muscle.3,5 Lactate is reconverted to pyruvate and consumed in the mitochondria of the liver, kidney, and other tissues, which have LDHB-rich LDH. The pathways include the Cori cycle, which generates glucose but consumes ATP in the liver and kidney (gluconeogenesis), as well as the tricarboxylic acid cycle and oxidative phosphorylation in the liver, kidney, muscle, heart, brain, and other tissues, which generate ATP when pyruvate is oxidized to carbon dioxide and water. Lactate consumption is subserved by intraorgan and interorgan lactate shuttles fa- cilitated by monocarboxylic acid transporters (MCTs), which mediate the influx and efflux of lactate and accompanying protons. Normally, the generation and consump-

n engl j med 371;24 nejm.org December 11, 2014 2309 The New England Journal of Medicine Downloaded from nejm.org on December 11, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved. The new england journal of medicine

tion of lactate are equivalent, which results in a Supplementary Appendix). Other disorders as- stable concentration of lactate in the blood.4,6 Lac- sociated with elevated epinephrine levels, such tate production can rise markedly, as exempli- as severe asthma (especially with overuse of

fied by its increase by a factor of several hundred β2-adrenergic agonists), extensive trauma, cardio- during maximal exercise,5 but it can also be rap- genic or hemorrhagic shock, and pheochromocy- idly consumed, as seen after cessation of exercise, toma, can cause hyperlactatemia through this seizures, or exogenous lactate loads.5,7 mechanism.9 In inflammatory states, aerobic gly- The bioenergetics of lactate generation can be colysis can also be driven by cytokine-dependent summarized as follows: stimulation of cellular glucose uptake10; in alkalemic disorders, it can be driven by stimulation of glucose + 2(ADP + inorganic phosphate) 6-phosphofructokinase.4 Aerobic glycolysis and tis- → 2 lactate + 2 H+ + 2 ATP. sue hypoxia are not mutually exclusive; under certain circumstances, both can contribute to hyper- Production of lactate ions by means of glycoly- lactatemia.4,9 sis is accompanied by the release of an equivalent Drugs that impair oxidative phosphorylation, number of protons from the hydrolysis of the such as antiretroviral agents and propofol, can generated ATP. Conversely, lactate consumption augment lactic acid production and on rare occa- removes an equivalent number of protons, there- sions cause severe lactic acidosis. Patients receiv- by maintaining the internal acid–base balance.4 ing these drugs should be monitored carefully. The liver accounts for up to 70% of whole- Hyperlactatemia body lactate clearance.11 In patients with sepsis, Hyperlactatemia occurs when lactate production even when they are hemodynamically stable and exceeds lactate consumption. It also signifies the have normal liver function, lactate clearance can addition of a number of protons equivalent to be reduced, possibly through inhibition of pyru- the number of excess lactate ions, regardless of vate dehydrogenase.12 Chronic liver disease exac- the prevailing acid–base status. Establishing the erbates hyperlactatemia due to sepsis or other pathogenesis of hyperlactatemia can be a valu- disorders,7,11 but in the absence of such disorders, able guide to therapy. even severe cirrhosis rarely generates blood lac- In tissue hypoxia, whether global or localized, tate levels that are more than minimally elevated. lactate is overproduced and underutilized as a However, hyperlactatemia is common in acute result of impaired mitochondrial oxidation (see fulminant liver disease, reflecting both reduced Fig. S1A and S1B in the Supplementary Appendix, clearance and increased production of lactate by available with the full text of this article at NEJM the liver,13 and is an important prognostic factor. .org).4 Even if systemic oxygen delivery is not low enough to cause generalized hypoxia, microcir- Effects on Cellular Function culatory dysfunction can cause regional tissue hy- The cellular dysfunction in hyperlactatemia is com- poxia and hyperlactatemia.8 Coexisting acidemia plex. Tissue hypoxia, if present, is a major factor. contributes to decreased lactate removal by the If the cellular milieu is also severely acidic, cel- liver; severe hypoxia and acidemia can convert the lular dysfunction is likely to be exacerbated. The liver into a net lactate-producing organ.4 latter factor alone can decrease cardiac contrac- Hyperlactatemia can also result from aerobic tility, cardiac output, blood pressure, and tissue glycolysis, a term denoting stimulated glycolysis perfusion; can sensitize the myocardium to car- that depends on factors other than tissue hypoxia. diac arrhythmias; and can attenuate the cardio- Activated in response to stress, aerobic glycolysis vascular responsiveness to catecholamines.14 is an effective, albeit inefficient, mechanism for In some studies, the severity of the acidemia rapid generation of ATP. In the hyperdynamic was a better predictor of cellular dysfunction and stage of sepsis, epinephrine-dependent stimula- clinical outcomes than the hyperlactatemia.15

tion of the β2-adrenoceptor augments the glyco- However, acidemia is often absent as a result of lytic flux both directly and through enhancement coexisting acid–base disorders.7,14 The interaction of the sarcolemmal Na+,K+-ATPase (which con- of systemic acidity and blood lactate and their sumes large quantities of ATP)9 (Fig. S1C in the effect on clinical outcomes require further study.

2310 n engl j med 371;24 nejm.org December 11, 2014 The New England Journal of Medicine Downloaded from nejm.org on December 11, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved. Lactic Acidosis

Whether hyperlactatemia itself has an effect ing acid–base disturbances.19 In lactic acidosis, − on cellular function remains unclear. In vitro stud- the ΔAG:ΔHCO3 ratio is often greater than 1, in ies suggest that lactate can depress cardiac con- part because the apparent space of distribution tractility,4,16 yet sodium lactate infusions that raised of protons exceeds that of lactate19,20; therefore, blood lactate to levels as high as 15 mmol per liter an increased ratio might not always suggest a did not depress hemodynamic measures in pa- coexisting acid–base disorder. tients after cardiac surgery.16 An elevated blood lactate level is essential for confirmation of the diagnosis. The lower limit Causes of the normal range for the blood lactate level, 0.5 mmol per liter, is consistent among clinical The major causes of lactic acidosis and their laboratories, but the upper limit can vary sub- presumed mechanisms are listed in Table 1. stantially, from as low as 1.0 mmol per liter to as Typically, they have been divided into disorders high as 2.2 mmol per liter.6,21,22 Therefore, the associated with tissue hypoxia (type A) and cutoff for abnormal values often differs among disorders in which tissue hypoxia is absent laboratories. Levels at the upper tier of normal (type B). However, the evidence of tissue hy- values have been associated with increased mor- poxia can be subtle, and hyperlactatemia can tality among seriously ill patients.21,23 Thus, blood be of both hypoxic and nonhypoxic origin.4,9,10 lactate concentrations at the upper tier of normal Cardiogenic or hypovolemic shock, severe heart values or slightly increased from a previous base- failure, severe trauma, and sepsis are the most line value, although remaining within the normal common causes of lactic acidosis, accounting range, can augur a poor outcome and call for for the vast majority of cases.17 monitoring of the patient. Previously, the definition of lactic acidosis in- cluded a blood pH of 7.35 or lower and a serum Diagnosis − 24 [HCO3 ] of 20 mmol per liter or lower. How- Evidence of severe cardiopulmonary disease, the ever, the absence of one or both of these features systemic inflammatory response syndrome, sep- because of coexisting acid–base disorders does not sis, severe trauma, or volume depletion offers im- rule out lactic acidosis. For example, coexisting portant clues for diagnosing lactic acidosis. An respiratory alkalosis can increase the blood pH elevated serum anion gap, particularly a value into the alkalemic range, whereas coexisting meta- higher than 30 mmol per liter, can provide sup- bolic alkalosis can result in both hyperbicarbona- portive evidence. However, other causes of a raised temia and alkalemia (Fig. 1 and 2). In contrast, anion gap, such as ketoacidosis and toxic alco- coexisting respiratory acidosis can cause severe hol ingestion, should always be considered.18,19 acidemia (Fig. 2). The increase in the anion gap (ΔAG) can mirror Lactic acidosis due to grand mal seizures is the blood lactate level, but a close relationship associated with normokalemia, because the con- might not always be found, since anions other current entry of lactate and protons into cells than lactate often contribute to the ΔAG. negates the need for potassium exit from cells to A normal anion gap does not rule out lactic maintain electroneutrality. However, hyperkalemia acidosis. In one study, 50% of patients with a is commonly observed in critically ill patients, be- serum lactate level of 5 to 10 mmol per liter did cause they often have renal failure. In addition, not have an elevated anion gap.18 Correction of potassium is released from damaged tissues. How- the anion gap for the effect of serum albumin ever, hypokalemia can occur when β2-adrenoceptor can improve its sensitivity, but many cases will stimulation drives potassium into cells. still escape detection. Therefore, the serum an- A serum osmolal gap of more than 20 mOsm ion gap lacks sufficient sensitivity or specificity per kilogram of water has been reported in some to serve as a screening tool for lactic acidosis. cases,25 probably reflecting the release of osmoti- Because a 1:1 relationship between the ΔAG cally active solute from ischemic tissues. However, and the decrease in serum bicarbonate concentra- other disorders characterized by an increased os- − − tion ([HCO3 ]), ΔHCO3 , is often found in ketoaci- molal gap and hyperlactatemia (e.g., exposure to dosis, deviations from this ratio suggest coexist- toxic alcohols) should be ruled out.

n engl j med 371;24 nejm.org December 11, 2014 2311 The New England Journal of Medicine Downloaded from nejm.org on December 11, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved. The new england journal of medicine

Table 1. Causes of Lactic Acidosis.*

Cause Presumed Mechanism or Mechanisms Comments

Cardiogenic or hypovolemic Decreased O2 delivery to tissues; epineph- With sepsis, these causes account for the majority of cas- shock, advanced heart failure, rine-induced β2-adrenoceptor stimula- es of lactic acidosis or severe trauma tion can be a contributory factor

Sepsis Epinephrine-induced β2-adrenoceptor stim- Evidence of decreased O2 delivery can be subtle; even in ulation with or without decreased O2 de- the absence of macrocirculatory impairment, dysfunc- livery to tissues; reduced clearance of lac- tion of microcirculation can be present tate even in hemodynamically stable patients

Severe hypoxemia Decreased O2 delivery to tissues Requires Pao2 <30 mm Hg

Carbon monoxide poisoning Decreased O2 delivery to tissues, interfer- Hyperbaric O2 therapy is recommended if pH <7.1 ence with oxidative phosphorylation

Severe anemia Decreased O2 delivery to tissues Requires hemoglobin level <5 g/dl

Vigorous exercise, seizures, Increased O2 requirements The decrease in pH and hyperlactatemia is transient; lactic or shivering acidosis can impair exercise performance Diabetes mellitus Mechanism unclear The risk of death in patients with ketoacidosis can be increased by coexisting lactic acidosis Cancer Increased glycolytic activity of tumor Lactic acidosis can be seen in association with lympho- − (Warburg effect), tumor tissue hypoxia, mas, leukemias, and solid tumors; HCO3 administra- decreased clearance of lactate with tion may increase lactic acid production; acidic micro- severe liver metastases environment is critical for tumorigenesis, angiogene- sis, and metastasis Liver disease Lactate clearance decreased Fulminant liver disease can cause substantial hyperlactatemia; hyperlactatemia is usually mild with chronic liver disease alone; lactate clearance can also be decreased when liver function is normal, in association with sepsis

Pheochromocytoma Decreased O2 delivery to tissues and epi- In rare cases, lactic acidosis is a presenting feature of nephrine-induced β2-adrenoceptor pheochromocytoma stimulation Metformin Interference with oxidative phosphorylation, This is usually seen in association with high plasma met- suppression of hepatic gluconeogenesis formin levels; treatment with dialysis is beneficial Nucleoside reverse-transcriptase Interference with oxidative phosphorylation Marked hyperlactatemia is uncommon in the absence of inhibitors other predisposing factors

Cocaine Decreased O2 delivery to tissues and epi- Marked hyperlactatemia is seen in some patients having nephrine-induced β2-adrenoceptor seizures or being stimulation restrained Toxic alcohols, methanol, ethyl- Interference with oxidative phosphorylation The increase in lactate is small; a small increase in the os- ene glycol, diethylene glycol molal gap (usually <20 mOsm/kg H2O) can be seen in some cases of lactic acidosis without toxic alcohols

Propylene glycol d-Lactate and l-lactate are normal products Lactic acidosis can occur in the absence of impaired oxida- of metabolism tive phosphorylation Salicylates Interference with oxidative phosphorylation Hyperlactatemia is usually minimal Cyanide Interference with oxidative phosphorylation Lactic acidosis is an important manifestation of poisoning

β2 agonists Stimulation of aerobic glycolysis This is most common with treatment of acute asthma; hypokalemia can result from enhanced cellular uptake of potassium Propofol Interference with oxidative phosphorylation Lactic acidosis can be seen with prolonged high-dose infusion Thiamine deficiency Impairment of pyruvate dehydrogenase This is most common in children or adults receiving par- activity enteral nutrition or those with fulminant beriberi

* Pao2 denotes partial pressure of arterial oxygen.

2312 n engl j med 371;24 nejm.org December 11, 2014 The New England Journal of Medicine Downloaded from nejm.org on December 11, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved. Lactic Acidosis

1 − 2 3 4 140 A 10 − A− A A− 18 120 − 20 − HCO3 36 HCO3 24 10 − − 100 HCO3 HCO3 4 30 80 Na+ Na+ Na+ Na+ 140 140 140 137 − 60 Cl− Cl Cl− 112 106 100 Cl− 40 87

Ion Concentration (mmol/liter) 20

0 Normal Severe Lactic Lactic Acidosis Lactic Acidosis Acidosis and Metabolic and Non–Anion Gap Alkalosis Metabolic Acidosis

PaCO2 (mm Hg) 40 15 43 26 pH 7.40 7.05 7.47 7.21 Lactate (mmol/liter) 1 18 8 6

Figure 1. Changes in Acid–Base and Electrolyte Composition in Patients with Lactic Acidosis with or without Coexisting Metabolic Acid–Base Disturbances. From left to right, the bar pairs depict normal acid–base status, severe lactic acidosis induced by hemorrhagic shock, lactic acidosis and metabolic alkalosis in a patient with advanced heart failure who is receiving diuretic agents, and lactic acidosis and non–anion gap metabolic acidosis in a patient with septic shock who has been re- suscitated with large quantities of normal saline. In bar pair 2, the increment in the anion gap (the anion gap is cal- culated by subtracting the serum concentrations of chloride and bicarbonate anions from the concentration of sodi- − um cations) exceeds the decrement in the serum bicarbonate level (ΔAG:ΔHCO3 ratio, 1.3), a common occurrence in lactic acidosis. In bar pairs 2, 3, and 4, the increment in the anion gap cannot be completely accounted for by the increase in the lactate concentration. Several studies have indicated that unmeasured anions other than lactate account for a substantial portion of the increased anion gap in lactic acidosis. A− denotes unmeasured serum anions, and Paco2 partial pressure of arterial carbon dioxide. The numbers associated with ions are ion concentrations in millimoles per liter.

Treatment tality in association with hydroxyethyl starch syn- thetic-colloid solutions provide evidence against Supporting the Circulation and Ventilation their use. If a colloid solution is indicated, albu- Restoring tissue perfusion after hemodynamic min should be used. Saline administration can compromise is essential in the treatment of pa- generate or exacerbate a non–anion gap metabolic tients with lactic acidosis. Vasopressors and ino- acidosis29 and reduce ionized calcium levels, fac- tropic agents should be administered as needed.26,27 tors that could depress cardiac function.30,31 Also, Acidemia blunts the response to catecholamines, chloride-rich solutions have been linked to acute thereby increasing the required dose.14 High dos- kidney injury.32 Crystalloids containing bicarbon- es of catecholamines can aggravate hyperlactate- ate or its precursors (balanced salt solutions), such mia by reducing tissue perfusion or overstimulat- as Ringer’s solution with lactate and Plasma-Lyte ing the β2-adrenoceptor; therefore, the dose should (Baxter International) with acetate and gluconate, be adjusted carefully. will not cause non–anion gap metabolic acidosis Crystalloid and colloid solutions are both ef- and may reduce the risk of acute kidney injury, fective in restoring tissue perfusion in patients but they can occasionally cause metabolic alka- with sepsis or hypovolemia.28 However, reports of losis.31,33 acute kidney injury, bleeding, and increased mor- A reduced need for renal replacement therapy

n engl j med 371;24 nejm.org December 11, 2014 2313 The New England Journal of Medicine Downloaded from nejm.org on December 11, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved. The new england journal of medicine

1 A− 2 3 4 5 140 10 A− A− A− 36 25 33 A− − 120 HCO3 36 − 24 HCO3 − − HCO3 100 HCO3 12 4 8 − HCO3 20 80 + + + Na+ Na Na Na Na+ 140 140 140 142 60 − − 132 Cl Cl− Cl Cl− 106 100 103 101 − 40 Cl 76

Ion Concentration (mmol/liter) 20

0 Normal Severe Lactic Lactic Acidosis Lactic Acidosis Lactic Acidosis, Acidosis and Respiratory and Respiratory Respiratory Alkalosis Acidosis Alkalosis, and Metabolic Alkalosis

PaCO2 (mm Hg) 40 15 16 42 30 pH 7.40 7.05 7.50 6.90 7.45 Lactate (mmol/liter) 1 18 10 16 17

Figure 2. Changes in Acid–Base and Electrolyte Composition in Patients with Lactic Acidosis with or without Coexisting Respiratory and Metabolic Acid–Base Disturbances. From left to right, the bar pairs depict normal acid–base status, severe lactic acidosis induced by hemorrhagic shock, lactic acidosis and respiratory alkalosis in a patient with sepsis, lactic acidosis and respiratory acidosis in a patient with trauma-associated lactic acidosis complicated by pulmonary insufficiency and progressive carbon dioxide retention, and lactic acidosis, respiratory alkalosis, and metabolic alkalosis in a patient with sepsis undergoing gastric drainage. The numbers associated with ions are ion concentrations in millimoles per liter.

has been reported in seriously ill patients receiving sive ventilation may also be required to prevent balanced salt solutions rather than saline,32,34 but hypercapnia, particularly if acidemia persists or opinions differ regarding which solution should worsens.26 be favored.28,31 Solutions containing a racemic mixture of d-lactate and l-lactate generate as much Improving the Microcirculation base as do solutions with an equimolar concen- Abnormalities of the microcirculation, if persis- tration of only l-lactate.35 Infusion of large quan- tent, can augur clinical deterioration and death.8,37 tities of Ringer’s lactate can increase blood lac- Several agents, including dobutamine, acetylcho- tate levels, but the increment is often small in the line, and nitroglycerin, have been shown to im- absence of abnormalities in lactate clearance.36 prove microvascular perfusion independently of Citrate-containing solutions can lead to the gen- systemic hemodynamics, to reduce hyperlactate- eration of microthrombi.33 Randomized, controlled mia, and even to improve the outcome.38,39 Mea- studies are needed to determine the most effec- sures to rescue the microcirculation are likely to tive and safe crystalloid.31,33 become a high priority in the future. Oxygen delivery to tissues depends on the cardiac output, regional blood flow, hemoglobin Initiating Cause-Specific Measures o concentration, and partial pressure of oxygen (P 2). Resuscitative efforts should be complemented by Red-cell transfusions should be administered to measures targeting the cause or causes of lactic maintain the hemoglobin concentration at a level acidosis. Such measures can include treatment of o above 7 g per deciliter. An adequate P 2 should be sepsis with the appropriate antibiotic agents; man- maintained by ensuring an appropriate inspired agement of arrhythmias, resynchronization ther- oxygen concentration, with endotracheal intuba- apy, and left-ventricular assist devices for advanced tion and mechanical ventilation as needed. Inva- heart failure; coronary intervention for acute myo-

2314 n engl j med 371;24 nejm.org December 11, 2014 The New England Journal of Medicine Downloaded from nejm.org on December 11, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved. Lactic Acidosis cardial infarction; surgery for trauma, tissue ische out increasing carbon dioxide — or, even better, mia, or toxic megacolon; dialysis for removal of that are capable of consuming carbon dioxide toxins or drugs; discontinuation of certain drugs; — is warranted to determine their potential role and reduction of tumor mass for cancer. in the treatment of metabolic acidosis.

Base Administration Potential Future Therapies Given the potentially deleterious effects of an The sodium–hydrogen (Na+–H+) exchanger NHE1 acidic environment, some clinicians recommend is activated during lactic acidosis, leading to therapy with intravenous sodium bicarbonate for deleterious sodium and calcium overload in the severe acidemia (blood pH, <7.2).14,40,41 However, heart; its inhibition reduces cellular injury. In the value of bicarbonate therapy in reducing mor- experimental models of lactic acidosis due to tality or improving hemodynamics remains un- sepsis, hypoxia, hemorrhagic shock, or cardiac proven.14,17,30,42 This absence of evidence that bi- arrest, NHE1 inhibitors attenuated the lactic carbonate therapy is beneficial has been attributed acidosis and hypotension, improved myocardial primarily to two adverse events that occur with performance and tissue oxygen delivery, enabled its administration: intracellular acidification due resuscitation, and reduced mortality.44,45 These to the accumulation of carbon dioxide after bi- promising results in animals call for controlled carbonate infusion and a pH-dependent decrease studies in humans. in levels of ionized calcium, a modulator of car- Cancer cells are programmed to use aerobic diac contractility.14,30 Intracellular acidification is glycolysis and lactate production as their main presumed to be more frequent and severe when energy source (the Warburg effect).46,47 The export large quantities of bicarbonate are administered of lactate and protons in the tumor microenviron- rapidly in patients with severe circulatory failure, ment has emerged as a critical regulator of can- which impedes the removal of carbon dioxide from cer development, maintenance, and metastasis. tissues and its excretion by the lungs. Circum- Inhibitors of LDH and MCT lactate transporters vention of these complications might allow the are being investigated as promising cancer ther- putative benefits of bicarbonate to be manifested. apies.3 These compounds might also prove effec- Using dialysis to provide bicarbonate can pre- tive in the management of tumor-induced and vent a decrease in ionized calcium, prevent volume other types of systemic lactic acidosis. overload and hyperosmolality (potential complications of bicarbonate infusion), and remove sub- Monitoring of Patients, Goals of Therapy, stances associated with lactic acidosis, such as and Prognosis metformin. Through an alkalinizing effect, base Measures to monitor and recommended goals of administration, whether by means of infusion or therapy are shown in Table 2. Close assessment dialysis, increases net lactic acid production.43 of hemodynamic, oxygenation, and acid–base sta- Substantial clearance of lactate can be achieved tus is paramount. with dialysis, although the quantity cleared is The detection of tissue hypoxia is important much lower than the quantity of lactate produced for assessing the effectiveness of resuscitation and in severe lactic acidosis. Continuous dialysis is the need for other procedures to match oxygen often favored over intermittent dialysis because delivery and demand. Although measurement of it delivers bicarbonate at a lower rate and is as- the blood lactate level is often used for this pur- sociated with fewer adverse events in patients with pose, hyperlactatemia does not always signify hemodynamic instability. Controlled studies of the tissue hypoxia.47,51,52 Central venous oxygen satu- effect of dialysis on lactic acidosis are warranted. ration has been suggested as a replacement or Other buffers have been developed to mini- complementary measure, with the goal being a mize carbon dioxide generation, including THAM value greater than 70%.27,51,52 A recent study of (tris-hydroxymethyl aminomethane)14 and Carbi- septic shock indicated that the use of this mea- carb (a 1:1 mixture of sodium carbonate and sure provided no additional benefit as compared sodium bicarbonate).14 Only THAM is currently with the usual therapy without central hemody- available for clinical use, but further study of these namic targets.48 Further study of this issue is and other, novel compounds that buffer acid with- needed. New methods of detecting tissue hypoxia,

n engl j med 371;24 nejm.org December 11, 2014 2315 The New England Journal of Medicine Downloaded from nejm.org on December 11, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved. The new england journal of medicine showed no addi - no showed indicated that indicated 48 48 saturation requires saturation 2 Comments carrying capacity by optimizing hemoglobin con - hemoglobin optimizing by capacity carrying 2 reduction of blood lactate by 20% every 2 hr for the first 8 hr 8 first the for hr 2 every 20% by lactate blood of reduction 49 saturation. Monitoring of central venous O venous central of Monitoring saturation. 2 In one study, one In 50 , 49 often termed microcirculatory distress syndrome. Documentation of improvement of Documentation syndrome. distress microcirculatory termed often study. under remains goal important an as flow microcirculatory in Peripheral venous and arterial values are interchangeable. The initial value has value initial The interchangeable. are values arterial and venous Peripheral prognosis for value more have measurements serial but significance, prognostic stud - some in beneficial been has therapy Lactate-guided therapy. for guiding and ies. domized, controlled studies and expert opinion. A recent study recent A opinion. expert and studies controlled domized, was associated with a decrease in morbidity and mortality. The use of lactate clear - lactate of use The mortality. and morbidity in decrease a with associated was investigation. under remains therapy and guide to monitor ance mia can be a sign of a poor prognosis. The use of central venous blood gases to gases blood venous central of use The prognosis. poor a of sign a be can mia should measurement but established, been not has status acid–base tissue assess ef - hemodynamic Adverse hypoperfusion. severe with patients for considered be acid–base of improvement for base of use The <7.2. pH at occur acidemia fects of benefit clinical of evidence of lack a of because controversial remains measures therapy. of complications and centration and O and centration insertion of catheters and use of special probes. A recent study recent A probes. special of use and catheters of insertion there is no additional value of this measure over the usual hemodynamic mea - hemodynamic usual the over measure this of value additional no there is sures. tional benefit of protocol-based therapy or use of central venous catheterization in catheterization venous central of use or therapy protocol-based of benefit tional the insights into new provide might studies Ongoing shock. septic of treatment func - renal indicator of adequate an is not alone volume Urine therapy. of goals creatinine. serum in changes consider also should clinician the and tion, Microcirculation abnormalities can persist despite improvement in systemic variables, systemic in improvement despite persist can Microcirculation abnormalities Blood lactate is a useful tool for screening, risk stratification, and prognosis. and stratification, risk screening, for tool useful a is lactate Blood The goal is to provide maximal O maximal provide to is goal The Measures reflect the adequacy of the macrocirculation. Criteria are derived from ran - from derived are Criteria macrocirculation. the of adequacy the reflect Measures Measures should be evaluated every few hours. Acidemia accompanying hyperlactate - accompanying Acidemia hours. few every evaluated be should Measures

2 co appropriate for appropriate 2 co ≥92% ≥70% partial pressure of carbon dioxide. 2 8–12 mm Hg mm 8–12 co >0.5 ml/kg/hr >0.5 65–70 mm Hg mm 65–70 <100 beats/min <100 Goal of Therapy ]; in lung-protective ventilation, Pa ventilation, lung-protective in ]; − 3 including proportion of perfused small vessels, small perfused of proportion including of heterogeneity and index, flow microvascular vessels small perfused status of patient; some recommendations are recommendations some patient; of status g/dl 10 approximately for [HCO ventilation is maintained at hypercapnic levels hypercapnic at maintained is Substantial improvement of microvascular indexes, microvascular of improvement Substantial Decrease to the normal range (<1 to 2 mmol/liter) 2 to (<1 range normal the to Decrease >7 g/dl, but can vary on the basis of cardiovascular of basis the on vary can g/dl, but >7 12–15 mm Hg in patients receiving mechanical receiving patients in Hg mm 12–15 Arterial blood pH, >7.2; Pa >7.2; pH, blood Arterial , and , 2 delivery 2 co saturation 2 ]) saturation − 2 3 tion, near infrared spectroscopy, infrared near tion, spec - polarization orthogonal and microvasculature of imaging tral [HCO venous blood pH, P pH, blood venous denotes partial pressure of arterial carbon dioxide, and P 2 Central venous O venous Central Hemoglobin level Hemoglobin Mean arterial blood pressure blood arterial Mean Arterial O Arterial Heart rate Heart pressure venous Central pressure wedge Pulmonary output Urine co Measures for Monitoring and Goals of Therapy in Patients with Lactic Acidosis.* Lactic with Patients in Therapy of Goals and Monitoring for Measures 2. Table Assessment of integrity of microcircula - of integrity of Assessment Blood lactate Blood Measure Hemodynamic measures Hemodynamic Acid–base measures (arterial and central and (arterial measures Acid–base Blood measures affecting O affecting measures Blood * Pa

2316 n engl j med 371;24 nejm.org December 11, 2014 The New England Journal of Medicine Downloaded from nejm.org on December 11, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved. Lactic Acidosis such as monitoring the concentration of hypoxia- of this target of lactate clearance was associated inducing factor in body fluids, could prove useful.53 with decreased morbidity and mortality.49 Evidence Measurement of the blood lactate level re- that in seriously ill patients even lactate levels at mains the cornerstone of monitoring for lactic the upper end of the normal range are associated acidosis. Lactate can be measured in arterial or with poor clinical outcomes argues for the nor- venous blood, since the values are virtually inter- malization of blood lactate as a primary goal of changeable.6 Devices for point-of-care measure- therapy.2,47,54 The usefulness of lactate-guided ther- ment are available. Although a single elevated apy and the level of lactate to target remain un- blood lactate level often predicts an adverse out- der investigation. come, sustained hyperlactatemia is associated with Given the central role of the microcirculation even worse prognoses. Transient hyperlactatemia in lactic acidosis,37 its evaluation both before and does not necessarily predict a poor clinical out- after various interventions can be useful. Hand- come.6,49 An interval of 2 to 6 hours has been sug- held devices allowing direct visualization of the gested for repeat lactate measurements,6 but this microcirculation have been developed, and their issue has not been examined rigorously. clinical role is undergoing study.26,60 Sustained hyperlactatemia in hospitalized pa- In patients with severe circulatory compromise, tients with diverse disorders is associated with a central venous blood more accurately reflects the large increase in mortality, regardless of status acid–base status of tissues than does arterial or with respect to shock or hypotension.6,47,49,54-57 Also, peripheral venous blood.14,61 However, it remains there is a dose–response relationship between lac- unproven whether monitoring central venous blood tate levels and mortality: the higher the level, the gases in patients with severe hypoperfusion im- greater the risk of death.2,6,7,56,57 Some studies have proves clinical outcomes. Monitoring of arterial indicated that acidemia accompanying hyperlac- blood gases is, of course, required for assessing tatemia increases mortality,15 but the role of sys- pulmonary gas exchange. temic acidity in affecting clinical outcomes remains to be elucidated. Dr. Kraut reports holding pending patents related to the use of selective NHE1 inhibitors in the treatment of metabolic aci- Changes in levels of blood lactate have been dosis (lapsed without the filing of a nonprovisional application) used to guide therapy.47,49-51,58,59 In a randomized, and systems, methods, and compositions for improved treat- controlled study, a reduction of at least 20% in ment of acidosis. No other potential conflict of interest relevant to this article was reported. serum lactate levels every 2 hours was targeted Disclosure forms provided by the authors are available with for the first 8 hours of resuscitation; achievement the full text of this article at NEJM.org.

References 1. Gunnerson KJ, Saul M, He S, Kellum 7. Bakker J, Nijsten MW, Jansen TC. Mild hyperlactatemia in stable septic pa- JA. Lactate versus non-lactate metabolic Clinical use of lactate monitoring in criti- tients is due to impaired lactate clearance acidosis: a retrospective outcome evalua- cally ill patients. Ann Intensive Care 2013; rather than overproduction. Am J Respir tion of critically ill patients. Crit Care 3:12. Crit Care Med 1998;157:1021-6. 2006;10:R22-R32. 8. Ince C. The microcirculation is the 13. Murphy ND, Kodakat SK, Wendon JA, 2. Nichol AD, Egi M, Pettila V, et al. motor of sepsis. Crit Care 2005;9:Suppl 4: et al. Liver and intestinal lactate metabo- Relative hyperlactatemia and hospital S13-S19. lism in patients with acute hepatic failure mortality in critically ill patients: a retro- 9. Levy B, Desebbe O, Montemont C, undergoing liver transplantation. Crit spective multi-centre study. Crit Care Gibot S. Increased aerobic glycolysis Care Med 2001;29:2111-8. 2010;14:R25. through beta2 stimulation is a common 14. Kraut JA, Madias NE. Treatment of 3. Doherty JR, Cleveland JL. Targeting mechanism involved in lactate formation acute metabolic acidosis: a pathophysio- lactate metabolism for cancer therapeu- during shock states. Shock 2008;30:417- logic approach. Nat Rev Nephrol 2012;8: tics. J Clin Invest 2013;123:3685-92. 21. 589-601. 4. Madias NE. Lactic acidosis. Kidney 10. Taylor DJ, Faragher EB, Evanson JM. 15. Lee SW, Hong YS, Park DW, et al. Lac- Int 1986;29: 752-74. Inflammatory cytokines stimulate glu- tic acidosis not hyperlactatemia as a pre- 5. van Hall G. Lactate kinetics in human cose uptake and glycolysis but reduce glu- dictor of in hospital mortality in septic tissues at rest and during exercise. Acta cose oxidation in human dermal fibro- emergency patients. Emerg Med J 2008; Physiol (Oxf) 2010;199:499-508. blasts in vitro. Circ Shock 1992;37:105-10. 25:659-65. 6. Kruse O, Grunnet N, Barfod C. Blood 11. Jeppesen JB, Mortensen C, Bendtsen 16. Mustafa I, Leverve XM. Metabolic and lactate as a predictor for in-hospital mor- F, Møller S. Lactate metabolism in chron- hemodynamic effects of hypertonic solu- tality in patients admitted acutely to hos- ic liver disease. Scand J Clin Lab Invest tions: sodium-lactate versus sodium chlo- pital: a systematic review. Scand J Trauma 2013 March 20 (Epub ahead of print). ride infusion in postoperative patients. Resusc Emerg Med 2011;19:74. 12. Levraut J, Ciebiera JP, Chave S, et al. Shock 2002;18: 306-10.

n engl j med 371;24 nejm.org December 11, 2014 2317 The New England Journal of Medicine Downloaded from nejm.org on December 11, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved. The new england journal of medicine

17. Jung B, Rimmele T, Le Goff C, et al. 32. Yunos NM, Bellomo R, Hegarty C, Sabiporide improves cardiovascular func- Severe metabolic or mixed acidemia on Story D, Ho L, Bailey M. Association be- tion, decreases the inflammatory re- intensive care unit admission: incidence, tween a chloride-liberal vs chloride- sponse and reduces mortality in acute prognosis and administration of buffer restrictive intravenous fluid administra- metabolic acidosis in pigs. PLoS One therapy: a prospective, multiple-center tion strategy and kidney injury in criti- 2013;8(1): e53932. study. Crit Care 2011;15:R238. cally ill adults. JAMA 2012;308:1566-72. 46. Swietach P, Vaughan-Jones RD, Harris 18. Iberti TJ, Leibowitz AB, Papadakos PJ, 33. Myburgh JA, Mythen MG. Resuscita- AL, Hulikova A. The chemistry, physiolo- Fischer EP. Low sensitivity of the anion tion fluids. N Engl J Med 2013;369:1243- gy and pathology of pH in cancer. Philos gap as a screen to detect hyperlactatemia 51. Trans R Soc Lond B Biol Sci 2014;369: in critically ill patients. Crit Care Med 34. Shaw AD, Bagshaw SM, Goldstein SL, 20130099. 1990;18: 275-7. et al. Major complications, mortality, and 47. Okorie ON, Dellinger P. Lactate: bio- 19. Kraut JA, Madias NE. Serum anion resource utilization after open abdominal marker and potential therapeutic target. gap: its uses and limitations in clinical surgery: 0.9% saline compared to Plasma- Crit Care Clin 2011;27:299-326. medicine. Clin J Am Soc Nephrol 2007;2: Lyte. Ann Surg 2012;255:821-9. 48. Yealy DM, Kellum JA, Huang DT, et al. 162-74. 35. Kuze S, Naruse T, Yamazaki M, Hiro- A randomized trial of protocol-based care 20. Madias NE, Homer SM, Johns CA, Co- ta K, Ito Y, Miyahara T. Effects of sodium for early septic shock. N Engl J Med 2014; hen JJ. Hypochloremia as a consequence L-lactate and sodium racemic lactate on 370:1683-93. of anion gap metabolic acidosis. J Lab intraoperative acid-base status. Anesth 49. Jansen TC, van Bommel J, Schoonder- Clin Med 1984;104:15-23. Analg 1992;75:702-7. beek FJ, et al. Early lactate-guided therapy 21. Khosravani H, Shahpori R, Stelfox 36. Didwania A, Miller J, Kassel D, Jack- in intensive care unit patients: a multi- HT, Kirkpatrick AW, Laupland KB. Occur- son EV Jr, Chernow B. Effect of intrave- center, open-label, randomized controlled rence and adverse effect on outcome of nous lactated Ringer’s solution infusion trial. Am J Respir Crit Care Med 2010;182: hyperlactatemia in the critically ill. Crit on the circulating lactate concentration: 752-61. Care 2009;13: R90. Part 3. Results of a prospective, random- 50. Jones AE. Point: should lactate clear- 22. del Portal DA, Shofer F, Mikkelsen ized, double-blind, placebo-controlled ance be substituted for central venous ME, et al. Emergency department lactate trial. Crit Care Med 1997;25:1851-4. oxygen saturation as goals of early severe is associated with mortality in older 37. Sakr Y, Dubois MJ, De Backer D, Cre- sepsis and septic shock therapy? Yes. adults admitted with and without infec- teur J, Vincent JL. Persistent microcircula- Chest 2011;140:1406-8. tions. Acad Emerg Med 2010;17:260-8. tory alterations are associated with organ 51. Fuller BM, Dellinger RP. Lactate as a 23. Bakker J, Jansen TC. Don’t take vitals, failure and death in patients with septic hemodynamic marker in the critically ill. take a lactate. Intensive Care Med 2007; shock. Crit Care Med 2004;32:1825-31. Curr Opin Crit Care 2012;18:267-72. 33:1863-5. 38. De Backer D, Creteur J, Dubois MJ, et 52. James JH, Luchette FA, McCarter FD, 24. Fall PJ, Szerlip HM. Lactic acidosis: al. The effects of dobutamine on micro- Fischer JE. Lactate is an unreliable indica- from sour milk to septic shock. J Intensive circulatory alterations in patients with tor of tissue hypoxia in injury or sepsis. Care Med 2005;20:255-71. septic shock are independent of its sys- Lancet 1999;354: 505-8. 25. Schelling JR, Howard RL, Winter SD, temic effects. Crit Care Med 2006;34:403- 53. Textoris J, Beaufils N, Quintana G, et Linas SL. Increased osmolal gap in alco- 8. al. Hypoxia-inducible factor (HIF1α) gene holic ketoacidosis and lactic acidosis. Ann 39. De Backer D, Donadello K, Sakr Y, et expression in human shock states. Crit Intern Med 1990;113:580-2. al. Microcirculatory alterations in pa- Care 2012;16:R120. 26. Vincent JL, De Backer D. Circulatory tients with severe sepsis: impact of time 54. van Beest PA, Brander L, Jansen SP, shock. N Engl J Med 2013;369:1726-34. of assessment and relationship with out- Rommes JH, Kuiper MA, Spronk PE. Cu- 27. Rivers E, Nguyen B, Havstad S, et al. come. Crit Care Med 2013;41:791-9. mulative lactate and hospital mortality in Early goal-directed therapy in the treat- 40. Adrogué HJ, Madias NE. Management ICU patients. Ann Intensive Care 2013;3: ment of severe sepsis and septic shock. of life-threatening acid-base disorders: 6. N Engl J Med 2001;345:1368-77. first of two parts. N Engl J Med 1998;338: 55. Vanni S, Viviani G, Baioni M, et al. 28. Severs D, Hoorn EJ, Rookmaaker MB. 26-34. Prognostic value of plasma lactate levels A critical appraisal of intravenous fluids: 41. Kraut JA, Kurtz I. Use of base in the among patients with acute pulmonary from the physiological basis to clinical treatment of severe acidemic states. Am J embolism: the thrombo-embolism lactate evidence. Nephrol Dial Transplant 2014 Kidney Dis 2001;38:703-27. outcome study. Ann Emerg Med 2013;61: January 23 (Epub ahead of print). 42. Kim HJ, Son YK, An WS. Effect of so- 330-8. 29. O’Dell E, Tibby SM, Durward A, Mur- dium bicarbonate administration on mor- 56. Andersen LW, Mackenhauer J, Roberts doch IA. Hyperchloremia is the dominant tality in patients with lactic acidosis: a JC, Berg KM, Cocchi MN, Donnino MW. cause of metabolic acidosis in the retrospective analysis. PLoS One 2013; Etiology and therapeutic approach to ele- postresuscitation phase of pediatric me- 8(6):e65283. vated lactate levels. Mayo Clin Proc 2013; ningococcal sepsis. Crit Care Med 2007; 43. Hood VL, Tannen RL. Protection of 88:1127-40. 35:2390-4. acid–base balance by pH regulation of 57. Kliegel A, Losert H, Sterz F, et al. Se- 30. Cooper DJ, Walley KR, Wiggs BR, acid production. N Engl J Med 1998;339: rial lactate determinations for prediction Russell JA. Bicarbonate does not improve 819-26. of outcome after cardiac arrest. Medicine hemodynamics in critically ill patients 44. Wu DM, Qi JS, Dai H, Doods H, Abra- (Baltimore) 2004;83:274-9. who have lactic acidosis: a prospective, ham WM. Resuscitation with Na+/H+ 58. Jones AE, Shapiro NI, Trzeciak S, Ar- controlled clinical study. Ann Intern Med exchanger inhibitor in traumatic haem- nold RC, Claremont HA, Kline JA. Lactate 1990;112:492-8. orrhagic shock: cardiopulmonary per- clearance vs central venous oxygen satu- 31. Gennari FJ. Intravenous fluid therapy: formance, oxygen transport and tissue ration as goals of early sepsis therapy: a saline versus mixed electrolyte and or- inflammation. Clin Exp Pharmacol Physi- randomized clinical trial. JAMA 2010;303: ganic anion solutions. Am J Kidney Dis ol 2010;37:337-42. 739-46. 2013;62:20-2. 45. Wu DM, Kraut JA, Abraham WM. 59. Murtuza B, Wall D, Reinhardt Z, et al.

2318 n engl j med 371;24 nejm.org December 11, 2014 The New England Journal of Medicine Downloaded from nejm.org on December 11, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved. Lactic Acidosis

The importance of blood lactate clearance genation in the clinical setting. Transfus between arterial and central venous as a predictor of early mortality following Apher Sci 2010;43: 79-94. blood. N Engl J Med 1989;320:1312-6. the modified Norwood procedure. Eur J 61. Adrogué HJ, Rashad MN, Gorin AB, Copyright © 2014 Massachusetts Medical Society. Cardiothorac Surg 2011;40: 1207-14. Yacoub J, Madias NE. Assessing acid–base 60. Sakr Y. Techniques to assess tissue oxy- status in circulatory failure: differences

n engl j med 371;24 nejm.org December 11, 2014 2319 The New England Journal of Medicine Downloaded from nejm.org on December 11, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved.