Lactic acidosis: high-dose or low-dose bicarbonate therapy?

occurs intracellularly, extracellularly, and in the cerebrospinal fluid (CSF). Adding sodium bicarbonate to the extracellular fluid causes a decrease in hydrogen ions with subsequent increase in carbon dioxide (CO2) concentration. Since CO2) but not bicarbonate, readily diffuses across cell membranes and into the CSF, this leads to an increase in intracellular and CSF CO2 concentration, with consequent rise in hydrogen ion concentration. Thus sodium bicarbonate, though causing an increase in pH extracellularly, also results in a decrease in pH intracellularly and in the CSF. Support for this hypothesis comes from studies on skeletal muscle," heart muscle, liver, 7|12 erythrocytes, and CSF." While acid-base balance is normally restored via increased respiratory CO2 excretion, 17 intracellular and CSF acidosis may persist if the rate of bicarbonate administration exceeds the rate of CO2 excretion. In this context it is important to note that sodium bicarbonate therapy in metabolic acidosis has been associated with the onset of stupor,' and that severe lactic acidosis may itself culminate in respiratory failure. Among the major arguments put forward in favor of the use of high-dose sodium bicarbonate therapy have been that metabolic acidosis in the perfused rat liver severely impairs hepatic lactate metabolism,' and that metabolic acidosis can result in a negative inotropic effect.' However, both these effects are thought to be mediated by intracellular acidosis'' and indeed hepatic intracellular acidosis is postulated as a fundamental factor in the pathogenesis of some types of lactic acidosis. Thus if, as postulated, high-dose sodium bicarbonate causes intracellular acidosis, it would be expected to worsen rather than improve hepatic lactate metabolism and cardiac output in lactic acidosis. Animal experiments are in support of this view. When given to animals with lactic acidosis, sodium bicarbonate causes increasing lactic acidemia, a fall in intracellular pH, cardiovascular collapse, and sometimes increased mortality.' Infused sodium bicarbonate has been used in the induction and maintenance of lactic acidosis in the diabetic rabbit. Infusion of bicarbonate in the ketoacidotic rat produces signs of tissue anoxia and a threeto fourfold rise in blood and liver lactate concentrations. Conversely, when the experiments on the effect of metabolic acidosis on the perfused rat liver were recently repeated under conditions of simulated partial respiratory compensation of metabolic acidosis no effect of acidosis on hepatic lactate metabolism was found. It is considered that the low PCO2 of the perfusate in these experiments drew readily diffusable CO2 out of the hepatic cells, shifting the intracellular bicarbonate equilibrium to the left to compensate for the acidosis and allow hepatic lactate metabolism to continue normally. Controlled studies in human patients, similar to those performed in animals, are clearly not practical. In practice retrospective analysis of the case report literature is the main source of patient information. High-dose sodium bicarbonate therapy has previously appeared to receive support from a retrospective analysis of biguanide-induced lactic acidosis case reports suggesting that patients whose arterial pH was deliberately restored to normal within 2-6 h, and kept there, fared considerably better than those in whom partial or no correction was achieved. However, in this study, the relationship between survival and arterial pH was clearer than a relationship between arterial pH and bicarbonate therapy or survival and bicarbonate therapy. Obviously the pH would return to normal in survivors as the authors have later accepted, making it difficult to come to a satisfactory conclusion. A study from France of 18 patients found that comparable amounts of sodium bicarbonate were given to those who died and those who survived but both groups were given