Causes of acidosis in hyperlactataemia

Though the term "lactic acidosis" is imbibed together with mother's milk in one's medical training, the very concept of lactate causing acidosis is mired in controversy. The question of whether or not lactate actually causes acidosis is itself debated; highly respected members of the intensive care community disagree on this rather important point. Less respected members are even more conflicted. Capture a senior intensivist, and corner them with the question: can infusion of Hartmanns cause lactic acidosis in an anhepatic patient? Why or why not? How can there be severe hyperlactataemia with a normal base excess? What is the relationship between lactate level and pH? One may be disappointed with the results of this experiment. The unprepared consultant may stagger, babble, or confidently produce something astonishingly stupid in response.

From this point, it follows that because the author of these lazy notes is not in fact a biochemistry expert, it would be highly inappropriate for him to vandalise the topic with his own opinions. Ergo, opinions of others will be recounted here, to illuminate this debate rather than bring it to conclusion. 

A high lactate concentration may not be the main cause of the acidaemia

When there is metabolic acidosis with raised lactate, the lactate may not be the main cause of the acidosis, and some might say it is a causally unrelated bystander. In his EmCrit article on lactate myths, Paul Marik's position is a refutation of the classic "lactate = acid" explanation which is taken for granted by otherwise quite reasonable people. His argument is that another biochemical mechanism explains metabolic acidosis associated with raised lactate.  In the conversion of pyruvate into lactate, a hydrogen ion is consumedand this should in fact alkalinise the intracellular fluid. "The presence of an acidosis in many (not all) patients with an increased lactate is ... an association and not causal", offers Marik. The cause of the metabolic acidosis in states usually related with raised lactate is actually the increased rate of ATP turnover which lowers pH.

Marik referred to Robergs et al (2004) in support of this, which is a study of exercising muscle. The authors performed a thorough review of lactate biochemistry, and ultimately concluded that lactate is an innocent bystander and that the hydrolysis of ATP is responsible for the acidosis. This has a rather solid foundation in biochemistry. If we are going to talk in terms of protons, then it has long been known that lactate synthesis in anaerobic glycolysis is not a source of them. The excellent article by  Zilva (1978) was pillaged and his molecular representation of ATP hydrolysis reproduced here with no permission whatsoever.  In short, there is no net H⁺ production in the anaerobic glycolysis equation (the products are lactate ATP and water), but every time ATP is hydrolysed to ADP a proton is produced. Ergo, in anaerobic metabolism, the total cost of energy production is in fact acidosis, but the production of lactic acid per se is not the cause of it.

The inorganic phosphate removed from ATP  is generally expected to buffer the "free proton" generated by this process, and it is then incorporated into aerobic metabolism to roduce more ATP. During normal metabolism, something like 150 moles of hydrogen ions are recycled in this manner over 24 hours (Kreb, 1975). However, as the phosphate ion is rapidly recycled to produce more ATP (particularly in "stressed" metabolically active tissues) this buffering and recycling does not occur. The net production of ATP (i.e. consumption of H⁺)  lags behind the anaerobic production of H⁺. Unbuffered intracellular protons leave the cell via the sarcolemmal Na⁺/H⁺ exchangers and change the pH of your blood gas sample. In summary, tissues which have a high rate of ATP turnover are going to generate an acidosis. Lactate production occurs by processes which are usually associated with increased ATP turnover, and may therefore be proportional to the acidosis, but it is not causally linked to it - so Robergs et al argue.  

Lactate can increase in the absence of increased ATP hydrolysis

It is pretty difficult to refute the point made by Robergs et al. If increased ATP hydrolysis and increased lactate production are caused by the same process, then metabolic acidosis and hyperlactataemia will occur together, at least some of the time. Logically, this position supports situations where there is raised lactate and no acidosis (because there is no increased ATP hydrolysis). Theoretically there should be disease states with increased ATP hydrolysis but no hyperlactatemia.

However,  states which are known to cause severe metabolic acidosis and hyperlactataemia aren't always associated with any sort of change in ATP hydrolysis. In fact there is good data that in severe sepsis ATP hydrolysis does not seem to increase. May's team (2012) could not demonstrate any major change of the ATP:Pi ratio in their septic sheep using MRI. The sheep were injected with E.coli and became quite sick, with MAP declining by 40mmHg (from the 90s down to the 50s), but unfortunately the authors did not measure lactate or pH during this period. Fortunately quiet a few other authors did. There is a significant amount of literature where investigators consistently fail to find an association between lactate, acidosis and bioenergetic failure. Choosing randomly from a massive pile of search results, one can identify highly cited articles such as the one by Hotchkiss and Karl (1992). Tons of septic rat data is presented where the rise in lactate was not associated with any cellular metabolic evidence of tissue bioenergetic failure. This old article pre-dates more modern data which suggests that hyperlactataemia in septic shock may be more related to the inhibitory effects of cytokines and endotoxin on pyruvate dehydrogenase activity (Crouser,  2004). 

Increased lactate can occur in the absence of significant acidosis

From the hypothesis that lactate = acidosis, it follows that an increase in lactate should be associated with a decrease in pH and base excess. In fact there should be a stoichiometrically predictable relationship between these variables. People who argue that lactate and acidosis are not causally related usually point to the fact that this relationship is not usually demonstrated by in vivo experiments. For instance, in his exchange with Weingart, Marik refers to  Lewis et al (2014) who raised the lactate of their asthmatic patients from  2.05 to 2.94 mmol/L using salbutamol with little effect on pH or bicarbonate.  Davenport et al (1991) dialysed their patients with lactate-buffered solutions, raising their serum lactate by 3-4mmol/L, also with little effect on pH and bicarbonate. Many similar examples exist. Overall, authors do not tend to find a consistent relationship between the lactate level and the severity of acidosis in mildly and moderately acidotic humans.

Biochemically, lactate molecules should always produce acidosis 

From the viewpoint of a quantitative approach to metabolic acidosis, lactate in solution should always produce a decrease in pH. Lactate is a strong anion with a pKa of 4.0, fully dissociated under normal physiologic conditions. By decreasing the strong ion difference, increasing lactate concentration should push the PCO2/pH buffer relationship in the direction of metabolic acidosis. This should happen no matter what the origin of the lactate. According to this theoretical framework, when a high lactate level is seen in the absence of acidaemia, a concurrent process must also be present which protects the pH (for example, a pre-existing metabolic alkalosis).

It is arguable whether this quantitative approach is any better than any other approach. Moreover, the demands of routinely using it tend to exceed the capacity of mortal humans. Arguments about this take place in the rarified stratosphere of academic intensive care, where gods hurl lightning bolts at each other across mountaintops. For example,  Bellomo Marik and Kellum commented on the NEJM lactate article by Kraut and Madias (2014) suggesting that a quantitative interpretation is a more satisfactory explanation for these processes than the classical model. In their answer, Kraut and Madias quoted themselves in  support of the classical model of proton-based acid base explanations, because why would you use Stewarts approach for anything (because didn't we already explain how useless it was?). 

The support for the assertion that lactate should always cause acidosis comes from in vitro studies such as this one by TJ Morgan and Hall (1999). Morgan's own donated blood was titrated with varying concentrations of lactate. The resulting relationship of base excess and lactate concentration was linear (the graph stolen from that article is presented here for reference and out of reverence). In essence, the authors were able to demonstrate that in the absence of normal mechanisms of compensation, increasing plasma lactate causes acidosis (and that this increase in acidosis is the same for any other strong anion). 

The other striking finding was the amount of lactate required to create a significant change. Note that on this graph, the base excess did not fall below the normal laboratory range (-3) until the lactate level was almost 10mmol/L. This may help explain the in vivo findings, where investigators present patients with moderately raised lactate levels but minimal change in acid-base balance. Remember the Lewis study  study quoted by Marik. The investigators had asthmatic patients whose lactate increased by 0.9 mmol/L, which - going by Morgan and Hall's data  - would have produced a base excess change of less than 0.3 mEq/L.

In summary

• The metabolic acidosis which is seen in association with high lactate levels may be the consequence of increased ATP hydrolysis, but this does not explain situations where ATP hydrolysis is normal and a severe acidosis coexists with hyperlactataemia.
• Lactate in solution should produce acidaemia, but relatively large lactate concentrations are required for this. Furthermore the Stewart model which is used for this explanation is not widely supported.
• Relatively large increases in lactate can occur in patients without significant acidosis because of effective buffering and pre-existing alkalosis