Evidence for monitoring lactate in critical illness

In July 2015, I wrote that though this topic has never been examined before, "it would be worthwhile to apparoach it in a "critically evaluate lactate monitoring" sort of way". Then, in Question 8 from the first paper of 2016 the college asked the candidates to "critically evaluate the role of monitoring blood lactate levels in the critically ill". It was easy to guess that this was going to come up at some stage. Probably the best published resource for a quick overview of this topic is this review article by Bakker et al (2013).

Rationale for lactate monitoring in critical illness

• Lactate is a product of anaerobic metabolism
• Anaerobic metabolism in human tissues is an abnormal physiological state, be it of poor tissue perfusion or impaired oxygen utilisation.
• Lactate clearance is rapid in the presence of normal hepatic function
• Serum lactate may therefore be useful as a biomarker, a measure of anaerobic metabolism.
• Conditions which result in a raised lactate may not be easy to identify in critically ill patients (eg. ischaemic gut, ischaemic limbs, sepsis, etc) and so lactate levels may be the only feature of such conditions

Advantages of lactate monitoring in critical illness

• A simple test, widely available, and can be performed as a part of blood gas analysis
• Venous and arterial lactate levels correlate conveniently
• Lactate is cheap to test - there may be a saving in healthcare costs
• Its clearance by kidneys and dialysis is proportionally poorer than hepatic clearance, which means as a biomarker it is neither obscured by dialysis nor falsely elevated in the presence of renal failure.
• It offers a quantitative marker to evaluate the success of resuscitation
• It is thus an effective surrogate endpoint in the evaluation of therapies aimed to improve tissue perfusion

Errors of lactate measurement

• Storage delay: metabolism by blood cells can cause an elevated lactate
• Collection error: A sample accidentally mixed with compound sodium lactate will cause a spuriously elevated lactate level.
• Measurement error: a "lactate gap" is generated when the lactate-sensing electrode in the b;lood gas machine confuses lactate with ethylene glycol

Errors of lactate interpretation

• Tissue perfusion may be normal, and lactate clearance may be impaired (as in liver disease)
• Lactate may be raised due to poor tissue oxygen utilisation, rather than poor perfusion (eg. in septic shock)
• Lactate may be raised in the absence of acidosis (eg. exogenous compound sodium lactate administration).
• Lactate may be raised because of increased production by abnormal metabolic processes in neoplastic tissue

Evidence for and against the use of lactate as a biomarker

In the college answer to Question 8 from the first paper of 2016, various statements were made regarding the evidence for lactate monitoring. Below, I have tried to track down the references from where those statements came from.

Lactate as a prognostic marker:

• It correlates well with mortality in sepsis: patients with a presenting lactate over 4.0mmol/L were almost five times more  likely to die (Mikkelsen et al, 2009)
• High levels may reveal underlying "occult" sepsis in otherwise stable patients (Rivers et al, 2001).
• It is associated with a poor prognosis in trauma. Specifically, failure to normalise the lactate over the first 2 days is associated with worse organ failure and increased mortality (Manikis et al, 1995)
• It predicts poor recovery from cardiac arrest. Survivors and patients with good neurological outcome had lower lactate levels at 0, 12 and 24 hours (Donnino et al, 2015 is the 100-patient study mentioned by the college)
• It acts as a trigger for liver transplatation: lactate predicts non-survivors from paracetamol toxicity earlier than the King's College Criteria (Bernal et al, 2002). However, somehow the effect of adding lactate to the King's College Criteria actually decreases their diagnostic odds ratio slightly, from about 27 to about 26. (Craig et al, 2010)

Lactate as a guide to therapy:

• It can be used to guide therapy in cardiac surgical patients (though neither Pölönen et al in 2000 nor Shrestha et al in 2015 were able to demonstrate a statistcially significant mortality benefit, owing to underpowered study design)
• It can be used to guide therapy in sepsis. There are several possible ways of using it in this context:
  • As a biomarker which reports the effectiveness of your resuscitation strategy, it is non-inferior to mixed venous saturation as a goal of early sepsis therapy (Jones et al, 2010).
  • To target the use of GTN when ScvO₂ normalised but lactate remained high: The literature on the subject is somewhat mixed. As an example, in 2010 Janssen et al used GTN to optimise the microcirculation along with several other EGDT-era manoeuvres (eg. the use of CVP to guide fluid responsiveness). GTN infusion was commenced if the lactate levels did not fall with conventional therapy even after the ScvO₂ had returned to normal. Eight hours of such lactate-guided therapy produced a mortality reduction from 43.5% to 33.9%, which is weird because there was no difference in the lactate levels between the groups.  Ther practice is also mentioned in the review by Bakker et al (2013), where an (unpublished?) study by Lima et al (2012) is discussed, using GTN to normalise lactate . Bakker (a co-author of Lima et al, 2012) was critical of this technique in his later review article, saying that the evidence did not support this use of lactate monitoring (or GTN infusion).