Understanding Hyperlactatemia in Human Sepsis: Are We Making Progress?

To the Editor:

We have significant concerns about the interpretation of the data presented by Gattinoni and colleagues (1). The reminders that lactic acidosis commonly coexists with renal acidosis and that metabolic acidosis does not necessarily mean acidemia are welcome. Indeed, one should dissociate hyperlactatemia from acidosis because hyperlactatemia can be of hypoxic origin even in the absence of acidosis, and of nonhypoxic origin even when there is acidemia (2). Although we agree with Gattinoni and colleagues that a pH measurement can be misleading, and that lactate concentrations should be measured directly, neither the presence or absence of metabolic acidosis nor the central venous oxygen saturation (Scv_O2) value can help identify the origin of hyperlactatemia.

Gattinoni and colleagues also reemphasize the well-known fact that hyperlactatemia can coexist with any value of $\stackrel{.}{\mathrm{V}}$o₂/oxygen delivery ($\stackrel{.}{\mathrm{V}}$o₂/Do₂) (or S[c]v_O2). This is in part related to timing, because an increase in Do₂ as a result of resuscitative efforts may result in a rapid increase in Sv_O2 but a much slower decrease in blood lactate levels. More importantly, a normal or high Scv_O2 does not necessarily indicate that tissue perfusion is adequate. It is well known that a high Sv_O2 can be a sign of disease severity and worse prognosis (3). However, a high Sv_O2 does not always imply a significant alteration in cellular metabolism, as high Sv_O2 values can be the result of microcirculatory alterations. In our early study demonstrating the occurrence of microvascular alterations in sepsis (4), Sv_O2 values were identical in patients with sepsis and in other ICU patients, but hyperlactatemia was observed only in patients with septic shock (4). Accordingly, it may be erroneous and even potentially harmful to limit resuscitation efforts in a patient with hyperlactatemia just because his or her Scv_O2 values are normal or high. When associated with signs of tissue hypoperfusion, an elevated Sv_O2 (or Scv_O2) does not mean that resuscitation efforts are no longer necessary. We previously reported that elevated Sv_O2 values did not exclude fluid responsiveness in patients with sepsis and signs of tissue hypoperfusion (5). Similarly, Monnet and colleagues showed that blood lactate and venoarterial Pco₂ differences, but not Scv_O2, predicted an increase in $\stackrel{.}{\mathrm{V}}$o₂ in fluid-responsive patients (6).

Hence, we do not think that the observations by Gattinoni and colleagues should influence the way in which patients with sepsis are managed. Hyperlactatemia associated with other signs of tissue hypoperfusion should encourage attempts to increase Do₂ with fluids, transfusions, and/or dobutamine administration, even in the absence of acidemia or when ScvO₂ is not reduced.