Here is a recent free full text review.
It points out the following:
Lactate is a semiquantitative indicator of illness severity and risk of mortality. Its elevation indicates need for immediate resuscitative efforts. Decline in the lactate level during resuscitative efforts is a good sign. Lactate elevation can reflect global tissue ischemia. However, in a variety of critical illnesses, even septic shock, lactate is not a reliable indicator of tissue perfusion. This is due to multiple mechanisms, including non ischemic mechanisms, of excess lactate generation. Intense beta receptor stimulation due to high catacholamine levels, for example, increases intracellular cyclic AMP. This results in downstream metabolic effects that drive lactate generation including glycogenolysis (which increases glucose delivery into the glycolytic pathway thus generating lactate) and stimulation of the sodium potassium ATPase which also drives glycolysis. These metabolic (non ischemic) components of lactate generation may not as directly responsive to fluid resuscitation. Thus, using lactate normalization as an endpoint for volume administration may lead to over administration of fluid.