At the outset, the anesthesiologist and the surgeon should agree on the objectives with respect to PaCO2 . Induction of hypocapnia is a time-honored part of the management of intracranial neurosurgical procedures. The rationale is principally that the concomitant reduction in CBF (see Fig. 21-4 ) and CBV will result in a reduction in ICP, or "brain relaxation." The rationale is valid. However, two considerations should influence the clinician's use of hyperventilation. First, the vasoconstrictive effect of hypocapnia has the potential to cause ischemia in certain situations. Second, the CBF-lowering effect is not sustained.
|Drainage of cerebrospinal fluid (ventriculostomy, brain needle)|
|Diuresis (usually mannitol)|
|Suppression of the cerebral metabolic rate (usually barbiturates)|
|Reduction in mean arterial pressure (if dysautoregulation is present) (Lobectomy)|
At first, clinicians were skeptical that hyperventilation could actually result in ischemia, and it does in fact appear that normal brain is unlikely to be damaged by the typical clinical use of hyperventilation. However, such may not be the case in certain pathologic conditions.
The available data        indicate that in normal subjects, ischemia will not occur at a PaCO2 over 20 mm Hg. This generalization appears to also apply during induced hypotension.    However, physiologic alterations, as evidenced by both metabolic and electroencephalographic (EEG) abnormalities, have been observed in human volunteers   and in normal animals   at severe hypocapnia (PaCO2 <15 mm Hg) and in dogs subjected to the combination of extreme hypocapnia (PaCO2 of 10 mm Hg) and severe anemia (hemoglobin content of 5 g/dL). In one of these studies, EEG abnormalities and paresthesias occurred in volunteers hyperventilating to PaCO2 values less than 20 mm Hg, and these effects were reversed by hyperbaric oxygenation, thus suggesting that they may truly have been caused by ischemia. In two separate investigations in cats at PaCO2 levels of 10 to 12 mm Hg,  modest reductions in brain phosphocreatine levels with increased brain lactate but normal adenosine triphosphate levels were observed. It has been suggested that the changes observed may in part reflect pH-related alterations in enzyme function (specifically, an increase in the activity of phosphofructokinase causing increased lactate formation) rather than ischemia. Accordingly, given that a PaCO2 of less than 20 to 25 mm Hg offers very little additional benefit in terms of improvement in intracranial compliance, it seems prudent to limit acute PaCO2 reduction to 25 mm Hg in previously normocapnic individuals. Normal brain will not be injured by this degree of hypocapnia.
Although preventing herniation, maintaining ICP under 20 mm Hg, minimizing retractor pressure, and facilitating surgical access remain priorities that may justify hypocapnia, evidence is also accumulating that hyperventilation is potentially deleterious              and should not be overused. In the setting of head injury, there is evidence that hyperventilation can result in ischemia,    especially when baseline CBF is low, as is commonly the case in the first 24 hours after injury.      An increased frequency of brain regions with very low CBF has been demonstrated in head-injured patients who were acutely hyperventilated.  In addition, from centers that monitor jugular venous oxygen saturation (SjvO2 ), there have been numerous observations that low SjvO2 values can be increased and that lactate levels in the jugular venous effluent can be decreased by reducing the degree of hyperventilation,    although the inability of SjvO2 monitoring to detect ischemia consistently has also been reported.  However, at present, there is little information to confirm a deleterious effect of hyperventilation. The closest thing to "proof" resides in a study of patients with moderate head injuries by Muizelaar and colleagues.  These authors divided patients into a near-normocapnic group in which PaCO2 was maintained at approximately 35 mm Hg, a hypocapnic group in which PaCO2 was maintained in the vicinity of 25 mm Hg, and a third group in which carbon dioxide tension was maintained at 25 mm Hg and the buffer tromethamine was administered. Tromethamine is a buffer that can cross the blood-brain barrier, and it has been theorized that tromethamine can attenuate the adverse effect of the reduction in bicarbonate levels in CSF and brain extracellular fluid that occurs with chronic hyperventilation. They examined outcomes 3 and 6 months after injury and observed a poorer status in a post hoc subpopulation of the hyperventilation group. That subpopulation included patients with the best initial motor scores, specifically, a subgroup in which the severity of injury was such that they merited intubation by conventional criteria but whose clinical condition may have been such that hyperventilation was not necessarily required for control of ICP and who therefore had little to gain from hyperventilation.
Accordingly, hyperventilation should not be an automatic component of every "neuroanesthetic." It should be treated like any other therapeutic intervention. There should be an indication for instituting it (usually elevated or uncertain ICP or the need to improve conditions in the surgical field, or both). Hyperventilation should be used with the knowledge that it has the potential for causing an adverse effect, and as is the case with any other therapeutic intervention, it should be withdrawn as the indication for it subsides. The concern regarding the hazards of hypocapnia, which evolved in the context of head injury, has influenced all of neurosurgery. In particular, it is now widely avoided in the management of SAH because of the postictal low-CBF state that is known to occur.  In addition, brain tissue beneath retractors can have a similarly reduced CBF. 
The effect of hypocapnia on CBF is not sustained. Figure 53-6 is a nonquantitative representation of changes in CBF and CSF pH occurring in association with a sustained period of hyperventilation. With the onset of hyperventilation, the pH of both CSF and the brain's extracellular fluid space increases, and CBF decreases abruptly. However, the cerebral alkalosis is not sustained. By alterations in function of the enzyme carbonic anhydrase, the concentration of bicarbonate in CSF and the brain's extracellular fluid space is reduced, and in a time course of 6 to 18 hours, the pH of these compartments returns to normal. Pari passu, CBF returns toward normal levels.  The implications are twofold. First, the clinician should ideally hyperventilate patients for only as long as a reduction in brain volume is required. Prolonged, but unnecessary hyperventilation may lead to a circumstance wherein subsequent clinical events call for additional maneuvers to reduce the volume of the intracranial contents. However, if carbon dioxide tension is already in the 23- to 25-mm Hg range, it would be difficult to impose sufficient additional hyperventilation to once again accomplish the original reduction in CBF without the hazard of pulmonary barotrauma. Second, in a patient who has been hyperventilated for a sustained period (e.g., 2 days in an ICU setting), rapid restoration of carbon dioxide tension from values in the vicinity of 25 to typical normal values (e.g., 40 mm Hg) should ideally be accomplished slowly. A sudden increase in carbon dioxide tension from 25 to 40 mm Hg in an individual who has been
Figure 53-6 Changes in PaCO2 , cerebral blood flow (CBF), and cerebrospinal fluid (CSF) pH with prolonged hyperventilation. Whereas the decreased arterial PaCO2 (and the systemic alkalosis) persist for the duration of the period of hyperventilation, the pH of the brain and CBF return toward normal over a period of 8 to 12 hours.
If hypocapnia has been required as an adjunct to brain relaxation during craniotomy, PaCO2 should also be allowed to rise once the retractors are removed (if dural closure requirements permit) to minimize the residual intracranial pneumatocele (see the later section "Pneumocephalus").