Previous Next

Seizure Surgery/Awake Craniotomy

Awake craniotomies are performed when tumors or epileptic foci lie close to the cortical areas required for either speech or motor function or close to the mesial temporal structures critical to short-term memory. Most patients will have so-called temporal lobe epilepsy. A structural lesion is commonly visible on magnetic resonance imaging (MRI). Sometimes there is a history of trauma. More commonly, the cause is presumed to be an asphyxial birth injury.

Presurgical Evaluation

Before resection, most patients will have undergone either or both a Wada test and videotelemetry. More recently, functional testing using MRI or positron emission tomography or both has also been introduced to the presurgical evaluation. The Wada test involves selectively anesthetizing the cerebral hemispheres by injection of amobarbital (Amytal) into the carotid artery to localize the hemisphere that controls speech or confirm that there is bilateral representation for short-term memory (or both). Speech is an issue when the lateral portions of the temporal lobe are involved, and memory is the concern when the involvement is medial.

Anesthesia for EEG Electrode Placement

Videotelemetry is performed to permit localization of the seizure focus that is responsible for the clinically problematic events. Previous placement of either subdural strip electrodes (through bur holes) or a subdural electrode grid (requiring a craniotomy) is generally required. Occasionally, electrodes are placed deep into parenchyma, usually within the temporal lobe (placed stereotactically through bur holes), or they are positioned so that they "look at" the inferior surfaces of the temporal lobe. The latter is commonly accomplished with so-called foramen ovale electrodes. These electrodes are placed by using a needle similar to an epidural needle. The point of entry is about 2 cm lateral to the angle of the mouth. The needle is passed through soft tissue, under the temporal process of the zygomatic bone and medial to the ramus of the mandible, up to the base of the skull in the vicinity of the


2160
foramen ovale. Typically, this procedure is performed as a "MAC," although small doses of induction anesthetic, most often propofol, are usually required at the time of stimulation of the periosteum at the base of the skull by the needle. After placement of the relevant electrodes, the patient's seizure medication is discontinued and the patient remains in an observation unit with EEG and patient behavior recorded continuously. In this manner, the EEG events associated with the clinically significant seizure events and their anatomic origin can be identified.

Preanesthetic Evaluation/Preparation

At the preoperative interview, the patient should be educated about the nature and duration of the procedure and the limitations on movement. One should obtain a description of both the aura and the seizures to facilitate recognition of them. One should ascertain whether the patient is subject to grand mal convulsions. If intraoperative electrocorticography to identify seizure foci is intended, it is common to discontinue or reduce the anticonvulsant dose by half according to the perceived risk of uncontrolled seizures. Premedicants with an anticonvulsant effect, such as the benzodiazepines, should not be used because they may interfere with intraoperative EEG localization.

Anesthetic Technique

The objectives of the anesthetic technique are to

  1. Minimize patient discomfort associated with the potentially painful portions of the procedure and with the prolonged restriction of movement
  2. Ensure patient responsiveness and compliance during the phases of the procedure that require assessment of either speech or motor/sensory responses to cortical stimulation
  3. Select anesthetic techniques that produce minimal inhibition of spontaneous seizure activity

There are probably many ways of providing sedation that are consistent with the aforementioned objectives, and many techniques are in active use. They range from minimal sedation approaches, through deep sedation during which intermittent unresponsiveness is achieved with spontaneous ventilation and an unprotected airway, to asleep-awake-asleep techniques with intermittent airway management with an LMA, sometimes with positive-pressure ventilation. However, at the outset, the clinician should appreciate two things. First, the essential element of an "anesthetic" for an awake craniotomy is the surgeon's local anesthetic technique. "Sedation" cannot compensate for inadequate anesthesia of the scalp, as accomplished with pin site infiltration and nerve blocks. Anesthesiologists must not get trapped into thinking that it is their responsibility to provide a general anesthetic equivalent in a spontaneously breathing patient with an unprotected and all but inaccessible airway. Second, most craniotomies for which an awake technique is relevant can be accomplished without the presence of anesthesia personnel. No anesthetist was present when Wilder Penfield made the observations during open-cranium procedures that defined the motor and sensory homunculi. During the anesthesiologist shortage of the nineties, awake craniotomies were performed as "straight locals" in at least one institution that we visited, again with no anesthesia personnel present. Accordingly, there can be little justification for morbidity arising from the routine administration of sedatives and analgesics.

Several centers have used a droperidol/synthetic narcotic combination [317] (e.g., droperidol, 2.5 to 7.5 mg; alfentanil, 5- to 10-mg/kg load, 0.5- to 1.0-mg/kg/min infusion; fentanyl, 0.7-mg/kg load, 0.7-mg/kg/hr infusion).[318] Others use principally propofol by either physician- or patient-controlled infusion.[318] [319] [320] Care should be taken when administering additional sedative agents, especially narcotics, whose respiratory depressant effects might be synergistic with propofol. This consideration is especially relevant when pin fixation is used. Pin fixation severely restricts the anesthesiologist's capacity to intervene quickly in the event of excessive respiratory depression or loss of patency of the airway. Propofol, if used, should be discontinued at least 15 minutes before EEG recording. In spite of prompt awakening, propofol leaves a residual EEG "footprint" characterized by high-frequency, high-amplitude beta activity that can obscure the abnormal activity that is being sought in the cortical surface EEG.[319] Various groups have reported use of the LMA, commonly with narcotic-propofol sedation and spontaneous ventilation during the craniotomy; administration of the sedative is discontinued and the LMA is removed once the brain surface is exposed.[321] [322] [323] One of these groups reported that complications occurred least often when the asleep phases were accomplished with infusion of remifentanil and propofol and positive-pressure ventilation.[322] More recently, the α2 -agonist dexmedetomidine has been used by some, both with and without intermittent use of the LMA.[321] [323] [324] The authors of these series report satisfactory conditions for functional testing, including brain stimulation for speech mapping, sometimes with the dexmedetomidine infusion ongoing at low dose (0.1 to 0.3 mg/kg/hr) during the testing.[321] However, delay in achieving satisfactory patient responsiveness has been reported.[324] [325] The experience with administration of dexmedetomidine during electrocorticography is very limited, and that experience, obtained during light general anesthesia achieved with the combination of sufentanil and dexmedetomidine, revealed suppression of sharp wave and spike activity relative to preoperative recordings. [326] It is obvious that many approaches can be effective, but the clinician should note that to our knowledge, LMA techniques have not been reported in patients in pin fixation.

Routine, noninvasive monitors are almost always sufficient. Reliable capnography to provide breath-by-breath confirmation of airway patency and respiratory drive is an essential component of the technique if deep sedation is intended for any portion of the procedure. These procedures are often lengthy, and attention to the details of patient comfort (warming blankets, a sheepskin, room temperature) will improve patient tolerance.

The uncomfortable phases of the procedure are placement of the pin head holder (not all groups use a pin head holder) and the craniotomy. Many patients also find manipulation of the dura painful, in particular, traction. The actual manipulation of supratentorial brain parenchyma is painless. The volume of local anesthetic used to infiltrate pin


2161
sites and perform a field block of the scalp can be substantial. It is appropriate for the anesthesiologist, in particular, an anesthesiologist working with surgical trainees, to keep track of the doses of local anesthetics used.

The anesthesiologist should participate actively at the time of head positioning. The more "sniff" that can be achieved before the final lockdown of the head holder, the wider the latitudes will be for sedating the patient while maintaining spontaneous ventilation and a patent airway. During positioning of the patient, attention should also be paid to the need to maintain visual access to the face. A clear line of sight to the face will be necessary both to present the patient with images to name as part of speech testing and to identify the occurrence of facial motor responses during mapping of the motor strip.

In general, after the dural opening is complete, cortical surface EEG recording is performed to locate the seizure focus. If no seizure activity is observed, provocative maneuvers may be requested.[327] Methohexital in a dose of approximately 0.3 mg/kg is in general safe and effective. Etomidate, approximately 0.05 to 0.1 mg/kg, has also been used. Localization of the seizure focus can also be accomplished during light general anesthesia, for example, N2 O/fentanyl/low-dose isoflurane. During general anesthesia, alfentanil in a bolus dose of 30 to 50 µg/kg,[328] [329] etomidate in doses of 0.2 to 0.3 mg/kg,[330] [331] and remifentanil as a bolus of 2.5 µg/kg[332] have been reported to be effective in activating seizure foci.

After localization by EEG, functional testing may be performed by stimulating the cortical surface electrically and observing for motor, sensory, or speech interruption effects. During stimulation, the anesthesiologist should be prepared to treat grand mal convulsions that are not self-limited. Thiopental in 1-mg/kg increments is appropriate. However, thiopental should be withheld until it is clear that the seizure is not going to terminate spontaneously because it may interfere with subsequent EEG localization of the seizure focus for some time.

A comprehensive review of the "anesthetic implications of epilepsy, status epilepticus, and epilepsy surgery" is available.[333]

Previous Next