Stereotactic Surgery with the Leksell Frame

Lars Leksell

It is well to reflect on the seminal contributions of Lars Leksell as we embrace new technology as a matter of course in the 21st century. A flurry of recent articles has appeared, for example, on the use of radiosurgery in trigeminal neuralgia. Few contemporary clinicians are aware that Leksell, in fact, pioneered this procedure in 1951, obtaining good results. He was also the first to perform intracavitary treatment of cystic craniopharyngiomas with phosphorus-32 [1]. Leksell is best known among stereotactic surgeons for developing the accurate and versatile frame that bears his name [2]. This was the result of ideas incubated under the influence of Spiegel and Wycis, with whom Leksell studied in 1947. He returned to Stockholm and furthered the already formidable reputation of the Karolinska Institute. During the next three decades, he continued to refine radiosurgery, developing the Gamma Knife with Borje Larson. His stereotactic frame was modified and updated until it became a standard neurosurgical tool. No less a luminary than Sugita felt that radiosurgery, stereotactic surgery and the operating microscope were the greatest neurosurgical technical advances of the 20th century. That Lek-sell pioneered two of these areas is a measure of his revolutionary influence and serves as sobering inspiration to those who wish to follow in his footsteps.

Frame and Components

Leksell’s eloquent description of his system, as quoted by Lundsford et al. [3], is difficult to improve on. Essentially, it consists of a semicircular arc with a movable probe carrier. The arc is fixed to the patient’s head in such a manner that its center corresponds with a selected cerebral target. The electrodes are always directed toward the center and, hence, to the target. Rotation of the arc around the axis rods in association with lateral adjustment of the electrode carrier enables any convenient point of entrance of the electrodes to be chosen, independent of the site of the target [4].


The model G base frame is rectangular and has dimensions of 190 mm X 210 mm (Fig. 1) [5]. A straight or curved front piece can be used; I prefer the latter, as it allows airway access in emergencies. Y coordinates are defined by supports attached to the y axes. Z coordinates are measured on vertical rings that are secured through the y supports. X coordinates are entered on the arc, which in turn rotates around the z rings (Fig. 2). X, y, and z axes on the frame recapitulate the CT and MRI axes. The frame center coordinates are 100, 100, 100, whereas a hypothetical frame origin (x, y and z = 0) resides in the upper posterior right side of the frame. The semicircular arc attached to the base frame has a radius of 190 mm. Probes that are attached to the arc probe carrier, therefore, have a working distance of 190 mm, to ensure target access. Target coordinates can be confirmed by a lateral X-ray through the Z rings, the probe tip terminating at their center.

Leksell G base frame.

Figure 4.1 Leksell G base frame.

Z rings and arc attached to base frame.

Figure 4.2 Z rings and arc attached to base frame.

A variety of technical aids are available to attach to the arc. A twist-drill craniotomy set is widely used. Biopy systems, hematoma evacuation kits, and lesion-generating devices are available. There are also accessories for microsurgery, including laser beam localizers, endoscopic adapters, and brachytherapy catheter arrays. A software program called SurgiPlan has been allied to a computer workstation, allowing simulation of probe trajectories and verifying their safety [5]. The cost of the system we use in Calgary was US $82,000 in 1996, without any of the above accessories.

Application

General preoperative precautions are followed as in any stereotactic operation [6]. In particular, normal coagulation profiles are clearly required. An- ticoagulated patients receive intravenous heparin for 5 days before surgery. The frame is placed orthogonally to the midsagittal plane by advancing the ear bars symmetrically through the lowest side holes into the external auditory meati. Salcman suggests placing the frame off center, with the ear contralateral to the target abutting the side bar of the frame, if increased lateral travel is needed on the arc [7]. The meati are routinely inspected for the presence of microhearing aids before ear bar insertion. Ear bar insertion may cause the patient considerable discomfort. This can be reduced if assistants are available to stabilize the frame for the surgeon. This obviates the need to maximize ear bar penetration and allows the frame to freely rotate. Application of foam strips (Reston, 3M) to the ear bars before their insertion also reduces pain considerably [5]. The frame is tilted 15 degrees to the orbitomeatal plane and is inclined 6 degrees from the horizontal plane to parallel the plane of the anterior commissure-posterior commissure (AC-PC) line. Balancing the frame on a finger placed on the tip of the patient’s nose is usually adequate approximation. Another useful guide to the AC-PC plane is a line drawn from the nasion to the inion.

The scalp is infiltrated with a 50:50 mixture of 1.0% lidocaine with 1:100 000 epinephrine and 0.5% plain bupivacaine. A 27-gauge needle is used to inject 3 to 5 ml of this solution through the pinholes. I have tried applying topical local anesthetic to the pin site areas 1 hour before scalp infiltration to minimize discomfort and have had an approximately 50% response rate so far. Mild sedation with midazolam or fentanyl may be needed, remembering that confusion should be avoided at all costs. Alternatively, propofol offers rapid emergence and is an excellent choice. It does require the availability of a dedicated anesthesiology team. After adopting the ear bar foam as championed by Lundsford et al., I apply the majority of frames without intravenous sedation. Of great practical importance is that the fiducial channels on the magnetic resonance (MR) localizer require fas-tiduous filling with a copper sulphate solution on a regular basis (Fig. 3). These channels are imperfectly sealed and lose enough fluid every few months to entrain air bubbles, which are capable of corrupting the fiducial markers on the MR images. It is imperative that the localizer be checked before application, as it is a costly and time-consuming exercise to discover degraded images after they have been generated. A table adaptor is also required for MR. This is secured to the base frame and fits into a slot in the MR table head. The objective is to stabilize the head and avoid excessive movement.

Unfortunately, this system does not work well for patients with thoracic kyphosis (not uncommon in elderly patients), as their limited neck extension precludes secure fixation in the table head slot. The best method to deal with this is to place one or more pillows under the patient’s buttocks and hips, effectively flattening the kyphus. The head coil is then slid over the adaptor. It is vital that the MR alignment beam lights are then symmetrically superimposed on the lateral and anterior fiducial channels. Rotation of the adaptor screws is often required to accomplish this.

MR localizer.

Figure 4.3 MR localizer.

MR STRATEGIES

In general terms, attempts should be made to minimize distortion. Any hidden clips buried in long hair should be identified and removed. Impulse generators for deep brain stimulators should have their amplitude reduced to zero to prevent side effects. Higher field magnets are preferable, especially 1.5 Tesla and above. A protocol should be in place with the radiology department to ensure frequent calibration to minimize field heterogeneities. The Leksell frame is engineered to very high standards, and Burchiel et al. have indicated that its metallic purity is such that it induces little distortion relative to other commercially available frames [8].

Specific MR sequences are available in many articles and are beyond the purview of this topic. T1-weighted sequences (thin cuts) reduce spatial distortion, and inversion recovery images sharply demarcate anatomical structures. Axial images are used to determine the x(lateral) and y(an-teroposterior) target coordinates, whereas coronal scans define the x and z(vertical) coordinates. Cartesian principles are, therefore, used to create a three-dimensional address for a specific target. In functional cases in particular, it is helpful to generate axial images parallel to the AC-PC plane. Ideally, an image that contains both the AC and the PC in the same plane is available for study. The distance between the middle and basal fiducials is measured, and a discrepancy of 2 mm between the two sides of the frame is tolerated. Greater distances mandate frame realignment to the gantry. Diagonal lines between opposing anterior and posterior fiducials are drawn on the MR console, their crossing point indicating the frame center. A cursor is placed at this point and the coordinates recorded. A target is then selected (tumor or physiological) and its coordinates noted. The x and y frame coordinates for the target are then calculated by subtracting the frame center from the target coordinates. The z coordinate is determined by adding a constant of 40 to the distance between the basal and middle fiducials at the target plane. Frame coordinates can be confirmed using SurgiPlan software or a digitizer (Elekta Instruments, Atlanta, GA).

Next post:

Previous post: