19.1 Introduction

Corpus callosotomy was first described by Van Wegenen and Herren in 1940 ¹ and has since developed technically and the indications have been refined and more precise. It is a palliative procedure in medically therapy-resistant epilepsy where the evaluation has found no seizure generator to resect. ² , ³ , ⁴ , ⁵ , ⁶ The most important indication is drop attacks. ² , ⁴ The results reported vary in different reports. ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ This patient group has often been evaluated for either VNS or callosotomy, but some differences have been reported in the effect of VNS and callosotomy. ¹¹ It is mandatory to use strict indications and to have realistic expectations from both the medical side and from the well-informed patient and the relatives to get an optimal result to satisfy the patient and the family. As in all surgical procedures, the techniques used must, of course, be as safe as possible to avoid unnecessary risks for the patient. However, we must also remember that while dealing with palliative procedures, any unexpected neurological deficit may be even more catastrophic than when the goal is to cure the patient, as the possible gain is much lower.

Most commonly, corpus callosotomy has included the anterior two-thirds of corpus callosum, but today more and more reports favor a complete callosotomy, either in a single-stage or in a multistage procedure. There are also intermediate approaches that leave only the splenium or half of the splenium. ¹² , ¹³ , ¹⁴ A complete callosotomy seems to yield a better outcome, ¹⁵ , ¹⁶ although a population-based study in which the absolute majority of patients underwent anterior two-thirds callosotomy reported a good long-term outcome. ⁶ ▶Fig. 19.1 illustrates a complete callosotomy.

Several publications reported a low risk of a major complications ¹⁷ , ¹⁸ , ¹⁹ ; most groups use advanced technical equipment such as neuronavigation ²⁰ , ²¹ and advanced planning algorithms such as MEG ²² , ²³ and perioperative EEG monitoring. ²⁴ There were also reports on the use of endoscopic-assisted ²⁵ , ²⁶ , ²⁷ , ²⁸ and robot-assisted techniques, ²⁹ including minimally invasive neurosurgical approaches. ²⁵ , ²⁹ In addition, the use of a laser was reported as a safe procedure. ³⁰ In this chapter, the classical open corpus callosotomy technique as performed at the Sahlgrenska University Hospital ³¹ using the ultrasonic aspirator, microscope, and neuronavigation will be described. The anatomy relevant for the procedure was recently very clearly described by Schaller and Cabrilo. ³² Other published techniques will also be described and discussed.

19.2 Open Classic Corpus Callosotomy

▶Fig. 19.2 illustrates the usefulness of neuronavigation and the most important landmarks: the rostrum, genu, and splenium. The head is fixed to a head holder; four-pin holders are preferred as they give higher accuracy in relation to neuronavigation. However, in small children, it may be easier to use a three-pin holder. The hemisphere used in the surgical approach for callosotomy is by default the one regarded as nondominant.

If there is a lesion, atrophy, or possible seizure generator in the dominant hemisphere which was not amenable for resection, callosotomy could be performed from that side. The nose-up position is preferred, although some groups described the procedure with the nose to one side. ³³ I find it easier to get a good relaxation and surgical space using the nose-up position. The skin opening and the craniotomy are adapted to the length of the intended callosotomy. The skin is opened as a flap, and retracted to the side. However, the skin opening is not crucial, and a lazy “s” or straight incision could be used according to the preference of the surgeon. The periosteum is usually spared and made into a separate flap, for eventual use in stopping possible bleeding coming from the midline after opening of the bone. Two-burr holes are placed strictly in the midline, and then the craniotome is used to open 3 to 4 cm laterally on the intended side and 1 to 2 cm laterally on the other side. Personally, I prefer to have margins regarding the length of the craniotomy, especially at the posterior end, as bridging veins may give you trouble. It is, of course, possible to minimize the bone opening with the help of the venous anatomy visualized on MRI. After performing dura-tacking sutures to the bone, the dura is opened and turned to the midline. The opening of the dura is tailored, if needed, to respect the bridging vein anatomy. ▶Fig. 19.3 illustrates the steps in positioning and opening.

To get surgical space, the use of mannitol and light hyperventilation is of good help. A lumbar drain is never used; you will anyway be able to get more space by cerebrospinal fluid drainage via the surgical opening. Using very light retraction of the hemisphere and one retractor on the falx cerebri, you can gently divide any adherences and work your way down to the corpus callosum with its typical white appearance, and also identify the pericallosal arteries that will serve as the main landmark during the anterior part of the procedure. Going strictly in the midline and respecting the pial surface of the cortex, there is only minimal risk for ending up in the wrong place. After identifying the corpus callosum, it is advisable to first divide any adherences all the way to the genu, where you can identify the pericallosal arteries turning down and around the genu. This is an important landmark. At this stage, you can start the callosal division using a dissector, ultrasonic aspirator, or bipolar coagulation. It may be an advantage to preserve the roof of the ventricle, although opening it may very seldom originate any complications. Anteriorly, the thickest part to divide is the genu. Using the pericallosal arteries as landmarks and neuronavigation information is mandatory at this point not to end up in one of the frontal lobes. Staying between the arteries you can identify the pial surface anteriorly and use that to go around the genu and reach the rostrum, which you must also divide. Division of the anterior commissure is not part of the standard procedure.

Having performed the anterior portion you may now continue to identify the posterior portion including the splenium. The pericallosal arteries will lose the landmark function, and you may rely more on the falx and neuronavigation. You may have to adjust the approach to the posterior part of corpus callosum to respect midline bridging veins; sometimes, it is also helpful to angulate the head position upward. The splenium may be very deeply seated, but this information is, of course, already obvious from the preoperative planning stage. The posterior part of the corpus callosum is thinner; when reaching the splenium, you would have once again more volume to divide. At this point, the pial surface is very useful as a landmark to secure a complete division. You may also visualize the large veins through the pial surface. ▶Fig. 19.4 illustrates some of the dissection steps, also illustrated in the short video (Video 19.1) of the procedure.