Preoperative workup is aimed at determining the best surgical approach and proximity to nearby structures in order to accurately assess surgical risks. CT is the best tool for assessing bony anatomy in regard to hyperostosis and calcifications, as well as allowing for preoperative determination of the surgical corridor and the degree of bony removal required. In the event of significant bony erosion by the tumor, the patient may require surgical fusion to prevent instability at the craniocervical junction. MRI remains the imaging modality of choice for assessing soft tissues, including the origin of the tumor and the involvement of the critical neurovascular structures located nearby [19].

Many advocate vascular imaging (CTA, MRA, or conventional angiography) preoperatively to evaluate arterial feeders, venous drainage, and the extent of vascular involvement. In particular, defining the V3 and V4 segments of the vertebral artery along with the origin of the PICA will aid in operative planning and the avoidance of complications [8]. Identification and analysis of vessels that are encased in tumor are particularly important. The presence of significant stenosis is suggestive of tumor invasion of the adventitia. In these cases, conventional angiography allows the surgeon to determine if vessel sacrifice is a feasible option based on the results of balloon occlusion test and the presence or absence of collaterals [4].

Patients presenting with FMMs are usually considered for surgery if the lesion is symptomatic, has experienced growth, or is causing mass effect on the brainstem. In these patients,radiosurgery is considered difficult due to the absence of a plane between the tumor and the brainstem and the likelihood of ongoing compression of the brainstem.

As most foramen magnum meningiomas are located ventrally, surgical resection via a far-lateral or extreme-lateral approach may be utilized (see chapter on thefar-lateral approach ). For midline posterior tumors that do not cross the plane of the dentate ligament, a midline posterior approach is indicated. Midline anterior tumors without significant spinal extension may appear at first glance to be best approached via an endoscopic endonasal route (EEA) , if the tumor is located on the anterior rim of the foramen magnum, the origin is medial to the hypoglossal canal and jugular foramen with posterior and lateral displacement of the lower cranial nerves. However, significant inferior extension would necessitate removal of the anterior arch of C1, the C2 odontoid process, and the ligamentous complex that provides stability at the craniocervical junction. Therefore these tumors are usually managed via aposterolateral approach [4]. The proponents of theEEA would state that it allows for tumor resection without retraction of brain tissue and does not necessitate crossing the plane of the cranial nerves. Inanteriorly based tumors , this approach allows for early access to the dural blood supply with improved visualization during tumor resection and decreased intraoperative blood loss. Also, involved bone and dura may be resected more easily as compared to a more lateral trajectory. Our objection to utilizing theEEA to intradural meningiomas is based on five categories of pitfalls. First, the bony removal in this approach necessitates reconstruction of the skull base and is accompanied by a significant risk of CSF leak [11]. Second, lower rates of complete resection are achieved. Similar to the case withtuberculum sella meningiomas , the dural tails, not well appreciated on preoperative MRIs, often extend lateral to the “presumed” margins of the tumor and are missed or not seen properly during an endoscopic approach. Third, the current, even state-of-the-art instrumentation available for endoscopic skull base surgery lags behind the more mature microscopic arsenal available to microsurgeons. Fourth, thenasal/nasopharyngeal mucosa postoperative morbidity is not trivial and often mischaracterized. Lastly, if a transoral approach is combined with a transnasal route, there is the risk of velo-palatine insufficiency [9]. Additionally, this approach requires a multi-disciplinary team and not all surgeons are familiar or adept with it.

Thevertebral artery is identified early in lateral surgical approaches, and the decision must be made to work around the artery or mobilize it. We agree with those surgeons who argue against mobilization of the vertebral artery in most cases [6,14,15,23], although, others routinely employ this tactic [16,20]. Encasement of the vertebral artery by tumor can be seen, and if identified preoperatively, the consequences of vessel sacrifice can be anticipated with balloon occlusion testing. Both extradural encasement and repeat surgery are associated with increased risk of vessel rupture as well as incomplete removal (41% and 51%, respectively ) [20,23].

Many surgeons recommend the use of somatosensory evoked potentials, brainstem auditory evoked potentials, and electromyographic monitoring of the lower cranial nerves (CN X, XI, XII), and this is good practice for this type of surgery [8]. Approaching via a transnasal route mandates monitoring CN VI motor and sensory evoked potentials as this cranial nerve will be encountered early during the approach to the tumor [4]. The senior author has experienced cases where the radicular artery traveling with the C1 nerve root would have been sacrificed (with devastating consequences) to improve exposure for a foramen magnum meningioma during a far-lateral approach, had it not been for a change in evoked potentials when a temporary clip was placed on the artery to test its contribution to the vascular supply of the upper cervical cord .

A detailed discussion of thefar-lateral approach is described in the chapter dedicated to that subject. Patient positioning, location of the incision, and drilling of the foramen magnum are addressed there. Once the approach to a foramen magnum meningioma has been completed, be it a far lateral or unilateral suboccipital, there are some general principles to be respected.

The dura is opened in a linear or C-shaped manner based laterally. The dentate ligament should be divided, with particular attention to not confuse it with a portion of the spinal accessory nerve (located posterior to the dentate). The other relational anatomy of relevance is that the V4 segment of the vertebral artery is anterior to the 12th nerve rootlets, which in turn are anterior to the 9/10/11th nerve complex, while the PICA originates at variable heights along the V4 and also courses in a variable direction between the nerves (Fig.10.2). Once the dentate ligament is divided, therostrocaudal extent of the tumor needs to be defined. A good practice is to lyse allarachnoidal fibers above and below the tumor, then posterior and medial to the tumor, to allow the cerebellum and other structures to “fall away” from the tumor with gravity. This helps define the “boundaries” of the resection and focuses the surgery.Self-retaining retractors are almost never used. We favor the use of nonstick Telfa strips to create the surgical boundaries .

Ideally, the next step should then be an exposure to the dural base of the tumor for earlybipolar devascularization (Fig. 10.3 ). This step is always straightforward in the case of a lateral origin of the tumor, but may be more problematic in the case of a large midline base covered by a large bulk of tumor, particularly when the vertebral artery and its perforators may be engulfed in the tumor. Here, a “test” resection of an accessible piece of tumor is done first to see how vascular it might be. Piecemeal resection can continue if the tumor is not too bloody, without further consideration given to reaching the base first (Figs.10.4,10.5). If on the other hand the vascularity is significant, then a tailored corridor to the dural base should be created through a careful and systematic partial debulking of the most accessible part of the tumor. Once part of the base is reached, the tumor can be devascularized, leading to incremental exposure of more of the base, more debulking, and so on with sequential steps of increasing returns. The tumor shell most adherent to neurovascular structures is naturally left for the end of the resection, when more space is available to tease it out safely. It has been our observation that no matter how large or fibrous or invasive a foramen magnum meningioma is, it never transgresses the pia of the medulla, which, for unclear reasons, is not the case for large petroclival meningiomas that unfortunately often invade the pons subpially .