Chapter 23 Transcranial Surgery for Pituitary Macroadenomas
Transcranial approaches are used in approximately 1% to 4% of pituitary tumors that require surgical management.1,2 The most common indications for transcranial surgery for pituitary tumors are large extrasellar tumor extension without sellar enlargement, marked frontal, middle fossa, or clival extension of tumors (particularly dumbbell shaped), very fibrous tumors not amenable to trans-sphenoidal resection (meningiomas), a persistent visual field deficit after incomplete decompression via the trans-sphenoidal approach, a coexistent aneurysm, loss of oculomotor function, ectatic midline carotid arteries, or sphenoid sinusitis.1–3 Treatment selection depends on the tumor characteristics and its associated findings. A thorough understanding of the advantages and disadvantages of each transcranial approach is critical when choosing the most appropriate surgical option.
History
In 1893, Caton and Paul recorded the first attempted, but unsuccessful, pituitary tumor resection using a two-stage lateral subtemporal decompression in a patient with acromegaly.4 This approach was suggested by Sir Victor Horsley, who from 1904 to 1906, used a subfrontal approach and a lateral middle fossa approach to operate on 10 patients with pituitary tumors.5 In comparison to the 50% to 80% mortality rate of his colleagues, Horsley reported a mortality rate of 20% using this approach. In 1904, Kilani showed an extensive bifrontal intradural approach using cadavers.6 In 1905, Krause demonstrated that it was possible to reach the sella turcica via a frontal transcranial approach in a living patient.7 In 1912, McArthur described an extradural approach via resection of the supraorbital ridge and orbital plate.8 These approaches formed the basic foundation upon which subsequent neurosurgeons improved and modified surgical treatment of these pathologies.8,9
Harvey Cushing played a central role in standardizing the preference for transcranial surgery over the trans-sphenoidal approach. Cushing’s initial experience with transcranial approaches to sellar neoplasms was very discouraging and as such, he adopted the trans-sphenoidal approach that was undergoing marked improvement at the time.10 Cushing had a mortality rate of 5.6% in his series of 231 patients from 1910 to 1925 who had undergone surgical resection via the trans-sphenoidal approach.11 The morbidity and mortality associated with the trans-sphenoidal approach at the time were primarily attributed to infection, most often associated with postoperative cerebrospinal fluid (CSF) leak, postoperative edema, and hemorrhage.9 These complications coupled with Cushing’s extreme interest in intracranial surgery prompted the development of a novel transcranial approach to sellar-based neoplasms. Cushing’s increased experience with transcranial procedures, and subsequently the reduced mortality he experienced utilizing these approaches, led to a preferential treatment of sellar tumors using an intracranial approach.9 The transfrontal craniotomy, performed via a direct midline approach, allowed Cushing to attain a more extensive resection, better visualization and decompression of the optic nerve and chiasm, lower recurrence rates, and better recovery of vision.9,12 The transfrontal approach also avoided the potentially fatal complication of infection seen with the trans-sphenoidal approach. For these reasons, Cushing discarded the trans-sphenoidal operation, and with his prominence in American neurosurgery, he helped to usher in a strong preference for the transfrontal approach.11,13,14
From the 1930s to the 1950s, transcranial approaches to the pituitary dominated neurosurgical practice and teaching. In the late 1950s and 1960s, the trans-sphenoidal procedure received renewed interest as many technological and surgical advances improved this surgical approach including: (1) development of a lighted speculum retractor by Dott in 1956,15 (2) enhanced surgical accuracy with the intraoperative radiofluoroscopy to define nasal passage anatomy and control the position of the instruments by Guiot,16 (3) development of antibiotics in the 1950s to reduce the surgical mortality associated with this approach,17 and (4) introduction of corticosteroids allowing safer surgeries to be performed on the pituitary gland.12 Some of these advances were utilized by Hardy in 1971 in his landmark paper that demonstrated the trans-sphenoidal operation on over 300 patients undergoing hypophysectomy.18 In his paper, he described the technical aspects of the trans-sphenoidal hypophysectomy and also reported a morbidity and mortality rate that was less than with transcranial approaches. Specifically, the risk of CSF leak was markedly reduced when the arachnoid was left intact. Since then, trans-sphenoidal surgery has dominated neurosurgical treatment of sellar-based lesions.
Specific Indications
While the trans-sphenoidal approach is standard for resection of sellar/parasellar region tumors, there are specific indications for transcranial approaches. The usual reason for using a transcranial approach is doubt about the diagnosis.19 For instance, if the lesion is not a pituitary adenoma, but instead perhaps a meningioma or craniopharyngioma, a craniotomy is advised because these lesions are more safely removed transcranially.19 For tumors with a wide extension on the cranial base, craniotomy remains a superior approach. Craniotomy is more effective in a variety of other surgical cases as well because it increases access to tumors, facilitates preservation of the surrounding neurovascular structures, improves visualization at the dural edge, and allows earlier identification of the cranial nerves and the feeding vessels.20
Failed Trans-Sphenoidal Surgery
A major indication for a transcranial procedure is a failed trans-sphenoidal surgery. Failure can occur for multiple different reasons. Any intrinsic tumor characteristic that does not allow it to fall into the sella compromises the efficacy of trans-sphenoidal surgery. For instance, if a pituitary adenoma is too fibrous, a trans-sphenoidal approach will fail. In those circumstances, it is advisable to take no more than a biopsy specimen and obliterate the sphenoid sinus. The latter action may seem unnecessary with a large tough tumor between the nasal cavity and the CSF. The consistency of the tumor, not the size, limits the effectiveness of trans-sphenoidal surgery.19
Failure of the suprasellar extension to descend is another reason for failed trans-sphenoidal surgery. The usual cause is that a component of the tumor is extending superiorly, laterally, or anteriorly. If the suprasellar extension has pushed vertically, it usually falls into the tumor cavity created by the trans-sphenoidal surgeon. If an extension of tumor hooks around the optic nerve or carotid artery, this extension prevents descent of the tumor, and the tumor extension above the carotid artery or chiasm is inaccessible.19 The anterior extension is particularly troublesome for the trans-sphenoidal surgeon because it is at the wrong angle for the line of approach.19 In addition, recurrent pituitary adenomas may not fall down satisfactorily because of adhesions between the tumor and the surrounding brain. In these circumstances, the combined transcranial/trans-sphenoidal approach may be the best method to completely remove the tumor. Thus, it may be justified if the tumor has recurred despite previous irradiation. Radical surgery is also indicated for hormone-secreting adenomas such as those that occur in patients with acromegaly (or Cushing syndrome), in which hormonal cure depends on total removal of the tumor.19 A trans-sphenoidal procedure is also insufficient when the surgery fails to remove adequate amounts of tumor, particularly enough to decompress the optic chiasm.19,21 The failure often manifests after incomplete decompression as a persistent visual field deficit or persistent loss of vision.
Para/Extrasellar Extension
In cases with large extrasellar tumor extension without sellar enlargement, the pituitary gland may be vulnerable to a trans-sphenoidal approach. Usually in this case, the suprasellar component either encases optic nerves or intracranial arteries or it spreads over the surface of the planum sphenoidale.21
Also, pituitary adenomas can invade one or both cavernous sinuses. While the trans-sphenoidal approach is limited to the compartment of the cavernous sinus medial to the C4 segment of the internal carotid artery, this approach can be used in patients with non-secreting adenomas in which tumor debulking and neurovascular decompression is sufficient.21 Transcranial approaches are warranted, however, if the adenoma is secretory, or if the parasellar extension ventures beyond the cavernous sinus into the middle fossa.21 Transcranial approaches are also indicated when restoration of oculomotor function is the goal.
Dumbbell-shaped pituitary adenomas that have a ‘narrow’ waist created by a thick diaphragm sellae and a small pituitary stalk opening warrant a transcranial approach over a trans-sphenoidal. One can predict the failure of trans-sphenoidal surgery for a dumbbell shaped lesion based on tumor compression into the sella by intraoperative lumbar intrathecal injection of sterile saline and endoscopic navigation of the opening in the diaphragm sellae.21
Other
Co-existent aneurysms, ectatic carotid arteries, and severe sinus infection are indications for transcranial surgery. Aneurysms are found concomitantly in approximately 1.1% of all pituitary adenoma cases.21,22 If an aneurysm adjacent to a pituitary adenoma (i.e. an aneurysm of the anterior cerebral artery) is detected preoperatively, both lesions can be potentially treated in the same operation. Co-treatment of these lesions is more applicable when an aneurysm will be affected by manipulation of the regional anatomy.21 Alternative treatment strategies include a staged-procedure, surveillance of either lesion or nonsurgical treatments such as endovascular coiling for the aneurysm and radiotherapy for the pituitary adenoma.21 In terms of ectatic vessels, the tumor’s relationship to the intercavernous carotid arteries should be taken into account when deciding on an appropriate approach. There is commonly a 1 to 3 mm separation between the medial margin of the internal cerebral arteries and the lateral surface of the pituitary gland.23 Ectatic carotid arteries can veer into the midline trajectory preventing a trans-sphenoidal approach.9 Also, if a sinus infection is severe and surgical delay will pose a threat of acute neurological deterioration, a transcranial approach is warranted.20,21
Diagnosis and Workup
The preoperative work-up for patients with signs or symptoms of a sellar neoplasm should include formal testing of the patient’s visual fields and pituitary function. A T1-weighted magnetic resonance imaging (MRI) scan with and without gadolinium in the sagittal and coronal planes should also be performed to allow proper visualization of the optic nerves, optic chiasm, carotid arteries, cavernous sinuses and surrounding soft tissue.19,24 A T2-weighted scan may be useful in determining the fibrotic nature of a tumor in this region to aid in surgical planning. Although no reliable predictor exists, evidence suggests that pituitary adenomas with a homogeneously isointense, as opposed to hypertense, signal on T2-weighted MRI are predicted to be firm and fibrotic, although this is not routinely used radiologic criteria.20,21,25
The position of the optic chiasm can be predicted on the sagittal MRI scan by finding the anterior communicating artery.19 While the length of the intracranial optic nerves varies from patient to patient, tumor extension influences the distance of the optic nerves to the tuberculum sella. If the suprasellar extension is thrust between the optic nerves, it pushes the chiasm upward and backward, allowing good surgical access. On occasion the suprasellar extension pushes the chiasm upward and forward, severely limiting surgical access.19 This is called a “prefixed chiasm.”
Craniotomy for tumors in this region carries the same general risks as any operation (i.e., deep vein thrombosis, bleeding, infection, anesthetic risks), however there are specific risks associated such as damage to the optic nerves (resulting in impaired vision or visual fields), damage to the internal carotid artery or anterior cerebral artery, and damage to the pituitary stalk (resulting in hypopituitarism and diabetes insipidus).19 The transcranial approach almost inevitably causes hypopituitarism because the normal pituitary tissue is pushed superiorly under the diaphragm sella, and this tissue is specifically coagulated and cut by the surgeon en route to the tumor.19 The potential for harm to the vascular structures in this area necessitates cross-matching of 2 pints of blood. Anticonvulsants should be started preoperatively because epilepsy is possible after transcranial surgery. Cerebrospinal fluid rhinorrhea or anosmia can occasionally occur, especially if excessive retraction of the frontal lobe is indicated.19
Pterional
The frontosphenotemporal or pterional craniotomy is the most commonly used transcranial approach to pituitary tumors (Fig. 23-1). It was popularized by Yaşargil as an approach to intracranial aneurysms and is now the most widely used transcranial approach in neurosurgery.26–28 It provides a direct path to the sella turcica and allows removal of large pituitary tumors with minimal brain retraction. It should also be the transcranial approach of choice for tumors when a prefixed chiasm is present as the tumor can be resected safely underneath the optic chiasm.27
Technique
After reflecting the skin flap and fat pad, the temporalis muscle is elevated. An incision is made in the temporalis fascia 1.5 cm posterior to the frontozygomatic process and continued posteriorly along the linea temporalis. A 1- to 2-cm cuff is left attached to the linea temporalis for reattachment of the temporalis flap during closure.29 Using subperiosteal dissection, the temporalis myofascial flap is elevated off of the skull.30 Once the myofascial flap is completely reflected anteriorly and inferiorly, it is held in place using fish hooks.
The frontosphenotemporal craniotomy may be created using varying numbers of burr holes and either a standard router with the footplate attachment or with the Gigli saw. We prefer to make our craniotomy using two burr holes: (1) in the squamous portion of the temporal bone just superior to the root of zygoma and (2) at the MacCarty keyhole. The ideal location for the MacCarty keyhole burr hole is to create it on the frontosphenoid suture approximately 5 to 6 mm posterior to the junction of the frontozygomatic, the sphenozygomatic, and the frontosphenoid sutures.31 A Penfield #3 dissector is used to dissect the dura from the inner table.A standard router with footplate on a pneumatic drill is utilized to turn the craniotomy. The supraorbital foramen serves as the medial border of the craniotomy. Once the bone flap it ready to be elevated, it is important to free the dura that remains attached to the inner table with a Penfield #3 dissector.
After the bone flap is removed, extradural bony removal continues prior to opening the dura. The frontal and temporal dura are dissected off of the ridge of the sphenoid bone with a Penfield #1 dissector. The bony sphenoid ridge is removed using rongeurs. The sharp edges of bone may be smoothed out with a pneumatic drill with a #2 or #3 diamond bit. It is important to constantly irrigate the drill to prevent thermal injury to the dura and underlying optic nerve. The optic canal is skeletonized and the anterior clinoid process completely removed. The underlying clinoidal segment of the internal carotid artery is then visualized. When approaching tumors with wide parasellar extension or that encroach the cavernous sinus, uncovering the foramen rotundum and foramen ovale gives additional needed exposure to minimize temporal retraction.26