Jules Hardy

Over the last 10 years, numerous articles have reviewed the historical background of the surgical treatment of pituitary tumors and confronted the challenge between the alternative of intracranial versus transsphenoidal approaches. They deserve careful reading for the understanding of the scientific evolution of the transnasal transsphenoidal approach, which has now become the standard procedure for more than 90% of the sellar lesions.

This introductory chapter will be confined to pertinent comments on the transsphenoidal approach with regards to significant progress in the methodology of therapy of pituitary tumors and emphasis on the technical improvements related to concurrent instrumental developments.

In his pioneering contribution, Schloffer used an upper nasal approach on his first patient on March 16, 1907. A review of the French literature reveals that on April 10, 1909, Lecène performed a supranasal transsphenoidal operation in an acromegalic patient. The procedure was then modified in 1910 by Hirsch, who used an inferior endonasal approach on his first live patient. The technique underwent several variants by other authors in cadaver studies to be finally adopted as a standard procedure in patients by Cushing’s, Dott, and Guiot, who remained faithful to the sublabial rhinoseptal midline approach. The lateral endonasal approach originally used by Hirsch remained an alternative procedure eventually revived by Griffith in 1987.

The major drawback of the open surgical approach through the nose was that it was in the past a blind procedure in a dark cavity; this had been overcome earlier by Cushing’s by using a headlight. Later, the Norman Dott bivalve speculum contained spotlights at the tip of each valve, thus providing a retractor combined with endoscopic illumination into the sphenoid cavity.

The direct approach from the nasal cavity avoids dissecting the rhinoseptal mucosa and allows direct entrance through the ostium sphenoidale into the sphenoid sinus, but the operative field is, in no way, wider than that from the rhinoseptal approach after bilateral submucosal elevation. The important issue is not where we are coming from but what we are doing once we are in the sella turcica, and how we deal with the pituitary tumor; that is the major point.

The first endoscope for transsphenoidal approach with an external light source mounted on a rigid shaft was first used by Guiot, with an original instrument devised by Fourestier-Vulmiére. In 1961, I was Dr. Guiot’s assistant when he performed the first procedure with this endoscope. Although Guiot did use the endoscope for the sellar approach in several cases, he abandoned it because of its time-consuming and cumbersome maneuvers. He then used the endoscope but only occasionally to explore the sellar cavity after removal of the tumor to seek residual tumor tissue hidden in the corner laterally or underneath the tuberculum sellae. Upon my return to Montreal in 1962, I similarly used the endoscopic approach in several cases but again only to explore the sella after tumor removal.

The argument that the endoscope provides a wide field illumination allowing a panoramic view seems somewhat overstated because it is not necessary to see the lateral aspect of the sphenoid sinus nor the carotid grooves on the sides of the sella turcica. I concur with Ed Laws’ aphorism: “ It is safe to jump into endoscopic pituitary surgery without previous experience in transsphenoidal surgery .”

There has been great controversy regarding the endoscopic approach as opposed to the open approach with the surgical microscope. In my opinion, this is not a major issue. A well-known surgeon (P.B.) hinted that endoscopic surgery is sometimes seen as a “marketing device to try to change traditional patterns of referral for pituitary surgery.” Advertising the preferential use of an endoscope to draw the attention of a referring physician does not guarantee the superiority of the surgical outcome on the treatment of the lesion, and a study on clinical and biological statistical results comparing the alternative methods has not yet been published.

I came to the same conclusion that it was not more useful than the spotlights at the tips of the Dott bivalve speculum, which I used during a few years before the introduction of the surgical microscope.

Image intensifier fluoroscopy , also introduced by Guiot, was the most important advance in the early 1960s. This contributed to a better definition of the tumor contour and a clear visualization of the instruments placed in the sella at the base of the skull during the major step of the surgical procedure. Televised fluoroscopy was the very first navigating system highly useful in the removal of macroadenomas with large suprasellar expansion. The injection of air in the subarachnoid space by lumbar puncture as a way to outline the superior contour of the tumor provided clear delineation of the intracranial contour of the tumor for monitoring the placement of instruments and the protection of the supratumoral nervous structures: optic nerves, chiasm, hypothalamus, and adjacent cerebral arteries. This provided significant advancement in the transsphenoidal approach for removal of large pituitary tumors and other lesions around the sella turcica. The major indication was the symmetrical midline suprasellar expansion, which could be removed in one or two stages, depending on the consistency of the lesion.

In my experience, this method of navigating toward the sella with image intensifier–guided fluoroscopy is the most secure and safest procedure. It provides a real time radiologic image of the tumor and sequential modification during the progressive descent of the tumor into the sella turcica. Such dynamic imaging is not yet available by the use of magnetic resonance imaging (MRI) guided system unable to provide a real-time picture. As the suprasellar mass is moving down into the sella turcica and the magnetic resonance image remains the same on the screen, it is no longer useful as a reference and is in fact potentially dangerous for inaccurate placement of the instruments beyond the virtual image. It is also useless in the identification of bony landmarks, such as the anterior nasal spine and the vomer bone, that guide the surgeon towards the sella. Only in cases of large invasive tumors, or in repeated surgery where the bony landmarks are destroyed, does midline identification with image guidance become as useful with MRI as it is with the computed tomography–guided assisted fluoroscopy. The use of an MRI system in transsphenoidal pituitary tumor surgery has therefore not yet proved to be superior to the actual standard image-intensified fluoroscopic method. If subarachnoid air injection via the lumbar route fails, the injection of soluble contrast material (omnipac-240: iohexol USP-52%) directly into the tumor with a 26-gauge needle is an alternative. The opaque perfusion of the entire tumor outlines its boundaries for placement of curettes into the tumoral cavity. Progressive excision of the opacified tumor serves as a marker for tumor removal and lasts about 10 minutes before being completely reabsorbed. Preoperative MRI identification of the residual normal gland helps the surgeon to preserve it while doing selective tumoral tissue removal.

The recently introduced intraoperative MRI-guided system is a promising instrument because it affords the combination of real-time image acquisition and integrated optical tracking capabilities.

The introduction of the surgical microscope constituted a major technical development with magnification coupled with bright illumination allowing the surgeon to distinguish clearly between normal and pathological tissues well visualized in three dimensions. This concept evolved from my experience with hypophysectomies for treatment of advanced painful metastatic breast cancer, while at the same time I was exploring patients with Cushing’s disease or acromegaly with a small sella. Becoming familiar with the appearance and texture of the normal hypophyseal tissue, the discovery of intrapituitary microadenomas and their selective removal became possible, and eventually this knowledge was applied to the treatment of hypersecreting pituitary disorders. This discovery has revolutionized the therapeutic concepts in neuroendocrinology—patients could not be submitted to a selective adenomectomy without resulting in hormonal deficit and replacement therapy. Furthermore, preoperative deficits could be corrected with resumption of normal functions. Refinements, in microscopic techniques were required to achieve selective microadenomectomy because of the small size of a lesion hidden inside the pituitary parenchyma.

Selective extraction of a microadenoma in the lateral wing of the gland (GH and PRL) requires delicate maneuvers with blunt-ended microenucleators to avoid perforating through the lateral dura of the sella, entering into the cavernous sinus, and lacerating the carotid artery. An invading lesion is removed by aspiration and gentle scooping with a curette, which may penetrate the cavernous sinus, thus producing a flush of blood along with the tumor tissue. The proper response then should be packing and not coagulation.

Laterally located lesions may require partial wing resection to ensure complete excision of the microadenoma. Posteriorly located lesions should not be pulled out but rather aspirated to prevent tugging on the inferior hypophysial artery and its ascending branch along the pituitary stalk and the anterior hypothalamic nuclei. Disruption of this vessel may result in permanent diabetes insipidus. Centrally located microadenomas, most frequently found in Cushing’s disease, are easier to enucleate but they can extend posteriorly into the neural lobe, which needs to be excised to achieve complete tumor removal and therefore biological control of the disease. Adrenocorticotropic hormone microadenomas not visible on the surface of the gland can be detected by ultrasound imaging, as originally described by Oldfield. During these procedures, accidental arachnoidal tearing produces a cerebrospinal fluid leak that must be repaired as described elsewhere in the book. The extended transsphenoidal approach for other lesions of the sellar and parasellar regions such as craniopharyngiomas, clivus chordomas, and meningiomas, had already been described in 1971. Further developments in the “classic” transsphenoidal approach using the most recent technical modalities are discussed in the following chapters.