Complete drainage was achieved in all but one patient, due to an unaccessible pouch separated from the main cavity. The mean surgical time was 126 min. There were no intraoperative complications and even the postoperative course was uneventful: surprisingly no chemical meningitis was observed and transient hyperpyrexia in three cases rapidly subsided. The mean hospital stay was 7 days. During the follow-up period (range 12–72 months; average 40 months), we observed two recurrences of the cystic portion. Both patients underwent successful repeat neuroendoscopy procedure. One died postoperatively due to unrelated causes (pulmonary embolism) while neuroradiologic examinations had shown complete drainage of the cyst. In the remaining patients, follow-up examinations have demonstrated persistent drainage and resolution of symptoms (Fig. 9.2). One patient (20-year-old man) underwent microsurgical removal of an enlarging solid nodule 9 months after endoscopy (Fig. 9.3).

Between January 1985 and June 2012, 21 patients (5 pediatric patients, age range 6–17 years, mean 11 years; 16 adults, age range 35–74 years, mean 59 years) underwent cystosphenoidostomy as surgical treatment for mainly cystic sellar craniopharyngioma. Six patients had preoperative diabetes insipidus requiring medical therapy, 16 patients had some degree of endocrine disturbance requiring hormonal substitution, and all patients had visual field defects that ranged from minimum peripheral deficits to monocular blindness in one case. All tumors had an intrasellar component, with sellar enlargement and supradiaphragmatic extension. In all cases the cyst was at least partially within the pituitary fossa. Indications for cystosphenoidostomy were recurrent tumor (6 cases), open surgery refusal (3 cases), unlikeliness of total removal basing on neuroradiological appearance (7 cases), and high anesthesiological risk in case of craniotomy (5 cases). No patients had preoperative hydrocephalus. Recurrent tumors had previously been treated with microsurgical technique either by a transcranial approach (2 cases) or a transsphenoidal approach (4 cases); 3 patients had undergone to radiotherapy as an adjuvant treatment.

Surgical technique consisted in a microsurgical or purely endoscopic transsphenoidal approach. A wide opening in the sellar floor was performed in order to expose the cyst wall. The cyst wall was incised, the contents were suctioned, and cyst walls were dissected as far as possible, taking care in avoiding any lesion to the arachnoid and cerebrospinal fluid leakage. The permanent communication with the sphenoid sinus was established through a self-securing X- or H-shaped silastic catheter, which was positioned over the sellar floor so that two arms lied within the pituitary fossa and the other two in the sphenoid sinus (Fig.9.4).

The results of direct surgical attacks on craniopharyngiomas have enormously improved over the last half century. Surgical mortality has dropped from 41 % [16] before replacement therapy became available to 2 % or less in recent series [1, 2, 17]. However, a closer insight raises several issues. First of all, radicality can be achieved in only 45.7 % of transcranial procedures [1] and an overall long-term good clinical outcome in 60.3 % of patients [17]. Endocrinological sequelae are the rule and 80 % of Yasargil’s patients required substitution therapy [2]. Postoperative diabetes insipidus must be accepted as a common sequela since its incidence increases 3.7 times after surgery [2, 18] and anterior pituitary function is only slightly but constantly affected [18]. More subtle neuropsychological deficits have been observed in surgically treated children in the long term [19]. Focusing on the subgroup of recurrent tumors, total removal is possible in only 21.1 % of craniotomies, and all procedures (craniotomies and transsphenoidal approach) are burdened by a 10.5 % operative mortality [1]. Finally, it must also be emphasized that these figures are quoted from some of the most successful series in the literature: craniopharyngioma is a rare tumor (1.3/million/year) [20] and it is reasonable to assume that less experienced hands would perform less brilliantly. A predominantly cystic tumor is observed in 60 % of craniopharyngioma patients [4]. Since mass effect rather than infiltration is responsible for the clinical syndrome, permanent/repeated cyst drainage seems a suitable compromise in most such cases. Fox has to be credited for the first implant of a sump drainage system in a cystic suprasellar tumor [7]. Similar experiences followed for Miles and Gutin et al. [8, 9]. In all their cases, a catheter connected to the subgaleal reservoir was positioned at craniotomy. Cysto-ventriculostomy, as proposed by Spaziante and de Divitiis, represents an evolution of these early experiences: the authors introduced the concept of marsupialization of the cyst into the ventricles, aiming at continuous dilution of the cyst’s fluid and resorption through the CSF pathways [4]. They reported clinical improvement in all patients and a good outcome in the six patients available for follow-up [4]. Periodic aspiration has also been implemented with intracavitary administration of antineoplastic agents in the attempt to overcome the palliative nature of sump drainage. Although β-emitting radiocolloids and bleomycin have proven effective under certain conditions, there have also been some major drawbacks related to isotope handling and radiation damage [10, 21] and the absence of unitary protocols and fatal toxicity. In both cases treatment failures are not uncommon [4, 6, 22–24]. Fluid-filled spaces represent an ideal battlefield for neuroendoscopists: since most craniopharyngioma cysts impinge or grow into the ventricular systems (types c–f according to Yasargil’s system), an endoscopic approach seems a suitable option. Minimum invasiveness, a radical cyst drainage, and wide marsupialization into CSF fluid spaces under direct visual control represent clear-cut advantages over other draining techniques. Indeed in the last decade, different institutes have published endoscopic treatment of cystic craniopharyngioma [3, 5, 12, 25] On the other hand, we never attempted nor in the light of our experience recommend an endoscopic resection of solid components due to the risk of uncontrollable damage to neurovascular structures. We prefer to remove as much of the relatively avascular dome and leave behind basal tumor remnants that can later be considered for radiation therapy. In our hands, the neuroendoscopic management of cystic craniopharyngiomas has proven to be safe, repeatable, and effective in the midterm, and it should be especially considered in recurrent tumors and in very young or very old patients. No intraoperative or postoperative complications related to surgery were observed, including chemical meningitis, probably due to careful rinsing of all debris. On the other hand, it must be emphasized that neuroendoscopy is oncologically palliative and not compatible with further intracavitary treatments, which require a sealed cavity. The recurrence rate is undetermined, but seems unrelated to surgical aggressiveness. In only two of our cases, reclosure and filling of the cyst led to clinical symptoms several months after the first procedure and was probably due to insufficient dome stripping. Systematic cisternostomy and stenting of the internal cyst create an additional outflow route to the intracystic CSF circulation and can limit the need for further procedures. Remote recurrence of craniopharyngiomas has been related to intraoperative transplant or CSF spreading [26], and therefore cysto-ventriculo-cisternostomy should in theory be at risk. However, seeding represents an exceptional event, with only eight cases reported to date in the world literature. All of these cases occurred after open surgery with the manipulation of solid, more cellular tumors. The surgical treatment of craniopharyngiomas is extremely challenging and often success requires a multimodal approach. Dealing with mostly cystic tumors can be deceiving since the microsurgical approach ensures drainage, but only seldom radicality, since stripping of the capsule away from the hypothalamus and major vessels is highly risky. Simple drainage has therefore been achieved by puncture and aspiration. The neuroendoscopic approach represents a major advance compared with classic draining techniques, since emptying is obtained under direct vision and operativity of the instrument allows a partial resection of the capsule for diagnostic purposes and especially wide marsupialization. Perforation into the basal cisterns and stenting are measures to prevent reclosure and recurrence, and other therapeutic modalities (microsurgery, radiosurgery, fractionated external radiation therapy) can later be scheduled to eradicate solid remnants, if indicated. A purely endoscopic approach as a first choice is an alternative to microsurgery, in selected cases. Case selection must be based on anatomo-surgical (cystic/solid rate, Yasargil’s type) and clinical (primary/recurrent, age, conditions) considerations: in such cases, if tumor control and satisfactory reintegration into the appropriate social setting are the goals, endoscopy cannot be overlooked.