Rathke’s cleft cysts have a high prevalence and stay mostly asymptomatic [11]. Complaints usually appear through mass effect and compression of surrounding structures when the cyst starts growing [33]. In cases of symptomatic RCCs, there is wide consensus on the necessity of surgical treatment in the current literature[1, 7, 34, 35]. While permanent damages of the visual, olfactory or endocrinological system need to be prevented [7]. In most neurosurgical centres, those resections are routinely performed via a minimally invasive, transsphenoidal microscopic or endoscopic approach, since in the context of other neoplasia like meningioma or pituitary adenoma, most surgeons have a high level of experience with it [9, 13, 19, 20, 25, 30, 36]. However, the transcranial keyhole approach is in general limited to patients whose cyst morphology or other individual characteristics count against the transsphenoidal approach and is therefore used only in few, often complex cases actually [9, 12]. Due to this lack of equivalent data, the question whether the transcranial or the endonasal approach shows better radicality, recurrence rate und postoperative outcome remained so far unanswered. In order to find the technique with the best postoperative outcome but also with the lowest complication rate, the two patient cohorts of this study were compared with regard to their preoperative symptoms, surgical data, postoperative outcome and complications. Thereby, the main criterion of the neurosurgeon for choosing the approach was the cyst localisation in preoperative imaging. A cyst lying at the bottom of the hypophyseal fossa will preferably be resected via transsphenoidal approach, whereas a completely suprasellar cyst will be reached easier via transcranial approach without manipulation of the gland tissue [12]. This principle helps to preserve the anatomical surrounding by choosing the shortest or most direct trajectory. Furthermore, the risk of postoperative new hormonal deficiencies should be minimized by a transcranial route [12, 26, 28].
The authors´ study aims therefore to find the best surgical approach in cases where both techniques are feasible in their retrospective analysis of 29 patients. Considering the last-named point and the retrospective character of our work, it is not surprising that the cyst morphology in the two cohorts differed. There were much more purely suprasellar cysts in the transcranial group, whereas all purely intrasellar cysts were resected via transsphenoidal approach. This is in line with previous reports of the literature [12]. The percentages of the individual cyst locations were comparable with those reported in previous studies [1, 8, 37, 38]. The higher number of suprasellar cysts in the transcranial group was attended by a higher incidence of preoperative visual deficits, which is based on the suprasellar location of the optic chiasm. The transsphenoidal group in contrast contained more cases of endocrinological dysfunction, which can be explained by the fact, that all of the cysts had a purely or partly intrasellar location neighbouring the pituitary gland. In addition, neurological symptoms like headache or nausea were more numerous in the transsphenoidal group which is probably due to its higher average cyst volume and a therefore distinct meningeal irritation.
In terms of postoperative outcome, the present study showed equality of the two surgical techniques. Both approaches allowed an adequate partial removal of the cyst wall with NTR of 15% in the transcranial cohort and of 31% in the endonasal cohort. The GTR rate differed significantly (85% vs 31%). However, the radicality did not influence the recurrence rate and the postoperative follow up. There was no recurrence in both cohorts within the long-term follow up (mean 5.7 years). Therefore, the authors stated, that the outcome did not depend on the extent of cyst wall resection, given that symptoms of RCCs are caused by irritation and compression effect on surrounding structures but not by invasive infiltration or endocrinological activity of the cyst tissue itself. Lu et al. concluded in their study, that an extended wall resection might lower the number of cyst recurrence [3]. This hypothesis is discussed controversially in literature, and the authors cohort did not reveal any confirmations of this hypothesis compared to the presented results of Chotai et al. [1, 30]. Also, it must be considered that due to the fact, that a part of the RCCs wall is part of the adenohypophysis, an entire resection is not possible without hypophyseal harm [1]. A gross total resection should therefore be sought but not be forced as it may increase the risk of postoperative endocrinological complications.
Whilst the two techniques were comparable regarding their postoperative outcome and recurrence rate, there were distinct differences in terms of complications. Especially the probability of persisting CSF fistula and consecutive rhinoliquorrhea was significantly higher in the endonasal cohort as expected due the selected surgical technique. However, the risk of intraoperative opening of the diaphragm by cyst removal and fenestration to the CSF space is much higher than in routine endoscopic intrasellar surgery [13]. This induces a higher risk of CSF fistulas in correlation to other series and pathologies treated endonasal endoscopically [13, 20, 21, 28]. The literature discuss the major disadvantage of endoscopic endonasal skull base approaches with significant risk of CSF leakage reported about 5–30% [39–44]. Although, using a sandwich-technique and lumbar drainage additionally after intraoperative CSF flow did not reduce the risk of persisting CSF fistulas in the authors´ series below the reported previous results. Therefore, some authors suggest a surgical technique of cyst fenestration in cases of RCCs via an endonasal approach [25]. In the authors´ opinion this surgical concept is not an adequate treatment because of the high recurrence rate and insufficient cyst opening. The radicality and recurrence rate of the authors series underlines the hypothesis that a partial resection of the cyst wall and communication of the RCC to the CSF space is necessary to prevent a recurrence in long-term follow up. An intraoperative perforation of the diaphragm with CSF flow was seen in about one half of the cases in the presented cohort. On the one hand, it paves the way for postoperative CSF fistulas and infections, but on the other hand, authors suppose that it might decrease the risk for RCC recurrence. Seen that after transsphenoidal surgery, the sphenoid bone and thereby the hypophyseal fossa will usually be closed, newly secreted cyst content might accumulate and put pressure on surrounding structures again [13, 20, 45]. A constant connection to the intracranial CSF system via a controlled opening of the diaphragm could prevent this mechanism. Further complications in the transsphenoidal cohort were nasal and endocrinological dysfunctions. These results are in line with previous reports [13–16, 28]. Previous research showed that nasal complaints can decrease patients’ quality of life significantly by limiting sleep quality and performance [15, 46, 47]. In the presented study, postinterventional treatment by an ENT physician was necessary in a fourth of the endonasal cohort. Comparable percentages were described in the literature [12, 15, 46, 47]. Additionally, the endonasal route can induce hypo- or anosmia. Postoperative endocrinological dysfunction was the most common complication in the presented cohorts. However, it has to be distinguished in to transient and permanent dysfunction. Since transient endocrinological complications were well treatable and disappeared after a few weeks, their relevance is negligible in decision making. However, permanent dysfunction goes along with lifelong medicamentous substitution and laboratory tests and should therefore be avoided. Although endocrinological complications were more common after transsphenoidal interventions in this study, data in further literature are not consistent. Fan et al. for example presented a comparable study design with comparison of the transsphenoidal and transcranial technique in the treatment of RCCs and reported, that hormonal dysfunction appeared more often after transcranial surgery [12]. Further studies regarding both techniques in the context of other pathologies like craniopharyngeomas or meningeomas showed ambiguous results [23, 28, 48]. Apart from one case with endocrinological complications, the main risk of the transcranial approach in the authors´ study were lesions of cranial nerves: two of these 13 patients showed transient hyposmia, hypaesthesia or a limited facial expression caused by an irritation of the olfactory nerve, trigeminal nerve or the temporal branch of the facial nerve during intervention. As all patients recovered within one year, these complications did not affect their quality of life in the long-term. Peng et al., who analysed a cohort of exclusively transcranial operated patients, had only two cases of transient postoperative Diabetes insipidus, but no other complications [6]. Their study confirms the authors´ results of a low complication rate of the transcranial keyhole approach. This result might be caused by a larger surgical corridor with better manoeuvrability in the suprasellar space. Additionally, the risk of destruction of the pituitary stalk and gland tissue is lower via transcranial route. On the other hand, the risk of manipulation of the chiasm and optic nerves is higher although the authors´ did not see any complication or worsening of visual function in their series. To minimize these risks and to increase the surgical results it is necessary to use an endoscopic assisted microsurgical technique via a keyhole approach in the authors´ opinion. Many disadvantages of the classical transcranial routes to the sellar region might be minimized using a supraorbital keyhole approach. Theoretical advantages are reduced brain retraction, smaller skin incision with reduced tissue dissection and therefore less postoperative related complications and a more direct approach to certain pathologies of the skull base. Another important advantage of the supraorbital approach with eyebrow skin incision is the wide opened surgical field in depth with increasing the distance from the craniotomy through a small opening. Thus, paradoxically, the superior surgical field at the anterior skull base and sellar region might appear larger than expected via minicraniotomy. Furthermore, such minimally invasive approaches can overcome complications such as temporalis muscle atrophy, mandibular pain and chewing problems [32, 49, 50]. The major advantage of a transcranial route to RCCs is the large opening with NTR or GTR without the high risk of CSF fistulas and new hormonal deficiencies of the different axes. Whereas the endonasal route might induce hypo- or anosmia the olfaction can be preserved using the supraorbital approach to sellar region. In addition, nasal complaints can be avoided completely via the supraorbital approach.
The endoscope is an essential tool to successfully finish the surgery using a keyhole approach. The use of the endoscope and identification of the anatomical structures and remnant tissue is the presented cohort underlines this hypothesis. Endoscope-assisted techniques in the field of neurosurgical skull base surgery have proven to be beneficial regarding the following aspects [32, 51, 52]: they provide increased light in the depth of the surgical field, provide a clear depiction of anatomical details in a close-up fashion and offer extended view also around the corner with angled telescopes. However, the exposure is not 3-dimensional and the surgeon has to be familiar with the special endoscopic devices and anatomic feature and needs a sufficient training for eye-hand coordination in the endoscopic view. The endoscope may limit the operating range of the instruments, depending on angle and depth of the selected approach to the RCC. An area where the endoscope is probably essential for resection control is the region under the ipsilateral optic nerve for inspection of diaphragm sellae or retrochiasmatic space as well as in the optocarotid window and laterally to the internal carotid artery.
In this context, the authors want to indicate the meaning of modern technical innovations who aim to continuously ameliorate and facilitate surgical work conditions. More precise instruments will help to lower also the transsphenoidal complication rate and make these operations even safe [6, 17, 18, 29, 53].
When considering the ideal approach to a RCC in a selected case, the neurosurgeon must filter out those complications of those which were discussed, that lead to a momentous or long-lasting limitation of the patients’ quality of life. In authors´ opinion, these include CSF fistulas with necessary reoperation, permanent endocrinological dysfunction and persisting nasal complaints. All of them were more frequent or even limited to the transsphenoidal approach. In combination with the equal postoperative outcome concerning radicality, recurrence rate and visual function, the transcranial keyhole approach should be recommended if both approaches are feasible.
It must be stated, that the patient cohort is relatively small due to the low incidence of symptomatic RCCs, and that a longer follow-up period up to 10 years ore even more would give us more specific information about recurrence rates and surgical outcome of both approaches. The mean follow up time period was more than 5 years in the presented study. However, there are reported recurrences of RCCs 5–10 years after surgery in the literature in some cases. Additionally, no blinded direct comparison between both surgical approaches has been performed. The results reflect an unavoidable personal preference of the operating neurosurgeon regarding the choice of approach in correlation to tumor size and localisation. However, the presented findings depend on the individual anatomy of the tumor. Therefore, a reevaluation of the presented data after further follow-up and further studies with more patients are necessary to substantiate the results.