Introduction

Initially pioneered as a means of accessing the pituitary gland, the transsphenoidal approach is an increasingly popular method to treat a wide variety of anterior skull base lesions [1,2,3,4]. Since the 1960s, the evolving popularity of minimally invasive approaches and technological advances have allowed for the development of the extended, transnasal approaches to anterior skull base lesions [1, 5]. This technique, augmented by stereotactic intra-operative guidance, intra-operative MRI, and advances in endoscopic technology, has emerged as a relatively safe and effective approach for the resection or debulking of selected anterior skull base meningiomas (Table 1) [6,7,8,9,10].

Table 1 Types of meningiomas which may be operable via the extended, transsphenoidal approach
Full size table

Historically, the transcranial approach to the anterior skull base was thought to be beneficial for most lesions in this region, due to a wider operative exposure which improved visualization and surgical freedom of access [11, 12]. Although transcranial approaches are still the preferred approach for the majority of anterior skull base meningiomas, the transsphenoidal approach has recently emerged as a viable alternative, related in part to reduced approach-related morbidity. Some authors advocate that this approach can result in comparable surgical outcomes in carefully selected patients [13,14,15].

The criteria for selecting patients with anterior skull base meningiomas for whom extended transnasal surgery is a viable option are a hotly debated topic in the neurosurgical community. A multitude of factors are thought to play a role in patient selection, but vary depending on institutional and surgeon experience [16,17,18,19,20,21]. In this study, we set out to define our institutional recommendations for patient eligibility for transsphenoidal resection of parasellar meningiomas, and to report our experience with extended transnasal approaches for these lesions over the course of 8 years. In particular, we aimed to discuss the important considerations of patient selection and risk stratification in order to optimize outcomes for patients with these difficult lesions, and we also include a series of considerations that should be reviewed during approach selection.

Methods

Medical records from Brigham and Women’s Hospital were retrospectively reviewed for all patients who underwent transsphenoidal surgery for pituitary disease with the senior author from April 2008 to March 2017 (938 procedures). Patient data, including demographics, pre- and post-operative symptoms, pathological analysis, and post-operative complications were collected. The extent of resection and tumor consistency were quantitatively determined based on the surgeon’s operative report and post-operative radiographic imaging, and tumor size was determined on pre-operative radiographic imaging. The Institutional Review Board of Brigham and Women’s Hospital approved this study.

Results

Seven patients (four women, three men) underwent transsphenoidal resection of anterior skull base meningiomas during the study period. The median length of follow-up was 19 months (mean 35.6, range 3–104). Five patients presented with visual field deficits, three presented with headache, two presented with hypopituitarism, and one woman presented with infertility. One patient was receiving steroid replacement preoperatively, and three were receiving thyroid replacement therapy. One patient had type-2 diabetes mellitus preoperatively, which was treated with oral medication.

The median maximum tumor diameter was 1.7 cm (mean 2.1 cm, range 1.4–4.2 cm). All seven tumors had suprasellar extension, four were primarily intrasellar, and five were parasellar (Fig. 1). No patient had lateral extension of the tumor beyond the anterior clinoid processes on coronal MRI. All patients underwent extended transnasal transsphenoidal resection (five endoscopic, one microscopic, and one hybrid endoscopic/microscopic). Six of seven patients received subtotal resection, four of whom had firm tumors intra-operatively and two of whom had loosely-organized tumors, while one patient received gross total resection (GTR) of a loosely-organized tumor (Fig. 2). On pathology all patients had meningioma, one of which was atypical and another of which was borderline, with a MIB-1 index of 4.5. The median MIB-1 index was 2.3 (mean 3.2, range 1–7.6). Six tumors were WHO grade I, and one was grade II (Table 2).

Fig. 1
figure 1

Pre-operative coronal (left) and sagittal (right) post-gadolinium T1-weighted magnetic resonance imaging of anterior skull base meningioma

Full size image
Fig. 2
figure 2

Post-operative coronal (left) and sagittal (right) post-gadolinium T1-weighted magnetic resonance imaging of anterior skull base meningioma

Full size image
Table 2 Demographics and presenting symptoms of seven patients undergoing transsphenoidal surgery for anterior skull base meningiomas
Full size table

Prior interventions among these patients included craniotomy and Gamma Knife radiosurgery in one patient, and two prior craniotomies with linear accelerator radiation therapy in another. Complications included two readmissions (one on POD11 for small bowel obstruction, one on POD48 for epistaxis), and the development of new onset thyroid deficiency and transient diabetes insipidus in one patient.

Two patients had reoperations by craniotomy for recurrence after 5 and 6 years, respectively. Two years prior to the initial extended transsphenoidal operation with the senior author, the first patient presented with visual loss and hypopituitarism at an outside hospital, and underwent a craniotomy, which was complicated by the new onset of blindness in one eye. The subsequent transsphenoidal operation with the senior author resulted in a subtotal resection of the 3.2 × 4.2 × 2.5 cm lesion due to the highly vascular nature of the tumor, and 5 years later the tumor recurred, requiring another craniotomy. The second patient presented with bilateral hemianopia and underwent transsphenoidal surgery with the senior author for a 1.7 × 1.0 × 1.7 cm tumor. A subtotal resection was achieved, and a craniotomy was performed after 6 years due to progressive vision loss caused by tumor recurrence.

Discussion

Gradual experience gained by skull base surgeons in the treatment of sellar lesions using endoscopic visualization in the last 2 decades have permitted the treatment of a variety of lesions previously considered inoperable via the transnasal corridor. Parasellar anterior skull base meningiomas represent one such challenging pathology. These lesions have a tendency to surround important neurovascular structures, invade adjacent skull base foramina, and often have unpredictable tumor consistency. Although standard transcranial approaches for these lesions are appropriate in the vast majority of cases, there are certain characteristics that make transsphenoidal surgery a viable or even preferred option for some lesions (Table 3). There remains much discussion in the literature on this issue, and reported rates of GTR and post-operative complications are heavily dependent on not only surgeon experience, but also patient selection [5, 19, 22].

Table 3 Considerations for surgical approach selection for patients with anterior skull base meningiomas
Full size table

Presenting symptoms and patient characteristics

Presenting symptoms of anterior skull base meningiomas can mimic other intrinsic pituitary tumors [5]. Mass effect on the pituitary gland by intrasellar lesions can result in hyposecretion and deficiency of pituitary hormones [23]. Patients for whom surgery is not expected to restore pituitary function may be more suited to a transsphenoidal approach given that the risk of iatrogenic damage to the pituitary gland is potentially greater due to the trans-sellar window. In addition to causing pituitary insufficiency, these lesions can result in mild hyperprolactinemia via the pituitary stalk effect, resulting in infertility, a presenting symptom in one patient in this series [24,25,26]. For this reason, female patients of child-bearing age may be better suited for transcranial approaches if it is expected that transsphenoidal surgery might result in permanent pituitary gonadotropin insufficiency.

Headaches are another common presenting symptom for patients with sellar lesions, as demonstrated in our series [27, 28]. The potential co-morbidities of each approach should be considered in patients who may be predisposed to headaches. For instance, patients who already have severe incapacitating headaches may be better suited for transsphenoidal surgery given the reduced approach-related morbidity associated with this exposure.

There were two readmissions in this series: one for post-operative small bowel obstruction and one for epistaxis 6 weeks after surgery. One patient also developed new onset thyroid deficiency and transient diabetes insipidus. These recognized complications of transsphenoidal surgery should be considered in case selection and weighed against potential complications of transcranial surgery [29, 30]. Furthermore, transnasal approaches may be best suited for patients with extensive co-morbid medical conditions. Several studies indicate that transnasal approaches carry a lower risk for perioperative medical complications and shorter postoperative recovery versus traditional transcranial approaches [31, 32].

Tumor location and size

In this case series, patients with anterior skull base meningiomas most commonly presented with visual field deficits. This is an unsurprising finding given that these lesions usually originate in proximity to the optic chiasm [3]. A transnasal corridor can provide a more direct route to the lesion versus transcranial approaches, the latter of which necessitate dissection around the optic chiasm. In cases with significant suprasellar extension and optic nerve/chiasm deformation, the transcranial route can often place the optic nerve at significant risk given its splaying over the tumor. The goal of surgery in these cases may be debulking rather than gross total resection, which may make transnasal approaches the more favorable initial approach given the retrochiasmal view provided by this corridor. None of the patients in this series had tumor extension lateral to the anterior clinoid processes, which we believe will increase the potential for optic canal invasion. Since the optic canal in these patients can be decompressed via opening of the falciform ligament in the transcranial approach, but less effectively by the transnasal approach, these patients may be better suited for craniotomies [14, 33].

Tumor location plays a critical role in patient selection. Five patients in this series had lesions with parasellar extension. This poses the risk of carotid artery injury, which can be difficult to manage from a transnasal corridor. The degree of neurovascular encasement by the tumor can also influence whether the goal of surgery is gross total resection or debulking [14]. It is imperative that close analysis of radiographic imaging be performed before deciding on surgical approach, with particular emphasis on vascular studies. If neurovascular structures are displaced laterally, making a lesion more difficult to access transcranially, then the transsphenoidal approach may provide better access [14]. Likewise, when the meningioma inferiorly abuts the hypothalamus, the transsphenoidal approach may offer more direct access to the lesion [34].

Tumor characteristics

Tumor consistency also affects the degree to which tumor can be resected, and should be considered in patient selection, and anticipated outcomes [35]. All four patients with firm tumors in this series underwent subtotal resection, while the three patients with loosely-organized tumors underwent either subtotal (two patients) or gross total (one patient) resection depending on tumor involvement with neurovascular structures. Advances in MRI imaging, including the development of effective MR elastography, may soon provide the ability to detect differences in tumor consistency; however, at present the best practice is to rely on operative reports from past surgery (if any) to determine probable tumor type [36, 37]. Firm tumors pose a greater risk of adherence to the optic nerves or neurovascular structures during resection, and the wider operative field of the transcranial approach may help in the resection of these lesions [28].

Surgeon experience

Surgeon experience is an important factor in determining mode of visualization. Although most neurosurgeons are adept at standard transcranial approaches and often work independently, transnasal approaches on the other hand require subspecialty training and often a team-based approach. A well-coordinated team effort is necessary in order to successfully execute extended transnasal approaches, often with the assistance of ear, nose and throat (ENT) surgeons, and surgical technicians and assistants adept at maintaining, setting up, and holding endoscopes. Furthermore, extended transnasal approaches to the ventral skull base require that the primary surgeon has gained experience incrementally over time in both the intraoperative and postoperative management of these complex patients. Several studies indicate that surgeon experience plays a key role in the safety and efficacy of transnasal approaches to the skull base [22, 38].

The fact that two patients had a prior craniotomy followed by radiation therapy and two other patients underwent a craniotomy within 6 years subsequent to transsphenoidal surgery underscores how challenging these lesions can be. In some particularly difficult cases, transsphenoidal surgery for debulking of an anterior skull base meningioma can be seen as a salvage therapy for patients who have previously failed craniotomy or radiosurgery. These cases also highlight how critical case selection is, and how a variety of factors, including patient characteristics and prior treatment, as well as the exact size, positioning, and consistency of anterior skull base meningiomas, can affect the optimum surgical approach.

Institutional recommendations

It is clear that we have significant institutional bias in our patient selection. The fact that over 9 years only seven patients with meningiomas were taken to the operating room for transsphenoidal surgery speaks to the care put into selecting patients for whom a successful outcome is expected. Only certain tumor types, which we list in Table 1, may be safely resected transsphenoidally. The institutional availability of highly skilled open skull base surgeons also can assist in recommending the optimal approach in each case.

All of the previously discussed factors play a role in deciding if a patient is a surgical candidate, however, the size of the lesion is particularly important, as it can drive or exacerbate other factors, e.g., neurovascular involvement. The only lesion in this series with a maximum dimension greater than 3.0 cm was one of the two which required reoperation, and the average maximum tumor dimension in this series was relatively small at 1.7 cm. In a somewhat related vein, tumors closer to midline and with less lateral extension increase the appropriateness of being candidates for a transsphenoidal resection. If available, previous operative reports are used to exclude patients with very fibrous tumors from receiving transsphenoidal operations. Still, this is an imperfect criterion, as four of the patients in this series had firm tumors.

Other series report wide ranges of rates of CSF leak from 0 to 40%, a small incidence of neurological deficits, and even death as complications of the extended transsphenoidal approach (Table 4) [14, 39,40,41,42]. In this series, we did not experience any of these complications, perhaps due to the relatively small size of the lesions. Thus, although we did experience complications not included in other series, including reoperation, and despite the small size of this series, we believe that we have demonstrated that transsphenoidal resection of these lesions can be safely performed given proper surgeon experience and patient selection.

Table 4 Comparison of outcomes with other case series in the literature [14, 39,40,41,42]
Full size table

Conclusions

Although more commonly treated transcranially, meningiomas are sometimes amenable to resection transphenoidally. Patient selection is critical, and multiple factors, including tumor size, consistency, and location, patient and surgeon preference, and presenting symptoms each play a role in determining the optimum surgical approach. Transsphenoidal surgery for a select sub-group of patients can be used to resect or debulk tumor safely and to alleviate mass effect on parasellar structures, including neurovasculature and the optic chiasm.