Impact of the Extent of Microsurgical Resection in Sphenopetro-clival Meningiomas Trough a Multistaged Approach: A Volumetric Analysis

Abstract

Objective  Sphenopetroclival meningiomas (SPCMs) represent a challenge for surgical treatment. The authors used an objective volumetric analysis to assess the effect of the grade of resection and different surgical strategies that may affect the outcome of this tumors.

Methods  Over a period of 4 years, patients with SPCMs were treated using a middle versus posterior fossa approach, or a two-stage surgery combining both approaches, based on the tumor location in relation to the petrous ridge and tumor volume. Retrospectively, all cases were analyzed with regard to tumor volume, extent of resection (EOR), functional outcome, and complications.

Results  Twenty-seven patients with SPCMs met the inclusion criteria, and the mean follow-up was 24.8 months. Eleven patients underwent a two-stage surgery, while 16 patients had their SPCMs resected via a single craniotomy. Mean EOR was 87.6% and gross total resection was achieved in 48% of patients. Patients with greater EOR had better functional outcomes ( r  = 0.81, p  < 0.01). Greater EOR was not accompanied by a significant increase in surgical complications. There was a trend toward lower postoperative volumes and better EOR with our two-stage approach (2.2 vs. 3.2 cm ³ , p  = 0.09; and 94.1 vs. 91.2%, p  = 0.49, respectively), without an increase in the rate of complications (18.7 vs. 18.2%, p  = 0.5).

Conclusion  Staging of the surgical resection of larger tumors may lead to greater EOR, and this strategy should be considered for larger tumors.

Introduction

Meningiomas are the most common benign intracranial tumors, accounting for up to a third of all primary central nervous system tumors. ¹ They arise from the arachnoid cap cells of the leptomeninges, and may occur anywhere arachnoidal cells are located (intracranially or inside the spinal canal). Based on imaging, 60 to 70% of meningiomas occur along the falx, the sphenoid ridge or over the convexity. ² Sphenopetroclival meningiomas (SPCMs) are considered to be a much rarer variant of petroclival meningiomas. ³ Generally speaking, the term SPCM is referred to those petroclival meningiomas extending into the cavernous sinus and sphenoid bone. ⁴ Beyond the series presented by Kawase et al, ⁵ the literature reporting surgical series of this rare category of meningiomas is scarce.

Simpson reported the relationship between the extent of surgical resection and the rate of recurrence for intracranial meningiomas. ⁶ The subsequent literature on the accuracy of those prognostic factors is vastly based on surgical case series and subjective assessment of postoperative magnetic resonance imaging (MRI). The impact of the extent of surgical resection in the group of SPCMs infers from clinical series that includes other than SPCM and use semiquantitative or subjective methods to determine the extent of resection (EOR). The current literature lacks, therefore, a quantitative analysis that compares different surgical strategies for approaching skull base meningiomas. ⁷

Volumetric measurement of brain tumors using MRI has been shown to be a precise tool that facilitates the assessment of tumor removal and residual tumor volume with good inter- and intraobserver reliability. ^{8 9} These methods are superior to other traditional ones, which are based on nonquantitative assessment of the tumor size. This is particularly relevant for intracranial meningiomas, since they may grow in various directions and not just horizontally or vertically. ¹⁰ Likewise, given the invasion of the cavernous sinus, most SPCMs are subtotally resected to ensure good functional outcomes, but with the cost of leaving behind some tumor residual around critical neurovascular structures. ^{11 12} Thus, calculation of the EOR based on volumetric analysis is a much more reliable outcome predictor, since this method is able to detect such small variations in the volume that are dismissed, when using other radiological assessment techniques, such as the ellipsoid method. ¹³

The goal of our study is to quantitatively analyze and compare functional outcome and surgical results, in regard to the extent of microsurgical resection. Likewise, we used the same volumetric analysis to determine the clinical and radiological impact of using a two-stage versus a single approach for resection of SPCM.

Material and Methods

Patient Sample

Medical records and diagnostic radiological examinations of patients who underwent surgery for the treatment of a skull base meningioma between March 2014 and March 2018 were retrospectively reviewed. Patients with SPCMs were selected for further analysis. SPCMs were defined as those skull base meningiomas with a tumor origin within the upper two thirds of the clivus at the level of the petroclival junction, lateral to the midline and medial to the Meckel's cave, that invaded the cavernous sinus and expanded to both the supratentorial and infratentorial compartment. ³ Identification of SPCM was based on the preoperative computed tomography (CT) and MRI.

Only patients with a preoperative and an early postoperative MRI (<3 months) were included in the statistical analysis. Patients are regularly assessed at 3 and at 6-month follow-up by a neurosurgeon and a rehabilitation therapist. Functional status is reported by using the modified Rankin's score (mRS). Patients without clinical and functional assessment at 6 months were excluded from the analysis. Patients who had undergone a previous surgical intervention at our institution or elsewhere or had received radiation therapy were also excluded.

The work was presented to the investigational review board of our institution, which waved the need for informed consent in light of the retrospective design of the study and the application of strict patient privacy regulations operating in our center.

Management Protocol

All cases were discussed at length at our neurooncology multidisciplinary meeting, and neuroimaging evaluation was included in the decision-making process. Middle fossa approaches, including pterional craniotomy and its variants, cranioorbitozygomatic (COZ) craniotomy with or without extradural anterior clinoidectomy, anterior petrosectomy, and/or section of the tentorium, were considered whenever the bulk of the lesion was located in the middle fossa. Alternatively, posterior fossa approaches, including retrosigmoid, translabyrinthine, retrolabyrinthine, or transcrural, as well as combined presigmoid–retrosigmoid craniotomies were considered, should these lesions lie predominantly in the posterior fossa. Those tumors with equal involvement of both the middle and posterior fossae and/or those causing mass effect above the petroclival junction and underneath the internal acoustic meatus were resected using a staged approach, performing the second stage of 4 to 8 weeks after the first procedure. Patients with a lesion. extending medial to the horizontal portion of the cavernous segment of the internal carotid artery (ICA), were also managed with a two-stage approach, including an expanded endoscopic, transsphenoidal approach to address the extension inside the cavernous sinus.

Once the optimal surgical strategy was chosen, a preoperative digital subtraction angiography (DSA) and a balloon test occlusion (BTO) were performed to assess the feasibility and safety of internal carotid artery (ICA) sacrifice in the unfortunate case of intraoperative injury. Preoperative tumor embolization was generally avoided, since the feeders of most SPCM arise from branches of the ICA and posterior cerebral artery (PCA), and the risks of preoperative embolization often outweigh the benefits. On the other hand, MRI and CT fusion were routinely used for surgical planning, along with neuronavigation.

Total intravenous anesthesia (TIVA) was routinely used. Intraoperative neuromonitoring included brainstem auditory-evoked potentials (BAEPs), median nerve somatosensory evoked potentials (SSEPs), and transcranial motor-evoked potentials (MEPs). Continuous multichannel electromiogram (EMG) recording was used throughout the entire procedure. Cranial nerves V to VII and IX to XII were also monitored intraoperatively. A hand-held monopolar stimulating probe was used to map the exact location of certain cranial nerves in relation to the tumor.

A lumbar drain was placed prior to incision to allow adequate drainage of cerebrospinal fluid (CSF) in all cases. CSF was further released intraoperatively by opening cisternal arachnoid membranes to promote brain relaxation. Our surgical strategy was maximal, yet safe tumor resection, often based on the intraoperative neuromonitoring findings. This included cranial nerve decompression either at the posterior fossa or at the cavernous sinus. Complete tumor resection with coagulation of its dural attachment was considered the most reasonable goal for these skull base tumors (Simpson's grade II), since their attachment to deep veins and/or the cavernous sinus itself made the complete tumor along with its dural attachment unsafe. If the tumor invading the cavernous sinus extended medial to the ICA, the strategy followed-up at our institution was to proceed with the complete resection of the extracavernous component only, liberate the nerves from the lateral wall of the cavernous sinus, and decompress the Meckel's cave if this was also invaded by the tumor. We did not consider removal of the intracavernous component of the tumor.

Patients were followed-up regularly with serial clinical examinations and appropriate follow-up imaging. In case of clinical worsening and/or radiological progression or recurrence of the SPCM, the patient was discussed at our neurooncology board. Overall, our strategy was to treat with fractioned radiotherapy (FR) those patients that refused to have a second surgical intervention or had a poor functional capacity, the ones that had tumors with atypical features or with an elevated K _i -67 (>5%) in the immunohistological analysis, or the ones that had more than 10% residual tumor.

Image Analysis

A digital volumetric analysis was retrospectively conducted, as follows: all digital imaging and communications in medicine (DICOM) images were analyzed using the Analyze software (version 9.0, Analyze Direct, Overland Park, Kansas, United States). The tumor volume was based on the analysis of a region of interest (ROI), as previously reported by Lagares's group. ⁶ In each slice of the contrast-enhanced, axial T1-weighted sequences (parameters: two-dimensional, spin echo, Time Echo (TE) = M-FUL, Time repetition (TR) = 500 ms, FOV = 24 cm, slice thickness = 5 mm, interleaved, image matrix = 256 × 192, NEX = 2), the ROI was identified using a threshold of density, and the sum of them was used to calculate tumor volume.

To avoid biases, an independent observer, who was unaware of the preoperative neurological deficits or postoperative functional status of the patients, as well as the surgical management chosen, performed all calculations.

Statistical Analysis

At the time of data analysis, the following categorical variables were considered: functional independence at 6 months of follow-up (mRS ≤ 2), tumor recurrence, surgical to the tumor (single vs. two-stage procedure), and presence of complications within the first postoperative month. The numeric variables analyzed included: mRS score, preoperative and postoperative volumes of the tumor, and EOR. The EOR percentage was calculated from the early postoperative MRI as follows: ([preoperative tumor volume – postoperative tumor volume]/preoperative tumor volume) × 100. Recurrence was defined as the need for additional treatment of SPCM with corticosteroids, surgery, or FR based on neurologic deterioration and/or radiologic progression versus regrowth of the tumor during the follow-up period. Tumor recurrence and local control were assessed using the MRI at last follow-up. Growth of more than 25% of the tumor volume at follow-up MRI was considered to be tumor recurrence.

All comparisons were made using nonparametric tests. To depict the association between EOR, pre- and postoperative volumes and tumor recurrence or other independent categorical variables, the Mann–Whitney U -test was utilized. The Spearman's rank correlation coefficient (Spearman's rho) was used to assess the relationship between tumor volumes/EOR and mRS at follow-up. Comparisons between approaches chosen (single vs. multistaged procedure) and volumetric parameters or functional status at follow-up were made by performing additional Mann–Whitney U -test. The threshold for statistical significance was p  < 0.05. All tests were calculated using SPSS version 20 (IBM, Armonk, New York, United States).

Results

Out of a total of 149 patients with the diagnosis of “skull base meningioma” operated between 2014 and 2018, 29 of them exhibited the anatomical characteristics compatible with SPCM, as described in the previous section. Two of those patients were excluded from the analysis, since no postoperative MRI at 3 months was available. The final analysis included 27 patients, out of which 16 underwent a single surgical approach, while 11 were treated in a two-stage fashion. As a result, data from 38 interventions were available for the 27 patients included in the present study. Mean follow-up was 24.8 ± 16 months. Mean age was 54.4 ± 10 (range: 32–69) years with some female predominance (female/male ratio: 4/3). Mean length of stay was 23.4 ± 18 (range: 7–64) days and mean length of surgical intervention (based on total anesthesia time) was 10.7 ± 3.2 (range: 4–18) hours.

Table 1 shows the distribution of the surgical approach selected for each treatment strategy. In the group of 16 patients who were treated with a single approach, 5 patients were treated through a combined presigmoid and retrosigmoid approach, 1 patient through a retrosigmoid approach, 8 through a pterional/minipterional/transzygomatic approach, and 2 through a COZ approach. In the group of 11 patients who were treated with a multistage approach, 1 patient underwent a COZ and a combined presigmoid/retrosigmoid approach, 3 patients underwent a pterional/minipterional approach and a retrosigmoid approach, and 7 patients were treated through a pterional/minipterional approach and a combined presigmoid/retrosigmoid approach.

Table 1. Treatment strategy and surgical approaches.

	Single-approach strategy a ( n  = 16)	Multistage strategy a ( n  = 11 patients, 22 craniotomies)
Middle fossa approach
PT/MPT/TZ	8 (50)	10 (45) b
OZ	2 (12)	1 (5) b
Posterior fossa approach
RS	5 (31)	4 (22) b
PRSI	1 (6)	7 (32) b

Abbreviations: MPT, minipterional; OZ, orbitozygomatic, PRSI, combined pre and retrosigmoid; PT, pterional; RS, retrosigmoid.

^aValues represent numbers of patients (%).

^bIn the multistage approach, the middle fossa approach was used first in all cases.

There were no significant differences in the total anesthesia time between patients who underwent one-single procedure and those with a multistaged approach (10.1 ± 3.1 vs. 11.8 ± 3.3, p  = 0.21). However, when analyzing the anesthesia time of each single procedure, patients who were treated with a multistage approach underwent shorter individual interventions (10.1 ± 3.1 vs. 5.9 ± 1.6, p  = 0.01).

Patients' Cranial Nerve Function

The effect of surgical resection on cranial nerve function is shown in Table 2 . Twenty-five patients presented with cranial nerve deficits (92.6%). The most common presenting symptom were diplopia (51%), hearing loss (33%), and visual loss (22%). Overall, 21 of 41 cranial nerve deficits (51%) improved after surgery, 14 (34%) got worse, and 6 (15%) remained stable. New cranial nerve deficits were observed in four patients (15%). Diplopia was the most common new-onset neuropathy after surgery (15%).

Table 2. Cranial nerve deficit before and after surgery ^a .

Types of deficit	Prior deficit	Improved	Unchanged	Worse	New
Visual loss	6	2 (33)	4 (66)	0 (0)	0/21 (0)
Diplopia	14	9 (64)	3 (21)	2 (14)	2/13 (15)
Facial numbness	1	0 (0)	1 (100)	0 (0)	2/26 (8)
Facial pain	4	3 (75)	1 (25)	0 (0)	0/23 (0)
Facial weakness	5	3 (60)	1 (20)	1 (20)	2/22 (9)
Hearing loss	9	2 (22)	4 (44)	3 (33)	1/18 (5)
Dysphagia	2	2 (100)	0 (0)	0 (0)	0/25 (0)
Total	41 b	21 (51)	14 (34)	6 (15)	4/27 (15) c

^aValues represent numbers of patients (%).

^b41 cranial nerve deficits were noted in 25 patients. 2 out of 27 patients presented with headache and other unspecific symptoms, but no cranial nerve deficits.

^cSeven new deficits were noted in four patients.

Surgery-Related Complications

There were no immediate perioperative complications, while early postoperative complications (within the first month after surgery) occurred in five patients (19% of patients, 13% of interventions). Those complications included superficial wound infection (one patient), posterior fossa hematoma requiring evacuation (one patient), and CSF leak (one patient). One patient had a corneal ulcer associated to a trigeminal neuropathy (hypoesthesia) and a facial palsy House–Brackman grade 4 that improved to grade 3 at the last follow-up. One patient experienced intraoperative laceration of the paraclinoid ICA that was managed with temporary clipping and suturing of the small defect. Postoperatively, the patient experienced no related neurological deficits.

Additional Treatments

During a mean period of follow-up of 24.8 ± 16 (range: 8–53) months, four patients experienced tumor recurrence that required a surgical revision. Two of these patients received adjuvant FR. One of these patients had a K _i -67 of 10%, while the other one presented some areas with atypical features on the histopathological analysis and a tumor regrowth medial to internal carotid artery that was not suitable for surgical resection.

Volumetric Analysis

Mean preoperative volume, postoperative volume, and EOR were 19.4, 2.3 cm ³ , and 87.6%, respectively. Gross total resection (GTR) of the extracavernous tumor portion was achieved in 13 patients (48% in this series), while an EOR > 90% was achieved in 22 patients (81% in this series). Patients experiencing recurrence were more prone to have larger pre- and postoperative volumes, and more extensive resections ( p  > 0.05, Table 3 ). Lower postoperative volumes and bigger EOR were correlated with better functional grades at 6 months-follow-up ( p  < 0.001 and 0.001, respectively, Table 4 ). Table 5 summarizes the differences in volumetric parameters in different cohorts of patients. Postoperative volume and EOR demonstrated significant association with independent status at 6-month -follow-up ( p  = 0.04 and 0.01, respectively). No differences in volumetric parameters were detected for surgical complications. No association was found between preoperative volume and any other independent variable analyzed.

Table 3. Relationship between extent of tumor removal and local control.

Variable	Local control a	Local recurrence a	p -Value b
Pre-op tumor vol. (cm 3 )	18.5 ± 14	22 ± 17.3	0.61
Post-op tumor vol. (cm 3 )	1.4 ± 1.6	5.1 ± 7.5	0.31
EOR (%)	91.2 ± 14.3	84.2 ± 15.1	0.18

Abbreviation: EOR, extent of resection; post-op, postoperative; pre-op, preoperative; vol., volume.

^aValues are mean ± standard deviation.

^b Statistically significant ( p  < 0.05, Mann–Whitney U -test).

Table 4. Volumetric analysis: univariate analysis for quantitative variables.

Volumetric parameter	mRS at follow-up
Pre-op tumor vol. (cm 3 )	0.31 ( p  = 0.18)
Post-op tumor vol. (cm 3 )	0.77 ( p  < 0.001) a
EOR (%)	−0.81 ( p  < 0.001) a

Abbreviations: mRS, modified Rankin's score; post-op, postoperative; pre-op, preoperative; vol., volume.

^a Statistically significant ( p  < 0.05, Spearman's rho test).

Table 5. Volumetric analysis: univariate analysis for discrete variables.

Volumetric parameter	Functional status			Surgical complications
	Dependent	Independent	p -Value	Yes	No	p -Value
Pre-op tumor vol. (cm 3 )	25.5 ± 13	20.7 ± 10	0.24	16.3 ± 19	22.4 ± 13	0.54
Post-op tumor vol. (cm 3 )	3.2 ± 3	0.7 ± 1	0.04 a	2.9 ± 3	2.8 ± 4	0.22
EOR (%)	84.4 ± 1.8	96.8 ± 3.1	0.01 a	72.4 ± 32.5	91.1 ± 9	0.11

(Continuation)

Abbreviations: EOR, extent of resection; mRS, modified Rankin's score; post-op, postoperative; pre-op, preoperative; vol., volume.

^a Statistically significant ( p  < 0.05, Mann–Whitney U -test).

Patients treated with a multistage approach ( Fig. 1 ) had smaller postoperative volumes and greater EOR than those treated with a single approach ( Fig. 2 ) (2.2 vs. 3.2 cm ³ and 94.1 vs. 91.2%, respectively), although this comparison reached no statistical significance ( p  = 0.09 and 0.49, respectively). Complications occurred in three patients that underwent a single-stage treatment, and in two patients that were treated with a multistaged approach. No significant difference in functional status (mean mRS at follow-up) or number of complications between these two strategies was noted ( p  = 0.21 and 0.4, respectively).

Fig. 1.

Case no. 3: voluminous sphenopetroclival meningioma in a 57-year-old woman. ( A ) The preoperative MRI showed an extra-axial mass invading posterior and middle fossa. Given the tumor volume and the wide extension through two different compartments, a two-step approach was selected to achieve the maximum grade of resection with lesser neurovascular injury risks. ( B ) First, a mini-Hakuba's approach, including middle fossa peeling and anterior extradural clinoidectomy via a minipterional craniotomy permitted decompression of the tumor located at the temporal fossa. After the peeling of the temporal fossa was performed (B́), a C -shaped dural incision was performed. Before removing the tumor, we performed a decompression of the oculomotor nerves located within the lateral wall of the cavernous sinus and that are compressed by the tumor (B́́́). B́́́́ shows the tumor cavity after the complete resection of the tumor. ( C ) The first preoperative MRI unveiled the complete resection of the extracavernous part of the tumor located at the temporal fossa and partial volume reduction of the tumor located in the posterior fossa, allowing a safer approach in a second time via a retrosigmoid approach ( D ). ( D ) Retrosigmoid approach, stepwise resection of the posterior fossa part of the tumor: opening the arachnoid cisterns permitted brain relaxation and optimal visualization of the posterior part of the tumor (D́), reducing brainstem retraction and giving access to the caudal portion of the posterior fossa meningioma (D́́). Section of the tentorium aided in the complete resection of the superior part of the tumor (D́́́ and D́́́́), and the tumor implantation base at the temporal bone is catheterized using bipolar (D́́́́). ( E ) The final MRI exhibits the complete resection of the initial tumor mass. MRI, magnetic resonance imaging.

Fig. 2.

Case no 9: sphenopetroclival meningioma in a 59-year-old female. ( A ) Preoperative MRI showed a petroclival tumor invading cavernous sinus, and equally extended through the middle and posterior fossa. ( B ) Middle fossa approach was chosen, including minipterional craniotomy, extradural anterior clinoidectomy, peeling (B́), and tumor resection (B́́). Middle fossa approach offers a wide working angle for SPC meningiomas with temporal fossa extension, as it is seen in the last picture after tumor removal (B́́́). ( C ) Postoperative imaging (CT) demonstrated removal of the extracevernous portion of the middle fossa and adequate decompression of the brainstem. CT, computed tomography; MRI, magnetic resonance imaging.

Discussion

Our results provide some evidence that a surgical strategy that consists of an initial middle fossa approach, followed by a posterior fossa approach after 4 to 8 weeks, provides a larger extent of resection without increasing the complication rate, in comparison to the most commonly performed technique of a single approach.

It is already known that a larger tumor residual after surgical resection correlates with a higher rate of recurrence. ^{10 14} However, the relationship between tumor volume and functional outcome at follow-up needs to be determined. Generally speaking, skull base meningiomas tend to grow faster than meningiomas in other locations. ¹⁵ Therefore, assumptions regarding the outcome of this subtype of meningioma cannot be deducted from larger series that include meningiomas that arise in locations other than the skull base. Additionally, the variability in tumor volume at initial diagnosis and the differences in surgical approaches used to address these lesions make the systematic analysis difficult, and may explain the lack of large series in the literature. ^{3 16} Beyond that, the anatomical complexity of the SPC region, the poor surgical outcomes reported in the past, often associated to injuries of critical neurovascular structures, and the emergence of radiotherapy have contributed to the current controversy in regard to the treatment of SPCM. ¹⁷

In spite of the development of the expanded endoscopic, endonasal, transsphenoidal approach, open surgical approaches are still the most commonly performed ones worldwide. ^{3 4 11 16 18} Observation, stereotactic radiosurgery (SRS), FR, and a combination of all the above surgical techniques are also part of a multimodality management of those complex patients. ^{13 19 20 21} Moreover, FR and SRS alter the rate of growth of the tumor and may eventually cause irreversible damages to surrounding neurovascular structures that may affect the functional outcome at follow-up. ^{22 23 24} For all these reasons, the treatment of SPCM should be decided on an individual basis, rather than following rigid protocols. Given the lack of large series including SPCM, and since the standardization of treatment protocols is affected by all factors discussed above, a thorough analysis of objective parameters is needed to establish the indications for surgery and maximize the clinical benefit for the patients. As such, we decided to focus our attention on volumetric analysis to better define the clinical behavior of SPCM and outline the relationship between the extent of microsurgical resection and other variables considered in this study.

No comparison of resection volumes among different surgical approaches to SPCM has been made to date. Morisako et al ¹³ found that larger postoperative volumes were associated with poor local tumor control. In the same study, a greater EOR was also associated with more perioperative complications. In contrast to these findings, our results, which purely describe the EOR in terms of volumetric analysis and are not impaired by other confounding factors, such as additional treatment modalities, showed that higher EOR is not necessarily associated with higher incidence of surgical complications. Despite our constant attempt to reach a maximal safe resection, it is also important to emphasize that complete excision was only achieved in two patients. Whenever the tumor extended beyond the cavernous segment of the ICA or the plane between the tumor and other important neurovascular structures was poor, some thin sleeve of residual tumor was left in place, to preserve neurological function. We would also like to emphasize that smaller postoperative volumes produce less mass effect on the surrounding structures and therefore less neurological deficits. ^{11 25 26 27} Patients with small or no residual tumors tended to have better functional outcomes at follow-up. Differences in tumor consistency and vascularity, or other intraoperative features, such as tumor adhesion to critical structures, likely led to worse functional results in patients with larger postoperative residual volumes.

The question then becomes how can anatomical information guide the surgical management on a case by case basis? It is uniformly agreed that large volume lesions preclude SRS or FR treatment as a standalone option. ^{23 24} Nonetheless, the use of radiotherapy as part of a multimodality management of skull base meningiomas became extremely popular in the last decade. ^{22 23 28} It is important to stress that the close relationship between the tumor and other important neurovascular structures implies an additional risk of severe neurological impairment, even with this treatment modality. ^{22 23 29} Moreover, despite the encouraging results of initial reports, the long-term effect of the radiation therapy in tumor control has not been as well described as the surgical excision. ⁶ Therefore, it does not offer a valid alternative to surgical resection in certain cases, such in large tumors or symptomatic young patients. ^{12 23 30} On the other hand, the option of observation with serial imaging tends not to be a great one either, since most skull base meningiomas tend to grow over time, and the long-term outcome appears to be worse in larger lesions. In fact, in a recent series of conservatively treated petroclival meningiomas, 76% of the lesions demonstrated progressive growth, out of which 63% had neurological deterioration after just 4 years of clinical and radiological follow-up. ²⁴ In our series, we also observed a more pronounced trend toward tumor recurrence in SPCM with larger preoperative volumes. Since preoperative tumor volume is likely the best predictor of surgical outcomes, ³ we believe that it is better not to wait until the tumor has reached a large volume to consider surgical excision. This is particularly critical in younger patients in whom tumor progression is expected, given their longer life expectancy.

One of the most remarkable findings of the present study was that the multistage surgical approach is a feasible alternative to the standard surgical approach of a single procedure. To achieve maximal safe resection, in 41% of our cases we opted a two-stage approach, consisting of an initial pterional or minipterional craniotomy, peeling and sectioning of the tentorium, followed by a delayed retrosigmoid-type approach that allowed us to remove the posterior fossa component. This strategy offers a good and practical alternative to a single-stage, combined intervention, and fits both goals of achieving greater EOR and better functional outcomes with lower complication rates, as shown by our results. In our opinion, a multistage approach offers some benefits. First, the reduced intraoperative time of each single procedure will also reduce anesthesia time and complications related to it. Improved surgeon performance is also an additional benefit of reduced operative times due to lower fatigue and stress. Prior surgical decompression via a middle fossa approach prevents an excessive manipulation of the brainstem during the second intervention via a posterior fossa approach. Likewise, the middle fossa approach might be directed toward relieving the compression of the cavernous sinus and oculomotor nerves affected by the tumor. ¹⁸ Edema induced by surgical manipulation and brain retraction is also relieved after waiting for a few weeks. Conversely, an excessive delay in the second surgery, beyond 6 to 8 weeks, might induce inflammatory changes and fibrosis, and make the second intervention more challenging by affecting the separation plane between the tumor and healthy neurovascular structures. A multistage approach typically increases the cost of care per patient. Hence, a multistaged approach should be reserved for patients, who cannot tolerate long operative times, such as those with significant comorbidities. A single approach is the mainstay of treatment for the vast majority of patients, especially the ones that have good functional capacity and tumors that can be safely removed through a single craniotomy.

The fact that we did not find significant differences in functional outcomes between single- and two-stage approaches might be related to two factors: the limited number of cases and a selection bias. Tumors selected for a multistage approach had higher degree of surgical complexity due to their location and extent. Sassun et al ³ used a combined approach (posterior and middle fossa) to treat three meningiomas with SPC extension. Among the three patients treated with this strategy, only one (33%) experienced new deficits and in two of them (66%), the EOR was greater than 90%. Once again, the limited number of patients in that study does not allow us to draw robust conclusions, but it is felts that a combined approach may be a safer alternative for the treatment of those skull base tumors, even in experienced hands.

Despite all emerging technologies that provide great tools for the skull base surgeon, nothing can replace thoughtful decision making and microsurgical dexterity, both essential in the treatment of complex skull base lesions. Since large, prospective clinical trials are rather difficult due to the small number of these lesions, the neurosurgical community might benefit from the development of cadaveric tumor models simulating the technical challenges encountered in real life. This would be a valid way to better assess the pros and cons of different approaches to SPCM. Moreover, we feel that careful presurgical planning and volumetric tumor analysis, as described in our study, likely represents the safest way to approach these complex lesions and collect useful insights for the skull base surgeon.

Limitations

Our study has certainly certain limitations due to its design. Along with the previously discussed drawbacks regarding the small number of patients and the existence of potential selection biases, the current study has several other limitations that are derived from the study design. As such, our findings led to conclusions drawn by a single neurosurgical center. Thus, the external validation of our findings is desirable.

Having said that, the EOR and neurological complication rates tend to be operator-dependent. Therefore, our outcomes and treatment strategies should be preferably compared with the ones established by other centers with similar case load. We believe that highly complex tumors, such as SPCM, should be ideally treated at high-volume centers, by surgeons who are comfortable with skull base approaches and the complicated and often altered intracranial anatomy.

Conclusion

SPCMs represent a neurosurgical challenge due to their location, involvement of neurovascular structures, and tendency to grow over time. Staging of the surgical resection of larger tumors may lead to greater EOR and better outcomes. The volumetric residuals influence the functional outcome and the need for adjuvant treatment, such as SRS or FR. Our study further suggests that more radical resections, while preserving critical neurovascular structures, provide better functional outcomes in patients with SPCM. As such, we advocate for maximal safe resection.