Combined Extradural and Intradural Pterional Transzygomatic Approach to Large Sphenoid Wing Meningiomas. Operative Technique and Surgical Results

Abstract

Objectives  Describe a unique and safe surgical procedure for the microsurgical management of large sphenoid wing meningiomas (SWMs) aimed to a radical resection of these tumors.

Design  A prospective series of 26 cases with SWMs larger than 3 cm in one of its main diameter is presented. All patients were studied following the same clinical and imaging procedures. The surgical approach was through a pterional transzygomatic craniotomy. The surgical procedure has the following steps: 1. Extradural tumor devascularization and resection of the hyperostotic and/or infiltrated bone and then intradurally; 2. Intradural tumor debunking; 3. Microdissection of vascular branches and perforators from the capsule; 4. Identification of the optic and oculomotor nerves and internal carotid artery; 5. Tumor capsule dissection and resection; 6. Dural resection or cauterization; 7. Dural and bone reconstruction and closing.

Results  All lesions were completely removed. Most complications were transient. The most relevant complication was a large middle cerebral artery infarct with permanent hemiplegia despite a decompressive craniotomy.

Conclusion  Large SWMs can be considered as a single pathology regarding the surgical approach and intraoperative microsurgical procedure strategies. The pterional transzygomatic approach allows an extradural devascularization of the tumor and an extensive bone resection that facilitates the intradural stage of tumor resection. The proposed approach allows a wide and radical resection of the duramater and bone that increases the Simpson grade. However, surgery does not control other biological or molecular prognostic factors involved in tumor recurrence.

Introduction

Meningiomas that grow from the dura mater of the sphenoid ridge are a classic subtype of the skull base meningiomas. This portion of dura mater is actually covering the lesser wing of the sphenoid bone from the anterior clinoid process (ACP) to the pterion. These meningiomas are denominated in the current literature as sphenoid wing meningiomas (SWMs).

In this paper we propose and describe a unique surgical approach useable for large SWMs whose objective is the radical resection of the tumor growth, its dural implantation and also the underlying bone with hyperostosis eventually infiltrated by the tumor, allowing a Simpson's grade 1 resection. ¹ We present the clinical results of our recent surgical series with the use of this technique in large SWMs to demonstrate the efficacy and the safety of the proposed approach.

Material and Methods

Patients

A series of 26 patients with SWMs, operated on by the author between 2009 and 2015 at the Hospital Clínico Universitario of Valencia is reviewed. Only lesions larger than 3 cm in at least one of their diameters have been included. In all cases, the same preoperative work-up study, surgical approach and technique, and postoperative control protocol were used. Epidemiological data include age, sex, relevant associated pathology, clinical symptoms, and neurological examination at diagnosis, and in the immediate postoperative period.

The preoperative imaging study includes computed tomography (CT) and magnetic resonance imaging (MRI). The CT study was done without and with intravenous contrast and with a protocol study with fine bone cuts. The MRI study always included the T1, T2, FLAIR (Fluid Attenuation Inversion Recovery), diffusion, and T1 with gadolinium sequences. T1 sequences with gadolinium were used to delineate the volume and extent of the tumor. The T2 sequence was used specially to define the relations of the greater vessels of the polygon of Willis's and their main branches with the tumor capsule. Vessels targeted were the M1 and M2 segments of the middle cerebral artery (MCA), the A1 segment of the anterior cerebral artery (ACA) and the internal carotid artery (ICA). Perilesional edema was evaluated with the T2 and FLAIR sequences. Cerebral angiography was only performed in case of suspicion of stenosis of the ICA or large intracranial vessels in the CT and/or MRI study. In the immediate postoperative period, a CT scan was performed to rule out postoperative complications. The degree of resection and the presence of tumor recurrence were evaluated with an MRI study at 6 months of surgery and later every year. All tumors were pathologically studied and classified in grades I-II-III, according to the WHO tumor classification of the year 2016. ²

Surgical Technique

All the patients were operated on following the same approach and microsurgical techniques. The patient is positioned in the supine position with the head attached to the table with a Mayfield's clamp. The head is elevated and left parallel to the ground plane. The surgical procedure is divided in several steps.

Pterional Transzygomatic Frontotemporal Craniotomy

This approach, as performed by the authors, has been described in the literature in other papers. ^{3 4} A soft tissue incision is made from the preauricular region that crosses the midline to the contralateral frontal region until the hair growth line. A soft-tissue flap that includes the periosteum is elevated and exposes the superficial aponeurosis of the temporal muscle. This superficial aponeurosis is sectioned 1 cm below the beginning of the orbital process of the frontal bone to separate its 2 layers and make an interfascial dissection. The superficial layer of the aponeurosis of the temporal muscle reaches the zygomatic arch and the deep layer continues to cover the temporal muscle. The zygomatic arch is cleaned to its full extent and sectioned with a fine vertical saw near its root and in the zygomatic bone ( Fig. 1A ). The first osteotomy is oblique and the second is right angled. The holes for the screws of the reconstruction miniplates are marked on the bone before the osteotomies. The bone segment can be removed or remains attached to the soft tissues. The temporal aponeurosis and muscle are then sectioned 1 to 2 cm below its insertion in the upper temporal line and the muscle is detached from the temporal bone while retaining the deep aponeurosis intact. This cuff of muscle and temporal aponeurosis remains for the subsequent reinsertion of the muscle in the closure. The temporal fossa is thus exposed to the crest that marks the beginning of the horizontal portion of the greater wing of the sphenoid bone ( Fig. 1B ). The soft tissues are held in position with the aid of multiple hooks.

Fig. 1.

Left transzygomatic craniotomy. ( A ) The entire zygomatic process (ZP) and zygomatic bone (ZB) have been exposed. The anterior and posterior osteotomy lines have been marked with dotted lines. The deep layer of the superficial fascia of the temporal muscle (TM) remains covered by the interfascial fat. ( B ) The zygoma has been resected and the temporal muscle (TM) has been separated and elevated, leaving the temporal fossa (TF) exposed. ( C ) The dura mater covering the frontal lobe (FL) and the temporal lobe (TL) is shown, where the middle meningeal artery (mma) is running. The sphenoid wing (SW) has been drilled to the superior orbital fissure (SOF), where the recurrent meningeal branch of the ophthalmic artery can be coagulated and divided.

Two burr holes are made in the pterion above and below the lower wing of the sphenoid bone and the bone between them is flattened with a burr. A frontotemporal bone flap is cut with a vertical saw that includes the temporal muscle cuff. An additional hole below the upper temporal line may be helpful for this purpose. A free bone flap is lifted elevating and breaking down the bone.

In the event of tumors that infiltrate the pterional bone or the external third of the sphenoid wing, it may be necessary to make the craniotomy around the involved bone, which is then removed by drilling or with a bone gouge. This is a pathological bone with reactive hyperostosis and/or tumor infiltration that must be removed, sometimes with profuse vascularization.

Extradural Stage

The objective of the extradural phase is the devascularization of the tumor and the resection of the pathological bone. To do this, the entire vertical portion of the greater wing of the sphenoid bone is drilled out and the middle meningeal artery (MMA) is coagulated and sectioned at its entry in the middle fossa through the foramen spinosum. The lesser wing of the sphenoid bone is also drilled out until the beginning of superior orbital fissure (SOF) ( Fig. 1C ). At this point the recurrent meningeal branch (RMB) of the ophthalmic artery is coagulated and divided. Then, the ACP can be total or partially removed extradurally. Even if the clinoidectomy is not complete, the splitting of the dura from the bone must be completed since this maneuver detaches the tumor from all the vascular supplies.

Intradural Stage

The dura mater is opened around the pterion and the tumor is exposed, which in the case of large tumors and regardless of its insertion point, widens the sylvian fissure. Therefore, the tumor is usually visible when the dura is opened. The arachnoid is divided around the tumor and the sylvian fissure is opened generously ( Fig. 2A ). The tumor is found devascularized and using the ultrasonic aspirator is quickly debunked intracapsularly ( Fig. 2B ), leaving around a capsule approximately 0.5 cm thick. The tumor capsule is initially separated from the temporal lobe and portions of capsule are sectioned to mobilize it and facilitate separation, dissection and further resection. At the level of the sylvian fissure the dissection of the capsule must be very careful since the arterial vessels may be attached to the capsule or embedded in the tumor. The study of MRI in T2 sequence can be helpful to recognize this situation. The perforating branches of segments A1 and M1 and all those that enter the anterior perforated substance must be carefully identified and respected. The optic nerve is then identified and separated from the tumor. Finally, the frontal portion of the tumor capsule is removed ( Fig. 2C ).

Fig. 2.

Left transzygomatic approach. ( A ) The dura mater has been opened after the extradural stage and the arachnoid has been opened from the sylvian fissure, exposing the tumor (SWM) between the temporal lobe (TL) and the frontal lobe (FL). A segment of the middle cerebral artery (mca) is seen. The dura mater is infiltrated by the tumor (*). ( B ) With the ultrasonic aspirator the tumor has been partially removed. ( C ) The intradural tumor has been resected and the internal carotid artery (ACI) and its main branches are exposed, as well as the oculomotor (III), optic (II) and the olfactory (I) nerves. The dura mater of the middle fossa (fm) has been exposed. ( D ) Completed the extradural and intradural stages the duramater is ready to be removed from the skull base of the anterior fossa and middle fossa and the lateral wall of the cavernous sinus.

Now the dura mater of the middle and anterior fossa is completely detached from the bone, and it is removed together with the remains of the tumor that attached to it. Finally, the 'peeling' of the dura mater of the lateral side of the cavernous sinus is done, upwards beginning on the third nerve root of the trigeminal nerve or downward beginning below the SOF ( Fig. 2D ). By separating the tumor from the lateral side of the cavernous sinus, there may be defects that bleed profusely and they must be obliterated with fibrin glue. When there is a true infiltration of the cavernous sinus wall by the tumor it is advisable to left this portion in place. Remnants may remain around the carotid or optic and oculomotor nerves that are carefully removed in small pieces ( Fig. 3 ). Again, if a true infiltration of the ICA or great vessels adventitia is found, it is judicious to leave in place these remains. Once the tumor and dura have been removed, the resection of the bone rests of the sphenoid wing or ACP is completed. Portions of dura that cannot be resected are coagulated profusely.

Fig. 3.

( A ) Intraoperative view of a tumor that displaces the optic nerve (II) and wraps the internal carotid artery to its bifurcation in the middle (mca) and anterior (aca) cerebral arteries. ( B ) After resection, the entire course of the internal carotid artery (ICA) and the oculomotor (III) and optic (II) nerves have been freed. Some remnants of the anterior clinoid process (acp), hyperostosic and infiltrated by the tumor remain to be removed.

Reconstruction and Closure

After the resection, a local vasodilator is applied to the arterial vessels. The dura mater is closed with the help of a large dural graft that is cranially sutured to the dura mater border and applied covering the floor of the middle and anterior fossa. The bone flap that is fastened with cranial fixators is repositioned. In the case of bone resection, an acrylic cement cranioplasty is constructed to cover the bone defect. The temporal muscle is sutured in the musculoaponeurotic cuff left on the bone flap. The zygomatic arch is repositioned with miniplates. The soft tissues are sutured in layers. No drainage is left.

Results

The series includes a total of 26 patients; 15 women and 11 men. The average age of the patients was 62.4 years, with a range between 34 and 81 years. The lesion was in the left side in 16 patients and in the right in the remaining 10. Two patients had been operated on for another intracranial meningioma and another 3 had other intracranial meningiomas at diagnosis. Two tumors had undergone very partial resections by other surgeons, showing growth in the follow-up imaging controls. The tumor of a patient had received radiosurgery, again with tumor growth. The average size of the lesion was 81.7 cc (range between 27 and 260 cc), calculated by the formula a.b.c/2, with a, b, and c being the largest diameters of the tumor in the three axes. This volume corresponds to a sphere of around 5.45 cm in diameter. 15 tumors showed a relevant perilesional cerebral edema with midline displacement in 16 patients (range between 3 and 11 mm). In the T2 sequence, the contact of the great vessels (ICA, M1, M2, A1) with the tumor capsule was observed in 16 cases, the invagination in the tumor in 7 cases and a total encasement in 3 cases. Some tumors had extension to the anterior fossa, middle fossa, cavernous sinus, optic canal, or extracranial temporal fossa.

Hyperostosis of some portion of the sphenoid wing in 15 cases and of the ACP in 8 cases was evidenced in the CT study. In 1 patient there was an invasion of the bone with extension of the tumor to the temporal muscle. One patient had both ACPs pneumatized by paranasal sinus extensions. Two patients were studied only with CT, 1 for morbid obesity and another for carrying a cardiac pacemaker. Only 2 cerebral angiographies were performed, showing vascular displacements by the tumor and pathological vascularization in 1 of them, without signs of true stenosis or deformity of the ICA. The clinical symptomatology at diagnosis was headache and other subtle symptoms suggestive of increased intracranial pressure in 10 patients, a first seizure fit in 7 patients, contralateral hemiparesis or speech disorder in 4 patients and loss of unilateral vision in 5 with a complete visual loss only in 1 patient. In 5 patients, diagnosis was done in imaging studies performed by studies unrelated to neurological symptoms or irrelevant causes. Some of these tumors were subjected to observation, being referred to surgery when the growth of them was evidenced.

Complete resection of the tumor was achieved in all cases. These are cases where there is no tumor remnant in the MRI of the 6 months. In all these cases the surgeon's perception was also that the removal had been complete, including resection of the underlying dura mater and removal or drilling of the bone of the lesser wing of the sphenoid bone with hyperostosis or tumor invasion. One patient showed some postoperative T1 gadolinium enhancement in the cavernous sinus suggestive of tumor remnant. This supposed remnant was left under observation without any growing during follow-up. The average duration of operating room neurosurgical time was 5.3 hours, with a range between 2.3 and 9 hours. No patient required intra or postoperative blood transfusion. There was no postoperative mortality. An incidental aneurysm of the anterior communicating artery complex was uneventfully clipped after the resection of the tumor in a patient.

Some patients suffered one or more relevant postoperative complications. Two patients were immediately reoperated in the postoperative period due to subdural hematoma and intracerebral hematoma. Two patients presented cerebrospinal fluid fistula, solved respectively with a temporary external lumbar drainage and ventriculo-peritoneal shunt. A transient paresis of oculomotor cranial nerve was observed in 4 patients. One patient suffered atrial fibrillation and another patient had nosocomial pneumonia. Two patients suffered cerebral ischemia. In one of them the cause was hemodynamic due to an intraoperative hypotension event, with dysphasia and good clinical evolution in a few weeks. The most serious complication was a malignant infarction of the middle cerebral artery that requiring decompressive craniectomy, with permanent hemiplegia.

All tumors were grade I of the WHO except 3 cases with grade II. The grade II cases were followed with the same protocol of observation by imaging, although every 6 months instead of yearly. No recurrences were identified in any case. The minimum follow-up was 3.5 years, with an average of 5.8 years. Two patients died earlier from causes unrelated to the procedure. In a patient with other intracranial meningiomas, the growth of 3 small lesions was observed in the middle fossa, outside the original tumor implant.

Discussion

Large SWMs are formidable lesions and they are a challenge for their radical and safe microsurgical removal. Although the SWMs were classically divided into 3 subgroups based on anatomical and clinical considerations and with subsequent implications in the neurosurgical approach, the fact is that they can currently be considered as a single entity from a neurosurgical point of view.

Cushing and Eisenhardt ⁵ divided the sphenoid ridge into 3 more or less equal portions and classified the sphenoid ridge meningiomas into 3 subtypes according to their origin in the inner third (deep, medial or clinoid meningiomas), middle third (alar meningiomas) or external third (pterional meningiomas) of the sphenoid ridge. For these authors the sphenoid ridge is the edge that separates the anterior fossa from the middle fossa in an endocranial view of the base of the skull covered with the dura mater. Therefore, the sphenoid ridge is not anatomically the same that the lesser wing of the sphenoid bone.

The SWMs can grow at any point of this ridge and the potential growing places continue medially with the meningiomas of the cavernous sinus and optic nerve canal, laterally with the meningiomas of the convexity, anteriorly with the meningiomas of the roof of the orbit and caudally with the meningiomas of the floor of the middle fossa. On the other hand, in their growth, true SWMs can extend to the anterior fossa, middle fossa, optic canal, orbit, paranasal sinuses, cavernous sinus and temporal fossa and infratemporal regions. The bone is often thickened by hyperostosis or infiltration of the tumor, and sometimes eroded by the growing.

All these multiple growing and growth forms explain the different clinical syndromes classically described. It is patent that more medial tumors have a higher risk of producing visual symptoms by compression of the optic nerve or optic pathway or by infiltration of the optic canal. It also seems clear that more lateral tumors hit more frequently with epilepsy. Finally, the meningiomas of the middle third, which block the venous drainage of the orbit through the SOF, and those that grow ‘en plaque’ or produce hyperostosis are more likely to produce exophthalmos. However, since the popularization of MRI techniques, the diagnosis of SWMs, and in general of most intracranial lesions, is made by imaging and not by clinical syndromes. In the diagnostic work up of the developed countries, a first neurological symptom avails to make an image study that leads to the diagnosis. Many tumors are true serendipities. In our series, most of the lesions were diagnosed quickly after the first neurological event and some were true findings. In a few cases it was already known tumors and observed by image where tumor growth was evidenced. However, it is striking that a high percentage of SWMs are of large size at diagnosis.

Another common feature of SWMs is that they are generally grade I tumors of the WHO and the same is true for most of the meningiomas of the skull base. This fact justifies the recommendation made by Sughrue et al ⁶ to be no aggressive in the resection of small tumor remnants in skull base meningiomas if this significantly increases the risk of neurological deficits for the patient. However, recently Voss et al ⁷ show that Simpson's grading and extent of resection were not similarly important in all locations. It is remarkable that the recurrence in the SWMs has been reported in the literature as very high. In medial meningiomas the recurrence rate at 10 years varies between 30% and 100% of cases, where the rates of complete resection of the meningiomas of this location are very low in the historical series, between 60% and 85%. ⁷

On the other hand, the new 2016 WHO classification modifies the diagnostic criteria for grade II meningiomas and with these criteria some grade I tumors have one or more pathological criteria for grade II meningiomas. ² In our series, cases of grade I meningiomas showing a rapid growth is not exceptional, as well as regional recurrences or the synchronic presence of multiple lesions. This shows that the grade is not a single prognostic factor in the recurrence either.

Many other factors involved in meningioma recurrence, in particular pathology, location, receptors and other subtle biological prognostic factors, could explain how meningiomas with the same WHO prognostic grade or Simpson's resection grade have different behavior. Our group works to identify biological prognostic factors of aggressive behavior in grade I meningiomas. ⁸ Unfortunately, the surgeon could not control any of these alleged prognostic factors, excepting by removing as radically as possible the tumor, including a macroscopic complete removal of the tumor along with the associated dura mater surrounding the tumor implantation and underlying bone. This is the only thing the surgeon can do to reduce the risk of recurrence and he/she must try to do it.

This goal of microsurgical treatment with maximum resection must be done while guaranteeing maximum safety for the patient. To this end, we have developed the described treatment strategy, which is based on the knowledge of microsurgical anatomy, the use of skull base approaches and techniques, and the systematization of the surgical procedure.

The systematization of the treatments allows a greater experience in the surgical team and it is demonstrated that there is a relationship between the case-volume and the favorable clinical results in the management of complex surgical procedures and diseases. In an administrative database study of the results of the surgical treatment of intracranial meningiomas, it was shown that the patients treated at high case-volume centers and by high case-volume surgeons had significantly lower rates of in-patient mortality and adverse outcome. ⁹ All this leads to design an everyday surgical work up for all of SWMs avoiding variability.

Skull base techniques improve the approach and facilitate bone resection and tumor devascularization but increase the complexity of the approach. However, the zygomatic osteotomy of our approach is very simple and fast. Others recommend more complex osteotomies or doing excessive surgery. Langevin et al ¹⁰ recommend a pterional transzygomatic approach similar to ours, but their zygomatic osteotomies are by far much more complex. Sughrue et al ¹¹ report the use in 36 cases an orbito-zygomatic (OZ) craniotomy to remove tumors of any location in their series of 56 SWMs, even in 9 tumors out of 13 arising from the lateral third of the sphenoid wing actually not very large. In our experience it has not been necessary to use these large or complex surgical approaches even for huge or medial tumors.

The fundamental steps of the surgical strategy described are the following: 1. Extradural tumor devascularization and resection of the hyperostotic and/or infiltrated bone and then intradurally; 2. Intradural tumor debunking; 3. Microdissection of MCA branches, and M1, A1, and perforators from the capsule; 4. Identification of the optic and oculomotor nerves and ICA; 5. Tumor capsule dissection and resection; 6. Dural resection or cauterization; 7. Dural and bone reconstruction and closing.

The resection of the zygomatic arch allows a wide exposure of the temporal fossa and a flat approach to the floor of the middle fossa. ^{3 4} The craniotomy allows a broad exposure of the frontal lobe and the temporal lobe, as well as the sylvian fissure in the center of the surgical field, occupied by large tumors regardless of their implantation point. This exposure allows, in the intradural phase, the removal of the tumor without any cerebral retraction.

Nevertheless, the extradural phase is crucial in the approach. First, the extradural phase allows drilling of the lesser wing of the sphenoid bone and the ACP, which provides a valuable additional space. A high percentage of meningiomas of the sphenoid wing are associated with bone hyperostosis. The cause of hyperostosis in meningioma has long been a matter of debate. As early as 1934, Echlin ¹² suggested a direct association between hyperostosis and tumor invasion of the bone. This topic has been discussed along decades. In a recent study, Goyal et al ¹³ show the presence of tumor cells in the bone underlying to meningioma in 20% of the cases and in cases showing hyperostosis, the tumor cells were present in 23.3% of cases. These results indicate that tumor invasion into the bone is present in a significant number of patients with meningioma, especially those showing hyperostosis. Therefore, to achieve higher Simpson's grade of tumor excision, surgeon should remove as much bone in contact with the tumor as possible whenever feasible, particularly the hyperostotic areas.

The most recondite portion for resection is the ACP. ¹⁴ When it comes to a clinoid without hyperostosis, extradural clinoidectomy is easy and direct, as well as the decompression of the optic canal. When the clinoid is hyperostotic it can be difficult to complete clinoidectomy extradurally. In this case it is resected as much as possible and the dura mater is separated from the bone surface to help the tumor devascularization. The clinoidectomy and optic canal decompression can be completed intradurally after the resection of the bulk of the tumor and the dura mater.

Second, the extradural phase achieves tumor devascularization. The SWMs are vascularized by branches of the external carotid artery through the MMA. The RMB also usually participates in the vascular supply. Here, it is the ophthalmic artery that supplies the territory of the MMA through its RMB, actually a branch of the lacrimal artery. This branch projects posteriorly through the SOF or through its own foramen. On occasion, it assumes supply of middle meningeal territory, via a meningeal branch of the sphenoid ridge running along the ridge. These 2 vascular contributions are taken and divided during the extradural phase of the approach, the MMA in the foramen spinosum and the RMB in or near of the SOF. The detachment of the dura mater from the bone allows eliminating multiple small direct vascular supplies. The tumor is almost completely devascularized, leaving only small arachnoid contributions or exceptionally recruited branches of small cortical arteries. For years, preoperative embolization of meningiomas was recommended to reduce intraoperative hemorrhage. ^{15 16 17} At present the interest in embolization has been reduced ¹⁴ and in our opinion, the best way to avoid intraoperative bleeding of SWMs is to eliminate extradurally the vascular supply early before resection.

A possible cause of complications is the encasement of the vessels by the tumor that increases the risk of vascular injury, which can produce intraoperative catastrophic hemorrhages or postoperative ischemia. In our study, we assessed the degree of encasement in the T2 sequences with the aim to evaluate the surgical complexity expected. McCracken et al ¹⁸ did a more exhaustive study of encasement with the MPRAGE (Magnetization Prepared Rapid Acquisition Gradient Recalled Echo) sequence of MRI and establish that 90% of patients have some degree of preoperative encasement, and in 76% of patients some degree of postoperative ischemia with a statistically significant correlation in the univariate analysis. We have found some degree of invagination of the vessels in the tumor in 7 out 26 cases and a complete encasement in 3 out 26 cases. However, none of our 2 patients with postoperative ischemic complications had invagination or encasement of the intracranial vessels in the tumor.

McCracken et al ¹⁸ also describe a 'deadly triad', when there is a 360 ° encasement of the supraclinoid ICA and of the segments M1 and A1. In this case they recommended a subtotal resection of the tumor. However, the imaging encasement does not mean that it is impossible to free the artery from the tumor. In this sense we consider more important the conservation or not of the arachnoid plane around each vessel as proposed by AlMefty. ¹⁹ In our opinion the lack of an arachnoid sheath is the only factor that prevents a complete release of the arteries of the tumor and this can only be determined during surgery. In our experience all great vessels were freed from tumor adhesions and no remains were seen in postoperative MRI studies.

Extensions to cavernous sinus, optic canal and other areas are resected using standard microsurgical techniques. The invasion of the cavernous sinus is usually in the form of globular masses, which can be resected in bloc with the 'peeling' of the dura mater of the lateral wall of the cavernous sinus. When it comes to a real infiltration of the tumor in the sinus wall or involving the cranial nerves, the recommendation is to leave a tumor remnant for observation or eventually treat with radiosurgery in case of growing.

The resection of the dura mater around the base of the tumor implantation and, if this is not possible, its coagulation is also important in the reduction of the risk of recurrence of the meningiomas. Decades ago Borovich and Doron ²⁰ already demonstrated in 1986 that there are macronodules of meningothelial cells, of 1 to 8 mm Ø, up to 3 cm of the tumor implant, as well as microscopic islets of intradural meningothelial cells around the meningioma. In SWMs, the resection of the dura should include the extensive removal of the dura mater from the frontal and temporal convexity, from the base of the anterior fossa, from the entire sphenoid wing and ACP, and from the lateral aspect of the cavernous sinus.

Our study has some limitations. The surgical procedure we describe is the result of the personal evolution of the personal experience of the author, although all patients included in the study have been operated on using the same protocol. The surgical technique described is particularly useful in large tumors, where all the steps described are necessary. It is also true that smaller tumors can be managed with more restricted approaches, such as the regular pterional approach and the limit of size of 3 cm is not absolute in any case. The clinical results are also the product of the surgical skills of the author and the series has not a control group.

In conclusion, large SWMs can be considered as a single pathology regarding the surgical approach and intraoperative microsurgical procedures strategies. The pterional transzygomatic approach allows an extradural devascularization of the tumor and an extensive bone resection, very often invaded by tumor nets, that facilitates the intradural stage of tumor resection. Angiography and embolization are no longer recommended in the preoperative work-up. The proposed approach allows a wide and radical resection of the dura mater and bone that increases the Simpson's grade. Surgery does not control other biological or molecular prognostic factors involved in tumor recurrence.