The lateral supraorbital approach, doable and cosmetic access to anterior skull base

The anterior skull base region can be reached through multiple corridors. The most frequently used are the pterional, bifrontal, unifrontal, and orbitozygomatic approaches. These approaches are more extensive and time consuming. The lateral supraorbital approach is a less extensive frontal modification of the classic pterional approach designed to manage tumors and aneurysms of this region. To evaluate efficacy, safety, and cosmetic results of the lateral supraorbital approach and to present some tricks to have an easy access to different pathologies at the anterior skull base through this approach. Analysis of 50 operations for anterior skull base pathologies, vascular and neoplastic, made through the lateral supraorbital approach in Neurosurgery Department, Tanta University Hospitals between January 2013 and April 2017. The basic steps in performing the procedure are described. Special tricks required in individual cases and how to identify their need from preoperative images are discussed. We operated on 50 patients having 34 neoplastic and 16 vascular lesions. The mean incision length was 11 cm, mean bone flap size was 3.5 × 5.4 cm, and mean craniotomy time was 20 min. All patients were satisfied with their scars; there were no limitations to this approach regarding nature or size of the lesion. We had 8 morbidities (4 transient and 4 permanent) and 2 mortalities; all morbidities and mortalities were not related to the approach. The lateral supraorbital approach is simple, fast, and effective corridor to anterior skull base pathologies without significant morbidities or mortalities related to the approach.


Introduction
The anterior skull base extends from the orbital ridge to the anterior clinoids and sphenoid ridges. It can be reached through multiple corridors; the most frequently used are the pterional, bicoronal, unifrontal, orbitozygomatic, transnasal, and endoscopic endonasal approaches [1].
The idea of the lateral supraorbital approach appeared because most of classic anterior skull base approaches are time consuming, extensive, and associated with bad cosmetic outcomes; at the same time, the temporal lobe does not obscure the anterior skull base, and its exposure is not needed in most of the cases. This approach is a less extensive frontal modification of the classic pterional approach [2].
We used the lateral supraorbital approach that was first described by Broke and Dietz. This approach can be used for management of all anterior circulation aneurysms except the pericallosal aneurysm. Anterior and middle skull base meningiomas can also be reached through this approach. The term lateral supraorbital approach was used to name a different procedure from the traditional supraorbital macrocraniotomy described by Krause in 1908 and modified by Delashaw et al. in 1993 till the supraorbital keyhole approach developed by Axel Pernesky since 1980 with the help of the endoscope to overcome limitation of small craniotomy [3][4][5][6][7][8][9].

Patients and method
All patients were subjected to history taking, complete physical examination, and routine laboratory investigations. CT and MRI of the brain were obtained for tumor cases. CTA was done in vascular cases.
After careful studying of patient clinical status and images, talking with patient relatives, the patient or one of his close relatives signed informed consent then we proceeded to surgery.
Perfect neuroanesthesia ensures low intracranial pressure while maintaining adequate cerebral perfusion. To have a slack brain, we routinely use an epidural catheter to be inserted in the lumbar intradural space, connected to a haemoset collection bag and clamped to be opened when needed to release 50-100 cc of CSF to decrease intracranial pressure and allow surgery without brain retraction.
Perfect positioning is the first key to operate comfortably through the lateral supraorbital approach. Generally, the patient's head is fixed to a Mayfield frame with about 20°head elevation above heart level, mild extension, mild rotation, and mild tilt.
Fine adjustment of patient head should be tailored case by case according to the 3D imagination of target pathology and its relation to the skull base. The operating table can be rotated intraoperatively to have comfortable working angle.
We do minimal shaving 2 cm behind the hairline then we make a curvilinear frontotemporal skin incision that stops 1 cm above the zygoma (Fig. 1a).
We make a small incision in the anterior superior part of the temporalis muscle (Fig. 1b); this way decreases the risk of disfiguring temporalis muscle wasting and postoperative mastication problems, then one layer skin muscle flap is dissected from the periosteum, elevated, and retracted anteriorly by fishhooks and rubber bands attached to a long arm fixed to the operating table till the supraorbital rim and zygomatic arch are exposed (Fig. 1c). This diminishes the risk of injury of frontotemporal branch of facial nerve.
The fine adjustment of craniotomy position and flap size depends on the pathology. A single burr hole is made at the superior temporal line, and bone flap is cut using the sidecutting drill with footplate (Fig. 1d). Then, the flap is detached from the dura and lifted (Fig. 1e).
The sphenoid ridge is drilled using the high-speed drill with cutting ball tip then with diamond ball tip without saline irrigation (hot drilling) to control oozing from the bone.
The dura is opened in curvilinear incision pointing anterolaterally and dural edges elevated by multiple stitches to prevent oozing from epidural space (Fig. 1f).
From this point, all surgery is performed under the operating microscope. To minimize brain retraction, the first goal during intradural dissection is to release CSF from the basal cisterns (if can be reached early) around the optic nerve, internal carotid artery, and the lamina terminalis especially in ruptured aneurysm cases to release significant amount of CSF from the third ventricle.

Results
Our series included 50 cases: 34 males and 16 females. Their age ranged from 17 to 70 years with mean age of 35 years. All patients had anterior skull base lesions, 16 aneurysms and 34 neoplasms operated through the lateral supraorbital approach in Tanta University Hospitals from January 2013 to April 2017. Example cases are shown in Figs. 2, 3, 4 and 5.
The mean incision length was 11 cm, mean bone flap size was 3.5 × 5.4 cm, and mean craniotomy time was 20 min.
All of our patients had satisfactory cosmetic results with no temporalis atrophy and no injury to the frontotemporal branch of facial nerve (Fig. 1g).

Relation between lesion pathology and surgical outcome
All aneurysms were totally clipped except one case of a bilobed AcoA aneurysm where there was neck remnant; another surgery was performed, and one more clip was applied. Subtotal resection was done in 2 craniopharyngiomas, 2 tuberculum sellae meningiomas, and one olfactory grove meningioma due to solid consistency and adherence to neurovascular structures ( Table 2).

Clinical outcome
We had 14 cases with preoperative visual affection, six of them improved, three cases remained stationary, and 5 cases deteriorated (2 olfactory groove, 2 tuberculum sellae, and one craniopharyngima); one more case of olfactory grove meningioma that had preoperative normal vision developed postoperative field defect. This visual deterioration was transient in 4 cases and permanent in 2 cases. Two cases with ruptured aneurysms had vasospasm, but both were presented by extensive SAH and disturbed conscious level. One case of ruptured AcoA aneurysm developed hemiparesis due to ACA vasospasm. One patient with olfactory groove meningioma had postoperative CSF Fig. 1 Steps of lateral supraorbital approach on the left side. An oblique frontotemporal skin mark is placed behind the hairline after minimal hair shave (a). Skin and temporalis muscle are incised together (In the photograph the muscle is partially separated from the flap just for demonstration.) (b). The one-layer skin-muscle flap is retracted frontally (c). A single burr hole is placed just under the superior temporal line, and the bone flap is cut (d). Bone flap is detached, and sphenoid ridge is drilled (e). Dura is opened and tented with sutures (f). One-year postoperative image showing minimal atrophy of temporalis muscle and excellent cosmetic outcome (g) rhinorrhea that stopped spontaneously. One case of Pcom aneurysm had persistent 3rd nerve palsy and did not improve. One case had subcutaneous CSF collection that disappeared after single tapping and bandage. One case had periorbital edema that subsided spontaneously, and one diabetic case had superficial wound infection that improved with daily dressing and antibiotics.
The clinical outcome of our cases was evaluated according to Modified Rankin Scale and Glasgow Outcome Scale as shown in Tables 3 and 4, respectively.

Final outcome
One patient with ruptured AcoA aneurysm had persistent postoperative hydrocephalus, and a ventriculoperitoneal shunt was inserted; one case of craniopharyngioma had recurrence of the cyst, and an Ommaya reservoir was inserted. One case of ruptured AcoA aneurysm did not recover from his coma and developed persistent vegetative state. We had 2 mortalities; one carotid ophthalmic aneurysm died from carotid vasospasm, and one craniopharyngioma died from diabetes insipidus.

Discussion
We can classify anterior skull base approaches according to bone flap size into macrocraniotomy and minicraniotomy approaches. Macrocraniotomy approaches like the bifrontal, bicoronal, orbitozygomatic, and classic pterional approaches. Minicraniotomy approaches like the supraorbital keyhole approach. The bifrontal and bicoronal approaches are time consuming, extensive with complications related to large craniotomy flap, more blood loss, superior sagittal sinus violation, frontal air sinus opening leading to postoperative CSF leak, and postoperative behavioral changes from brain edema of bilateral brain retraction. They are sometimes associated with bad cosmetic results. The classic pterional approach of Yasargil was associated with temporalis muscle atrophy leading to postoperative mastication problems and bad cosmetic outcome due to complete subperiosteal temporalis dissection beside the risk of injury of frontotemporal branch of facial nerve [10][11][12][13].
Minicraniotomy approaches like the supraorbital keyhole approach had some limitations such as difficult application for vascular lesions due to difficult proximal control, impossible to operate in case of intraoperative aneurysm rupture, or brain swelling which occlude the surgical field. Also, difficulty to perform on large-sized neoplastic lesions, Hammad et al. operated on 22 patients in Ain Shams University, Egypt. They used the keyhole approach only in tumor sizes less than 4 cm. The eyebrow incision used frequently in this approach is historically associated with loss of supraorbital sensation or palsy of frontal branch of facial nerve [14,15].
We operated on 50 patients having anterior skull base lesions (16 vascular and 34 neoplastic) through the lateral supraorbital approach, a medium-sized approach, more sub-frontal, and less temporal exposure. This approach is a less extensive modification of the classic pterional approach. The craniotomy size was fair enough to decrease the risk of large craniotomy associated problems but at the same time sufficient to reach the entire anterior skull base [16].
Incomplete clipping was done in one bilobed AcoA aneurysm where there was remnant neck; another surgery was performed, and one more clip was applied. Subtotal resection was done in 5 tumors (2 craniopharyngiomas, 2 tuberculum sellae meningiomas, and one olfactory grove meningioma). This was not related to operating through the lateral supraorbital approach neither for vascular nor large neoplastic lesions. We believe that many other factors had affected the surgical outcome such as aneurysm morphology, tumor consistency, anatomical extension, and attachment to neurovascular structures.
Mean flap size in our series was 3.5 × 5.4 cm while in the series of Prof. Hernesneimi it was 3 × 5 cm because in the early beginning we used larger flaps due to lack of confidence and need of assistant hand in the field especially in aneurysm cases. In large-sized tumors, we preferred to extend the bone flap 1 cm posteriorly [8,9,13].
The lateral supraorbital approach takes short time; this shortens the overall operative time. Mean approach time in our series was 20 min while in the series of Prof. Hernesneimi it was 14 min because he is a well-known fast surgeon; he has some facilities that saves the time needed for hemostasis as the Sugita frame used for head fixation with strong fishhook retractors, using rani clips in skin edge hemostasis and fibrin glue in epidural space hemostasis [8,9,13].
We had one case of postoperative CSF rhinorrhea that stopped spontaneously. One case had subcutaneous CSF collection that disappeared after single tapping and bandage. One case had periorbital edema that subsided spontaneously, and one diabetic case had superficial wound infection that improved with daily dressing and antibiotics.
The cosmetic results were assessed according to the patient and surrounding personnel opinion in follow-up visit despite depending on cultural background and expectation before surgery; the scar was accepted in all patients.
We used a curvilinear skin incision situated behind the hair-line (better for healing and can be sutured intradermally with small invisible linear scar) with better cosmetic results. The skin incision stops 1 cm above the zygoma with less trauma to temporalis muscle. The frontotemporal branch of the facial nerve is safe in this approach because we do not do superficial dissection. The temporalis muscle is split only in its anterior and superior parts, which decreases the incidence of postoperative temporalis wasting and mastication problems. Our cosmetic results were similar to Prof. Hernesneimi because we used the same technique. We did not use the eyebrow incision, which is historically associated with loss of supraorbital sensation or palsy of frontal branch of facial   [12,17,18]. Sánchez-Vázquez et al. reported hypesthesia of the frontal region appeared in all patients; however, it disappeared completely in all patients as early as the 2nd month after surgery. Inability to raise the eyebrow immediately after surgery was a complaint in all patients; however, in five, the defect disappeared during the 1st month after surgery; in eight others, it disappeared after the 2nd month, and in the remaining patients, it disappeared completely after the 3rd month following surgery [19].
All morbidities and mortalities in our approach were related to lesion pathology or clinical condition of our patients, and there were no approach-related morbidities or mortalities.
The lateral supraorbital approach cannot be done in the presence of angry brain with high intracranial pressure so we paid all efforts to have a slack brain from perfect positioning, perfect neuroanasthesia, and CSF release either from the lumbar subarachnoid space or basal cisterns with minimal brain retraction. The Sylvian fissure appears at the posterior edge of the dural flap and can be easily opened when necessary [8,9,13].

Conclusion
The lateral supraorbital approach is a doable, cosmetic and effective corridor to anterior skull base pathologies without significant morbidities or mortalities related to the approach.

Limitations of the study
The number of cases is relatively small. A comparative study between different approaches for the anterior skull base could reveal more consistent results.