Lateral Transorbital versus Endonasal Transpterygoid Approach to the Lateral Recess of the Sphenoid Sinus: A Comparative Anatomical Study

Background The lateral recess of the sphenoid sinus (LRSS) is a common site of spontaneous cerebrospinal (CSF) leak with or without concomitant meningoencephalocele. Its treatment faces technical difficulties due to the deep location of the osseous defect, commonly at the anterolateral triangle of the middle fossa, between the maxillary (V2) and mandibular (V3) branches of the trigeminal nerve. Traditionally, a transcranial middle fossa approach has been recommended, but over the last decade, the endoscopic endonasal transpterygoid approach (EETA) has been introduced as a more direct and less invasive surgical access. The EETA, however, may require transection of the ipsilateral vidian nerve (VN), which can cause decreased lacrimation and could have significant impact on patients with pre-existent V1 injury or dry eye. In this study, we investigated the feasibility of a lateral transorbital approach as an alternative pathway to the LRSS that avoids the injury of the VN.

Method Anatomical specimens with significantly pneumatized LRSS were preselected by inspecting their pre-dissection CT scan. Five lightly embalmed heads with intravascular colored-silicone injection were included for dissection on five sides (three left). Neuronavigation was employed to assist with the procedures and to take measurements. The lateral transorbital approach was completed first and then the EETA was performed for comparison. The two approaches were compared on working distance and neurovascular structures being sacrificed.

Results The lateral transorbital approach was divided into three stages: (1) Lateral orbitotomy: a 1.5-cm incision was made at the lateral canthus followed by subperiosteal dissection exposing the lateral orbital rim, the temporalis muscle was detached from the lateral orbital rim and sphenoid wing, and a lateral orbitotomy was then performed; (2) Infratemporal fossa exposure: the temporalis muscle was dissected inferiorly and detached to expose the inferior orbital fissure and infratemporal crest; (3) Middle fossa floor drilling: the bone of the infratemporal crest was drilled under endoscopy to advance medially and posteriorly following the inner cortical bone of the middle fossa floor in an extracranial route to reach the LRSS. The average working distance was 51.24 (± 3.84) mm for the transorbital approach and 71.66 (± 5.73) mm for the EETA (p = 0.0002). The VN and sphenopalatine artery were encountered and sacrificed during the EETA in all five specimens. No neurovascular structures were found in the trajectory of the transorbital approach, and minimal retraction of the orbit was required.

Conclusion The lateral transorbital approach to the LRSS is technically feasible, with minimal risk of injuring any neurovascular structures, such as VN and V2. It also offers a shorter working distance when compared with the EETA. The transorbital approach, however, offers no options for vascularized mucosal flap reconstruction but immediate access to muscle grafting. The lateral transorbital approach, as proposed here, may serve as an alternative to the EETA for treating spontaneous CSF leaks of the LRSS when preservation of vidian nerve is a major concern.