Optimizing Endoscopic Approaches for Sphenoid Lateral Recess Cerebrospinal Fluid Leaks: Proposal of a New Algorithm.

Manogaran Ravi Sankar, Das K Nidhin, Keshri Amit, Jaiswal Awadesh Kumar, Srivastava Arun Kumar, Mehrotra Anant, Das Kuntal Kanti, Bhaisora Kamlesh, Angrish Prerna, Subramaniam Dheeneshwar, Nehra Sukriti, Raj Reshma, Kumar Sanjeev, Roy Shrimoyee, Hameed Nazrin, Chidambaram Kalyan, Aqib Mohd, Sinha Mohit, Virk Ramandeep Singh

Neuro-Otology, Department of Neurosurgery, SGPGIMS, Lucknow , Uttar Pradesh , India.

Department of Neurosurgery, SGPGIMS, Lucknow , Uttar Pradesh , India.

Oper Neurosurg. 2025 Jul 3. doi: 10.1227/ons.0000000000001694.

BACKGROUND AND OBJECTIVES

Sphenoid lateral recess cerebrospinal fluid (CSF) leaks are a distinct subset of spontaneous CSF leaks, often linked to idiopathic intracranial hypertension. Despite their clinical significance, a structured management algorithm tailored for them is lacking. This work proposes a comprehensive algorithm for managing surgical interventions in sphenoid lateral recess CSF leaks, which is based on radiological classifications and anatomic parameters.

METHODS

This ambispective study, conducted from 2017 to 2024, integrated cadaveric dissections with clinical data from 33 patients presenting with sphenoid lateral recess CSF leaks. Cadaveric models were used to explore the feasibility of transcaruncular approach to lateral recess of sphenoid with respect to V 2 -Vidian distance (VVD). Spiral computed tomography scans with 1-mm slices provided a precise classification of defects. Radiological and intraoperative parameters were analyzed, including the VVD. Patient outcomes and surgical difficulty, scored using a Likert scale, were statistically evaluated.

RESULTS

The study classified defects into 4 types: Type I (0%), Type II (30.3%) with 12 cases, Type III (42.4%) with 14 cases, and Type IV (27.3%) with 7 cases. The mean VVD was 6.28 ± 1.9 mm, significantly influencing surgical difficulty. Modified transpterygoid and transpterygoid approaches were predominant for Types III and IV. Cases with VVD <4 mm had higher rates of Vidian nerve and sphenopalatine artery sacrifices ( P < .05). Surgical success was 100%, with no CSF recurrence. Complications included transient cheek paresthesia in 12% of patients, resolved on follow-up.

CONCLUSION

Based on defect classification and anatomic parameters, the proposed algorithm optimizes the surgical management of sphenoid lateral recess CSF leaks. The study underscores the role of radiological evaluations in guiding approach selection and highlights the importance of preserving neurovascular structures to achieve improved outcomes.