Introduction: Endonasal placement of spinal instrumentation is a novel topic of investigation that may eventually eliminate the need for separate-stage posterior stabilization following odontoidectomy. While previous investigative and clinical studies have reported methods for endonasal C1-2 stabilization following odontoidectomy, currently no such method for stabilization of the craniocervical junction (O-C1) through the endonasal corridor exists; such a technique would require the ability to instrument both the clivus and C1. This study sought to investigate the feasibility of anterior C1 (AC1) screw placement through the endonasal corridor for purposes of craniocervical stabilization.

Methods: Eight adult, injected cadaveric heads were studied for the placement of 16 AC1 screws through the endonasal corridor. The heads were registered with Medtronic®S7®System using thin-cut CT imaging. The atlanto-axial joint was exposed by turning a standard inferior U-shaped nasopharyngeal flap. With neuronavigation guidance, 4mm x 22mm A1C screws were placed in the C1 lateral mass using a zero-degree driver. Post-placement CT scans were obtained to determine: (1) site of entry, as measured from the anterior tubercle of C1, (2) screw angulation in axial and sagittal planes, (3) proximity of screws to the nearby central canal, foramen transversarium, and other critical structures.

Results: Mean (SD) site of entry of entry was 16.41 (3.00) mm lateral, 2.20 (1.33) mm rostral, and 5.70 (1.83) mm deep to the anteriormost portion of the C1 ring. Mean (SD) angulation in the axial plane was 19.08 (5.88) degrees lateral to the midline, as measured at the level of C1. Mean (SD) angulation in the sagittal plane was 13.86 (8.09) degrees inferior to the palatal line, as measured from the hard palate to the opisthion. The foramen transversarium was partially breached in 2 of 16 screws; the central canal was not violated by any screws; the O-C1 joint space was violated by 1 of 16 screws. Mean (SD) minimum screw distances from the unviolated foramen transversaria and central canal were 2.32 (1.33) and 4.00 (1.79) mm respectively. Bicortical purchase was achieved in 13 of 16 screws.

Conclusions: Anatomic placement of anterior C1 screws through the endonasal corridor using snapshot neuronavigation is feasible. Given consistently small intervals from screw trajectory to the foramen transversaria, central canal, and O-C1 joint space, development of methods to facilitate real-time neuronavigation during screw placement should be investigated. Future biomechanical studies are necessary to evaluate the stability of AC1 screws and ventral constructs compared to traditional posterior fusion methods.

Contact information: Daniel McGough, mcgougdp@mail.uc.edu, (310) 683-9889 Key words: craniocervical instability, C1 lateral mass, ventral fusion, endonasal, neuronavigation