This study was approved by the Institutional Review Board of corresponding Pohang Semyeng Christianty Hospital, and informed consent was obtained from all study participants. This study was a retrospective review of medical records and radiologic images. Between March 2013 and June 2014, 32 patients (64 screws) who underwent PS insertion in C7 using our technique were enrolled. All operations were performed under a surgical microscope by a single surgeon (the corresponding author), without the aid of fluoroscopy.

This study was conducted only on the patients who complied fully with the following inclusion and exclusion criteria. The inclusion criteria were: (1) patients with compression of cervical spinal cord due to multilevel cervical spondylotic myelopathy or ossification of posterior longitudinal ligament, which were diagnosed using cervical spine radiographs and magnetic resonance imaging (MRI) that corresponded to clinical manifestations and physical examinations; and (2) patients who underwent multilevel posterior cervical fusion with PS fixation at C7 by author's technique. Those who met any of the following criteria were excluded: (1) infection or tumors in the cervical spine; (2) patients with hemorrhagic disorders, such as hemophilia and thrombocytopenia; and (3) patients judged unsuitable by the corresponding author.

All operations were performed with Mayfield tongs, with the patient in the prone position. A typical midline skin incision was made over the affected level of the cervical spine. In general, screw insertion was carried out after the main procedure, which included posterior decompression, decortication, and bone grafting to the facet joint. Prior to preparation and insertion of the PS in C7, careful periosteal dissections were performed to better expose the cortical bone and C6–7 facet joint. To achieve optimal placement of the PS in C7 without fluoroscopy using our technique, three critical elements were determined preoperatively: (1) the determination of the ideal PS entry point; (2) the convergent angle distance (CVAD); and (3) the cranial-caudal angle distance (CCAD), which were preoperatively determined with an Infinitt picture archiving and communication system (Bracknell, Berkshire, UK) and V-works three-dimensional (3D) simulation software (Cybermed Inc., Reston, VA, USA) based on preoperative fine-cut CT images. First, to choose the ideal PS entry point, a bony mark near the C6–7 facet joint, such as a tiny groove or protruded osteophyte, was initially identified at the 3D-simulated images based on preoperative CT images. Most cases had suitable bony marks near the facet joint (Fig. 2). By identifying the relationship between the bony marks and the ideal entry point and preoperatively simulating the ideal point, the actual position of the ideal entry point was identified in the surgical field. Second, the CVAD was measured preoperatively using preoperative axial CT images, which was defined as the distance between the tip of C7 spinous process and the extended line passing through the pedicle axis at the same height (Fig. 3). Finally, the CCAD was also measured preoperatively at sagittal CT images, which was defined as the distance between the tip of C7 spinous process and the extended line passing through the pedicle axis at the same height (Fig. 4). To validate the values of the CVAD and CCAD in the actual surgical situation, determination of the two distances was facilitated by the use of strict instrument such as K-wire, which aided to identify the distance from the spinous process tip of C7 to the ideal pre-determined screw trajectory measured preoperatively, as described above (Fig. 5). The ideal entry point was marked with a 1.8 mm burr, which was then advanced to a depth of approximately 1 cm with the preservation of the two pre-determined distances (Fig. 5). The determination of the ideal entry point for PS, CVAD, and CCAD can be carried out more easily with 3D-simulated images and preoperative multidirectional CT images. After verifying the absence of a cortical breach by using a ball-tipped probe, the hole was deepened for PS placement using a straight Lenke probe, which preserved the pre-determined angles. After carefully confirming the absence of a cortical breach with a ball-tipped probe in the direction of the pedicle, the PS was inserted under a surgical microscope, but without aiding fluoroscopy (Fig. 5).

Demographic, clinical, and radiologic data were retrospectively collected from medical records and radiologic images. Radiologic evaluation was performed postoperative cervical radiographs and 1.0-mm interval CT scans using a Mx8000 IDT device (Philips Medical Systems, Best, Netherlands). Special attention was paid to the presence of a cortical breach of the PS, or the encroachment of the PS into the spinal canal or the vertebral foramen (Fig. 6). A cortical breach was defined as any violation of the exterior margin of the cortical bone at the pedicle of C7. The location of a cortical breach was classified as medial, lateral, superior, or inferior to the pedicle. The following data were collected for each operation: whether or not the PS invaded the exterior margin of the cortical bone of the pedicle in any direction, and the relationship between the encroachment of the PS and its postoperative clinical manifestations. All radiologic images were analyzed by one surgeon who was not otherwise involved in this study.

Of the 32 patients (64 screws) enrolled in this study, 15 were men and 17 were female. The mean patient age was 67.8±13.0 years (range, 38–86 years). Preoperative diagnoses included 18 cervical spondylotic myelopathies, 12 diagnoses of ossification of the posterior longitudinal ligament, and two diagnoses of cervical spondylotic radiculopathy. All patients underwent a posterior fusion operation with a screw-rod system, including PS insertion into C7, by the author's technique.

Of 64 screws, cortical breach was observed in three screws, including two screws in the right PS and one screw in the left PS. The three breaches occurred in the first five patients. Outside of the initial placements, there was no cortical breach in any cases. The three breached screws violated only the lateral cortex of the pedicle. Postoperative neurologic deterioration was not observed in any of enrolled patients, including the three screws with a cortical breach.