Summary Statement. Pedicle screws provide robust fixation and rigid immobilization. There has been no attempt to correlate the anatomic dimensions of thoracic and lumbar pedicles with the accuracy of navigated insertion. This study demonstrates that comparable accuracy using this technique. Introduction. Pedicle screws provide robust mechanical fixation, which makes their use attractive; their use enables fixation of the three spinal columns. There remains concern about the potential both for misplacement; various investigators have studied the accuracy of pedicle screw insertions, comparing different techniques. What is not clear, however, is whether there is any relation between the variables of pedicles’ anatomic dimensions, screw dimensions and accuracy. This study aims to elucidate the relationship between these variables. Patients & Methods. We conducted a retrospective review of consecutive pedicle screws that were inserted in the thoracic and lumbar spine at our institution. Screws were inserted using the navigated method (Stealth Station® TREON™, Medtronic, Louisville, CO). The accuracy of the screw insertion was measured using the classification system developed by Gertzbein and Robbins; pedicle dimensions were measured from post-operative computed tomography scans. The corresponding pre-operative scans were then used to measure the pedicle dimensions at the other levels. The magnitude of a cortical breach in the pedicle was represented by a letter: A (no breach), B (<2mm), C (>2mm, <4mm), D (>4mm, <6mm) and E (>6mm). In addition, measurements were made of the anatomic dimensions of the pedicles. The combination of these two measures allowed for direct correlation to be made between the accuracy of screw insertion, screw dimensions and pedicle anatomy. We then computed the proportion of each pedicle (width) occupied by a screw. Results. A total of 765 screws were reviewed, 493 were in the thoracic spine and 272 in the lumbar spine. Of the screws in the thoracic spine, 472 (96%) were either fully in the pedicle or less than 2mm beyond the cortex (within the A+B classification); when considered separately, 323 (66%) were completely within the pedicle (A) and 149 (30%) were less than 2mm beyond the cortex. A total of 21 (4%) screws were beyond 2mm but within 6mm (C+D). In the lumbar spine, 270 (99%) were either completely within the cortex or less than 2mm exposed (classified as A or B). The nadir of pedicle width was at T4. From L1 to L5, measured pedicle width also rose. This pattern was followed, although it was less profound, when screw diameter was measured in the lumbar spine (and even less so in the thoracic vertebrae). The height of pedicles was noted to progressively increase, peaking at the thoraco-lumbar junction. The mid thoracic region was associated with screws occupying the greatest proportion of pedicle diameter. Discussion/Conclusion. The use of pedicle screws in the thoracic and lumbar spine remains relatively safe. The accuracy of navigated insertion was found to compare well to previous series’. Although there is some association between the anatomical dimensions of pedicles and the dimensions of screws, this doesn't seem to be a strong association. Based on the findings in this series, future studies that relate the long term outcome (e.g. failure or screw loosening) with proportion of pedicle diameter taken up by a screw may be warranted
