Thoracic pedicle screw insertion in Asian cadaveric specimen: does radiological pedicle profile affect outcome?

PURPOSE: Pedicle screw instrumentation has superior biomechanical as well as clinical outcome. Thoracic pedicles show great variation in different population groups, particularly in Asians who have been shown to have smaller pedicle dimensions. Although plain radiographs are widely performed prior to spine surgery, no studies have been done so far to investigate whether the thoracic pedicle profile on plain radiographs affect thoracic pedicle screw insertion. Therefore, this is a cadaveric study aimed to determine the relationship between plain radiographic thoracic pedicle profile in Asians and the outcome of pedicle screw insertion in the thoracic spine.
METHODS: A pre-insertion radiograph with an enlargement reference scale was performed and surgeons were blinded to the plain radiographic morphometry of the thoracic pedicles. From the pre-insertion radiograph, the normalized pedicle width and height (which controls for any magnification error) as well as the pedicle width:body width (PWBW) and pedicle width:pedicle height (PWPH) ratio was derived. 240 pedicle screws were inserted in ten Asian cadavers from T1 to T12 using the funnel technique. 5.0 mm screws were used from T1 to T6 while 6.0 mm screws were used from T7 to T12. Perforations were detected by direct visualization via wide laminectomies after pedicle screw insertion. The outcome of thoracic pedicle screw insertion was correlated with the radiological profile using independent t-test. Pearson correlation coefficient was used to investigate the correlation between the ratios and the normalised pedicle width and height.
RESULTS: The narrowest pedicle width is from T3 to T6 determined from normalized measurement of the pedicle width. T5 pedicle width is the smallest measuring 4.1 ± 1.3 mm. The overall perforation rate is 10.4% (25 perforations). There is only one significant perforation. There were twice as many lateral and inferior perforations compared to the medial perforations. 48% of the perforations occurred at T1, T2 and T3. Pedicles <4.0 mm in width and upper thoracic pedicles are risk factors for pedicle perforation. The normalised pedicle width has a high degree of linear correlation with PWBW and a normalised pedicle width of 4.0 mm correlated with a PWBW ratio of 0.3.
CONCLUSIONS: Plain radiographic thoracic pedicle morphometry has an influence on the outcome of thoracic pedicle screw insertion in Asian cadavers. Pedicle width <4.0 mm is associated with higher risk of pedicle perforation. This critical value corresponds to a PWBW ratio of 0.3. Identification of such pedicle profile warrants full evaluation of the morphometry of the thoracic pedicles and possibly extra-pedicular techniques should be employed to avert the risk of critical pedicle perforation.