Minimally invasive percutaneous cannulated polyaxial screw and rod fixation thoracolumbar and lumbosacral vertebrae: cadaveric evaluation of accuracy and safety of a fluoroscopic-guided technique, in large breed dogs

Objective To develop and report the accuracy and safety of a technique for percutaneous pedicle screw-rod fixation of the thoracolumbar and lumbosacral regions using fluoroscopic guidance in dog cadavers, comparing two levels of surgical experience. Study design Cadaveric methods comparative study. Animals Cadaveric vertebral columns of dogs ( n = 3). Methods Pre-operative computed tomography (CT) scans facilitated in-silico surgical planning. Ideal implant insertion points and trajectory were recorded, and the image was subsequently reconstructed to mimic the appearance of intra-operative fluoroscopy. After surgical preparation from T8 to S2, the study was divided in two parts: part one to evaluate safety and accuracy of percutaneous “end-on” fluoroscopic drill hole position technique (EOFG); and part two to determinate the feasibility of thoracolumbar (T13-L1) and lumbosacral (L7-S1) percutaneous cannulated pedicle screw and rod fixation (PC-PSRF). The procedures were performed by one of two experienced ECVN diplomates, and one ECVN neurology resident (inexperienced in spinal fixation surgery). Following post-operative CT imaging, drill-hole safety was established using a modified Zdichavsky classification and optimal placement was compared between surgeons. Drill-hole accuracy was calculated based on the deviation from planned angle and bone purchase, between the two levels of surgeon experience. Data sets were analysed at both univariable and multivariable levels with logistic regression analysis. Results No unsafe holes (0/34) were drilled by experienced surgeons, whereas the inexperienced surgeon drilled 4/34 (11.8%) unsafe holes. Optimal drill-hole placement was significantly associated with surgeon experience ( p < 0.001, OR = 9.66), whereas cadaver number and spinal region did not reach significance. Drill-hole angle and bone depth deviations were significantly different between experienced and inexperienced surgeons ( p = 0.04 and p < 0.001, respectively). Angle deviations were not significantly different between experience levels by region, whereas mean bone purchase deviations were significantly lower for experienced surgeons in the thoracic region ( p = 0.002), but differences were not significant in the lumbar and sacral regions. The mean time taken to drill the hole was significantly longer for the inexperienced surgeon ( p < 0.001). PC-PSRF using EOFG was successfully demonstrated at T13-L1 and L7-S1. Conclusion The technique was feasible and allowed for percutaneous fixation of the thoracolumbar and lumbosacral segments. Previous experience in open instrumented spinal surgery is beneficial for safe and accurate application. Clinical significance The PC-PSRF technique using EOFG guidance warrants future clinical studies to investigation a potential role in reducing pain, hospitalisation length, muscle injury and need for surgical revision.