BACKGROUND
Given the aging population and increasing prevalence of degenerative diseases of the spine, neurosurgeons need to perform more lumbar fusions with pedicle screw fixations. As these interventions are currently carried out percutaneously, surgeons are dependent on intraoperative imaging to replace the anatomical landmarks found in open procedures. In our institution, we recently acquired a neuronavigation system based on intraoperative 3D fluoroscopy. OBJECTIVES The study was aimed at comparing pedicle screw insertions using intraoperative 3D fluoroscopy based-neuronavigation compared to intraoperative 3D fluoroscopy alone in terms of accuracy, postoperative complications, surgery duration, and operative irradiation. Finally, a review of the literature on the subject was also performed.
METHODS
Non-randomized prospective study. From October 2009 to September 2016, 438 percutaneous pedicle screws were inserted in 102 adult patients without neuronavigation (Group F3D) and 248 percutaneous pedicle screws were inserted in 59 patients with neuronavigation (Group F3DNav). The precision of the screws was determined on intraoperative 3D fluoroscopy or postoperative computed tomography (CT) using the Wang classification (Grade 0: no cortical breach, Grade 1: breach <2mm, Grade 2: breach of 2-4mm, Grade 3: breach >4mm) before and after intraoperative repositioning. Only complications related to the surgical procedure were reported. Surgery duration (in minutes) and irradiation (in μGy.m²) data were acquired after surgery. For the literature review, we selected in the MEDLINE database the studies that compared the accuracy of percutaneous lumbar pedicle screws between 3D fluoroscopy without neuronavigation vs. 3D fluoroscopy with neuronavigation.
RESULTS
The demographic characteristics of the F3D and F3DNav groups were comparable. In our study, 84.2% and 96.8% of screws were Grade 0 in the F3D and F3DNav groups, respectively (p<0.001). Intraoperative 3D control allowed the repositioning of 11 screws in the F3D group and of 5 in the F3DNav group. In total, 14 complications occurred in the F3D group as compared to 4 in the F3DNav group (p>0.05). Most of them (10/18) were dural tears. The mean surgery duration was 270+/-91 minutes in the F3D group and 257+/-71 minutes in the F3DNav group (p>0.05). The mean operative irradiation per patient was 11,950.40 μGy.m² in the F3D group and 10,500.53 μGy.m² in the F3DNav group (p>0.05). For the literature review, no papers meeting the initial criteria were found.
CONCLUSION
Neuronavigation appears more accurate than 3D fluoroscopy alone. However, we did not find any differences regarding complications, surgery duration, and patient irradiation.