[Clinical accuracy of fluoroscopic navigation at the thoracic and lumbar spine]

The aim of the present study was the verification of the accuracy of 2-D fluoroscopy-based navigated pedicle screws at the thoracic and lumbar spine in a case series of traumatised patients. Within 36 months 111 pedicle screws in 29 patients were instrumented using C-arm based navigation, 60 at the thoracic and 51 at the lumbar spine. All screw positions were evaluated postoperatively by a routine thin slice CT scan using multiplanar reconstruction. The position of a screw in relation to its pedicle was classified in a) screw completely intraosseous, b) screw perforated less than thread level and c) screw perforated more than thread level. In 34 thoracic (56.7%) and 32 lumbar (62.7%) screws complete intraosseous placement was observed, 14 thoracic (23.3%) and 14 lumbar (27.5%) screws perforated less than thread level. Perforations more than thread level were found in 12 thoracic (20%) and 5 lumbar (9.8%) screws. Only medial and lateral perforations of the pedicle were documented (without neurological signs), cranial or caudal ones did not occur. Segmentation of the C-arm navigation into two comparable treatment periods showed a learning curve with a reduction of perforations in the second sequence (after 57 pedicle instrumentations) of about 15%, this was found to be not statistically significant. The fluoroscopic navigation of pedicle screws is a safe procedure at the lumbar spine with equal accuracy compared to the non-navigated conventional instrumentation. Application of C-arm navigation at the thoracic spine showed more inaccuracies, so that 3-D-based navigation seems to be advantageous in this region.