RESUMO
AIM: To evaluate the feasibility and accuracy of minimally invasive, transpedicular screw placement in cervicothoracic fractures with the help of computed tomography (CT)-controlled guidewires. MATERIALS AND METHODS: Two hundred and ninety-three guidewires were inserted in 35 patients (42.9±21.2 years) under CT fluoroscopy (286 thoracic, seven cervical). There were 28 traumatic cases, three pathological fractures, three fractures due to infectious infiltrations, and one osteoporotic fracture. In 151 pedicles, screw placement was performed in the CT room. CT images were reviewed regarding accuracy and cortical violations using the popular 2 mm increment deviation classification of Gertzbein and Robbins. RESULTS: Guidewire implantation resulted in only 28 cortical contacts. Minor encroachments of the pedicle wall by inserted screws occurred in 39.1% (59 of 151) and in 23.8% if taking unavoidable encroachments into account (30 of 59). Pedicular isthmus width correlated to cortical guidewire contacts (r=-0.449; p=0.077) and pedicle violations (all graded "A") by the inserted screws (r=-0.581; p=0.049). Total procedural duration was 138.6±44.2 minutes, representing 14.5±11.6 minutes for each pedicle, while showing a significant correlation against higher vertebral levels (r=-0.849; p=0.0002) and the occurrence of pedicle violations (r=-0.641; p=0.027). CONCLUSIONS: The treatment of vertebral fractures with a guidewire-based pedicle screw insertion technique under CT imaging results in very high accuracy and a low complication rate.