2:20 Accuracy of computed tomography in assessment of thoracic pedicle screw

Autor: Jason Pennypacker, Steven Ludwig, Amir H. Fayazi, Daniel Gelb, Ronald Hugate
Rok vydání: 2002
Předmět:
Zdroj: The Spine Journal. 2:88-89
ISSN: 1529-9430
DOI: 10.1016/s1529-9430(02)00351-0
Popis: Purpose of study: Transpedicular instrumentation in the thoracic spine is fraught with danger because of close proximity of vital structures and morphological variability of thoracic vertebrae. Given the potential implications, a number of cadaveric and in vivo studies have been performed to evaluate the efficacy of different insertion techniques. In most cadaveric studies, open dissection has been the standard in evaluating the position of the pedicle screw, whereas computed tomography has been employed to evaluate the position of in vivo thoracic pedicle screws. Despite its common use, the accuracy of computed tomography (CT) assessment of thoracic pedicle screw position has not been adequately investigated. Methods used: Nine cadaveric thoracic spines were stripped of their soft tissue posteriorly and scanned using spiral CT. The pedicle diameter at each level was then measured radiographically. Each pedicle was then instrumented with 5 × 35 mm titanium pedicle screws using either routine fluoroscopic technique or with help of computer-assisted devices (BrainLAB VectorVision2 or Stealth Station Treon). Postinstrumentation CT images of each specimen was obtained and evaluated by an independent blind observer for evidence of thoracic pedicle or vertebral body breech. Open dissection was performed to confirm the position of each pedicle screw. The costovertebral joints were not explored during the dissection. Each pedicle screw was graded as either contained (fully contained within the pedicle and the rib head) or out (any breech of the medial, inferior or anterior/lateral body). of findings: A total of 194 pedicles were instrumented and evaluated radiographically and by means of open dissection. Radiographic evaluation consisted of spiral CT (1.1-mm pitch) followed by reconstruction into 2.5-mm axial slices and 3-mm sagital and coronal slices. Using CT evaluation, 130 of the pedicle screws were contained, whereas 64 screws were malpositioned. Using open dissection, 149 of the screws were contained, whereas 45 were malpositioned. Assuming that open dissection is the gold standard, computed tomography was 45% sensitive and 88% specific in diagnosing a malpositioned screw (55% positive predictive value, 88% negative predictive value). CT was also 78% sensitive and 64% specific in diagnosing a contained screw (88% positive predictive value, 55% negative predictive value). The overall accuracy of CT was 73% in this study. Relationship between findings and existing knowledge: The use of pedicle screw instrumentation in the thoracic spine has become increasingly popular. Thoracic pedicle screws offer many advantages in the treatment of spinal pathology. However, potential for complications exist. Surgeons need reliable means to evaluate the placement of these screws. Likewise, they need to understand the limitations of any given technique of evaluation. Based on the results of this investigation, CT scan was found to be relatively insensitive in the diagnosis of malpositioned screws. Overall significance of findings: CT alone may not be sufficiently accurate to serve as the sole method of evaluation of thoracic pedicle screw position. In the face of postoperative neurologic deficits, surgical exploration and hardware removal should remain as the "gold standard" of care. Disclosures: Device or drug: pedicle screws. Status: investigational. Conflict of interest: No conflicts.
Databáze: OpenAIRE
Externí odkaz: