Accuracy and efficacy of thoracic pedicle screws in curves more than 90 degrees

Retrospective study of large-magnitude thoracic curves (> or =90 degrees ) treated with pedicle screw constructs.

To evaluate the results of pedicle screw constructs for thoracic curves > or = 90 degrees in terms of sagittal and coronal correction/efficacy, as well as accuracy and safety of thoracic pedicle screw placement.

Thoracic pedicle screw constructs continue to become increasingly more common; however, the debate continues about the safety and efficacy of these constructs because of their perceived increased risk of neurologic injury and the increased cost of spinal instrumentation.

Since 1998, all patients with adolescent idiopathic scoliosis, or adult progression of adolescent idiopathic scoliosis, a thoracic curve > or = 90 degrees and a minimum 2-year follow-up who were treated with pedicle screw constructs were included in this study. Standing anteroposterior (or posteroanterior), lateral and bending preoperative radiographs, and anteroposterior (or posteroanterior) and lateral postoperative radiographs were evaluated for curve magnitude, flexibility, and postoperative correction to assess the efficacy of these constructs in the immediate postoperative period and at latest follow-up. Postoperative CT scans were evaluated for screw accuracy using established 2-mm increments (intrapedicular, 0-2 mm breach, 2-4 mm breach, > 4 mm breach). Preoperative plans were also reviewed to evaluate the ability to place a pedicle screw at each planned level in these large-magnitude curves.

Twenty patients with thoracic curves > or = 90 degrees and an average follow-up of 3.3 years (range, 2.0-5.2 years) were included in the study. All patients underwent a posterior spinal fusion with a pedicle screw only construct. The average preoperative main thoracic curve measured 100.2 degrees (range, 90 degrees -133 degrees ), with an average side-bender of 71.6 degrees (29% flexibility). The average postoperative main thoracic curve was 32.3 degrees (68% correction). A total of 352 thoracic screws were placed in the 20 cases (17.6 screws/case). Screw accuracy (either intrapedicular or <2 mm breach) was 96.3% (339 of 352 screws) by postoperative CT scanning. Ten screws were considered to have a breach between 2 and 4 mm (3 medial, 7 lateral), while three screws were > 4 mm (2 medial, 1 lateral). The two medial screws were the only placed screws that were removed (0.57%). Overall, 94% of planned screws (352 of 374 screws) were placed according to the preoperative plan. There were no incidences of screw or instrumentation failure. Of note, there was a temporary decrease in motor-evoked potentials during curve correction in 2 cases; however, there were no identifiable neurologic complications.

Thoracic pedicle screw constructs can be safely used for large-magnitude curves. Curve correction (68%) is powerful for these curves, which are stiff and difficult to manage. Correction should be performed carefully with consideration given to convex compression for cases with concomitant hyperkyphosis for these "at risk" spinal cords. Screw accuracy (96.3%) was excellent in this review. The authors found that screws can consistently be placed according to the preoperative plan even in these large-magnitude curves.