CT Scan Verification of Long Screw Placement in 10 Idiopathic Scoliotic VDS Patients With 10 or More Year Follow-Up

Purpose: Since 1984, approximately 320 idiopathic scoliotics have been operated on at our institution utilizing anterior screw placement and ventral derotation with fusion (VDS). To date, there are no reported vascular /visceral injuries resulting from screw placement accomplished through an open thoracotomy and mobilisation of the parietal pleura and great vessels with bicortical screw placement. No patient is symptomatic. After reports by Sucato , et al regarding CT investigation revealing long screws and there relationship to vital structures after thoracoscopic anterior spinal instrumentation ( SRS 2001), we retrospectively reviewed 10 patients with 10 or more years follow-up and long screws on plain radiographs to evaluate screw proximity to vascular /visceral structures and the possible clinical significance thereof.

Methods: 10 right thoracotomy VDS patients (8 female, 2 male) operated on between September 1987 and April 1993 were evaluated by CT to examine screw placement after review of x-rays which showed 2 or more ventrally placed screws measuring 5 mm or more. Mean age at surgery was 17 years 11 months (range 13 years 4 months - 26 years 6 months). Follow-up ranged from 10 years 8 months to 16 years 4 months (mean 13 years 11 months). All patients were evaluated with spiral CT scans with 3mm thick slices, reduced radiation dosage through the operated region. Bone and soft tissue windows were evaluated. CT scans were reformatted on the 3 D workstation and reviewed by the radiologists. Screw tip penetration of the far cortex and screw tip relationship to vascular /visceral structures was noted.

Findings: A mean of 7.3 levels were instrumented in 10 patients with a total of 73 screws being placed. 36 screws were noted to be 5mm or longer on x-ray 49 %). 38 were noted to be 5mm or longer on CT ( 52%). Long screws on x-ray did not necessarily correlate with long screws on CT. On CT, 18 screws (47% of long screws) closely approximated vascular/ visceral structures. 16 screws (42% of long screws) closely approximated the aorta. 3 of these screws, 1 in each of 3 patients, appeared to impress upon the aorta. (These 3 patients are scheduled to undergo transesophogeal ultrasound and possible angiography). In addition, one screw tip that measured only 4mm on CT, ended adjacent to the aorta.

There were no pseudoaneurysms. 2 screws ( 5% of long screws) ended adjacent to lung parenchyma. 20 of the long screws (52.6%) had no relationship to vascular/ visceral structures.

Summary: Ventral screw placement in the treatment of idiopathic scoliosis is a safe procedure when accomplished through an open thoracotomy with mobilisation of the parietal pleura and great vessels.

Potential for great vessel, lung and visceral injury should be recognized. It is likely reduced as the result of pleural/great vessel mobilization. Plain radiographic measurement of long screws does not necessarily correlate with CT measurement likely secondary to residual vertebral body rotation and sensitivity of the diagnostic device. Transesophgeal ultrasound followed, as needed by angiography, is recommended in cases of impression of the vessels.