Proximal Thoracic Pedicle Screw Complications: Fractures With Spinal Cord Injury

• d - Medronic Sofamor Danek

Thoracic pedicle screws have gained acceptance as an excellent fixation method for thoracic deformities. However, long-term studies of their use are limited. We have encountered seven patients who sustained fractures at the cephalad end of spinal constructs placed for significant deformities. Two of these patients suffered severe neurologic loss.

Biomechanical studies have proven thoracic pedicle fixation strength far exceeds that of hook constructs. Substantial correctional forces can be transferred to the thoracic spine with these implants. Excellent correction of coronal and sagittal deformities can be obtained with screws. However, failure of constructs using thoracic screws has not been previously described. This retrospective review analyzes seven patients who underwent posterior fixation for significant sagittal imbalance and sustained vertebral fractures through the upperinstrumented levels at the site of the thoracic screws. Two of these patients sustained “sheer-type” fractures resulting in incomplete spinal cord injuries.

All patients were female, ages 67-74 with significant osteopenia. Five were substantially overweight. Initial post-operative restoration of coronal and sagittal balance was obtained through anterior/posterior fusions or pedicle subtraction osteotomies. No peri-operative neurologic deficits occurred, and all patients were ambulatory post-operatively.

Delayed neuro deficits occurred in two patients. Case 1, a 69-year-old woman underwent an uneventful staged anterior/posterior T3-sacrum fusion. After three weeks of ambulation in rehab, the patient developed significant right leg weakness which progressed to T6 paraplegia over 24 hours. X-rays, CT scans, myelograms, and cervical MRI were read as normal but visualization of upper/mid thoracic levels proved difficult. Urgent thoracic laminectomies were done, no abnormality found. Follow-up myelogram revealed a reduced but mobile “sheer-type” fracture through the T3 screws which were removed and cephalad implants placed. Some neuro recovery has occurred.

All other patients were found to have fractures occurring within six weeks of surgery. One sustained a T10 “sheer-type” injury after a bad fall with neurologic damage. The others underwent uneventful hook fixation proximal to their fractures without any neurologic loss.

Hook construct mode of failure is usually pull-out, rarely with neurologic injury. Thoracic pedicle fixation improves correctability, but may lead to “sheer-type” failure in osteopenic bone with potential neurologic injury.

Our recommendation is to place supra-laminar/infra-laminar hook claw at the cephalad end of long constructs in these patients. If failure occurs, the hooks likely will fracture the lamina and decompress the canal. Since utilizing this technique we have not encountered failure nor neurologic loss. Current hook/screw designs are being made to address this problem. Further studies are needed on the use of thoracic pedicle fixation in osteopenic patients.

• If noted the author indicates something of value received. The codes are identified as: a-research or institutional support; b-miscellaneous funding; c-stock or stock options; d-royalties; e-other financial or material support including consulting.