Spinal motion after cervical fusion. In vivo assessment with roentgen stereophotogrammetry

A roentgen stereophotogrammetric (RS) technique using metallic markers was used to determine in vivo three-dimensional (3-D) motion data in the post-operative cervical spine. Two patients were examined with RS following fusion for up to 12 months at scheduled intervals.

The study was designed to develop a technique and provide an in vivo assessment of the fusion process following surgery. Besides the researchers and surgeons, this type of information is of particular interest to the FDA and implant manufacturers.

The high accuracy of the roentgen stereophotogrammetric (RS) technique using metallic markers to predict displacements across body joints, including spinal joints, has been well established in the literature. Its applications in the lumbar region to assess the fusion process as a function of time following surgery have also been reported. Similar in vivo applications (and results) dealing with the cervical region are lacking.

A roentgen stereophotogrammetric (RS) technique using metallic markers was developed. Appropriate in vitro studies were undertaken to assess its accuracy for in vivo applications in the cervical spine region. Following this, three Vitallium beads were surgically implanted in anatomically appropriate positions in of each the exposed vertebrae of two patients at the time of surgery. The patients were followed with RS serially for 12 months after surgery at scheduled intervals. At each RS session, roentgen stereo pairs of the cervical spine in neutral, maximum voluntary flexion, and maximum voluntary extension were obtained using a biplanar radiographic system. The metallic beads on the radiographs were digitized using an in-house software package to determine vertebral motions across the fused segments.

In patient 1 (atlantoaxial fusion), the flexion-extension range of rotational motion decreased with time while the corresponding AP translation at the fusion level increased from 6 mm at 3 months post-op to 13 mm at 6 and 12 months postoperatively. The abnormal AP translation at 6 months and beyond was evident on extension/extension lateral radiographs, but the AP translation at 3 months was not visually evident. In patient 2 (anterior discectomy and interbody fusion), motions of small magnitude were observed at the fused level in all three axes. These motions actually increased over the one year observation period, but were never large enough to be visually detectable on lateral extension/extension radiographs. Patient 2 had a good clinical result despite these small motions.

Roentgen stereophotogrammetry may detect motion in the in vivo cervical spine with a sensitivity heretofore unavailable. In patient 1, we were able to detect motion before it was visually manifest. In patient 2, small motions were detected at the level of a successful anterior cervical fusion, the significance of which remain uncertain.