Study design: Prospective comparative randomized clinical and radiologic study.
Objective: This study was conducted to compare the short-term effects of rigid versus semirigid and dynamic instrumentation on the global and segmental lumbar spine profile, subjective evaluation of the result, and the associated complications.
Background data: Lumbar spine fusion with rigid instrumentation for degenerative spinal disorders seems to increase the fusion rate. However, rigid instrumentation may be associated with some undesirable effects, such as increased low back pain following decrease of lumbar lordosis, fracture of the vertebral body and pedicle, pedicle screw loosening, and adjacent segment degeneration. The use of semirigid and dynamic devices has been advocated to reduce such adverse effects of the rigid instrumentation and thus to achieve a more physiologic bony fusion.
Materials and methods: This study compared 3 equal groups of 45 adult patients, who underwent primary decompression and stabilization for symptomatic degenerative lumbar spinal stenosis. The patients of each group were randomly selected and received either the rigid (Group A), or semirigid (Group B), or dynamic (Group C) spinal instrumentation with formal decompression and fusion. The mean ages of the patients who received rigid, semirigid, and dynamic instrumentation were 65 +/- 9, 59 +/- 16, and 62 +/- 10 years, respectively. All patients had detailed roentgenographic study including computed tomography scan and magnetic resonance imaging before surgery to the latest follow-up observation. The following roentgenographic parameters were measured and compared in all spines: lumbar lordosis (L1-S1), total lumbar lordosis (T12-S1), sacral tilt, distal lordosis (L4-S1), segmental lordosis, vertebral inclination, and disc index. The SF-36 health survey and Visual Analogue Scale was used before surgery to the latest evaluation.
Results: All patients were evaluated after a mean follow-up of 47 +/- 14 months. Both lumbar and total lordosis correction did not correlate with the number of the levels instrumented in any group. Total lordosis was slightly decreased after surgery (3%, P < 0.05) in Group C. The segmentallordosis L2-L3 was increased after surgery by 8.5% (P < 0.05) in Group C, whereas the segmentallordosis L4-L5 was significantly decreased in Group Aand C by 9.8% (P = 0.01) and 16.2% (P < 0.01), respectively. The disc index L2-L3 was decreased after surgery in Group A and C by 17% (P < 0.05) and 23.5% (P < 0.05), respectively. The disc index L3-L4 was increased in Group C by 18.74% (P < 0.01). The disc index L4-L5 was after surgery decreased in all 3 groups: Group A by 21% (P = 0.01), Group B by 13% (P < 0.05), and Group C by 13.23% (P < 0.05). The disc index L5-S1 was significantly decreased in Group B by 13% (P < 0.05). The mean preoperative scores of the SF-36 before surgery were 11, 14, and 13 for Groups C, B, and A, respectively. In the first year after surgery, there was a significant increase of the preoperative SF-36 scores to 65, 61, and 61 for Groups C, B, and A, respectively, that represents an improvement of 83%, 77%, and 79%, respectively. In the second year after surgery and thereafter, there was a further increase of SF-36 scores of 19%, 23%, and 21% for Groups C, B, and A, respectively. The mean preoperative scores of Visual Analogue Scale for low back pain for Groups C, B, and A were 5, 4.5, and 4.3, respectively, and decreased after surgery to 1.9, 1.5, and 1.6, respectively. The mean preoperative scores of the Visual Analogue Scale for leg pain for Groups C, B, and A were 7.6, 7.1, and 6.9, respectively, and decreased after surgery to 2.5, 2.5, and 2.7, respectively. All fusions healed radiologically within the expected time in all three groups without pseudarthrosis or malunion. Delayed hardware failure (1 screw and 2 rod breakages) 1 year and 18 months after surgery without radiologic pseudarthrosis was observed in 2 patients in Group C. Asymptomatic radiolucent areas were shown around pedicle screws in the pedd pedicle screws in the pedicles L5 and S1 in 2, 3, and 4 cases in Group C, A, and B, respectively. There was no adjacent segment degeneration in any spine until the last evaluation.
Discussion and conclusion: This comparative study showed that all three instrumentations applied over a short area for symptomatic degenerative spinal stenosis almost equally after surgery maintained the preoperative global and segmental sagittal profile of the lumbosacral spine and was followed by similarly significant improvement of both self-assessment and pain scores. Hardware failure occurred at a low rate following dynamic instrumentation solely without radiologically visible pseudarthrosis or loss of correction. Because of the similar clinical and radiologic data in all three groups and the relative small number of patients that were included in each group, it is difficult for the authors to make any recommendation in favor of any instrumentation.