Purpose:
Lumbar spinal stenosis (LSS) can be treated surgically by decompressing the affected nervous structures, either by removing or preserving the midline structures. There is no conclusive evidence demonstrating the superiority of either surgical technique in the treatment of LSS. Our purpose was to compare the posterior midline-preserving techniques with midline-removing techniques separately for central stenosis and lateral recess stenosis to evaluate whether either technique leads to superior postoperative results in terms of functional outcome or pain reduction, by using data from the Finnish Spine Registry (FinSpine).

Methods:
A total of 7577 patients underwent decompression surgery for central (n=5670) or lateral recess (n=1907) stenosis in the lumbar spine between 2015 and 2022. In the central stenosis group 3025 patients were operated using midline-preserving techniques and 2645 using midline-removing techniques. In the lateral recess stenosis group, the corresponding numbers were 1536 and 371, respectively. Patients with less than 3mm degenerative spondylolisthesis were included. We recorded the baseline information, and the primary outcomes were the between-group differences in improvement in functional outcome and back and leg pain at one, two and five years postoperatively. Oswestry disability index (ODI) and visual analogue scale (VAS) for back and leg pain were used to assess clinical outcomes. We compared midline-preserving and midline-removing techniques separately in central and lateral recess stenosis groups.

Results:
Regardless of the surgical technique, patients improved in terms of functional outcome and pain during the 5-year follow-up. In the central stenosis group, there were no statistically significant differences between groups in primary outcomes. However, in the midline-preserving group, there were more new operations during the follow-up period as a secondary outcome. In the lateral recess stenosis group, there was a statistically significant improvement in the midline-preserving group in the change of ODI; 7.9 (95% C.I. [1.7, 14.1], p=0.01), and in the change in VAS leg; 12.8 (95% C.I. [0.5, 25.0], p=0.04) at the 5-year follow-up, compared to the midline-removing group.

Conclusions:
Based on a nationwide registry, the majority of patients improved in the primary outcomes and were satisfied with the operative treatment. Our study suggests that concerning central stenosis both techniques lead to good results, but midline-preserving techniques may lead to new operations more often than after midline-removing decompression in 5 years. Alternatively, in case of lateral recess stenosis, a question is raised whether surgery by midline-preserving decompression might benefit patients in terms of disability and leg pain.