To our knowledge, this study is the first to evaluate the characteristics and transitions of LBP in patients with LSS treated with decompression surgery. A detailed VAS scoring system for various lumbar disorders was first introduced by Aoki et al. and found to be useful. In an analysis of patients with nonspecific LBP, elderly patients showed significantly lower LBP VAS scores while sitting compared with that in a group of young patients, indicating the possibility of a discogenic factor in younger patients [
10]. In early-stage spondylolysis, the LBP VAS score while in motion was significantly higher than that found while standing or sitting [
11]. By contrast, although there are some studies that evaluated bilateral LEP VAS scores divided into those on left and right sides [
12], there are no studies that measured LBP VAS scores bilaterally.
The noteworthy point of the present study are the findings that LBP in patients with LSS before surgery was significantly greater while standing and the pain was reduced by ULBD surgery, with LBP (and LEP and LEN) improving equally on the approached and opposite side. These findings suggest, at least in part, that radicular LBP is improved by bilateral nerve root decompression surgery, supporting the report by Toyone et al. that LBP is improved by lumbar nerve root decompression surgery [
13]. Furthermore, LBP improvement after surgery to decompress the cauda equina may also be found, although signs of cauda equina syndrome do not generally result in LBP. In addition, we speculate the mechanism of the postural and symmetrical LBP improvement as being that the loads to the facet joint and disc bilaterally were improved because of the bilateral LEP relief by the decompression of the bilateral nerve root and cauda equina. Minimally invasive surgery particularly can also result in a favorable outcome for the treatment of patients with chronic LBP and spinal stenosis and our present results also support this finding [
14]. By contrast, greater preoperative LBP results in a poorer surgical outcome after the decompression surgery [
15]. However, in the decompression surgery just mentioned, a conventional laminectomy was performed and ULBD surgery may result in a more favorable outcome. A study comparing the ULBD with the conventional laminectomy indicated ULBD was superior in VAS of LEP and perioperative opioid use [
16]. Another study that compared unilateral with bilateral laminectomy indicated that unilateral laminectomy induces less translational motion because of the preservation of paraspinal muscle and facet joint of opposite side [
17]. According to these studies and our present results, ULBD is superior to bilateral conventional laminectomy in its clinical results and X-ray instability due to the preservation of paraspinal muscle and the facet joint on the opposite side. Generally, it is well known that LBP in LSS is multifactorial and that it includes facet pain, discogenic pain, radicular pain, and psychogenic pain [
18]. Our detailed three posture LBP VAS suggests that stronger LBP while standing than that while sitting reflects radicular or cauda equina symptoms such as intermediate LBP. By contrast, LBP while in motion may reflect the facet pain, discogenic pain, and radicular pain due to dynamic factors. Ultimately, in this study, LBP in the three postures improved to the same level after ULBD surgery. This may reflect that the intermediate LBP improved by ULBD surgery. In addition, the bilateral VAS scores improved equally on the approached and opposite sides indicating that the loads to facet joint and disc bilaterally were improved because of the bilateral LEP relief by decompression of the bilateral nerve root and cauda equina. For residual LBP, ULBD has less of a clinical outcome benefit than bilateral laminectomy, and this may result from the asymmetrical approach [
19]. However, in the present study, ULBD surgery did not worsen the residual LBP, but produced sufficient improvement of LBP on the approached side despite unsymmetrical invasion of the paraspinal muscles and facet joints. Non-neuropathic factors contribute to the LBP of patients with LSS [
20]. It is well known that the cause of LBP is multifactorial: deranged discs, facet joints, nerve roots, para-spinal muscles, and psychogenic factors [
18]. In the present study, patients with LBP complicated with DLS were excluded to avoid confounders of LBP caused by instability of discs and facet joints. To consider these types of pain, further investigations will be needed.
In the present X-ray image evaluation, severe instability or hypermobility such as severe worsening of local ROM or translation was not found because the spinous ligaments were preserved by ULBD surgery as shown in a representative case in Fig.
6 [
5]. Furthermore, a weak negative correlation between residual LBP during motion (2 years after surgery) and changes of local ROM was found. The smaller the residual LBP in motion, the more local ROM was obtained. This finding supports a report that spinal motion was significantly increased in flexion after multilevel fenestration [
21]. By contrast, changes of translation were not correlated with residual LBP, especially LBP in motion, which may reflect facet pain unrelated to changes of translation, and may also reflect asymmetrical invasion of paraspinal muscles and facet joints unrelated to the residual LBP.
The present study has some limitations. First, this study is observational and we did not evaluate the bilateral VAS scores of patients who underwent conventional laminectomy. Ideally, a randomized clinical trial that compares conventional laminectomy and ULBD should be conducted to demonstrate causality. However, conventional laminectomy may lead to the residual LBP because of the postoperative instability of decompression segment and ULBD is superior in the clinical outcome [
4,
5,
7,
8]. Thus, such study is precluded by ethical constraints. Second, because of its small sample size, the present study includes both single level and multilevel cases. The invasion of the paraspinal muscles and facet joints is different in single level and multilevel cases. However, LBP VAS scores before surgery are unrelated to the surgical procedure and therefore it is feasible that the significantly strong LBP that exists while standing will be improved by ULBD surgery. Study with a larger sample will be needed to improve statistical power. Third, the present study excluded patients with DLS who have unstable vertebra because we wanted to avoid LBP caused by instability of discs and facet joints. If these patients wanted, we performed ULBD surgery for DLS (4 cases in this study). However, we usually recommend transforaminal lumbar interbody fusion for patients with DLS. Further investigation will be needed to evaluate the characteristics and changes of bilateral LBP VAS scores in patients with DLS. Fourth, the present study did not evaluate sagittal alignment. Sagittal imbalance such as pelvic incidence and lumbar lordosis mismatch may be related to postoperative LBP [
22]. Using detailed VAS scores, Aoki et al. indicated that sagittal imbalance after the short segment fusion surgery resulted in residual LBP while standing [
23]. The main purpose of the present study was to evaluate the characteristics and transitions of LBP using detailed bilateral VAS scores. Thus, we did not evaluate the correlation between residual LBP and sagittal alignment. Investigation of the changes of sagittal alignment will be needed to evaluate further the characteristics of residual LBP.