Patients with lumbar disc herniation may have sciatic scoliosis, which is called nonstructural scoliosis secondary to nerve root irritation. Scholars generally believe that this is the compensatory behavior produced by the body to alleviate nerve root stimulation [3, 16], and some scholars have pointed out that the hyperactivity of paraspinal muscles is related to the occurrence of sciatic scoliosis [17–19].

There are different reports on the incidence of sciatic scoliosis in patients with lumbar disc herniation. Kim et al. [5] reported that the incidence of sciatic scoliosis in 164 adult patients with lumbar disc herniation was approximately 18%. Ozgen et al. [20] and other reports pointed out that the incidence of sciatic scoliosis is higher in adolescent patients with lumbar disc herniation, approximately 47%. Zhang et al. [15] reported that among 1087 patients with lumbar disc herniation, the incidence of sciatic scoliosis in adolescents and adults was approximately 23.8% and 12.2%, respectively. There is no report on the incidence of sciatic scoliosis in patients with single-segment and double-segment lumbar disc herniation. In this study, a retrospective analysis of 495 patients with lumbar disc herniation was conducted, and a total of 74 cases were found to have sciatic scoliosis, with an overall incidence of 14.9%. Among them, 53 and 21 patients had single-segment and double-segment lumbar disc herniation, respectively, and the incidence of single-segment and double-segment lumbar disc herniation with sciatic scoliosis was 10.7% and 4.2%, respectively. The incidence in double-segment patients is much lower than that in single-segment patients.

Kim et al. [5] reported that L4-5 intervertebral disc herniation is a risk factor for sciatic scoliosis. In this study, sciatic scoliosis was more common in patients with L4-5 lumbar disc herniation, which is consistent with previous reports [5, 8] and may be related to the anatomical characteristics of L4-5. The bilateral iliolumbar ligament starts from the transverse process of L4 and L5, ends at the sacroiliac joint and iliac crest, restricts the movement of the vertebral body and maintains its stability. However, L4-5 is not limited to the pelvic cavity and is more active than L5, which leads to greater shear force and longitudinal pressure in L4-5, and is more prone to degeneration. Therefore, L4-5 intervertebral disc herniation is more likely to be secondary to sciatic scoliosis.

For patients with ineffective conservative treatment, surgical treatment is the first choice [21–24]. Compared with traditional open surgery, PELD has the advantages of avoiding excessive exposure of the nerve root, less bone resection, less injury to the facet joint and muscle ligament structure, a low postoperative infection rate, and reducing the incidence of postoperative complications such as adjacent spondylosis caused by traditional open surgery [11, 25–29]. In addition, PELD can ablate the new blood vessels and granulation tissue after the rupture of the annulus fibrosus to reduce the inflammatory response [30].

Previous studies have reported that the clinical effect of PELD is similar to that of open surgery [9]. However, previous studies were limited to single-segment patients, and there was no report on the curative effect of double-segment patients. Double-segment lumbar disc herniation with sciatic scoliosis is often considered as a contraindication of PELD because of its many protruding segments, severe nerve compression and difficult operation. Double-segment patients often receive traditional open decompression and fusion internal fixation, which requires full dissection of paraspinal muscles, resection of some articular processes and bone grafting and fusion of the responsible segment, which is invasive and can lead to many complications such as bleeding during operation, intractable low back pain after operation, back muscle injury and accelerated degeneration of adjacent segments [31]. The clinical effect of PELD on double -segment patients needs further study.

In this study, during the follow-up of the two groups, the back VAS score, leg VAS score, JOA score and ODI index were significantly improved compared with those before the operation. Sciatic scoliosis patients have the characteristics of a relatively straight sagittal position, reduced thoracic kyphosis and lumbar kyphosis, and some patients with sagittal imbalance [32]. According to previous studies [6], sagittal imbalance is defined as SVA ≥ 40 mm. In this study, the preoperative TK and LL of single-segment patients were 12.8 ± 8.2° and 35.9 ± 8.6°, respectively, and the preoperative TK and LL of double-segment patients were 11.9 ± 7.1° and 36.8 ± 8.4°, respectively. The preoperative SVA of the two groups was 50.1 ± 21.8 mm and 49.7 ± 18.2 mm, respectively. After PELD treatment, at the last follow-up, the TK and LL of single-segment patients improved to 30.1 ± 13.4° and 51.9 ± 10.5°, respectively, and the TK and LL of double-segment patients improved to 31.2 ± 8.0° and 53.2 ± 9.4°, respectively. The SVA of the two groups improved to 3.9 ± 2.1 mm and 12.2 ± 3.0 mm, respectively. The sagittal imaging parameters were significantly improved compared with those before the operation, and all patients reached the standard of sagittal balance. This shows that after PELD treatment, nerve compression is relieved, the sagittal curve can be changed, and thoracic kyphosis and lumbar kyphosis can be increased. However, the patients in the double-segment group had a longer operation time, more intraoperative fluoroscopy and more intraoperative bleeding. During the follow-up, there was no significant difference in back VAS score, leg VAS score, JOA score or ODI index between the two groups at the same follow-up time. This shows that PELD can achieve satisfactory short-term effects and long-term effects for both single-segment and double-segment lumbar disc herniation with sciatic scoliosis. PELD for double-segment patients can reduce the related complications caused by traditional decompression, fusion and internal fixation.

This study found that at the last follow-up, both groups reached the coronal and sagittal balance criteria, but the AVT, CBD and SVA in the double-segment group were higher than those in the single-segment group, indicating that the long-term recovery of coronal and sagittal balance in the double-segment group was worse than that in the single-segment group. The author believes that this may be related to the lower PI of the double-segment group in this study. PI angle is a key anatomical parameter for the stability of spine-pelvic sagittal plane, which is related to sacral inclination and spinal curvature. In theory, patients with small PI angle are less able to compensate for the imbalance [33]. In the double-segment group, the PI value was low, the range of pelvic rotation around the femoral head was also lower, the pelvic parameters were smaller, the lumbar kyphosis was relatively flat, the center of gravity moved forward, and the compensation ability of sagittal pelvic retroversion was limited [34–36]. The self-compensating ability of the spine and pelvis to the curvature changes caused by lumbar degenerative diseases was weak, and the recovery of sagittal balance was poor after nerve compression was relieved. In addition, in this study, the proportion of patients with preoperative Pfirrmann grade IV in double-segment patients was higher than that in single-segment patients, which indicated that in double-segment patients, there were more serious degeneration of intervertebral disc, more imbalance of extracellular matrix catabolism, less number of nucleus pulposus cells, more apoptosis, lower cell density, active proliferation and functional protein synthesis, and lower proteoglycan, type II collagen and water content [37]. Type I and III collagen fibers are higher, resulting in more loss of intervertebral disc height, lower elasticity and tension of intervertebral disc, smaller nerve root volume, stronger chronic inflammatory reaction and more serious intervertebral instability [38, 39]. The intervertebral disc is not enough to support the large-scale activity of the patient, which is not conducive to the recovery and reconstruction of the coronal and sagittal plane after operation.

In addition to pain relief, patients are often concerned about whether the scoliosis posture can be corrected. Previous studies have shown that sciatic scoliosis is reversible. After nerve compression is released, the scoliosis posture can be corrected by itself. Zhang et al. [15] reported that six months after surgery, the resolution rates (RRs) of scoliosis in adolescents and adults were 85.71% and 92.68%, respectively. Kim et al. [5] and other studies have shown that 6 months after surgery, the resolution rate of scoliosis is more than 50%. Tu et al. [13] performed a retrospective analysis of 42 patients with sciatic scoliosis and found that the average Cobb angle improved from 18.4° to 8.7° three months after the operation. In this study, at the last follow-up, the average Cobb angle of single-segment group and double-segment group were 2.4 ± 0.6 °and 2.2 ± 1.0 °, AVT were 1.9 mm ± 0.4 mm and 5.2 ± 2.3 mm, CBD were 1.1 mm ± 1.6 and 5.1 mm ± 1.0 mm, respectively, which were significantly better than those before operation. At the last follow-up, scoliosis subsided in 94.3% of single-segment patients and 95.2% of double-segment patients. Early operation is beneficial to the correction of scoliosis posture and can avoid the development of sciatic scoliosis to structural scoliosis.

There are some limitations in this study. In the process of imaging parameter measurement, there may be some measurement bias due to the existence of osteophytes or the quality of X-ray films. The curve recovery of sciatic scoliosis is a dynamic process, and the 12-month follow-up cannot fully reflect the clinical outcome of patients after PELD. The effects of PI and the degree of intervertebral disc degeneration on the recovery of coronal and sagittal curves need to be further studied.