The retrospective study is designed according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement[10].(Supplyment S1)
Patient data
The inclusion criteria were as follows: 1. Single-segment degenerative disc herniation and spinal canal stenosis with neurological symptoms. 2. Age over 65 years. 3. More than one radiographic examination(X-ray, computed tomography (CT),Magnetic Resonance Imaging
(MRI) or Diffusion tensor imaging(DTI))confirming nerve root compression. 4. Good general condition: blood pressure after intervention < 160 mmHg systolic and <100 mmHg diastolic[11]; intraoperative blood glucose levels <10 mmol/l(The Society for Ambulatory Anesthesia)[12]. Cardiopulmonary function, assessed by the anesthesiologist, is able to tolerate general anesthesia.
The exclusion criteria were as follows: 1. More than one segmental disc herniation. 2. Lumbago and no clear nerve root symptoms. 3. Advanced age (over 95 years). 4. Tumors. 5. Serious postoperative complications.6. morbid obesity. 7. Systemic disease or ane insufficiency.
Lumbar disc herniation was combined with stenosis in L4/5 in 47 patients, L3/4 in 3 patients and L5/S1 in 11 patients. Group A received bilateral decompression via a unilateral approach surgery, and group B received conventional transforaminal lumbar interbody fusion approach. All patients underwent lumbar X-ray, three-dimensional CT, and magnetic resonance imaging (MRI) before surgery. After surgery, all patients were followed up for 26.2 months, with a range of 20–36 months.
Surgical methods
Bilateral decompression via unilateral approach (BDUA) group
Each patient was placed in prone position and intubated under general anesthesia. A paravertebral incision was made in the lesion intervertebral space. The paravertebral muscle space was obtusely separated, With the help of mini-retractor designed by ourselves[13], multifidus and the longissimus muscles were separated and the pedicle entry point was exposed clearly(Figure 1 A,B,C).The pedicle screw was inserted into the target vertebra, and the articular process was removed with bone biting forceps. The vertebral plate after c-arm X-ray fluoroscopy confirmed that the reduction was satisfactory. During this process, the nerve root and dural sac were protected. Then the inferior facet and approximately 1/3 of the superior facet of symptomatic side were removed, the spinal canal was exposed, and the upper and lower laminar margins were removed depending on the specific conditions of spinal stenosis. Then the ipsilateral ligamentum flavum was completely removed. The contralateral view were obtained by tilting the operating table. Resection of the contralateral junction of lamina with the spinous process was performed in order to expanded spinal canal. At this point, the contralateral ligamentum flavum was excised. The soft tissue and osteophytes of the contralateral subarticular zone was excised to decompress the contralateral nerve root. Meanwhile, the protruded nucleus was removed, the intervertebral space was opened, and the cartilage of the vertebral endplate was removed for use in the bone graft fusion. The extracted articular process and lamina were used for granular packing in the intervertebral space, and the cancellous bone was compressed and placed into the intervertebral fusion cage. Wiltse's approach was used to implant a contralateral pedicle screw. Finally, the incision was sutured after a negative pressure flow tube was placed. Figure 2A,C shows an intraoperative photographs and schematic diagram of this surgical approach.[14]
Control group: conventional approach transforaminal lumbar interbody fusion.
Each patient was placed in prone position and intubated under general anesthesia. A standard midline incision and subperiosteal exposure is made out to the tips of the transverse processes and the longissimus and multifidus are separated from the posterolateral gutter.The pedicle screws were placed into the upper and lower vertebral bodies. The spinous process and bilateral lamina and ligaments were removed with bone biting forceps, whereas the ligamentum flavum and the medial edge of the articular process were removed according to the specific conditions of the disease. The nerve root canal and lateral crypt were expanded, and the dural sac and nerve root were protected intraoperatively. Then the annulus fibrosus was cut open, the nucleus pulposus was removed, upper and lower cartilage endplates were removed, and autologous bone particles were implanted between the vertebral bodies. The dural sac and nerve roots were then explored. Finally, the incision was sutured after a negative pressure flow tube was placed. Figure 2B, D shows a schematic diagram and intraoperative photographs of this surgical approach.
Postoperative management
All patients in the two groups had the drainage tube removed within 72 hours and rested in bed for 3 days. Then, they were allowed to ambulate with the assistance of a lumbar brace within at least the next six weeks. All patients were followed up every three months, and X-ray, CT scans were reviewed. Besides ,MRI were followed up in 3 and 6months.
Clinical and radiological assessment
All patients were assessed with Japanese Orthopedic Association (JOA), Visual Analogue Scale (VAS) and Oswestry Disability Index(ODI)scores before and after surgery. In addition, all patients were followed up every 3 months after surgery with X-ray, CT of the lumbar spine. MRI were followed up in 3 and 6months. X-ray and CT were used to calculate the lumbar spine fusion rate through Lee’s radiographic criteria[15, 16].X-ray was also used to measure L1-S1 lumbar lordosis of the standing position[17](Figure 3 A, B). As shown in Figure 4 A,B, MRI was used to detect the fatty degeneration and muscle/vertebral body ratio[18]. VB represents the vertebral body size, CSA represents the cross-sectional area, and SC indicates subcutaneous fat. The calculation of the muscle/vertebral body ratio is also based on the cross-sectional area (CSA) and the vertebral body size (VB). The cross-sectional areas of the vertebral body and paraspinal muscles were outlined and measured by authors using Image J software (National Institutes of Health, MD, USA). Due to the cross-sectional area of the vertebral body would hardly change, the ratio of muscle/vertebral body can reflect the atrophy of paravertebral muscle. The gray-scale range of the CSA and SC areas was also analyzed in Image J software, as shown in Figure 4C, D. The grayscale value of the CSA region overlap with the SC region(Figure 4E)was used as an index of fatty degeneration of the multifidus muscle. In addition, the CSA/VB ratio indicated the degree of multifidus muscle atrophy. The surrounding layers of the lumbar fusion area were chosen to avoid metal interference.In order to unify the standard, the upper edge layer (inferior vertebral endplate) of the intervertebral disc in the upper segment of the fusion segment were selcected for the measurement.
Statistical analysis
All data were analyzed in SPSS 22.0 software (IBM Corporation, NY, USA) and are presented as mean ± standard deviation. Differences between groups were tested by unpaired t test or Man-Whitney U test. Categorical variables were compared via chi-square test or Fisher exact test. Paired t test was used to compare affected side and opposite side within groups.A P-value of < 0.05 was considered statistically significant.