A retrospective study on safety and clinical outcomes of unilateral biportal endoscopic technique for spinal degenerative diseases

Patient population and study design

Patients with clinically and radiographically confirmed SDD in our center from November 2021 to April 2022 were retrospectively reviewed. Pain intensity and functional disability were assessed to evaluate the pathological degree. Preoperative MRI and CT data were obtained for all candidates to confirm the surgical indications and select surgical strategies. The inclusion criteria were preoperative back/leg pain intensity ≥5 on Visual Analogue Score (VAS), intervertebral disc herniation/protrusion, spinal canal stenosis, spondylolisthesis and spondylosis undergoing UBESS, request of minimally invasive therapy, and follow-up with sufficient clinical data. Patients were excluded from the studies for the following reasons: lost to follow-up, contraindications, refusal of surgery, incomplete clinical data, and conservative treatment. Finally, a total of 105 patients were included in this retrospective study.

This research was approved by the Naval Hospital of Eastern Theater ethics committee (number: DHL202105), and informed consent was obtained from all participating patients or their delegated guardians. We adopted the method of having participating patients or their delegated guardians fill out informed consent forms and collected all the consent forms.

Surgical technique

After general anesthesia, the patient was placed in a prone position on the radiolucent table to ensure that the target pathological region was perpendicular to the ground under C-arm fluoroscopy. The initial target anatomical site was the junction of the spinal process and lamina, and the inferior margin of the upper lamina was then confirmed under fluoroscopy and marked with a horizontal line. Two skin incisions were made roughly on the medial margin of the upper and lower pedicles, 1−1.5 cm cranial or caudal to the marked line. The lower incision was for the working portal, and the upper one was for the endoscopic portal. The core dilators were inserted from the working and endoscopic portal through interfascicular space to the junction of the spinal process and lamina. Serial dilators and periosteal strippers were used to expand the corridor and space. Once the triangulation for observation and instrument channel was made, the endoscope combined with the irrigation system and radiofrequency was placed within the established corridor and space. The soft tissue on the lamina and facet was peeled off to expose the inferior margin of the upper lamina and articular process. Sustained hemostasis by radiofrequency was conducted during operation to maintain a clear view. Next, partial laminectomy was performed to remove part of the inferior lamina of the upper lamina until the upper free margin of the ligamentum favum was exposed. The lower margin of the ligamentum favum was detached from the lower lamina with pituitary forceps. Part of the articular process was resected with an electrical drill, Kerrison punch and osteotome, which is necessary for nerve protection (Fig. 1).

For patients with lateral recess stenosis, the lateral recess and nerve root canal were decompressed by ipsilateral flavectomy. For patients with bilateral stenosis, it is necessary to remove the lamina sufficiently to the contralateral side, and the ligamentum flavum to expose the contralateral facet joint and lateral recess for contralateral decompression (Fig. 2).

For patients with lumbar/cervical disc herniation, the ruptured disc was removed routinely, the soft disc particle was removed with pituitary forceps, and the calcified disc material was removed with a Kerrison punch. Internal disc decompression was recommended to reduce the risk of recurrence and chemical inflammation. After discectomy, annuloplasty was performed with a disposable fiber loop suture device (Jinxinxing, Inc., Beijing, China) for a certain group (Fig. 3).

For patients with lumbar spondylolisthesis, the intervertebral space was processed with a reamer and pituitary forceps after discectomy. Then the cartilage endplate was curetted for interbody bone grafting through an infundibular bone graft device. Autologous bone, or allografts were used to promote bony fusion. An interbody cage of appropriate size was placed under endoscopic visualization. After that, the percutaneous screws were implanted into bilateral pedicles under C-arm fluoroscopy. Finally, the pedicle screw position was confirmed and the drainage tube was indwelled before closure (Fig. 4).

Data collection

The data were collected from the hospital admission records up to May 2023. The general information included gender, age, height, weight, body mass index (BMI), symptom duration, cardinal symptom, and follow-up duration. Perioperative data included diagnosis, operative approaches, the number of operative segments, operative level, operation time, postoperative immobilization time, drainage tube indwelling time, postoperative hospital stay, complications, postoperative blood loss (PBL), TBL, IBL, HBL, hematocrit (Hct), Hb, CK and CRP.

Laboratory testing was regularly performed. Preoperative Hb, Hct, CK and CRP were recorded at admission. Postoperative Hb, Hct, CK and CRP were detected on the 2nd day after operation. No blood transfusion was required during or after operation, and thus TBL was calculated based on the perioperative change of Hct (Nash et al., 2016; Gross, 1983). ${\displaystyle \mathrm{TBL}\left(\mathrm{ml}\right)=\text{Patient blood volume}\left(\mathrm{PBV},\mathrm{ml}\right)\times \frac{\left(\text{Hctpre}-\text{Hctpost}\right)}{\left(\text{Hctpre}+\text{Hctpost}\right)}\times 2.}$ In this calculation, Hct_pre and Hct_post are short for pre- and postoperative Hct, respectively.

$\mathrm{PBV}\left(\mathrm{ml}\right)=\left(\mathrm{k}1\times \text{height}{\left(\mathrm{m}\right)}^3+\mathrm{k}2\times \text{weight}\left(\mathrm{Kg}\right)+\mathrm{k}3\right)\times 1000$. (k1= 0.3561, k2= 0.03308, and k3= 0.1833, for women; k1= 0.3669, k2= 0.03219, and k3= 0.6041, for men) (Nadler, Hidalgo & Bloch, 1962).

IBL was determined by the sum of volume differences between suction fluid and irrigation liquid, and the amount of blood absorbed by gauze and towels using visual estimation (Ali Algadiem et al., 2016). PBL comprised the drainage flow and the blood loss visually estimated by soaked gauze. The drainage tube was removed when the drainage flow was less than 50 ml. HBL was calculated by Sehat’s method (Sehat, Evans & Newman, 2004). Therefore, HBL=TBL-IBL-PBL.

Clinical outcomes, such as pain control, functional modification and patient satisfaction, were evaluated by VAS, Oswestry Disability Index (ODI), Japanese Orthopedic Association scores (JOA) and modified MacNab Score, respectively. Back and leg VAS (0–10), and ODI (0–100%) for lumbar disorders were investigated at admission, 1 month, 6 months and 1 year after operation. The improvement rate (%) of JOA (0–17) for cervical diseases was calculated at the final follow-up [improvement rate (%) = (postoperative score –preoperative score)/(17- preoperative score)]. Plain radiography and CT were conducted 1 year after ULIFD. Interbody fusion was assessed via radiographical imaging at the final follow-up. Grade I and Grade II were identified as spinal fusion according to the Bridwell grading system (Bridwell et al., 1995). Patient satisfaction was rated according to the modified MacNab score at the final follow-up, with excellent and good defined as clinically satisfactory.

Basic demographics and perioperative information

Demographic information of the 105 patients who underwent UBESS in our center is shown in Table 1. They included 64 males (60.95%) and 41 females ranging in age from 20 to 79 years, with a mean of 50.57 ± 17.04 (median 51) years. 17.1% of the patients were younger than 30 years, and 12.4% of the patients were older than 70 years. The height and weight of the patients were collected routinely as basic admission information. The mean height was 1.69 ± 0.08 m, and the mean weight was 68.55 ± 9.78 kg. Forty-four patients (41.9%) with the complaints of pain and weakness were admitted as inpatients; 35 patients (33.33%) complained of pain and paresthesia and 26 patients (24.76%) with pain were outpatients of our hospital. The mean symptom duration was 4.03 ± 5.66 years. All the patients with progressive symptoms sought further treatment owing to the failure of conservative treatment.

The surgical information was obtained from the medical records. Of the 105 included patients, 77 (73.33%) were affected by lumbar disc herniation (LDH), 12 (11.43%) with spinal stenosis (SS), 10 (9.53%) with lumbar isthmic spondylolisthesis (LIS), three (2.86%) with cervical spondylotic radiculopathy (CSR), two (1.90%) with lumbar spinal stenosis (LSS), and one (0.95%) with recurrent LDH (RLDH) (Table 2). In consideration of the operative site and disease type, LDH, LDH with SS, LSS and RLDH were considered as the lumbar degenerative disorder (LDD) group. CSR and LIS were classified as an independent group. There were 68 patients with single-level LDD, 22 with two-level LDD, two with three-level LDD, 10 with single-level LIS, and three with single-level CSR. Five operative approaches were employed, including UBE-assisted lumbar discectomy+laminectomy+syndesmectomy (UDLS-L), UBE-assisted lumbar discectomy+laminectomy+syndesmectomy+annuloplasty (ULSA), UBE-assisted laminectomy+syndesmectomy (ULS), UBE-assisted cervical discectomy+laminectomy+syndesmectomy (UDLS-C), and UBE-assisted lumbar interbody fusion+discectomy (ULIFD), of which three (UDLS-L, ULSA and ULS) were applied for LDD, and two (UDLS-C and ULIFD) were used for CSR and LIS.

Compared with single-level LDD, the mean operation time (mOT) was shorter for multilevel LDD with UDLS-L or ULS, and longer for single-level LDD with ULSA than UDLS-L or ULS. All UBESSs were performed by an experienced spinal surgeon.

Regarding complications, pleural effusion occurred in one patient, small dural tear in one patient, and lateral femoral cutaneous nerve injury in one patient in LDD group, the complication rate being 3/92. All the patients with complications in this group received non-surgical treatment for the symptoms, and close follow-up was warranted. In the LIS group, cerebrospinal fluid leakage, wound infection, nerve root compression and pedicle penetration by screws were found in three patients, the complication rate being 3/10. Potent antibiotic, 3rd generation cephalosporin, was given for incision infection after bacterial culture and antimicrobial susceptibility tests. Revision surgery by UBE was performed to adjust pedicle screws and decompression. The complications occurred in the early period when UBESS was firstly carried out. All patients were discharged from our center after recovery. The detailed information is listed in Tables 3, 4 and 5.

Comparisons of perioperative serology and blood loss

Overall, serological changes and blood loss in multilevel LDD were significantly higher than those in single-level LDD, and the serological parameters and blood loss were increased in the treatment of LIS with ULIFD as compared with UDLS-L, ULS, UDLSA and UDLS-C (Tables 6 and 7). There were approximate mean CK and CRP level changes in CSR and single-level LDD except ULS. The perioperative Hb and Hct differences were slightly obvious in LDD with UDLSA, but similar change was found between LDD with UDLS-L and with ULS. The mean TBL in ULS was 273.98 ±  42.59 ml, suggesting that it may be least invasive surgical approach for single-level LDD. The discrepancy was also observed in multilevel LDD. HBL was the major component to be reckoned with, which accounted for more than 70% of TBL in each group. The results of perioperative serology and blood loss are listed in Tables 6 and 7, respectively.

Clinical outcomes of UBESS

The clinical outcomes and patient satisfaction are shown in Table 8. The mean follow-up duration of the series available was 13.14 ± 0.90 months. Postoperative back and leg VAS, ODI and JOA were significantly improved in each disease group. Improvements in pain intensity and lumbar dysfunction one month after operation were remarkable. The improvement rate was 100%, 100% and 66.67% for the three CSR patients at the final follow-up. Evaluation of interbody fusion using the Bridwell grading system showed a fusion rate of 100%. At the last follow-up, the clinical satisfaction rate according to modified MacNab Score was 91.18% for UDLS-L, 92.85% for ULS, 90.00% for UDLSA, 66.67% for UDLS-C, and 90.00% for UDLIF.