Abstract
Objective To compare the clinical efficacy and perioperative outcomes of percutaneous interlaminar endoscopic discectomy (PIED) and unilateral biportal endoscopic discectomy (UBED) in treating L5 to S1 lumbar disc herniation.
Methods This prospective comparative study with quasirandom allocation included 59 patients with diagnosed isolated L5 to S1 disc herniation who underwent either PIED (n = 30) or UBED (n = 29) between January 2023 and June 2024. Surgical time, drain placement duration, drainage volume, postoperative hospitalization duration, visual analog scale (VAS) scores for low back and leg pain, and Oswestry Disability Index (ODI) scores were recorded pre- and postoperatively. Clinical outcomes were assessed using the modified MacNab criteria at the final follow-up.
Results Both PIED and UBED groups demonstrated significant postoperative improvements in lumbar and leg pain VAS scores and ODI scores (P < 0.001). The PIED group showed significantly shorter operative time (58.2 ± 3.1 minutes vs 109.6 ± 4.3 minutes, P < 0.001) and hospital stay (1.0 ± 0.2 days vs 3.1 ± 0.8 days, P < 0.001) compared with the UBED group. Additionally, the PIED group had lower lumbar pain VAS scores at final follow-up (0.6 ± 0.6 vs 1.2 ± 0.9, P < 0.05). No postoperative drainage was required in the PIED group, while the UBED group required a mean drain duration of 2.7 ± 0.7 days with a mean output of 71.6 ± 38.3 mL. No significant differences were observed in final ODI scores or overall clinical outcomes based on the modified MacNab criteria (96.6% vs 86.2%, P = 0.297).
Conclusion Both UBED and PIED are safe and effective techniques for treating L5 to S1 disc herniation. However, PIED is associated with shorter operative time, faster postoperative recovery, and less early postoperative lumbar pain. These findings support the preferential use of PIED in selected patients with isolated L5 to S1 disc herniation.
Clinical Relevance This study provides comparative evidence to guide surgical decision-making for L5–S1 disc herniation. The advantages of PIED—including reduced operative duration, accelerated recovery, and decreased early postoperative pain—may translate into improved patient satisfaction, reduced healthcare resource utilization, and earlier return to daily activities or work. Surgeons can consider PIED as a favorable minimally invasive option for appropriately selected patients.
Level of Evidence 2.
- unilateral biportal endoscopic discectomy
- percutaneous interlaminar endoscopic discectomy
- L5–S1
- lumbar disc herniation
Introduction
A variety of minimally invasive surgical approaches have been developed for the treatment of lumbar disc herniation at the L5 to S1 level, including microdiscectomy under a tubular retractor, percutaneous transforaminal endoscopic discectomy, percutaneous interlaminar endoscopic discectomy (PIED), and unilateral biportal endoscopic discectomy (UBED).1–4 Given that the L5 to S1 level possesses the widest interlaminar space in the lumbar spine and the relatively greater mobility of the S1 nerve root, PIED has been widely adopted, demonstrating safety and efficacy in the treatment of lumbar disc herniation.5,6 Its advantages include minimal injury to the paraspinal musculature, excellent surgical visualization, and effective neural decompression. PIED at the L5 to S1 level can be performed under local, spinal, or general anesthesia. It typically does not require drainage and is associated with a short operative time, high success rate, and low complication rate, allowing patients to be discharged on the same or following day.7,8 These features make it a popular technique among spine surgeons specializing in endoscopic minimally invasive surgery.
UBED has emerged in recent years as another widely adopted endoscopic technique. By separating the viewing and working portals, UBED provides a significantly larger operative field and allows for the use of conventional open surgical instruments under endoscopic guidance.2,4 This expanded working space makes UBED particularly advantageous for treating lumbar spinal stenosis and performing endoscopic-assisted fusion procedures. Nevertheless, UBED requires a larger working corridor than PIED, which may result in greater soft tissue disruption and slower postoperative recovery.4 For L5 to S1 disc herniation, there is ongoing debate regarding the optimal endoscopic technique. In this prospective study, we compared the clinical efficacy and safety of PIED and UBED in the treatment of L5 to S1 disc herniation by evaluating postoperative outcomes in patients undergoing these 2 procedures.
Methods
This study was a prospective comparative study with quasirandom allocation and approved by the Institutional Review Board of Ningbo No. 6 Hospital (IRB No. 202334K). All procedures were performed by a single spine surgeon (G.L.), who is experienced in both techniques. Prior to study initiation, the surgeon had performed approximately 600 PIED procedures and 100 UBED procedures.
Patients
The inclusion criteria were as follows: the presence of unilateral radicular leg pain, ineffectiveness of conservative treatment for more than 2 months, and magnetic resonance imaging (MRI) confirmation of an L5 to S1 central or paramedian type disc herniation consistent with clinical symptoms and neurological signs. Exclusion criteria included foraminal or extreme lateral type disc herniation, lumbar spinal canal stenosis, history of previous lumbar spine surgery, lumbar instability or spondylolisthesis, and lost to follow-up.
Patients were enrolled between January 2023 and June 2024. All eligible participants were informed about the study and provided written informed consent. Baseline assessments included lumbar spine radiographs, computed tomography, MRI, standardized physical examinations, and neurological evaluations. Pain and disability were evaluated using the visual analog scale (VAS) for low back and leg pain and the Oswestry Disability Index (ODI). After completing baseline evaluations, participants were allocated to either group based on the weekday of admission (quasirandom allocation). Patients hospitalized initially on Monday, Thursday, and Friday were assigned to the PIED group, and patients on the other days were assigned to the UBED group. Allocation was determined by weekday of admission and was not concealed; therefore, this was quasirandom rather than true randomized assignment.
Outcome Measures
Surgical time, drain retention time, postoperative drain output, and length of postoperative hospitalization were recorded. VAS scores for low back and leg pain were assessed preoperatively, on postoperative day 2, and at the last follow-up. ODI scores were measured preoperatively and at the last follow-up. The ODI was calculated using a questionnaire comprising 10 scored questions on pain, sensation, function, and impact on daily life; the maximum total cumulative score was 50 points, with a higher ODI indicating worse functionality. The modified MacNab criteria were used to evaluate clinical efficacy at the final follow-up, with outcomes categorized as excellent (complete resolution of symptoms with full restoration of preoperative work and daily activities); good (mild residual symptoms with slight activity limitation that did not impact work or daily life); fair (partial symptom alleviation with significant activity limitation that impaired normal work and daily life); or poor (no symptomatic improvement or even clinical deterioration compared with the pretreatment condition).8
Techniques
UBED
After successful tracheal intubation under general anesthesia, the patient was placed in the prone position on a spine frame. Under C-arm fluoroscopic guidance, the surgical level was confirmed, and surface markings were made to identify the cranial and caudal portals according to established protocols. For the L5 to S1 level, the disc space was aligned perpendicular to the floor using lateral fluoroscopy; adjustments to the surgical bed or padding were made as needed.
Taking the left-sided approach as an example, the cranial portal (approximately 5 mm) was used for the endoscope and continuous saline irrigation. In comparison, the caudal portal (approximately 8 mm) was designated for outflow and surgical instruments. Guidewires and sequential dilators were inserted under fluoroscopic guidance to ensure their tips converged at the junction of the S1 lamina and spinous process. After confirming proper positioning, the guidewires and dilators were removed, and the endoscope (4 mm diameter, 0° field of view, Fangrun Endoscopic System, Shanghai Fang Run Medical Technology, China) was introduced.
Under continuous irrigation, an electrocautery probe was used to expose the surface of the facet joint, lamina, and ligamentum flavum. A high-speed drill and Kerrison rongeur were used to remove parts of the lamina and medial facet joint, followed by resection of the ligamentum flavum to expose the nerve root and herniated disc. The nerve root was carefully retracted and protected while removing the herniated disc and nucleus pulposus, ensuring complete neural decompression. Meticulous hemostasis was achieved, and a negative-pressure drainage tube was routinely placed via the caudal portal to reduce the risk of postoperative hematoma. The wound was then closed in layers (Figure 1).
Preoperative sagittal (A) and axial (B) T2-weighted magnetic resonance images (MRIs) showing a paracentral L5–S1 disc herniation. The patient underwent percutaneous interlaminar endoscopic discectomy under general anesthesia. Anteroposterior C-arm fluoroscopy confirmed accurate positioning of the endoscopic working sleeve (C). Follow-up MRIs (D and E) at 12 months showed complete removal with no evidence of recurrence during the follow-up period, and minimal postoperative changes in paraspinal muscle signal were observed. Intraoperative endoscopic view confirmed adequate decompression of the S1 nerve root (F).
PIED
After successful tracheal intubation under general anesthesia, the patient was placed in the prone position. The surgical level was localized with C-arm fluoroscopy, and the target entry point was identified at the outer boundary of the interlaminar window of L5 to S1. The surgical field was prepared and draped, and under fluoroscopic guidance, a guidewire, working sleeve, and spinal endoscope (6.3 mm outer diameter, 3.7 mm working channel, and 30° field of view; TESSYS Endoscopic System, Joimax Inc., Germany) were inserted.
Under endoscopic visualization, residual muscle and fat overlying the ligamentum flavum were removed. At the lateral margin of the interlaminar window, the ligamentum flavum was incised to expose the S1 nerve root, which was carefully protected. The herniated disc was approached from the superior lateral aspect of the nerve root. If the interlaminar window was too narrow or the lamina obstructed the endoscopic path, part of the lamina and facet joint was removed using a high-speed burr or bone chisel to enlarge the working space. After adequate exposure, the herniated nucleus pulposus was removed, and complete decompression was confirmed. Meticulous hemostasis was performed before the removal of the endoscope and working sleeve. The incision was then closed in layers (Figure 2).
Preoperative sagittal (A) and axial (B) T2-weighted magnetic resonance images (MRIs) showing a paracentral L5–S1 disc herniation. Intraoperative C-arm fluoroscopy in the anteroposterior view demonstrated accurate positioning of the guidewires for the endoscopic viewing and working portals during the unilateral biportal endoscopic discectomy procedure (C). Follow-up sagittal (D) and axial (E) T2-weighted MRIs at 12 months postoperatively showed removal of the herniated disc, with moderate postoperative changes in paraspinal muscle signal intensity compared with preoperative imaging. Intraoperative endoscopic view demonstrated compression of the S1 nerve root by a herniated disc (F).
Postoperative Care
Patients in both groups were allowed to ambulate with a lumbar brace starting on postoperative day 1. A drainage tube was placed in the UBED group and removed once the 24-hour drainage volume was less than 30 mL. No drainage was required in the PIED group.
Statistical Analysis
Data were analyzed using IBM SPSS (version 18.0, IBM Corp., Armonk, NY, USA). For noncontinuous variables, the χ 2 test was used to assess statistical significance. Continuous variables were expressed as mean ± SD. The 2-sample t test was performed to compare variables and results between the 2 groups. Analysis of variance was performed to compare results at different follow-up time points. The least-significant difference method was conducted for post hoc testing. All P values were 2-sided, and those <0.05 were considered statistically significant.
Results
Seventy patients underwent either PIED or UBED for L5 to S1 disc herniation during the study period at our institution. Of these, 59 patients (30 in the PIED group and 29 in the UBED group) met all predefined inclusion and exclusion criteria and were included in the final analysis cohort (Figure 3, Table 1). The reasons for exclusion of the 11 patients were foraminal or extreme lateral disc herniation (n = 3), lumbar spinal canal stenosis (n = 2), history of previous lumbar spine surgery (n = 2), loss to follow-up (n = 3), and discontinued intervention (n = 1). The 2 groups were comparable in terms of age, body mass index, disease duration, preoperative diagnosis, herniation type, and follow-up duration.
The flow diagram illustrates the progression of participants in the 2 groups through enrollment, treatment allocation, follow-up, and data analysis.
Demographic data of the studied patients.
Compared with the UBED group, the PIED group had a significantly shorter operation time (58.2 ± 3.1 vs 109.6 ± 4.3 minutes, t = −9.359, P < 0.001). Postoperative hospital stay was also significantly shorter in the PIED group (1.0 ± 0.2 vs 3.1 ± 0.8 days, t = −12.46, P < 0.001). Furthermore, the UBED group had a mean drain placement duration of 2.7 ± 0.7 days, with a postoperative drain output of 71.6 ± 38.3 mL. In contrast, no postoperative drainage was required in the PIED group.
The patients included in this study achieved overall good to excellent functional outcomes (Table 2). No severe perioperative or postoperative complications were observed during the follow-up period, such as major dural tears, infection, hematoma, neurological deficit, reoperation, readmission, or thromboembolism, except for 1 intraoperative case in each group involving a minor epineurial injury to the nerve root, which did not result in any new postoperative neurological symptoms.
Comparison of lumbar function and neurological recovery between the 2 groups.
The mean preoperative ODI and VAS scores did not differ significantly between the 2 groups. Both groups demonstrated significant postoperative improvement in ODI and VAS scores (P < 0.001). In the PIED group, the mean VAS score for lumbar pain decreased significantly from 3.5 ± 1.1 preoperatively to 1.2 ± 0.9 on postoperative day 2 and further to 0.6 ± 0.6 at the final follow-up (P < 0.001). The mean leg pain VAS score also decreased significantly from 6.0 ± 1.3 to 1.0 ± 0.8 at the final follow-up (P < 0.001). In the UBED group, the mean lumbar pain VAS score decreased from 3.5 ± 1.3 preoperatively to 2.0 ± 0.9 on postoperative day 2 and further to 1.2 ± 0.9 at the final follow-up, while the leg pain VAS score decreased from 6.0 ± 1.1 to 1.0 ± 0.7 (P < 0.001).
There were no significant differences between the 2 groups in leg pain VAS scores on postoperative day 2 or at the final follow-up. However, postoperative lumbar pain VAS scores were significantly lower in the PIED group compared with the UBED group (1.2 ± 0.9 for PIED vs 2.0 ± 0.6 for UBED, P < 0.05).
No significant difference was observed in postoperative ODI scores between the 2 groups at the final follow-up. Based on the modified MacNab criteria, both groups achieved good to excellent outcomes, with no statistically significant difference in success rates (96.6% vs 86.2%, P = 0.297). These findings suggest that although the UBED group tended to experience more postoperative low back pain than the PIED group, it did not appear to impair lumbar function significantly.
Discussion
Both groups demonstrated favorable postoperative lumbar functional outcomes, with no observed severe complications or recurrences during a mean follow-up of 13.5 months (range from 12 to 17 months). The primary findings of this study indicate that both PIED and UBED are effective surgical options for the treatment of L5 to S1 lumbar disc herniation. No significant difference was observed in postoperative ODI scores and patient satisfaction rates between the 2 groups. However, the PIED group had significantly shorter operative times, reduced postoperative hospital stays, and lower postoperative lumbar pain VAS scores compared with the UBED group. These findings suggest that PIED may offer superior early postoperative recovery and reduced surgical invasiveness compared with UBED for appropriately selected patients with isolated L5 to S1 disc herniation.
PIED is a single-channel endoscopic technique that permits direct access to the ligamentum flavum. Following its incision, the S1 nerve root is exposed, enabling targeted discectomy. Due to the relatively large interlaminar window at the L5 to S1 level, averaging 27.7 ± 4.0 mm, endoscope placement is typically achieved without the need for additional facet joint resection.6 Conversely, UBED requires exposure of both the lamina and facet joints, which broadens the surgical field and increases soft tissue dissection, potentially leading to greater paraspinal muscle trauma. Its advantage lies in the expanded operative workspace, allowing the use of standard open surgical instruments and making it especially suitable for cases with spinal canal stenosis or a narrow interlaminar space. For routine disc herniation at L5 to S1, however, PIED may offer greater procedural efficiency and patient comfort.
Our findings agree with previous comparative studies.2,4,9,10 Wang et al9 reported no significant differences in ODI scores at postoperative day 1, 3 months, and final follow-up between the UBED and PIED groups, nor were there notable differences in severe complications or recurrence rates during a mean follow-up of 13.5 months. Similarly, Yin et al10 observed comparable ODI scores and patient satisfaction rates according to the modified MacNab criteria. Wei et al4 also reported no significant differences in ODI scores at 1-year follow-up when comparing PIED and UBED in patients with L4 to L5 or L5 to S1 disc herniations.
However, unlike those studies, we observed significantly lower lumbar pain VAS scores in the PIED group at final follow-up (P < 0.05), despite no between-group difference in leg pain VAS scores. Notably, on postoperative day 2, the lumbar pain VAS score in the PIED group was significantly lower than in the UBED group (1.2 vs 2.0, P = 0.001), suggesting greater surgical trauma to paraspinal tissues in the UBED cohort.
This interpretation is supported by Wang et al,9 who found elevated creatine phosphokinase levels in the UBED group on postoperative day 1, as well as greater paraspinal muscle damage on MRI at both postoperative day 1 and 3 months. Unlike PIED, which exposes the nerve root by incising the ligamentum flavum without lamina or facet resection, UBED often requires partial facetectomy. Yin et al10 noted that nearly all patients undergoing UBED required partial lamina fenestration or facet joint resection, which may lead to iatrogenic instability and chronic low back pain.
Additionally, the PIED group in our study had significantly shorter operative times (58.2 vs 109.6 minutes) and postoperative hospital stays (1.0 vs 3.1 days) compared with the UBED group—findings consistent with Liu et al. In contrast, Wang et al9 reported similar operative times (83.6 vs 80.2 minutes) and hospital stays (2.1 vs 2.0 days) between the 2 groups, with longer times reported for PIED, possibly reflecting differences in surgical expertise. Yin et al10 also reported similar times and stays (80.9 vs 89.4 minutes; 5.5 vs 5.2 days), suggesting that surgeon proficiency may affect these parameters. Because all operations were performed by a single surgeon (G.L.), intersurgeon variability was minimized; however, differential cumulative experience between techniques may have influenced perioperative parameters, and learning-curve effects may partially explain the longer operative time and greater use of postoperative drainage in the UBED cohort.11,12
Postoperative drainage is routinely required following UBED. In our cohort, the UBED group had a mean drain duration of 2.7 ± 0.7 days and an average postoperative drainage volume of 71.6 ± 38.3 mL, consistent with findings by Qian et al.2 In contrast, no drainage was required in the PIED group. Despite these differences, UBED is more effective in treating lumbar spinal canal stenosis due to its superior decompressive capability, while PIED is more suited for isolated disc herniation.13–15 All patients in this study had single-level disc herniations without significant canal stenosis.
Limitations
This study has several limitations. First, allocation by weekday may introduce selection bias related to scheduling, referral patterns, or perioperative logistics. Second, the sample size was relatively small, which may limit the statistical power and generalizability of the findings, although the total sample (n = 59) is reasonable for a single-center prospective series and is consistent with prior comparative studies.16 Third, the follow-up duration was relatively short, with a mean of 13 months. Although no cases of postoperative recurrence at a mean follow-up of 13.5 months were observed in either group during the follow-up period, longer-term outcomes—including recurrence rates and potential delayed complications—remain to be elucidated. Fourth, although surgeries were performed by experienced spine surgeons, potential variability in surgical technique and postoperative care protocols could have introduced confounding factors that may affect the outcomes. Single-surgeon design may limit generalizability. Last, the study focused exclusively on patients with isolated L5 to S1 disc herniation, which may limit the applicability of the findings to patients with multilevel disc disease or concomitant lumbar spinal stenosis. Further prospective multicenter studies with larger patient cohorts and longer follow-up durations are warranted to confirm these findings, evaluate long-term outcomes, and refine the selection criteria for these 2 minimally invasive techniques.
Conclusion
Both UBED and PIED are safe and effective techniques for treating L5 to S1 disc herniation. However, PIED is associated with shorter operative time, faster postoperative recovery, and less early postoperative lumbar pain. These findings support the preferential use of PIED in selected patients with isolated L5 to S1 disc herniation.
Footnotes
Funding This study was financially supported by the Ningbo Clinical Research Center for Orthopedics, Sports Medicine & Rehabilitation (2024L004).
Declaration of Conflicting Interests The authors report no conflicts of interest in this work.
- This manuscript is generously published free of charge by ISASS, the International Society for the Advancement of Spine Surgery. Copyright © 2026 ISASS. To see more or order reprints or permissions, see http://ijssurgery.com.
References
- 1.↵Choi KC , Shim HK , Hwang JS , et al . Comparison of surgical invasiveness between microdiscectomy and 3 different endoscopic discectomy techniques for lumbar disc herniation. World Neurosurg. 2018;116:e750–e758. 10.1016/j.wneu.2018.05.085
- 2.↵Qian J , Lv X , Luo Y , Liu Y , Jiang W . Unilateral biportal endoscopic discectomy versus percutaneous endoscopic lumbar discectomy in the treatment of lumbar disc herniation linked with posterior ring apophysis separation: a retrospective study. World Neurosurg. 2025;193:957–963. 10.1016/j.wneu.2024.09.102
- 3.↵Park DY , Upfill-Brown A , Curtin N , et al . Clinical outcomes and complications after biportal endoscopic spine surgery: a comprehensive systematic review and meta-analysis of 3673 cases. Eur Spine J. 2023;32(8):2637–2646. 10.1007/s00586-023-07701-9
- 4.↵Wei WB , Dang SJ , Liu HZ , Duan DP , Wei L . Unilateral biportal endoscopic discectomy versus percutaneous endoscopic interlaminar discectomy for lumbar disc herniation. J Pain Res. 2024;17:1737–1744. 10.2147/JPR.S449620
- 5.↵Fu M , Li Q , Xu Y , et al . Variation in spatial distance between the lumbar interlaminar window and intervertebral disc space during flexion-extension. Surg Radiol Anat. 2021;43(9):1537–1544. 10.1007/s00276-021-02809-3
- 6.↵Sakçı Z , Önen MR , Fidan E , Yaşar Y , Uluğ H , Naderi S . Radiologic anatomy of the lumbar interlaminar window and surgical considerations for lumbar interlaminar endoscopic and microsurgical disc surgery. World Neurosurg. 2018;115:e22–e26. 10.1016/j.wneu.2018.03.049
- 7.↵Liu G , Zhang J , Zhang L , Yuan L , Wang X , Tursunmamat D . Comparison of spinal anesthesia and local anesthesia in percutaneous interlaminar endoscopic lumbar discectomy for L5/S1 disc herniation: a retrospective cohort study. BMC Musculoskelet Disord. 2024;25(1):774. 10.1186/s12891-024-07898-w
- 8.↵Liu G , Zhao J , Yuan L , Shi F , Zhang L . Spinal anesthesia for L5-S1 interlaminar endoscopic lumbar discectomy: a retrospective study. BMC Musculoskelet Disord. 2023;24(1):818. 10.1186/s12891-023-06956-z
- 9.↵Wang L , Li C , Han K , Chen Y , Qi L , Liu X . Comparison of clinical outcomes and muscle invasiveness between unilateral biportal endoscopic discectomy and percutaneous endoscopic interlaminar discectomy for lumbar disc herniation at L5/S1 level. Orthop Surg. 2023;15(3):695–703. 10.1111/os.13627
- 10.↵Yin J , Gao G , Chen S , Ma T , Nong L . Comparative study between unilateral biportal endoscopic discectomy and percutaneous interlaminar endoscopic discectomy for the treatment of L5/S1 disc herniation. World Neurosurg. 2025;194:123526. 10.1016/j.wneu.2024.11.109
- 11.↵Gadjradj PS , Rubinstein SM , Peul WC , et al . Full endoscopic versus open discectomy for sciatica: randomised controlled non-inferiority trial. BMJ. 2022;376:e065846. 10.1136/bmj-2021-065846
- 12.↵McNamee C , Keraidi S , McDonnell J , et al . Learning curve analyses in spine surgery: a systematic simulation-based critique of methodologies. Spine J. 2024;24(10):1789–1810. 10.1016/j.spinee.2024.05.014
- 13.↵Jiang H-W , Chen C-D , Zhan B-S , Wang Y-L , Tang P , Jiang X-S . Unilateral biportal endoscopic discectomy versus percutaneous endoscopic lumbar discectomy in the treatment of lumbar disc herniation: a retrospective study. J Orthop Surg Res. 2022;17(1):30. 10.1186/s13018-022-02929-5
- 14.↵He D , Cheng X , Zheng S , et al . Unilateral biportal endoscopic discectomy versus percutaneous endoscopic lumbar discectomy for lumbar disc herniation: a systematic review and meta-analysis. World Neurosurg. 2023;173:e509–e520. 10.1016/j.wneu.2023.02.087
- 15.↵Meng H , Su N , Lin J , Fei Q . Comparative efficacy of unilateral biportal endoscopy and micro-endoscopic discectomy in the treatment of degenerative lumbar spinal stenosis: a systematic review and meta-analysis. J Orthop Surg Res. 2023;18(1):814. 10.1186/s13018-023-04322-2
- 16.↵Ye XF , Wang S , Wu AM , et al . Comparison of the effects of general and local anesthesia in lumbar interlaminar endoscopic surgery. Ann Palliat Med. 2020;9(3):1103–1108. 10.21037/apm-20-623
