Propensity matched outcome analysis following microdiscectomy versus interlaminar endoscopic discectomy for L5-S1 disc herniation

Abstract

Purpose

The main purpose of the study is to perform a propensity-matched functional outcome analysis following microdiscectomy (MD) versus interlaminar endoscopic lumbar discectomy (IELD) for L5-S1 disc herniation. Although many studies have compared endoscopic lumbar discectomy and microdiscectomy, few have compared the outcomes of microdiscectomy (MD) and interlaminar endoscopic discectomy (IELD) at the L5-S1 level.

Methods

This is a propensity-matched analysis of 100 patients (50 MD patients, 50 IELD patients) based on baseline covariates with a minimum of one-year follow-up. Patient-reported outcome measures were obtained from EMR during follow-up visits. Back pain and sciatic pain were assessed by the Visual Analogue Scale (VAS-B and VAS-L). Functional outcome was assessed using Oswestry Disability Index (ODI) Score and 12-item Short Form Survey (SF-12) score. Data were obtained at baseline (pre-op) and at 0, 1, 3, and 12 months post-operatively.

Results

Mean operative time was significantly lower (p < 0.001) in the IELD group (44 min) compared to the MD group (59 min). Mean VAS-B at the immediate and 1-month postoperative period was significantly (p < 0.001) lower in the IELD group (0.36, 0.24) when compared with the MD group (1.74, 1.16). There was no significant difference between IELD and MD groups with regard to improvement in sciatic pain (VAS-L). ODI scores at 1 month and 3 months post-operative period were significantly (p < 0.001) lower in the IELD group (30.1, 23.2) when compared with the MD group (41, 27.5). However, there was no significant difference between the two groups with regards to VAS-B, ODI, and SF-12 at 1-year follow-up.

Conclusion

Our findings indicate that the IELD group achieved better immediate and early postoperative outcomes despite no significant difference at one-year follow-up.

Highlights

• •Although many studies have compared endoscopic lumbar discectomy and microdiscectomy, few have compared the outcomes of microdiscectomy (MD) and interlaminar endoscopic discectomy (IELD) at the L5-S1 level.
• •Moreover, there is no propensity-matched study comparing endoscopic and microdiscectomy at the L5-S1 level in the literature.
• •Hence, we aim to study a propensity-matched outcome analysis following MD versus IELD for L5-S1 disc herniation.

1. Introduction

Lumbar disc herniation (LDH) is a common degenerative disorder of the spine with a reported incidence of 49% at L4-L5 and 45.2% at L5-S1 levels.¹ Microdiscectomy (MD), is considered the gold standard surgical procedure for LDH in patients not responding to conservative treatment.² However, with changing trends, minimally invasive procedures like endoscopic discectomy have now gained acceptance. Studies have compared these two procedures with variable outcomes. In a systematic review of 1004 patients by Renato et al., clinically better outcome was observed in endoscopic discectomy compared to microdiscectomy with regard to sciatic pain3. In a meta-analysis by Zhang et al., 1527 patients were analyzed, which revealed no difference in sciatic pain and ODI score between microdiscectomy and endoscopic discectomy.⁴

Transforaminal endoscopic lumbar discectomy (TELD) is preferred in patients with lumbar disc herniation at L4-L5 and above levels. However, elevated iliac crests, the angle of sacral inclination, foraminal anatomy at L5–S1, and the hypertrophic transverse process of L5 make TELD technically challenging at L5-S1.⁵ Hence, interlaminar endoscopic lumbar discectomy (IELD) is favored over TELD at the L5-S1 level due to its various advantages. The IELD uses the posterior interlaminar approach, which is more in line with the way most spine surgeons operate and makes it easier to identify the anatomic orientation. It can dissect the extruded or prolapsed disc tissues under complete direct vision and completely decompress the central- and paracentral-type disc herniation. Moreover, IELD can aim the L5-S1 disc space easily through the large interlaminar window under simple fluoroscopy guidance, and provide better mobility to remove the sequestered or dislocated fragments without the limitation of the bony foramen and blockage of the pelvis.^6,7

In view of the paucity in the literature comparing these two procedures at the L5-S1 level, we felt it was necessary to perform a propensity-matched study comparing MD and IELD for LDH at L5-S1 to assess the immediate, early, and long-term outcomes.

2. Materials and methods

This single-center study was conducted after obtaining approval from Institutional Review Board. Informed consent was waived due to the retrospective nature of the study. Patients who had undergone single-level discectomy for failed conservative management of L5-S1 disc herniation formed the study group. One hundred fifty-seven discectomy procedures were performed in our hospital between October 2019 to March 2021 for L5-S1 disc herniation. To avoid selection bias between the 2 groups, we set up several inclusion and exclusion criteria. Patients were included if they met the following criteria: (1) they had symptomatic radiating lower limb pain that was more prevalent than back pain, and a positive straight leg-raising test; (2) their CT or MRI scans showed a single-level soft disc herniation at the L5-S1 level, which was consistent with their clinical findings; (3) they had tried the conservative treatment for 6 weeks without seeing significant improvement; and (4) they had no prior lumbar surgical history. People who fit one or more of the following criteria were excluded: (1) a hard (calcified) disc herniation on an MRI or CT scan, (2) a foraminal or extraforaminal disc herniation, (3) a history of prior lumbar spine surgery, (4) instability or spondylolisthesis, or (5) an inability to complete the pre-and post-operative questionnaires or inadequate medical records.

Demographic variables such as age, sex, height, weight, and BMI were recorded for all study patients which were retrieved from the electronic medical records (EMR) system. Propensity score matching (PSM) was used to match the two groups in the proportion of 1:1 to facilitate the comparison of age, gender, BMI, and comorbidities (hypertension, diabetes mellitus, and hypothyroidism).⁸ Based on the above criteria we included 100 patients (50 patients of the MD group, and 50 patients of the IELD group) aged between 18 and 70 years, with low back pain and unilateral radiculopathy with MRI confirmation of LDH at L5-S1 level (Fig. 1).

Fig. 1.

Flowchart depicting the patient selection process.

2.1. Outcome measures

Primary outcome measures were back pain and sciatic pain. Back pain was assessed by the Visual Analogue Scale for back pain (VAS-B) and sciatic pain was assessed by the Visual Analogue Scale for leg pain (VAS-L). Functional outcome was assessed using Oswestry Disability Index (ODI) score (ranging from 0 to 100, with higher scores indicating more disability) and a 12-item Short Form Survey (SF-12) Questionnaire.^9,10 Data were obtained at baseline (pre-op) and at 0, 1, 3, and 12-month post-operative periods. Time to discharge (in days) and duration of the procedure (in minutes) were noted for all patients. Patient-reported outcome measures were obtained from EMR during follow-up outpatient visits.

2.2. Surgical techniques

The surgeries were performed by a single surgeon formally trained and highly experienced to perform IELD and MD procedures.

Interlaminar endoscopic lumbar discectomy (IELD): Under general anesthesia, patients were positioned prone on a bow-type frame that was firmly fixed to the operating table in order to widen the interlaminar window. The surgical segment's lumbar process was superficially found to indicate the posterior midline under anterior-posterior fluoroscopy. The skin incision was made as close to the medial in the craniocaudal middle of the interlaminar window as possible. At the location where the puncture needle entered, a 0.8 cm incision was created using a pointed scalpel. A dilator is inserted through the incision and advanced laterally until the facet joint is reached. The anterior-posterior and lateral radiographs are used to locate the point of the dilator after contacting the facet joint. The multifidus muscle is dissected blindly from the cranial and caudal lamina with a dilator. To reach the lateral periphery of the nerve root, the trajectory of the approach is not vertical to the spinal canal, but rather oblique to the base of the medial facet joint. After dissecting the multifidus muscle with the dilator, a beveled-type working tube with the open side facing the medial side is inserted along the dilator. The endoscope is positioned within the working passage. The border between the capsule of the facet joint and the LF is identified endoscopically by colour, palpation, and coagulation. The border is incised, and the working tube is advanced along the medial border of the facet joint. Due to the angle of the facet joint and the location of the nerve root below, the surgical trajectory of the working tube and endoscope are angled medially, and the LF is separated from the medial facet joint using a dissector and the working tube's blade. A dissector was used to divide the ligamentum flavum vertically, and the working channel was inserted into the epidural space to identify the fat tissue, dural sac, and S1 roots. Under direct visual control and continuous irrigation, the working channel was rotated and inserted into the shoulder or axilla of the nerve root to act as a nerve hook. The extruded disc was then visible, and a sequestrectomy was done. Annulus fissure coagulation and hemostasis were performed using the bipolar radiofrequency at 15 W for coagulation. After removal of the scope, skin was sutured with a single stitch.

Microdiscectomy: A 2.5–4.0 cm midline skin incision was made, followed by periosteal dissection and the application of a Mcculloch retractor. After performing a partial laminectomy and removing the ligamentum flavum under microscopic guidance, the ruptured disc was removed, either with or without disc space evacuation. After inserting a drain, if necessary, the incision was finally stitched layer by layer.

Equipment selection, bone preparation, and intervertebral disc evacuation in both groups were based on preoperative planning and/or intraoperative findings.

2.3. Statistical analysis

Data were analyzed using SPSS v21. Using the multivariate logistic regression model, a propensity score was computed. The descriptive statistics of demographic and outcome variables were presented as mean ± SD for continuous variables and proportion and frequency for categorical variables. Continuous variables were analyzed using unpaired t-test and categorical variables were analyzed using either the Chi-square test or Fisher's exact test. One-way ANOVA test was used as the test of significance for comparison between groups for continuous data and LSD test was used for post-hoc analysis. P-value <0.05 was considered statistically significant.

3. Results

3.1. Demographic characteristics

The mean age of patients in the IELD group was 38 years and in the MD group was 39 years. There were 30 males and 20 females in the MD group, and 33 males and 17 females in the IELD group. A significant difference in mean operation time (p < 0.001) was noted between the MD group (59 min) and the IELD group (44 min) (Fig. 2).

Fig. 2.

Variation in mean operation time between two groups.

3.2. Outcome measures

Mean Preop VAS-B was 2.78 and 2.46 in the IELD and the MD group respectively. On the first postoperative day, the mean VAS-B was 0.36 in the IELD and 1.74 in the MD group which was statistically significant (p < 0.001). There was a significant difference (p < 0.001) in VAS-B at 1-month post-op between IELD (Mean VAS-B = 0.24) and MD group (mean VAS-B = 1.16) (Table 1). Mean VAS-B continued to be lower in the IELD group (1.46) compared to the MD group (1.58) although not significant at 12 months follow-up (Fig. 3).

Table 1.

Comparison of VAS-B between two groups.

	MD (n = 50)		IELD (n = 50)		P value
	Mean	±SD	Mean	±SD
Pre VAS-B	2.78	0.95	2.46	1.08	0.22
Post op VAS-B	1.74	1.12	0.36	0.53	<0.001
1 month VAS-B	1.16	0.91	0.24	0.52	<0.001
3 month VAS-B	0.94	1.1	0.58	0.8	0.23
12 month VAS-B	1.58	1.5	1.46	1.4	0.92

Fig. 3.

Mean VAS-B at each assessment point between the two groups.

Sciatic pain was measured using the Visual Analogue Scale for leg pain (VAS-L). The mean pre-op VAS-L between the two groups was similar with an average of 7.55 (Table 2). Postoperatively, mean VAS-L was lower in both groups (mean = 1.02) and it continued to be lower at 12 months post-surgery. There was no statistically significant difference between the IELD and the MLD group with respect to VAS-L. (Fig. 4)

Table 2.

Comparison of VAS-L between two groups.

	MD (n = 50)		IELD (n = 50)		P value
	Mean	±SD	Mean	±SD
Pre VAS-L	7.54	0.5	7.56	0.61	0.6
Post op VAS-L	0.72	1.05	1.32	0.65	0.07
1 month VAS-L	0.32	0.68	0.76	0.85	0.06
3 month VAS-L	0.36	0.88	0.82	1.16	0.25
12 month VAS-L	0.9	1.58	1.26	1.84	0.60

Fig. 4.

Mean VAS-L at each assessment period between two groups.

Functional outcome was assessed using ODI Score. The mean preop ODI score was 67 in both groups (Table 3). There was a statistical difference (p < 0.001) in the first month (IELD = 30.1; MD = 41) and 3rd month ODI score (IELD = 23.2; MD = 27.5) between the two groups. (Fig. 5). However, no such difference was noted at 12 months follow-up. The mean duration of hospital stay was 3 days in the IELD group and 3.62 days in the MD group.

Table 3.

Comparison of ODI scores between two groups.

	MD (n = 50)		IELD (n = 50)		P value
	Mean	±SD	Mean	±SD
Pre ODI	67.1	3.5	67.6	3.7	0.07
1 month ODI	40.9	5.8	30.15	3.2	<0.001
3 month ODI	27.44	4.4	23.23	3.0	<0.001
12 month ODI	12.76	8.6	13.71	8.4	0.89

Fig. 5.

Mean ODI score at each assessment period between two groups.

SF-12 (12-item short-form health survey) was calculated at 1-year follow-up. The mean SF-12 score was 30.8 in the IELD group and 30.48 in the MD group, which was not statistically significant (p = 0.52) (Fig. 6).

Fig. 6.

SF-12 between two groups at 1 year follow up.

3.3. Complications

In the MD group, we had one dural tear and one recurrence. The dural tear was managed successfully by dural repair with duragen application and watertight closure. One patient with recurrence at 4 months underwent a revision microdiscectomy. In the IELD group, we had one recurrence at 3 months managed by revision discectomy.

4. Discussion

It is well-known that microdiscectomy (MD) is a conventional surgical procedure for LDH and has adequate supporting data that has been published since the late 1970s.¹⁰ On the other hand, interlaminar endoscopic lumbar discectomy (IELD), a comprehensive endoscopic spine procedure for disc herniation at the L5-S1 level, has gained popularity around the globe as a minimally invasive option for MD since the mid-2000s.¹¹

Due to the technology utilized and different anatomic perspective, endoscopic surgery differs significantly from traditional microsurgery. The use of a narrow and magnified 2-dimensional endoscopic view, the inability to reach the interlaminar space, the use of a percutaneous small one port, the presence of a vague or obscured view due to epidural bleeding, the worry of dura tearing and neural injury, or uncertainty of successful decompression and disc space evacuation are some of the challenges faced the operating surgeon performing IELD at the L5-S1 level. Additionally, there are certain differences between the surgical indications for the two surgical methods; for example, IELD has limited indications than microdiscectomy.¹² Foraminal/extraforaminal disc herniation, upward disc migration, and hard disc with calcification are a few contraindications of IELD. Due to the aforementioned factors, the surgeon may have concerns when doing IELD, such as surgical failure or conversion to microsurgery, the steep learning curve, problems at the novice stage, and a higher recurrence rate than with MD.¹³

There are currently no conclusive findings on the differences between IELD and MD for the surgical treatment of symptomatic LDH at the L5-S1 level. The findings of our study suggest that both IELD and MD are effective and safe surgical procedures for the management of LDH. Compared with MD, IELD is associated with shorter incisions, lesser intraoperative blood loss, better early pain relief, and functional recovery.

The results of our study highlight that the mean operative time in the IELD group (44 min) was significantly shorter than the MD group (59 min). In microdiscectomy, layer-by-layer surgical dissection followed by stepwise wound closure and the occasional need for removal of bony lamina might result in longer operation time.¹³ However, studies by Yu et al. and Kim et al. reported that there was no significant difference in the operative time between endoscopic discectomy and microdiscectomy.^14,15

In our study, the IELD group had a significant (p < 0.001) decrease in back pain in the early postoperative period compared to those who had undergone MD which can be attributed to the minimal tissue dissection and the absence of muscle retraction in the IELD group⁵. At 1 year follow-up, back pain continued to be lower in the IELD group compared to the MD group though not statistically significant. Our findings were similar to the study by Ruetten et al. in which endoscopic technique had advantages in back pain, rehabilitation, and less traumatization.⁷ Similarly, Song et al. compared interlaminar endoscopic discectomy and microdiscectomy at L5-S1 and observed comparatively low postoperative back pain in the endoscopic group.¹⁴ In a randomized clinical study, Meyer et al. compared percutaneous endoscopic lumbar discectomy to conventional microdiscectomy in terms of postoperative pain, disability, and complications.¹⁵ Their study's findings showed that, during the first three months following surgery, endoscopic discectomy was associated with less severe postoperative back pain than microdiscectomy. Sciatic pain improved from baseline levels in all study patients in the immediate postoperative period. There was no significant difference in outcome between the two groups with regard to improvement in sciatic pain which was similar to our study results. In a retrospective study by Yang et al., there were no intergroup differences in leg VAS scores over the 24- month follow-up, which was similar to our study.¹⁶

Our findings showed no significant differences between the 2 groups for the long-term clinical outcomes, such as relief of back or leg pain and patient functional outcome. However, the IELD group performed better than the MD group in terms of immediate postoperative back pain and early functional outcome following surgery. A small wound, measuring 0.5 cm in the IELD group versus 2.5–4.0 cm in the MD group, and minimal muscle or soft tissue damage in the IELD group may be the reason for differences in immediate postoperative clinical outcomes between the two groups.

We observed a marginally shorter duration of hospitalization in the IELD group compared to the MD group. Our hospital is a tertiary center; patients come from distant places, and the duration of hospitalization is based on the patient's convenience. Hence not much relevance was given to this observation. Similarly, a meta-analysis by Bin Zhang et al. comparing endoscopic discectomy with microdiscectomy in 1527 patients reported that the length of hospital stay was shorter in the endoscopic group.¹⁷

ODI, a daily life assessing tool, is widely used to evaluate patients with spinal disorders, and the correlation of ODI scores with pain scores has been demonstrated by previous studies.¹⁸ In our study, there was a gradual improvement in the ODI score in both the MD and IELD groups. There was a statistical difference (p < 0.001) in the 1st and 3rd monthly ODI scores between the two groups. The IELD group had a better functional outcome in the early postop period, as evidenced by a considerably larger reduction in ODI score that was seen up until three months after surgery. This correlated with a study by Yu et al. in which similar outcomes were noticed between endoscopic discectomy and microdiscectomy groups.¹⁹

In our study, no patient in the IELD group underwent conversion to open surgery owing to the surgeon's sufficient endoscopic experience, proper patient selection, and specific radiographic examination. No serious complications such as dural tears or nerve root injury occurred in the IELD group. However postoperative hypoesthesia was noted in 10% of patients in the IELD group which improved gradually over a period of three months. Similarly, in their study, Yang et al. reported an incidence of postoperative leg numbness in 16.2% for endoscopic discectomy, which recovered over a period of 6–10 months. The probable cause for postoperative hypoesthesia in patients undergoing endoscopic discectomy could be related to irritation of the spinal ganglion or nerve root during surgery due to repeated hemostasis by bipolar coagulation and compression of the nerve root by the working channel.¹⁶

We performed propensity-score matching of the baseline characteristics between the two groups to minimize the effects of confounding caused by the non-randomized assignment of the surgical procedures. This was done to reduce the impact of selection bias and allow relevant variables to be evenly balanced between the two groups. However, there are many limitations to our study. One major limitation is the innate flaw existing with retrospective studies in which information may be not accurately recorded. Second, because of the smaller sample and the higher probability of error, the clinical efficacy lacks concurrent controls, and the data have a certain limitation. Third, the study did not include the radiographic parameters, which could have impacted the study's outcome. Fourth, the follow-up period is relatively short, and long-term results require further observation. Finally, propensity matching can never fully simulate the conditions of a randomized-controlled study because the matching algorithm can only use variables that have already been observed. Even after matching, unobserved (or unreported) variables may still be out of balance and not accounted for.

5. Conclusion

Interlaminar endoscopic discectomy and microdiscectomy achieve favorable clinical results for lumbar disc herniation at the L5-S1 level. Given the benefits of reduced hospitalization, quicker pain relief, and early functional rehabilitation, IELD should still be regarded as superior to MD. However, the outcomes of this study's findings may be further validated by more prospective randomized controlled trials.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Institutional ethical committee approval

The project was performed in accordance with the ethical principles mentioned in the Declaration of Helsinki. Ethical committee approval was obtained for the study.

Credit author statement

Ajoy Prasad Shetty: Conceptualization, Writing- Review & Editing. Thirumurugan Arumugam: Data Curation, investigation, Writing- Original draft. Karthik Ramachandran: Formal analysis, Writing- Original draft.K. S. Sri Vijay Anand: Writing- Review & Editing. Jalaj Meena: Data Curation, investigation. Rishi Mugesh Kanna: Resources, Data curation. Rajasekaran Shanmuganathan: Supervision, Writing- Review & Editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is/are the supplementary data to this article.