Pang Hung Wu and Eugene Tze-Chun Lau contributed equally to this study as co-first authors.
There is a lack of literature on indirect decompression in uniportal endoscopic posterolateral transforaminal lumbar interbody fusion (EPTLIF). Our aim is to evaluate the dimensions of the spinal canal and contralateral foramen before and after EPTLIF.
This is a retrospective study of patients who underwent EPTLIF in a tertiary spine centre over a 2-year period. The cross-sectional area of the spinal canal and the contralateral foramen at the level of fusion were measured on magnetic resonance imaging scan at 1-day postoperation and at the final follow-up. Patients were grouped according to the decompression performed as per the clinician’s judgement.
One hundred fifty-two levels of fusion were performed in 120 patients. There was a statistically significant clinical improvement in visual analogue scale and Oswestry Disability Index scores postoperation. The measurements of the spinal canal area were 106.0 mm2, 138.8 mm2, and 195.5 mm2; while contralateral foraminal area were 73.2 mm2, 104.4 mm2, and 120.7 mm2 at preoperation, 1-day postoperation, and at the final follow-up, respectively (p < 0.001). For the subgroup analyses, spinal canal area measurements for the bilateral decompression cohort (n = 35) were 57.0 mm2, 123.9 mm2, and 191.8 mm2; for the ipsilateral decompression cohort (n = 42) were 89.3 mm2, 128.9 mm2, 183.3 mm2; and for the cohort without any decompression and only cage inserted (n = 75) were 138.3 mm2, 151.2 mm2, and 204.1 mm2 (p < 0.001). Contralateral foraminal area measurements were 73.3 mm2, 106.4 mm2 and 120.4 mm2 in the bilateral decompression cohort; 69.5 mm2, 99.0 mm2, 116.9 mm2 in the ipsilateral decompression cohort; and 75.1 mm2, 106.5 mm2, 122.9 mm2 in the cohort without any decompression (p < 0.001).
Indirect decompression of both the spinal canal and the contralateral foramen can be achieved via EPTLIF. Decompression on an asymptomatic contralateral side is not necessary.
Lumbar spinal fusion surgeries have been the mainstay of treatment for many degenerative lumbar spinal conditions such as lumbar spinal stenosis and spondylolisthesis with dynamic instability [
In cases with foraminal stenosis, direct decompression requires additional bone and soft tissue (on top of those required for fusion alone) would need to be removed. This may increase the instability of the adjacent level, resulting in a theoretical increased risk of adjacent segment disease [
There is a lack of literature on indirect decompression in uniportal endoscopic posterolateral transforaminal lumbar interbody fusion (EPTLIF). Our aim is to evaluate the dimensions of the spinal canal and contralateral foramen before and after EPTLIF, allowing direct assessment of the viability of indirect decompression using this technique.
This is a retrospective study of all patients who underwent EPTLIF in a tertiary spine centre by a single fellowship-trained surgeon from 2020 to 2022. The inclusion criteria are patients who underwent EPTLIF for degenerative lumbar conditions such as lumbar spinal stenosis and spondylolisthesis with dynamic instability, without any contralateral radiculopathy. Cases who had previous spinal surgeries, spinal trauma, suspected spinal malignancies, inflammatory spinal conditions and spinal infection were excluded from this study.
All patients underwent a preoperative radiographs and magnetic resonance imaging (MRI) scans of the lumbar spine for preoperative assessment. Patients with corresponding MRIs and sufficient indications for lumbar fusion were counselled for surgery. The steps and techniques for surgery was as described in a technique paper published previously [
Demographic parameters including patient’s age, sex, and the level of fusion were recorded. Clinical parameters such as the visual analogue scale (VAS) score, the Oswestry Disability Index (ODI) and the McNab criteria were measured at the preoperative review and postoperative review at 1 week, 6 months, and at the final follow-up. Additional MRI scans were also performed at the 1-day postoperative mark and at the final follow-up. Computed tomography scans were also performed at the 1-year mark for assessment of fusion at the level operated.
The preoperative and the 2 postoperative MRIs were reviewed in detail. The cross-sectional area of the spinal canal was measured from the axial cuts parallel to the adjacent end plates at the level of the disc fused. The cross-sectional area of the contralateral foramen was measured from the parasagittal cuts at the centre of the contralateral pedicle at the adjacent levels.
EPTLIF had been described by Wu et al. [
Similar steps to EPTLIF with facet resection with cage alone insertion without decompression are taken initially to remove facet. In edition for unilateral decompression, we drill the insertion of the ipsilateral ligamentum flavum at the proximal and distal insertion. We removed the ipsilateral flavum completely to expose ipsilateral traversing nerve root and disc endplate preparation and cage insertion is similar to EPTILIF with cage alone cohort (
In addition to the steps in EPTLIF with unilateral decompression. We perform lumbar endoscopic unilateral laminotomy with bilateral decompression with over-the-top decompression of contralateral ligamentum flavum and medial tip of superior articular process of the contralateral side to decompress the contralateral lateral recess and the foramen. After bilateral bony decompression is completed, both flava are removed and disc preparation is continued similar to other 2 cohorts of EPTLIF (
All collected data were tabled using IBM SPSS Statistics ver. 23.0 (IBM Co., Armonk, NY, USA) with statistical significances set at p < 0.05. Baseline characteristics and radiographic parameters as described above for the entire cohort were tabulated and shown in
The patient cohort was divided into 3 groups—those who underwent bilateral foraminal decompression, ipsilateral decompression, or cage insertion alone without any decompression. Subgroup analyses were performed within each group, and paired t-test were used for comparison to VAS and ODI at baseline. Changes in the clinical and radiographic parameters were also studied and shown in
All procedures performed in studies involving human participants were in accordance with the ethical standards of the Ethics Committee of Nanoori Gangnam Hospital (2022-007) and the national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. All patients had given their informed consent for photographs, videos, and images for publication.
A total of 120 patients were recruited into this study. 152 levels of fusion were performed. The average age was 65.2 years. The mean follow-up period was 13.6 months. The most fused level was L4/L5 followed by L3/L4. There were 46 males and 106 females in the cohort. There was no significant difference between the groups when tested using chi-square test (
Looking at the VAS and ODI measurements, there were no significant differences among the 3 groups at baseline. In the entire cohort as well as within the subgroups, there were significant improvements in the VAS as well as the ODI measurements at 1-week postoperation with continued trends of improvement at 6-week postoperation and at the final follow-up (
Looking at the MRI measurements, patients who underwent bilateral decompression had a smaller cross-sectional area of the spinal canal when compared with ipsilateral decompression; and patients who had cage insertion alone without any decompression have the largest cross-sectional area (p < 0.001). There were no significant differences in the contralateral foraminal cross-sectional area among the 3 groups at baseline. In the entire cohort as well as within the subgroups, there were significant improvements in the cross-sectional area of the spinal canal as well as the contralateral foramina at 1-day postoperation and these changes persisted and remained significant at the final follow-up (
The authors standardized the width, length and angle of the cages which are 12 mm wide, 28 mm long, and 4° angle cages. There are variable height dimensions depending on the disc height of the patients. We used variable heights of cages with the mean of 11.23–11.52 mm height in the 3 groups. Comparing the 3 cohorts with analysis of variance test, there is no statistical difference in cage height used in the 3 groups. There is statistically significant increment of cross-sectional spinal canal area in bilateral decompression more than ipsilateral decompression which is in term more than nondecompression group at postoperative day one (66.95 ± 58.39, 39.56 ± 47.62, 12.91 ± 34.37 mm2) and final follow-up (134.77 ± 52, 94 ± 40.34, 65.84 ± 33.15 mm2). However, there is no significant changes in contralateral foramen among the 3 groups at postoperative day one and final follow-up (
Unilateral TLIF has also been shown to cause a contralateral radiculopathy with an incidence of up to 5.3% [
Newer techniques—especially with lateral approach techniques such as lateral lumbar interbody fusion, oblique lumbar interbody fusion (OLIF), and extreme lateral interbody fusion have been shown to achieve sufficient indirect decompression for foraminal stenosis [
Castellvi et al. [
Based on our results, this is the first time that such an effect can be seen in unilateral EPTLIF using similar cage techniques (in terms of cage sizing and placement) as well as not aggressively chasing the creation of lumbar lordosis. This becomes a balance of indirect decompression of the neural elements with obtaining sufficient correction of the sagittal imbalance. At the immediate postoperative MRI scans, there was a significant increase in the cross-sectional area of the spinal canal and the contralateral foramen regardless of whether any decompression was performed at all. The improvements in the radiological parameters were not only maintained but continues to improve with time due to subsequent remodeling as shown in the MRI scans at the final follow-up. There is significant change increment in axial cross-sectional cut in final follow-up compared to postoperative day one. In bilateral decompression, the axial cross-sectional area doubled in final follow-up. In ipsilateral decompression group, the cross-sectional area almost tripled in final follow-up and there is 5× increment in nondecompression group due to remodeling of spinal canal after EPTLIF. While foraminal height also remodeled with an average of 1.5× increment across all 3 groups. This remodeling pattern is concordant to the corresponding pattern in the patient-reported VAS and ODI scores at 1-week postoperation with continued trends of improvement at 6-week postoperation and at the final follow-up. Therefore, this technique can be used for indirect decompression of the central canal stenosis and contralateral foraminal stenosis and thus, able to reap the benefit of MIS approaches to preserve the anatomy of the spine and reduce the risk of adjacent segment disease while simultaneously avoiding iatrogenic injuries to the neural elements, dural tears, and epidural haematoma. This approach also allows for direct visualization of the contralateral foramen should the need arise and there are changes in neuromonitoring after the insertion of the interbody cage.
While there are studies on remodeling of spinal canal with the induced change of ligamentum flavum in ALIF [
In patients presented predominantly with claudication with minimal back pain, endoscopic foraminotomy is an alternative, minimally invasive technique compared to fusion designed to deal with foraminal stenosis [
One of the potential advantages of ipsilateral decompression or no decompression alone over bilateral decompression is the reduction of operative time. However, we did not find any statistical difference in surgical timing in our comparison of the 3 groups. This demonstrates the complexity of EPTLIF as a multiple step procedure which additional of 5–20 more minutes of additional bilateral decompression may not significantly add more surgical time overall. Although we find a statistical difference in final VAS and ODI with better clinical results in bilateral decompression compared to ipsilateral and no decompression, there is no clinically significant difference as the difference is minor. We suggest to perform bilateral decompression when MRI demonstrated severe central and bilateral lateral recess and foraminal stenosis and ipsilateral or no decompression in patients with MRI demonstrates predominantly ipsilateral degenerative changes.
Uniquely, looking at the cohort who had cage insertion without any dural decompression, there is a significant improvement in both patient-reported outcomes (VAS and ODI scores) as well as radiographic parameters (cross-sectional area of the spinal canal and the contralateral foramen). Such indirect decompression result has only been previously shown in anterior and lateral approach techniques and this is the first time it has been reported in posterior approach endoscopic techniques. This is due to the possibility of using a large cage for indirect disc height restoration with this technique. Direct visualization of the end plate for denudation and the use of a large 3D-printed cage also allowed a high rate of fusion contributing to the maintenance of the intervertebral space stability and resulting in the subsequent remodeling and enlargement of the radiographic parameters recorded. The use of large cages also allows greater lordosis to be achieved especially in the lower lumbar segments thus, improve the lordosis distribution index and reduce the risks of revision surgery in the future [
There is an increasing interest in 3D-printed titanium cages in lumbar spinal fusion. We used 3D-printed titanium cages in our cohort of patients. One of the limitations of minimally invasive fusion is cage subsidence, the use of 3D-printed cage was associated with lower rates of subsidence [
There are a few limitations of this study. The surgeon was not blinded to the patient’s symptoms and were given the option to decide if the patient requires bilateral decompression, ipsilateral decompression or cage insertion alone without any decompression during the surgery. At subsequent reviews, both the patient and the surgeon were not blinded as well when recording the patient-reported outcomes. Variabilities in measurements of the cross-sectional area of the spinal canal and the contralateral foramen is inescapable due to human error. Majority of our patients presented with spondylolisthesis, there is limitation in finding a difference during analysis in central and foraminal expansion between spondylolisthesis and spinal stenosis group. Our study is predominantly an MRI study, fusion rate and subsidence are not reflected in the study. Lastly, we did not include patients with bilateral lower limb radiculopathy that may benefit from bilateral decompressions. Future avenues of study could focus on patients with not only sagittal imbalance but also coronal deformity, as well as patients who underwent 3 or more level surgery.
Indirect decompression of both the spinal canal and the contralateral foramen can be achieved via EPTLIF. This radiological finding is supported by patient-reported outcome scores. Initial improvement (immediate postoperation) is not only maintained at the final follow-up, but there is continued improvement due to subsequent remodeling. Decompression on an asymptomatic contralateral side is not necessary unless it is accompanied by a very severe spinal stenosis due to the increased risks of injury in contralateral decompression.
Dr. Pang Hung Wu as first co-author declared his spouse is the director of Singapore based company Endocare PTE Ltd. which distributes orthopaedic and spine products including BESS, NSK drill, and Bonss energy system. No other co-authors have conflict of interest.
This study received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Conceptualization: PHW, HSK; Data curation: PHW, HSK; Formal analysis: PHW, HSK; Methodology: PHW, HSK; Project administration: PHW, HSK; Visualization: PHW; Writing - original draft: PHW, ETL, HSK; Writing - review & editing: PHW, HSK, GG, ITJ.
We would like to acknowledge scientific team members Ms. So-Jung Yoon, Se-Won Lee, and Mr. Kyeong Rae Kim for providing assistance in statistical support, acquiring full text articles and managing digital works.
Cartoon and intraoperative picture demonstrating the 3 types of technique of EPTLIF. (A) Cartoon demonstrated EPTLIF with cage insertion and conservation of ipsilateral ligamentum flavum overlying traversing nerve root and contralateral ligamentum flavum (dotted yellow arrow). (B) Intraoperative picture demonstrated interbody cage placed lateral to ipsilateral ligamentum flavum. (C) Cartoon demonstrated EPTLIF with cage insertion and complete removal of ipsilateral ligamentum flavum overlying traversing nerve root while preserving the contralateral ligamentum flavum (dotted red arrow). (D) Intraoperative picture demonstrated interbody cage placed lateral to ipsilateral traversing nerve root with ligamentum flavum removed. (E) Cartoon demonstrated EPTLIF with cage insertion and complete removal of both ligamentum flava (dotted red and blue arrows). (F) Intraoperative picture demonstrated contralateral decompression with removal of contralateral ligamentum flavum prior to placement of interbody cage on the ipsilateral side. EPTLIF, endoscopic posterolateral transforaminal lumbar interbody fusion.
(From left to right) Preoperative baseline magnetic resonance imaging (MRI), 1-day postoperative MRI, and MRI scans at the final follow-up of a patient who underwent endoscopic posterolateral transforaminal lumbar interbody fusion.
(From left to right) Preoperative baseline, 1-day postoperative, and final follow-up axial and right parasagittal MRI in nondecompression left EPTLIF of L4/5. EPTLIF, endoscopic posterolateral transforaminal lumbar interbody fusion.
Bar chart comparing the cross-sectional area of the spinal canal at preoperative baseline, 1-day postoperation, and at final follow-up. *p < 0.001, significant difference when compared to preoperative baseline.
Bar chart comparing the cross-sectional area of the contralateral foramen at preoperative baseline, 1-day postoperation, and at final follow-up. *p < 0.001, significant difference when compared to preoperative baseline.
Baseline demographic data, clinical parameters and radiographic measurements
Variable | Total cohort | Subgroup analyses |
||
---|---|---|---|---|
Bilateral decompression | Ipsilateral decompression | Cage insertion alone without any decompression | ||
No. of patients | 152 | 35 | 42 | 75 |
Mean age (yr) | 65.2 | 66.4 | 66.6 | 63.9 |
Sex | ||||
Male | 46 | 8 | 18 | 20 |
Female | 106 | 27 | 24 | 55 |
Mean follow-up period (mo) | 13.6 | 16.9 | 12.3 | 14.8 |
Level of fusion | ||||
L2/L3 | 11 | 0 | 6 | 5 |
L3/L4 | 31 | 10 | 10 | 11 |
L4/L5 | 82 | 23 | 22 | 37 |
L5/S1 | 28 | 2 | 4 | 22 |
Preoperative VAS scores | 7.6 ± 1.3 | 8.0 ± 1.2 | 7.5 ± 1.3 | 7.6 ± 1.4 |
Preoperative ODI scores | 70.9 ± 8.8 | 73.6 ± 6.9 | 70.0 ± 9.4 | 70.5 ± 9.4 |
Preoperative cross-sectional area of the spinal canal (mm2) | 95.4 ± 57.9 | 57.0 ± 38.0 | 89.3 ± 63.2 | 138.3 ± 65.1 |
Preoperative cross-sectional area of the contralateral foramen (mm2) | 71.9 ± 27.1 | 73.3 ± 26.6 | 69.5 ± 26.4 | 75.1 ± 28.8 |
Values are presented as number or mean±standard deviation.
VAS, visual analogue scale; ODI, Oswestry Disability Index.
Subgroup analyses looking at clinical parameters and radiographic measurements at baseline and on subsequent follow-up
Variable | Bilateral decompression | Ipsilateral decompression | Cage insertion alone without any decompression | |
---|---|---|---|---|
VAS measurements | ||||
Preoperative baseline | 8.0 ± 1.2 | 7.5 ± 1.3 | 7.6 ± 1.4 | |
1-Week postoperation | 3.3 ± 0.6 |
3.5 ± 0.7 |
3.4 ± 0.7 |
|
6-Month postoperation | 2.4 ± 0.7 |
2.7 ± 0.8 |
2.6 ± 0.9 |
|
At final follow-up | 1.9 ± 0.8 |
2.2 ± 0.8 |
2.2 ± 0.9 |
|
ODI | ||||
Preoperative baseline | 73.6 ± 6.9 | 70.0 ± 9.4 | 70.5 ± 9.4 | |
1-Week postoperation | 32.7 ± 4.2 |
33.6 ± 7.1 |
32.9 ± 5.7 |
|
6-Month postoperation | 27.1 ± 3.8 |
28.7 ± 5.2 |
27.8 ± 5.6 |
|
At final follow-up | 24.1 ± 4.5 |
26.0 ± 4.7 |
25.4 ± 5.2 |
|
Cross-sectional area of the spinal canal (mm2) | ||||
Preoperative baseline | 57.0 ± 38.0 | 89.3 ± 63.2 | 138.3 ± 65.1 | |
1-Day postoperation | 123.9 ± 63.3 |
128.9 ± 47.0 |
151.2 ± 65.6 |
|
At final follow-up | 191.8 ± 51.2 |
183.3 ± 44.9 |
204.1 ± 67.0 |
|
Cross-sectional area of the contralateral foramen (mm2) | ||||
Preoperative baseline | 73.3 ± 26.6 | 69.5 ± 26.4 | 75.1 ± 28.8 | |
1-Day postoperation | 106.4 ± 38.2 |
99.0 ± 36.3 |
106.5 ± 36.8 |
|
At final follow-up | 120.4 ± 35.5 |
116.9 ± 33.6 |
122.9 ± 34.8 |
Values are presented as number or mean±standard deviation.
VAS, visual analogue scale; ODI, Oswestry Disability Index.
p<0.001, significant difference when compared to preoperative baseline.
ANOVA test comparison of EPTLIF with bilateral decompression, ipsilateral decompression, and cage insertion without any decompression
Variable | Bilateral decompression | Ipsilateral decompression | Cage insertion alone without any decompression | p-value |
---|---|---|---|---|
Height of cage used in the surgery (mm) | 11.37 ± 1.26 | 11.52 ± 1.33 | 11.23 ± 1.56 | 0.558 |
Surgical timing of cage insertion (min) | 139.29 ± 19.37 | 134.41 ± 16.86 | 137.33 ± 20.67 | 0.534 |
POD 1 MRI axial cut cross-sectional area increment of spinal canal area in (POD 1–Preop) (mm2) | 66.95 ± 58.39 | 39.56 ± 47.62 | 12.91 ± 34.37 | < 0.001 |
POD final MRI axial cut cross-sectional area increment of spinal canal area in (POD final–Preop) (mm2) | 134.77 ± 52 | 94 ± 40.34 | 65.84 ± 33.15 | < 0.001 |
POD 1 MRI sagittal cut increment of foraminal area of contralateral foramen (POD1–Preop) (mm2) | 33.07 ± 28.69 | 29.54 ± 29.04 | 31.31 ± 25.94 | 0.853 |
POD final MRI sagittal cut increment of foraminal area of contralateral foramen (POD final–Preop) (mm2) | 47.12 ± 27.19 | 47.4 ± 28.81 | 47.77 ± 31.86 | 0.994 |
POD 1 week improvement of VAS (Preop–POD 1 VAS) | 4.63 ± 1.4 | 3.95 ± 1.48 | 4.16 ± 1.39 | 0.106 |
POD 6 months improvement of VAS (Preop–POD 6 months VAS) | 5.54 ± 1.52 | 4.79 ± 1.26 | 5.03 ± 1.42 | 0.059 |
POD final improvement of VAS (Preop–POD final VAS) | 6.06 ± 1.47 | 5.24 ± 1.32 | 5.41 ± 1.57 | 0.042 |
POD 1 week improvement of ODI (Preop–POD 1 ODI) | 40.86 ± 8.24 | 36.38 ± 11.52 | 37.6 ± 10.26 | 0.145 |
POD 6 months improvement of ODI (Preop–POD 6 months ODI) | 46.51 ± 8.18 | 41.29 ± 8.94 | 42.72 ± 9.38 | 0.035 |
POD final improvement of ODI (Preop–POD final ODI) | 49.49 ± 8.94 | 43.95 ± 8.84 | 45.11 ± 10.04 | 0.028 |
Values are presented as number or mean±standard deviation.
ANOVA, analysis of variance; EPTLIF, endoscopic posterolateral transforaminal lumbar interbody fusion; POD, postoperatve day; Preop, preoperative; MRI, magnetic resonance imaging; VAS, visual analogue scale; ODI, Oswestry Disability Index.
p<0.05, statistically significant differences.