1 Introduction

Thoracic canal stenosis due to ossification of the ligamentum flavum (OLF) has been known as one of the major causes of thoracic myelopathy.^1,2 It has a higher incidence among the Asian population and has been reported in Japan, Korea, and China.^3,4 The occult nature of the disease, often misdiagnosed as a lumbar or cervical problem, delays treatment. The duration and severity of pre-operative symptoms are important prognostic factors.⁵ Therefore, accurate diagnosis and early treatment are pivotal.

Posterior surgical decompression remains the gold standard.^4,6–8 However, the clinical efficacy and outcome after traditional surgery are often unsatisfactory, with a complication rate from 35% to 50%, including intraoperative dural tear, neurological deficit, bleeding, screw breakage or loosening, and OLF at the adjacent proximal level.^1,4,9

Percutaneous unilateral biportal endoscopic (UBE) decompression is the latest novel minimally invasive procedure for managing OLF. Studies have shown the advantages and reliabilities of this procedure, such as minimized soft tissue and facet destruction, thus avoiding spinal fusion, and high-definition endoscope visualization with clear and magnified surgical field unrestricted by tube or working channel.^10,11 Currently, uniportal endoscopic surgical technique has been documented as another option for minimally invasive surgery in treating thoracic OLF with good satisfaction.^9,12,13

There are few reports of endoscopic decompression for the treatment of thoracic OLF, and despite some reported data in current literature, whether this novel technique can be used in OLF has not been well-examined. In this study, we aimed to detail our surgical technique in performing UBE for thoracic OLF and to evaluate the efficacy and safety of UBE decompression compared with conventional open surgery. This study has been made and written according to STROBE International Guideline.

2 Material and methods

We conducted a retrospective case-control study of 35 patients with single or two-level thoracic OLF managed operatively by conventional open surgery or percutaneous UBE procedure from January 2018 to December 2022. Subjects were screened, based on the following inclusion criteria: (a) symptomatic myelopathic patient with complaints such as pain, weakness, gait disturbance, or any upper motor neuron problem; (b) diagnosed by pre-operative computed tomography (CT) and magnetic resonance imaging (MRI), which confirmed the level and morphology of the OLF. The exclusion criteria for the patient as follow: (a) asymptomatic case; (b) accompanied by other spinal lesions; (c) surgical comorbidities or refusal of surgery; (d) dural ossification found on CT; all such were excluded from this study. Patients performing UBE were categorized as the intervention group, while open decompression with instrumentation was performed on the control group.

2.2

2.2 Radiological evaluation

Whole spine non-contrast CT scan and MRI were scanned for all patients before surgery to identify OLF characteristics and localized decompression level and to rule out any stenosis or pathology on another vertebral segment (Fig. 1). On the basis of pre-op axial CT scan, we classify OLF morphology onto fused and non-fused type. Dural ossification was also searched and ruled out by identifying the “comma” sign or the “tram track” sign on axial CT. A post-operative CT scan was performed to determine the impact of decompression on the OLF and affected spinal canal.

Close

Fig. 1 Radiographic axial and sagittal view of ossified ligamentum flavum compressing neural canal in the thoracic area.

Export to PPT

2.4

2.4 Surgical technique

All surgeries were performed by one experienced senior spine surgeon. The patient underwent general anesthesia and was positioned prone. The conventional open surgery used a standard midline approach and total laminectomy with or without instrumentation. In the intervention group, we used the biportal endoscopic spine surgery technique. Decompression level and portal incision site were determined using an image intensifier in the anteroposterior and lateral views. We marked the viewing portal at the cranial pedicle level and the working portal at the caudal pedicle level. We made a 5–10 mm stab incision for the viewing and working portal. The flow of continuous water irrigation with pressure around 30–40 mmHg through the working portal must be assured to prevent high pressure inside the working space.

The procedure was started by identifying the inferior border of the lamina and facet joint (Fig. 2). A high-speed burr was used to drill the lamina from the caudal to the cranial direction until we found the central fat. In case of severe ossification of the flavum without a central cleft left, we drilled the lamina and ossified the flavum ligament until it was paper thin. The superior border of the caudal lamina and the medial border of the superior articular process also had to be drilled until the flavum was naturally detached from the insertion (Fig. 3). The remaining calcified flavum was lifted using a Penfield elevator from the surface of the dura mater. At this stage, the surgeon must be aware of any adhesion between the ossified flavum and dura mater. When the adhesion was very severe, it was better to leave the calcified fragment floating above the dura mater by ensuring that the remaining flavum or ossified flavum was not attached to the origin or insertion. The end of the procedure was confirmed by palpating the medial border of both pedicles and seeing the pulsation of the dura mater.

Close

Fig. 2 Intraoperative endoscopic view of OLF decompression using UBE technique. The right side is the cranial side and the left side the caudal side. Decompression began by removing the cranial side lamina (A). Ipsilateral lamina and part of the facet were removed (B, yellow arrow). Contralateral lamina and part of the facet were removed (C, yellow arrow). After lamina was removed circumferentially (D), flavum was detached from its insertion (E) and exposed dural sac underneath (yellow triangle). Flavum and dural adhesion were often found and carefully separated (F) until OLF lesion “floated” freely (G, yellow arrow). OLF lesion was removed, leaving uninjured nerve fully decompressed (H–I). Ossified flavum varied in size and structure after removal (J).

Export to PPT

Close

Fig. 3 Illustration of bilateral OLF bilateral decompression using unilateral biportal endoscopy.

Export to PPT

3 Results

A total of 35 patients consists of 16 male and 19 female, underwent surgical treatment for symptomatic OLF. A biportal endoscopic decompression was performed on 18 patients, and the rest were managed by conventional open surgery. The mean age at surgery was 55.2 years old (range 42–75 years). Symptom duration ranged from 3 to 24 months with a mean of 10.5 months. Two patients in each group suffered stenosis on two consecutive thoracic levels, while others had single-level stenosis only. Both groups had seven patients with fused-type OLF with more severe neurologic impairment. Post-operative evaluation and clinical data collection were performed as scheduled, with an average final follow-up of 40.7 months (13–71 months). Subject demographic data can be seen in Table 1.

Pre-operative Frankel grade and mJOA score were assessed as described in Tables 2 and 3. Pre-operative mJOA score between the UBE and open-surgery groups was similar with mean of 4.9 ± 0.34 and 5.2 ± 0.31, respectively. Moderate symptom was found most frequent (61% on UBE, and 70% on open surgery). There was no statistically significant difference in the mJOA score during follow-up at 1 and 6 months between the two groups (p > 0.05). On the contrary, at 1-year follow-up, the mJOA score in the UBE group showed more notable improvement with an average of 8.2 ± 0.18 compared to 6.8 ± 0.24 in the open group (p = 0.015). More than 90% of patients underwent UBE with previous moderate symptoms successfully improved to mild mJOA score at 1-year follow-up while only 50% did so on the open-surgery group. The RR calculation 1 year after treatment resulted in significant recovery in patients treated with UBE (RR-UBE 43.2 ± 17, RR-open 26.3 ± 15.3) with p < 0.05.

Both groups exhibited comparable improvements in the average mJOA scores at each subsequent follow-up period (Table 3). Patients who underwent either UBE or open surgery had better mJOA scores on the 6-month and 1-year follow-up compared to the condition before surgery. Additionally, the Frankel grade also showed improvement with 4 of 7 (57%) patients who underwent UBE upgraded their Frankel from D to E as well as two patients from the open-surgery group (22%) and nine patients (5 UBE, 4 open surgery) converted their Frankel grade from C to D on 1-year follow-up (Table 2). Two patients in the UBE group showed superior outcomes than the others with 2 points Frankel grade improvement after surgery, yet none was found in the other group (Fig. 4).

Close

Fig. 4 Frankel grade improvement graphic after minimum 1-year follow-up. (a) Amount of point increased displayed in x axis. (b) Sum of subjects in each group displayed in y axis. (c) Blue bar indicated UBE group and orange bar indicated open-surgery group.

Export to PPT

The mean operation time for UBE was 62.5 min (range 45–100) compared to 180 min (range 90–210) on open surgery. The mean intraoperative blood loss comparison between UBE and open surgery was 50 ml (ranged 20–100 ml) and 250 ml (ranged 100–800 ml), respectively. The LOS average was 4 days in the UBE group and 6 days post-open surgery. Report data showed a significant reduction in surgery time, blood loss, and post-operative LOS in the UBE decompression group with p < 0.001 on all three variables.

Intraoperative dural tear occurred in three patients with fused-type OLF, one patient in the UBE group (5.5%), and two patients in the open-surgery group (11.7%). A severe adhesion between calcified material and dura mater was found in patients with a dural tear. One patient underwent primary repair while the other was patched with a collagen-based dural patch. We maintained the post-op drainage tube longer on all dural tear patients until post-operative day 3. No hyperalgesia, head pain, or neck pain was found among patients with a dural tear. One patient with CSF leakage in the open-surgery group developed delayed wound healing, yet no infection occurs during until final follow-up with good recovery. Overall, besides the dural tear, we found no other complications such as sudden neurological deterioration and deep venous thrombosis from either group.

4 Discussion

According to our findings, the surgery duration in the UBE group was significantly less than that of the open-surgery group (69.7 ± 3.9 vs. 156.4 ± 9.1, P < 0.001). Regarding operative duration, patients with multisegmented ossification were still included in this study and distributed fairly to both groups. In a previous study, the length of the surgery for multisegmented OLF was substantially greater than that for single-segment operations.¹⁴ On the contrary, we found that a well-executed UBE surgery for OLF decompression, even in multi-level lesions, can still significantly reduce operation time as conventional open surgery might encounter intraoperative difficulties or complications such as dural tear needing repair and bleeding management that requires more time. Adding subsequent posterior instrumentation during open surgery also prolongs operation time in all open-surgery patients in this study.

Additionally, a substantially larger blood loss was found in the open-surgery group (52.7 ± 5.3 vs. 291.1 ± 42.7, P < 0.001), which is suspected due to major paraspinal muscle dissection and uncontrollable venous plexus injury around the dural sac. Due to the abundance of blood vessels in the epidural area, intraoperative bleeding may occur quickly, be difficult to control, and result in blurry vision and increased intraoperative blood loss, if overlooked. One disadvantage of open surgery is that hemostasis will be challenging in cases of substantial bleeding, which increases the risk of post-operative hematoma and persistent OLF.⁹ We routinely pre-controlled potentially injured blood vessels during UBE with a radiofrequency ablator to prevent unwanted bleeding, which is easier to perform on endoscopy visualization. Water pressure on endoscopy could also add compression effect on vessel injury to assist hemostasis. Most importantly, increased water pressure will increase CSF pressure, leading to head and/or neck pain after surgery as reported by the previous preliminary study.¹ Nevertheless, no UBE patients in our study presented with complications as such.

Post-operative hospital LOS is often determined by how well the incision heals.¹ The UBE group showed no signs of incision infection or delayed healing. Consequently, the mean post-operative hospital stay of the UBE group (3.8 days) was significantly lower than that of the open-surgery group (mean 7.5 days). The number is almost similar to the one reported by Ye et al. with a mean duration of 4.5 days.¹⁴ Open posterior laminectomy is already known for several possible complications including a prolonged recovery period, which some thought was associated with healing of the incision site.^1,10,15

The patient underwent UBE showing significant improvement gradually in the mJOA score on each follow-up period (all p < 0.005) (Fig. 5). Larger soft tissue dissection is one disadvantage of conventional surgery compared to a 7–10 mm incision with the biportal technique, increasing the time for soft tissue healing and thus prolonging recovery and post-surgery mobilization.^9,13 This may explain the superiority of UBE in this study, which enabled the surgeon to decompress adequately with minimal incision. At 1-year follow-up, UBE significantly improved the mJOA score compared to the open group, as shown by the percentage of RR (p < 0.005). Gradual improvement usually occurred and recovery time can be prolonged to later observed as increased RR.² The average RR in this study was 43.2%, which was similar to previous studies.^1,13

Close

Fig. 5 Case of female, 57 years old, with weakness of both lower extremities due to OLF on Th 10–11. Underwent UBE decompression after 6 months of worsening symptoms. Directly post-operative MRI showed adequate canal decompression. Six-month follow-up showed remarkable motor recovery in both legs.

Export to PPT

On technical aspects, clear magnification of the surgical field in UBE is one of the main advantages compared to the traditional approach. Complete decompression of the thoracic thecal sac can be achieved only through a small portal incision on the ipsilateral side. Direct visualization of the OLF enables the surgeon to distinguish clearly between ossified and healthy flavum. It can be harder to evaluate on thick OLF. It also enabled both ventral and contralateral decompression simultaneously, which cannot be achieved in open surgery. On the contrary, adequate decompression using an open technique requires extensive posterior ligamentous structure sacrifice and removal of a large part of the facet joint, leading to instability in which instrumentation is mandatory.^2,14

Dural tear is the most common complication occurring during OLF decompression, more often on open technique.^1,16 Only one patient in the UBE group had a dural tear (5.5%), which was lower compared to other similar studies.^1,2,13 Three cases with a dural tear in this report were found in patients with fusion-type OLF. A higher chance for intraoperative dural tear was already found and correlated with dural ossification, severity of ossification in the axial plane (bridge or fused type), or morphology on the sagittal plane (beak or round type).⁶ The intraoperative finding on the fused type in our study found a larger area of adhesion below the ossified ligamentum flavum. In this case, we recommend the floating method as we do not try to remove the OLF completely from the dural structure below but make sure that the ossified flavum is disconnected circumferentially on all sides.¹⁶ Notably, the dural rupture was not repaired, as the tear size was incredibly small. An et al., using uniaxial scope and Endo-Kerrison punches used under an endoscopic view had an outer diameter of only 2.0 and 3.0 mm with a smaller footprint, left the dural tear unsutured without any further complication.¹³ With a similar approach in microendoscopic decompression, CSF leakage due to a small tear did not require repair and was left with drainage and water-tight sutures.¹⁷ OLF decompression is the most challenging part of the surgery. Therefore, we encourage extreme caution throughout the thoracic laminectomy and spinal cord decompression procedure on both endoscopic and open surgery. No complication related to CSF leakage was found in the cases reported in this study.

In our study, the morphology variation of the ossified ligament on the sagittal plane and dural calcification was not accounted for. The form of OLF seen on sagittal CT scans, such as the beak type, was found to be more correlated with functional outcomes.¹⁸ We categorized the lesion based on only fused or non-fused types. Hence, the result may be biased in cases with more severe ossification and dural compression. Other studies also focused only on a single-level OLF while in our study we included multi-level lesions (two levels at most) on both the UBE and open-surgery groups. This also may become a bias as more level requires a longer duration of surgery, more bleeding, and increased difficulties, which can lead to more complications or inadequate decompression. Contrary to limitations in other studies, we see this as the point of versatility in the endoscopic technique of our study. Using UBE, decompression of two-level OLF without wide posterior dissection is feasible. Even though we did not compare statistically between 1- or 2-level surgery on both groups, a 2-level decompression with the UBE technique still had shorter surgery time compared to single-level open surgery with less blood loss and favorable clinical score.

A one-year follow-up cutoff time may be considered a short period. Bo et al. found that the lamina closure on the defect after decompression indicated a bony ongrowth on 1–2 years follow-up.¹³ Longer observation is recommended to evaluate the decompression area for dural herniation or recurrence of the ossified lesion and re-evaluation of its impact on clinical outcome and nerve regeneration after decompression. We also suggested a multi-centered randomized control trial with a larger sample to evaluate a more representative result for the general population.

5 Conclusion

In conclusion, we suggest the UBE technique as a choice of treatment for symptomatic OLF with favorable clinical outcomes, less surgical time and bleeding, and minimal complication rate compared to conventional open surgery. However, proper indication and meticulous execution are paramount due to the technically demanding surgery and a steep learning curve for surgeons to maximize the potential of this technique.

Ethics approval and informed consent

This study had been approved by the ethics committees at Fatmawati General Hospital - Jakarta and Eka Hospital - Tangerang. All participants' privacy and personal identity information were protected in accordance with the Declaration of Helsinki. Written consent was acquired from all subjects included in this research.

Consent for publication

This manuscript does not contain any personal data. Consents for publishing any image in this manuscript had been taken by author and co-authors.

Funding

This study received no particular grants from public, commercial, or non-profit funding agencies.

Statements and Declarations

The authors declare that no financial, competing, nor other conflict of interest in this study.

CRediT authorship contribution statement

Asrafi Rizki Gatam: Writing, Methodology, Reviewing and Editing. Erwin Ardian Noor: Writing, Writing – original draft, Data curation, Visualization. Luthfi Gatam: Conceptualization, Investigation, Supervision.