Comparison of an innovative endoscopic posterior/transforaminal lumbar interbody fusion technique and a minimally invasive transforaminal lumbar interbody fusion technique in the treatment of lumbar spinal stenosis: a retrospective analysis

Clinical data

This study retrospectively reviewed 45 patients who underwent Endo-P/TLIF (n = 22) and MIS-TLIF (n = 23) for single-segment LSS between October 2018 and July 2020. No significant differences were found in sex, age, and surgical level between the two groups (P > 0.05) (Table 1).

Inclusion and Exclusion Criteria

The inclusion criteria were patients diagnosed with single-segment LSS through imaging examination whose condition had not improved after 3 ~ 6 months of strict conservative treatment. The exclusion criteria were as follows: (1) multilevel LSS, grade ≥ II lumbar spondylolisthesis based on the Meyerding scale, or spondylolysis; (2) lumbar tuberculosis or past lumbar surgery; (3) significant paravertebral muscle injury or atrophy; and (4) patients with other serious systemic diseases. All patients provided informed consent for the surgery.

Surgical Techniques

X-ray and neurophysiological monitoring were used during all surgeries.

Endo-P/TLIF Surgery (L4/5 as an example).

Patients were placed in the prone position under general anesthesia. X-ray fluoroscopy was used to view the intervertebral space, and the pedicle positions of the lumbar bridges L5 and L4 were adjusted. The midline and bilateral spinous processes were marked. Under the guidance of electromagnetic navigation, percutaneous punctures for bilateral pedicle screw channels were performed, and four guidewires were installed. A right guidewire incision at L5 was used to make a 1.5-cm oblique incision “from the outside down to the inside up,” approximately 2 cm from the midline of the spinous process. The first guidewire was inserted into the oblique incision, extending to the articular surface. After a satisfactory position was confirmed by fluoroscopy, expansion casing was inserted step by step, the working channel was inserted, and an outer ring saw was fixed onto the bone surface. The position was determined again by fluoroscopy, and then a Delta endoscope system was connected (Joimax, Germany) (Fig. 1A, B).

Using a microscope, a trephine was used to block surrounding soft tissues, and radiofrequency electrocoagulation was used to stop bleeding. Soft surface tissues were cleaned using an endoscopic grasp, and the articular process and lamina were gradually exposed (Fig. 2A). The L4 inferior articular process and part of the lamina were removed using a trephine and an endoscopic grinding drill (Fig. 2B). The ligamentum flavum was progressively exposed in the proximal-lateral-distal direction (Fig. 2C, D, E) and was resected off (Fig. 2F), and the thecal sac and nerve root were exposed (Fig. 2G). A medial margin of the L5 superior articular process was resected, and a spinal canal and nerve root canal were performed. The bone resection was restored with an intervertebral space bone graft. The roots of nerves involved in walking were probed and decompressed (Fig. 2H).

Then, the working channel was rotated to block the nerve root and thecal sac out of the channel (Fig. 3A). The annulus fibrosus was cut using an endoscopic bone chisel (Fig. 3B), and a discectomy was performed using a detachable endoscopic square reamer and nucleus pulposus forceps (Fig. 3C, D, E). Endplate cartilage was removed with a special endoscopic endplate scraper from the upper and lower surface of the lamina, causing slight bleeding. The anterior longitudinal ligament was assessed (Fig. 3F). After microscopically confirming the completion of the intervertebral space treatment, autologous bone and bone repair materials were mixed and implanted into the intervertebral space. Then, a metal expandable intervertebral fusion device was implanted (Fig. 3G).

The position of the cage wires was rechecked, and the nerve root and thecal sac were detected using a microscope (Fig. 3H). The endoscope and working channel were then removed. A bilateral hollow pedicle screw and longitudinal connecting rod were percutaneously inserted along the reserved guidewire and tightened to restore the normal angle of the lumbar bridge. After the positions of the fusion cage and pedicle screw were satisfactory, a drainage tube was installed and the incision was sutured.

Statistics Analysis

All statistical analyses were performed using the Statistical Package for Social Sciences (SPSS), version 21.0, software. Values are presented as means ± standard deviations. Independent t tests were used for the comparison between the groups. We used the χ2 test or Fisher’s exact probability method for the comparison of numerical data. A two-tail rank sum test was used for the comparison of grade data (α = 0.05).

The operation times in the Endo-P/TLIF group were significantly longer than those in the MIS-TLIF group (P < 0.05). Intraoperative blood loss, surgical incision length, postoperative drainage volume, and postoperative time spent bedbound in the Endo-P/TLIF group were significantly less than those in the MIS-TLIF group (P < 0.05, Table 2).  

The average follow-up time was 13.9 ± 2.7 months for the Endo-P/TLIF group and 12.6 ± 2.4 months for the MIS-TLIF group. Six months postoperatively, there were 10 cases of grade Ⅱ fusion and 12 cases of grade Ⅲ fusion in the Endo-P/TLIF group. There were 11 cases of grade Ⅱ fusion and 12 cases of grade Ⅲ fusion in the MIS-TLIF group. No significant difference was found in the rate of interbody fusion between the two groups (P > 0.05). Postoperatively, VAS scores for lumbar and leg pain in both groups were significantly lower at 7 days, 3 months, 6 months, and 1 year than those preoperatively (P < 0.05). However, VAS scores for low back pain in the Endo-P/TLIF group were significantly lower than those in the MIS-TLIF group at 7 days and 3 months after surgery (P < 0.05), but not at 6 months and 1 year. The postoperative ODI scores showed significant differences at 3 months and 6 months and at final follow-up between the groups (P < 0.05)(Table 3). The transversal areas of multifidus and erector revealed no statistical differences between the two groups (P > 0.05). The transversal area of the paravertebral muscles in the Endo-P/TLIF group 1 year after surgery was significantly greater than that in the MIS-TLIF group (P < 0.05). Six months postoperatively, there were no significant differences in fusion rates of definite grades between both groups (P > 0.05).

In this study, Endo-P/TLIF achieved satisfactory clinical results in patients with single-segment LSS. In a specific subset of patients, spinal fusion showed good clinical results as a management method in the treatment of degenerative spinal conditions [12]. Although traditional PLIF techniques have been widely used, they are associated with paraspinal muscle atrophy and long-term low back pain. It was found that multifidus muscle atrophy is commonly associated with postoperative low back pain [4]. Therefore, several scholars sought for an alternative minimally invasive surgery. Foley et al. [13] first introduced MIS-TLIF technology and used the natural muscle space in the spine to insert channels for surgery. This technique has a similar efficacy as traditional open surgery, but avoids the destruction of complex posterior muscle-ligament structures, causing less tissue damage and less intraoperative bleeding [5].

Nevertheless, MIS-TLIF is limited by a narrow operating space and poor field visibility. This can cause nerve root injury, which is a serious complication of endoscopic surgery, especially injury to the exiting nerve roots [14]. Moreover, MIS-TLIF exposes patients and surgeons to radiation. Therefore, an alternative minimally invasive technique was required to overcome these limitations of lumbar fusion surgery.

This study demonstrates that Endo-P/TLIF technology achieved clinical efficacy similar to MIS-TLIF, when treating patients with single-segment LSS. Furthermore, Endo-P/TLIF-treated patients had earlier improvements in low back pain and could generally get out of bed 1–2 days postoperatively. This indicates that Endo-P/TLIF technology causes less tissue injury than MIS-TLIF technology, consequently allowing patients an earlier return to normal life.

The fusion rate of lumbar spine surgery is impacted by various factors, including endplate preparation, bone graft area, bone graft type, contact with endplate surface, and personal constitution. The surgical area was expanded by adjusting the lumbar bridge before surgery to increase the lamina space and by removing part of the facet joint. Nerve roots and the thecal sac were protected from using an endoscopic cannula, which also improved the safety of fusion device implantation. The use of an expandable cage provided adequate intervertebral support and indirect decompression of the spinal canal, with smaller resection pathways for bony structures. Furthermore, Zhang et al. [15] have confirmed that the surface of an expandable cage apparatus provides stable and good osseous bonding in endoscopic surgery. Nevertheless, MIS-TLIF is limited by a poor field visibility and it may be difficult to view the deeper surgical field [8]. An endoscopic reamer and curette were used to treat cartilaginous endplates under direct vision, which was better than MIS-TLIF. Additionally, bone autografting and adequate bone graft sizing were used to increase the fusion rate. Zhao et al. [16] have suggested that pedicle screw fixation and installation and retention of a posterior spinal tension band can reduce complications, increase the rate and stability of intervertebral fusion, and reduce the occurrence of intervertebral fusion displacement. This is consistent with the results of our present study.

Recently, it has been reported that the volume of the multifidus dramatically decreases after open posterior lumbar fusion [4]. Kawaguchi et al. have measured paravertebral tissue images of patients receiving posterior lumbar surgery and found that the degree of muscle damage in patients receiving MIS-TLIF significantly declined when compared to that in patients who underwent open surgery [17]. It has been reported that the paraspinal muscles are the key to support the extension of the spine, maintain lumbar lordosis, and achieve dynamic spinal stability [18]. Our study found that the early curative effect of single-segment LSS with Endo-P/TLIF is satisfactory. In this study, the single-segment Endo-P/TLIF surgical incisions included four incisions of approximately 1 cm. Decompression and percutaneous pedicle screw placement were achieved through progressive expansion of soft tissue to establish the working channel, without extensive injury to the muscle. At 1-year follow-up, the cross-sectional area of the postoperative paravertebral muscles in the Endo-P/TLIF group was larger than that in the MIS-TLIF group, confirming that Endo-P/TLIF can effectively avoid postoperative paravertebral muscle atrophy caused by intraoperative dissection.

At present, a major difficulty with Endo-P/TLIF technology is that it is highly technical. It requires a surgeon to have extensive experience in endoscopic spinal surgery and conventional open surgery. Additionally, Endo-P/TLIF surgery requires more time to complete for doctors who are applying the technique for the first time. Our experience with Endo-P/TLIF technology is summarized as follows. First, subperiosteal dissection helps to reduce intraoperative bleeding. Second, blurred field vision caused by intraoperative bleeding is a major concern in endoscopic lumbar fusion surgery. The concept of “pre-hemostasis” should be emphasized to reduce surgical bleeding. For peripheral soft tissue bleeding, it is recommended to use an external endoscopic ring saw to block bleeding. Radiofrequency can be used to gradually clean the surrounding soft tissue and expose the bone surface, with the tip of the articular process as the center. Additionally, bleeding occurs on the surface of the bone after the articular process and lamina were resected, especially in patients with osteoporosis in which bleeding is more severe. The use of radiofrequency cauterization is recommended to reduce this bleeding. Moreover, bleeding can occur in small venous plexuses from the nerve roots and intervertebral disc surfaces. Radiofrequency cauterization should be used for pre-hemostasis to reduce bleeding of this kind. Third, it is recommended to gradually expose the insertion of the ligamentum flavum in the proximal-lateral-distal direction. The ligamentum flavum should be removed after the insertion is fully exposed because it is easy to separate the ligamentum flavum from the thecal sac using water pressure. Fourth, the recommended endoscopic procedures are excision of the inferior articular process and part of the superior articular process, exposure of the upper and lower lamina, excision of the ligamentum flavum, and decompression of the lateral recess and nerve root, followed by intervertebral space-bone grafting and the implantation of a metal expandable interbody fusion device.

Limitations

The limitation of this study should be addressed. Firstly, the sample size of our study was relatively small and the follow-up time was only 1 year. Secondly, our study lacked large-sample multicenter randomized controlled studies.