INTRODUCTION
It is well-established that AOSpine type A3 or A4 fractures involve the endplate and posterior wall of the vertebral body, and retrograde entry of bone fragments from the posterior wall of the vertebral body into the spinal canal may lead to neurological deficits. Surgical intervention is indicated once neurological impairment occurs [1,2]. In such cases, the goal of surgery is neural decompression and restoration of the spinal alignment and stability [3]. Conventional open surgery has long been the mainstay of treatment for these types of fractures but is associated with substantial blood loss and destruction of paravertebral soft tissue [4,5].
Endoscopic spine surgery represents the most contemporary form of minimally invasive spine surgery associated with minimal collateral tissue damage, decreased intraprocedural hemorrhage and fast functional recovery [6,7]. Endoscopic spine surgery was initially designed for discectomy and nerve root decompression via a transforaminal approach [8]. With the development of posterior approaches and improvements in instrumentation for endoscopic spine surgery, decompression of the central spinal canal in the lumbar, thoracic, and cervical spine is nowadays feasible [9-11]. These advances have brought to light the possibility of minimally invasive surgery for cases of thoracolumbar fractures with neurological deficits.
Spinal endoscopy can be used for neural decompression [9-11], percutaneous pedicle screw fixation (PPSF) can be used to restore the alignment and stability of spinal fractures [4,5], which can meet the surgical goals of patients with spinal fractures combined with neurological deficits, respectively, while reducing surgical trauma. On this basis, the authors have established a minimally invasive treatment approach for A3 or A4 thoracolumbar fractures with neurological deficits, involving endoscopic decompression combined with PPSF. Although the use of endoscopic spine surgery for treating traumatic unstable fractures has traditionally been considered contraindicated, this present retrospective cohort study broke the contraindication as the first to compare the clinical outcomes of endoscopic minimally invasive surgery (EMIS) and conventional open surgery (OS) in treating AOSpine A3 or A4 thoracolumbar fractures with neurological deficits, aiming to elucidate the respective advantages and disadvantages of these 2 methods and provide novel evidence for the application of minimally invasive endoscopy treatment.
MATERIALS AND METHODS
1. Study Design
This retrospective cohort analysis of prospectively collected data compared various outcomes between patients undergoing EMIS and OS. In compliance with the informed consent requirement, these patients were provided comprehensive information regarding the distinctions between EMIS and OS prior to surgery, enabling them to make surgical method choices based on their individual preferences. In order to create a relatively homogeneous study population and thus reduce potential confounding effects on outcome variables, inclusion and exclusion criteria were applied to both cohorts. Inclusion criteria: (1) AOSpine A3 or A4 single-level traumatic thoracolumbar burst fracture [1]; (2) thoracolumbar AOSpine injury score (TL AOSIS)≥5 points [1,2]; (3) American Spinal Injury Association (ASIA) grade A–D [12]. Exclusion criteria: (1) the presence of osteoporosis; (2) multiple organ injury or cardiopulmonary dysfunction precludes surgical intervention within a 24-hour timeframe [13]; (3) severe mental disorders; (4) presence of a reverse cortical sign in the fractured vertebral body was confirmed by computed tomography (CT) examination [14], which suggests that the application of percutaneous indirect reduction could potentially exacerbate neurological deficits; (5) posterior longitudinal ligament rupture was confirmed by magnetic resonance imaging (MRI) examination. In endoscopic surgery, the reduction of the protruding bone block primarily relies on the intact posterior longitudinal ligament. However, when the posterior longitudinal ligament is severely compromised, endoscopic surgery may be insufficient to achieve complete reduction of the protruding bone block; (6) lost to follow-up.
Based on the inclusion and exclusion criteria, 231 of 275 patients with AOSpine A3 or A4 thoracolumbar fractures and neurological deficits who underwent surgical treatment at our center between June 2019 to July 2021 were included in this study. There were 107 patients in the EMIS cohort and 124 patients in the OS cohort. Fig. 1 shows the details of the study selection process.
The study was approved by Honghui Hospital, Xi’an Jiaotong University Biological Research Ethics Committee (202405017) and was completed in accordance with the principles of the Declaration of Helsinki and the statement on STROBE (Strengthening Observational Studies in Epidemiology). The requirement for patient informed consent was waived due to the retrospective study design.
2. Data Collection
Prospectively recorded data from previous routine diagnosis, treatment, and follow-up were extracted and reviewed from the electronic medical record system by an unbiased observer who was not directly involved in the care of the patients. These comprehensive data encompass demographic characteristics, initial clinical characteristics, perioperative outcomes, Imaging data, patient-reported outcomes, neurological function status outcomes, and complications.
Demographic and initial clinical characteristics included sex, age, injured level, AOSpine thoracolumbar spine injury classification, and TL AOSIS.
The perioperative outcomes encompass surgical duration, incision length, intraoperative blood loss volume, postoperative plasma drainage volume, and duration of hospitalization.
Patient-reported outcomes encompass visual analogue scale (VAS) and Oswestry Disability Index (ODI) scores preoperatively, postoperatively, and at the final follow-up to assess back pain and spinal function, respectively. VAS scores ranged from 0 (no pain) to 10 (extreme pain), with higher scores indicating more severe pain. The ODI consists of 10 items rated on a 0 to 5 points scale, and the percentage is calculated. The higher the percentage, the more severe the dysfunction.
Imaging data included the outcomes of CT examination before and after surgery and at the last follow-up. Two independent spine surgeons used PACS (Picture Archiving and Communication System) to measure the spinal canal erosion rate (CER), sagittal Cobb angle (CA) [15], and percentage of anterior vertebral height (AVH) in these imaging data, which were used to assess decompression, spinal alignment recovery, and deformity correction, respectively. The measurement approach is shown in Fig. 2.
The outcome of neurological status included the ASIA classification of patients before operation and at the last follow-up. Grade A indicated complete neurological deficits, grade B–D indicated incomplete neurological deficits, and grade E indicated normal nerve function.
Iatrogenic vascular injury and dural sac injury, postoperative infection and implant breakage were judged as complications.
3. Surgical Procedures
After the administration of general anesthesia, the patient was positioned in a prone orientation. The surgical procedure was performed with the assistance of neurophysiological monitoring.
1) Endoscopic Minimally Invasive Surgery
Step 1: Primary reduction, fixation, and indirect decompression
Initially, under the guidance of C-arm fluoroscopy, percutaneous puncture was performed on the fractured vertebra and the pedicle of its upper and lower vertebral bodies. The placement order of percutaneous pedicle screws and rods can be adjusted according to the decompression method, but the final form is that multiaxial pedicle screws are placed in the injured vertebra, and uniaxial pedicle screws are placed in the upper and lower vertebrae of the injured vertebra. If only endoscope-assisted unilateral laminectomy decompression is required, pedicle screws should not be placed on the decompressive side of the fractured vertebra for the time being (Fig. 3). If endoscope-assisted bilateral laminectomy decompression is required, pedicle screws should not be placed on either side of the fractured vertebra for the time being (Fig. 4). The prefabricated rod is inserted percutaneously into the tail of the pedicle screw, and reduction of the fractured vertebra is achieved using specialized external distraction compression instruments (Fig. 3B). The accompanying ligamentotaxis technique will indirectly alleviate the compression of the central spinal canal.
Step 2: Direct decompression by uniaxial endoscope
The target and scope of decompression were determined based on preoperative imaging findings and the patients’ neurological status. Typically, the decompression target was observed in the center of the severely compressed area on imaging. If preoperative CT or MRI confirmed unilateral or central compression of the spinal canal, meanwhile, the noncompressed side exhibited no significant neurological symptoms, endoscopic unilateral laminectomy decompression was performed. In cases where preoperative imaging results indicated severe compression on both sides of the spinal canal, endoscopic bilateral laminectomy decompression was conducted. Regardless of the degree of compression observed on imaging, if a patient exhibits severe neurological symptoms in both lower extremities, endoscopic bilateral laminectomy decompression was also performed. With the base of the spinous process as the inner boundary, the width of each side of the laminectomy was about 1 cm (1 to 1.5 trephine diameters), and the length was the distance from the end of the laminae to the tip of the head. Under the guidance of C-arm fluoroscopy, a guide needle was inserted through the endoscopic channel into the lamina or inferior facet process near the decompression target. A skin incision of less than 1 cm was made with the guide needle as the center. Dilating cannula were successively inserted over the guide wire until a working cannula was in place, then the spinal endoscopic system is connected. A trephine is used to remove part of the lamina and less than 50% of the inferior facet process, at which point the dural sac will be seen for the first time. The laminectomy was repeated by changing the trephine position until decompression was complete. Burrs, forceps and Kerrison punch can also be used alternately for different tissue structures during decompression (Fig. 5). The endoscopic procedure was performed with about 30 to 50 mmHg of irrigation fluid pressure to ensure safe and optimal endoscopic visual field clarity [11]. If the pedicle screw or rod obstructs the movement of the endoscopic decompression channel, it can be temporarily removed. When there was no bleeding in the spinal canal and the pulsation of the dural sac resumed, the endoscopic system was withdrawn. For patients requiring bilateral decompression, the procedure was repeated on the other side. If a preexisting rupture of the dura mater sac is identified during surgery or if the surgical procedure itself causes such a rupture, it becomes imperative to supplement the open surgical repair by means of an auxiliary incision measuring approximately 2–3 cm in length. This incision facilitates the placement of a tubular dilator for gentle soft tissue stretching, followed by meticulous suturing and repair of the dura mater sac under microscopic guidance.
Step 3: Repeat reduction and fixation
Pedicle screws that were temporarily uninserted or internal fixators that were temporarily removed because of blocking the decompression pathway were inserted. C-arm fluoroscopy was used to confirm the restoration of spinal alignment. External traction and compression tools were applied again if necessary. Finally, a drainage tube was placed, and the wound was rinsed and closed.
2) Open Surgery
The conventional posterior open approach was implemented utilizing a midline incision about 8–12 cm. Pedicle screws and rods were inserted to restore spinal alignment and stability, and the criteria for fixation segments and screws were the same as in the EMIS cohort. The standard for laminectomy decompression in the OS cohort was consistent with that in the EMIS cohort, and the spinous process and supraspinous ligament were preserved in both cohorts. Similarly, a drainage tube was placed, and the wound was rinsed and closed.
4. Postoperative Treatment
Following surgery, all patients were required to wear a brace for a period of 8–12 weeks and underwent neurological rehabilitation for 6 months under the guidance of a rehabilitation physician. The patients were clinically and radiologically monitored for at least 2 years.
5. Statistical Analysis
All statistical analyses were performed using IBM SPSS Statistics ver. 25.0 (IBM Co., Armonk, NY, USA). The Shapiro-Wilk test was used to assess the normality of the data. Measurement data conforming to the normal distribution were expressed as mean±standard deviation, and the independent samples t-test was used to compare between cohorts. The enumeration data were expressed as percentages (%), and the chi-square test was used to compare between cohorts. The comparison of complications was conducted using Fisher exact test. A 2-sided p-value less than 0.05 was statistically significant.
RESULTS
1. Demographic and Initial Clinical Characteristics
A total of 231 patients were enrolled in this study and classified into EMIS (n=107) and OS (n=124) cohorts. There were no significant differences in demographic and initial clinical characteristics between the 2 cohorts (p>0.05) (Table 1).
2. Perioperative Outcomes
The EMIS cohort exhibited superior outcomes compared to the OS cohort in terms of incision length, intraoperative blood loss, postoperative plasma drainage, and hospitalization stay (p<0.05). However, the operative time of the EMIS cohort was significantly longer (p<0.05) (Table 2).
3. Patient-Reported Outcomes
The preoperative VAS and ODI scores did not show any significant difference between the 2 cohorts (p>0.05). However, postoperative and at the last follow-up, the EMIS cohort exhibited superior VAS and ODI outcomes compared to the OS cohort (p<0.05) (Table 3).
4. Imaging Outcomes
There were no significant differences in CER, AVH, and CA between the 2 cohorts before operation, after operation, and at the last follow-up (p>0.05). Additionally, there was no significant difference in postoperative correction loss between the 2 cohorts (p>0.05). These findings suggest that both EMIS and OS approaches yield comparable effects in terms of decompression, spine alignment recovery, and deformity correction (Table 4). The representative radiological outcomes of the EMIS cohort are shown in Fig. 6.
5. Neurological Status Outcomes
Regarding the ASIA grade, there was no statistically significant difference observed between the 2 cohorts (p>0.05). When assigning a 1-point improvement to each ASIA grade, the scores for neurological function recovery status in both the EMIS and OS cohorts were found to be 0.90±0.57 and 0.98±0.49, respectively (p>0.05). These findings suggest that there is comparable neurological recovery between the EMIS cohort and the OS cohort (Table 5).
DISCUSSION
The AOSpine thoracolumbar injury classification system (TLICS) integrates critical elements of the Magerl classification system and the TLICS [1]. It has been reported that the thoracolumbar AOSpine injury score established on this basis can guide clinical treatment, and surgical treatment is indicated for cases with scores greater than 5 [2]. Surgical treatment has the advantages of early mobilization, shorter hospital stay, fewer pulmonary complications, and better correction of sagittal balance compared with nonsurgical treatment. In the present study, all patients exhibited thoracolumbar AOSpine injury scores ranging from 5 to 9.
For patients with A3 or A4 fractures presenting with neurological deficits, surgery aims to decompress the nerve, reconstruct the spinal alignment and stability. Common surgical approaches include anterior, posterior, and combined approaches, although no consensus has been reached on the optimal surgical approach [3,16,17]. It is widely acknowledged that the anterior approach facilitates superior reconstruction of the anterior column of the spine, enabling direct and comprehensive decompression of the spinal canal; however, it does entail increased surgical trauma and a higher incidence of surgical complications. An overwhelming literature substantiates that the posterior approach yields comparable radiological and functional outcomes to the other 2 approaches, while simultaneously reducing operation time, bleeding, and complications [16-19].
During conventional posterior OS, instrumentation is often placed 1 or 2 levels above and below the injured segment, and spinal canal decompression is performed at the level of injury [18,19]. Therefore, the majority of incisions and soft tissue dissections are performed to facilitate the insertion of pedicle screws. In this study, the authors established a minimally invasive approach combining posterior endoscopic decompression with PPSF to treat AOSpine A3 or A4 thoracolumbar fractures with neurological deficits. The results showed that the EMIS cohort yielded superior outcomes compared to the OS cohort in terms of incision length, intraoperative blood loss, and postoperative drainage volume, at the same time, the improvement of VAS and ODI in the 2 cohorts met the requirements of minimal clinical important difference [20], but the improvement of VAS and ODI in the EMIS cohort was better than that in the OS cohort in the early postoperative period, although this advantage gradually decreased with time.
Timely neurodecompression represents a crucial step in the management of thoracolumbar fractures associated with neurological defects, and it can be categorized into indirect and direct during posterior surgical intervention. Indirect decompression is mainly achieved through ligamentotaxis [21], while direct decompression primarily involves the removal of the posterior lamina, ligamenta flavum and other associated structures. Although a small number of studies have shown that indirect decompression can achieve similar neurological improvement compared with direct decompression [22], indirect decompression can only restore about 50% of the spinal canal capacity, and even is ineffective for some cases [23]. Therefore, to ensure adequate neurodecompression, most surgeons opt for incorporating direct decompression alongside indirect decompression during posterior surgery. In this study, the EMIS cohort underwent direct decompression through a spinal endoscopic system involving laminae and ligamenta flavum removal. At the final follow-up, there were no significant differences observed in decompression-related CER or ASIA grade outcomes between both cohorts. Furthermore, despite the reduced surgical trauma compared to open decompression, endoscopic decompression also presents potential drawbacks. On review of the surgical records, we identified 3 and 5 patients in the EMIS and OS cohorts, respectively, who had noniatrogenic dural sac ruptions, most of which were caused by violence at the time of injury. Repairing these ruptures in the OS cohort was easier because there was a greater exposure field conducive to suturing. Unfortunately, in the EMIS cohort, the use of endoscopy to repair the dural sac under the existing conditions in the region is very challenging. Therefore, an additional small open incision is required to suture and repair the ruptured dural sac. However, with continuous advances in equipment and technology, we believe that this challenge will be overcome in the future.
Restoration of spinal alignment and stability constitutes a crucial aspect in the management of thoracolumbar fractures with neurological deficits, with open pedicle screw fixation being extensively employed in posterior surgical procedures. Compared to open pedicle screw fixation, PPSF demonstrates comparable mechanical strength and deformity correction capacity while minimizing surgical trauma, thereby benefiting patients [4,5]. Simultaneously, numerous studies on short segment fixation for this fracture type have reported favorable radiological outcomes. Therefore, short segment fixation was adopted in this study [13,21,24-27]. In studies involving the application of PPSF for the treatment of spinal fractures, including types A3 and A4, correction of CA ranged from 7.85° to 12.9° and AVH improved from 29.16% to 45% [24-27]. Consistent findings regarding CA and AVH were observed in the present study within the EMIS cohort. Moreover, the EMIS cohort did not yield significant differences in outcomes compared to the OS cohort.
The A3 and A4 thoracolumbar fractures are burst fractures and are generally considered unstable. If posterior surgery is used to treat this type of spinal fracture, additional bone fusion is thought to provide long-term stability and prevent recurrence of kyphosis. However, a growing body of prospective and retrospective evidence confirms the effectiveness of posterior internal fixation without bone fusion in the treatment of these unstable thoracolumbar burst fractures [13,28-30]. There is no difference in radiological results between posterior internal fixation without bone fusion and posterior internal fixation with bone fusion, but non-fusion surgery has less trauma and faster functional recover [28,30]. These results support the rationale for treating A3 or A4 thoracolumbar fractures with PPSF without assisted bone fusion. In studies evaluating the use of posterior internal fixation without fusion for the treatment of A3 or A4 thoracolumbar fractures, CA loss has ranged from 1.5° to 8.3° [13,28-30]. The EMIS cohort with no bone fusion surgery in this cohort experienced a CA loss of 2.25° at the last follow-up, which showed no difference from the OS cohort that underwent bone fusion (p>0.05). One case of screw breakage occurred in each cohort, and a comprehensive review of patient history showed that these patients resumed intense physical work after complete neurological recovery, causing a serious burden on the spine.
Endoscopic decompression and PPSF are considered different forms of minimally invasive spinal surgery [6], and few studies have reported combining these 2 approaches for AOSpine A3 or A4 thoracolumbar fractures with neurological deficits. Tubular retractors or microscopic decompression combined with PPSF have been used in previous studies applying minimally invasive approaches to manage similar diseases [24-27]. Unlike these procedures, surgical trauma with endoscopic decompression is likely to be minimal [6,7,9,10,31]. Water irrigation during endoscopic surgery may reduce the infection rate, clarify the operative field, and reduce epidural blood supply damage.
However, the limitations present in this study should be acknowledged. Firstly, patients selected a surgical method based on their individual preferences, which may introduce selection bias. Differences in patient background factors such as age, sex, and fracture levels could have influenced the results. However, no statistically significant differences were observed in the baseline data between the 2 groups, suggesting comparability. Therefore, we did not adjust for these background factors. We believe that the potential bias from these factors would have minimal impact on the study outcomes. Secondly, physicians and patients are often exposed to greater radiation doses due to PPSF, which robotic or navigation-assisted devices could remedy [32]. Thirdly, the measurement of the cross-sectional area of the dural sac by MRI can better reflect the decompression effect, but imaging errors caused by metal artifacts cannot be avoided with currently available detection equipment. Fourthly, none of the fracture patients experienced dislocation in this study, consequently, we did not assess inter-group differences in alignment restoration. However, evaluating the restoration of spinal alignment is crucial for the effective treatment of fracture patients, and thus will be a critical focus in our future research. Furthermore, there is currently no standardized protocol for the removal of internal fixation devices following fracture healing [33]. Consequently, this investigation did not assess radiographic changes subsequent to the extraction of internal fixation devices.
CONCLUSION
In conclusion, endoscopic decompression combined with PPSF in minimally invasive treatment of type A3 or A4 thoracolumbar fractures with neurological deficits has the same efficacy as traditional OS in terms of decompression, reduction, improvement of clinical symptoms and neurological function, while minimizing surgical trauma.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
