INTRODUCTION
The cervicothoracic junction (CTJ) is a complex anatomical region characterized by the transition from the lordotic and flexible cervical spine to the kyphotic and rigid thoracic spine [1]. It is standard practice to extend supportive constructs across the CTJ, as terminating fixation at the junctional level may lead to instability [2-4]. This is especially important in cases where CTJ lesions are accompanied by instability, such as pathologic fracture caused by spinal metastasis. Consequently, it is imperative to develop reliable methods for effectively stabilizing the CTJ.
Traditionally, anterior corpectomy (AC) (Fig. 1A), posterior long-level fusion, and a combined anterior and posterior approach have been used for CTJ stabilization. In the posterior approach, a lateral mass screw (LMS) with 3.5-mm rods is used for the subaxial cervical spine, whereas a pedicle screw with 5.5-mm rods is inserted at the thoracic spine. These 2 screws are connected using either a domino (Fig. 1B) or a tapered rod (Fig. 1C). While the anterior approach facilitates direct pathologic decompression, it provides less biomechanical stability and faces anatomical barriers including the sternum, critical vascular, and neural structures [5,6]. Posterior instrumentation of the CTJ is often preferred because of its biomechanical advantages, especially for long-level fusion [7]. However, the posterior approach has drawbacks, including longer operative times and fixation of normal segments.
Recently, cervical pedicle screw (CPS) has been adopted to overcome the disadvantages of posterior long-level fusion. The CPS allows robust support with the fixation of only short segments.8-11 Particularly, the use of intraoperative computed tomography (CT) navigation has made it relatively safe to insert CPS into the subaxial cervical spine [12]. Advancing this approach, our group has developed a novel technique that employs large screws typically used in the lumbar region for the subaxial cervical spine, enhancing support with more robust 5.5-mm rods (Fig. 1D). In our previous study, we compared the technique of inserting CPS with 5.5-mm rods and the technique of inserting CPS with the traditional 3.5-mm rods for the treatment of metastatic cervical spinal tumors extending from C2 to the CTJ and demonstrated the effectiveness of CPS with 5.5-mm single-diameter rods [13]. In this study, we evaluated the usefulness of posterior pedicle screw fixation with 5.5-mm rods (PPSF5.5) in cases limited to CTJ metastasis and compared it to AC. We propose PPSF5.5 as a novel method for stabilizing CTJ metastasis.
MATERIALS AND METHODS
This study adhered to the Helsinki Declaration and received approval from the Institutional Review Board (IRB) of Asan Medical Center. The IRB also granted a waiver for informed consent required for this research (AMC IRB 2024-0878).
1. Patient Population and Treatment Groups
From January 2000 through December 2023, 68 consecutive patients diagnosed with metastatic CTJ spine tumors were treated with anterior and posterior fusion surgery at a single center. A metastatic CTJ spine tumor was defined as a tumor located between C7 and T2 that required fixation of the C7–T1 segment. Surgery was considered for patients with a life expectancy exceeding 3 months. Inclusion criteria specified patients with a minimum follow-up duration of 3 months who underwent either AC or PPSF5.5. Exclusion criteria encompassed patients with a follow-up period of less than 3 months, those who received LMS fixation using a domino or tapered rod, and those who underwent combined anterior and posterior fusion surgery. Consequently, 11 patients with inadequate follow-up, 8 patients receiving LMS with a domino or tapered rod, and 14 patients undergoing combined procedures were excluded, resulting in a total of 35 included patients (Fig. 2). Of these, 12 underwent AC and 22 underwent PPSF5.5. The specific surgical methods for each group are described in subsequent sections.
2. Surgical Procedures
AC was primarily decided when the tumor was localized to the vertebral body. The level of corpectomy was determined to sufficiently remove the tumor to ensure that the spinal cord was fully decompressed. The patient was placed in a supine position with the head extension. A transverse incision in one-level corpectomy or oblique skin incision beyond 1-level corpectomy was performed. After the anterior surface of the vertebral body was exposed, Casper Distraction Screw (SURTEX, London, England) were inserted into the upper and lower vertebral body, which were adjacent to the corpectomy region. Tumor removal was conducted until the normal vertebrae above and below the tumor were exposed. A titanium mesh cage (TMC, Medtronic Sofamor-Danek, Memphis, TN, USA) filled with artificial bone graft was inserted into the interbody space, followed by the application of a plate and variable screw fixation. Finally, muscles and skin were closed tightly.
PPSF5.5 was primarily performed when the tumor was found to involve both the anterior and posterior column and the tumor center was in T1 or T2. The fixation procedure can be summarized as follows. The patient was placed in a prone position while continuously monitoring motor-evoked potentials, except in emergencies. Screw entry points were established based on CT scans. A pilot hole was drilled, followed by a curved pedicle probe to delineate the cancellous channel. Next, a straight pedicle probe, tap, and appropriately sized Legacy pedicle screws (Medtronic Sofamor-Danek, Memphis, TN, USA) were employed. The use of Legacy screws designed for the thoracolumbar area in the cervical area was considered off-label use. The procedures were executed using a freehand technique aided by intraoperative CT navigation. After screw placement, the laminae were excised to decompress the spinal cord. If necessary, the ventral osteolytic lesion of the tumor was removed through a transpedicular approach, and a 3D-printed cage (Genoss, Suwon, Korea) was implanted to maintain structural stability. The 5.5-mm rods were positioned, and the instrumentation placement was verified using intraoperative CT scans. Posterolateral fusion was accomplished using allograft bone chips.
3. Representative Case of PPSF5.5
A 75-year-old woman with neck pain and lower extremity weakness was incidentally found to have thyroid cancer with metastasis to the C7 vertebrae. Emergency CTJ lesion surgery was planned to improve neurologic deficits. Preoperative CT scans showed an osteolytic lesion with a pathological fracture at C7 (Fig. 3A), and T1-enhanced magnetic resonance imaging revealed a tumor confined to the vertebral body with epidural spinal cord compression scale grade 3 spinal cord compression (Fig. 3B). Dynamic x-ray analysis did not show significant instability (Fig. 3C). Initially, an anterior approach and C7 corpectomy were considered; however, since the tumor extended into the lower part of the C6 body and extending the corpectomy could compromise anterior support, a posterior approach was selected. Given the patient’s old age and comorbidities, PPSF5.5 was chosen. Additionally, the tumor’s ventral portion was removed through the C7 transpedicular approach, and a 3D-printed cage was inserted to provide anterior support. Finally, posterolateral fusion was performed using artificial bone chips. After surgery, the lower extremity weakness improved; postoperative x-ray analysis confirmed the instrument was firmly fixed (Fig. 3D), and CT scans showed the 3D titanium cage providing anterior support was well-secured (Fig. 3E).
4. Outcome Assessments
Patients were monitored at 1-, 3-, 6-, and 12-month postsurgery. Evaluations included the preoperative Spinal Instability Neoplastic Scale (SINS) score, pre- and postoperative visual analogue scale (VAS) scores for neck or back pain, McCormick scale, Nurick grade, and Eastern Cooperative Oncology Group (ECOG) performance status [14-17]. Radiological assessment of the segmental Cobb angle was conducted via x-ray analysis at the preoperative, immediate postoperative, and final follow-up stages. Changes in segmental Cobb angle (difference between last follow-up and immediate postoperative measurements) were calculated and described as reduction loss. A reduction loss exceeding 10° was defined as vertebral collapse. Furthermore, cases requiring revision surgery were categorized based on factors such as instrumentation failure, tumor regrowth, wound infection, and minor complications.
5. Statistical Analysis
All statistical analyses were conducted using IBM SPSS Statistics ver. 22.0 (IBM Co., Armonk, NY, USA). Demographic data are presented using descriptive statistics including frequency, percentage, range, mean, and standard deviation. The normality of variables was tested using the Shapiro-Wilk test. Continuous variables were compared using an independent t-test, whereas categorical data comparisons between the AC group and the PPSF5.5-mm group were made using the chi-square test or Fisher exact test. Multivariate linear regression analysis was conducted using the backward stepwise method, adjusted for variables that might influence the outcomes. The independent variables included age, sex, tumor center, fusion level, use of a cage, and approach. The dependent variable was the change in the segmental Cobb angle (postoperative – last follow-up), indicating kyphosis progression. Statistical significance was set at p< 0.05 (2-tailed).
RESULTS
Baseline demographic characteristics are summarized in Table 1. The mean age for the AC group was 53.0± 9.4 years, with 6 males (50%). In the PPSF5.5 group, the mean age was 57.4± 14.0 years, with 12 males (54.5%). Overall, lung cancer was the most common primary cancer (7 cases, 20.5%), though the distribution of primary cancers between the groups was not significantly different (p= 0.948). Tumors in the AC group were primarily located at C7 and T1, whereas those in the PPSF5.5 group were uniformly distributed from C7 to T2 (p < 0.001). According to the Weinstein-Boriani-Biagini classification system, tumors in the AC group were mainly located in anterior elements, whereas in the PPSF5.5 group, they were distributed across both anterior and posterior elements (p= 0.009). Fusion levels typically ranged from 2 to three in the AC group and from 2 to 6 (p= 0.001) in the PPSF5.5 group. Preoperative embolization was performed in 3 patients in the AC group and 4 patients in the PPSF5.5 group. The operative duration did not differ significantly between the AC and PPSF5.5 groups (272± 107 minutes vs. 264± 40 minutes, p= 0.769). The length of hospital stay also did not differ between the 2 groups (AC group, 12.0± 4.4 days vs. PPSF5.5 group, 11.1± 3.6 days, p= 0.523). The follow-up duration was 12.0 ± 14.0 months in the AC group and 21.2 ± 17.5 months in the PPSF5.5 group (p= 0.141).
Clinical outcomes for both the anterior and posterior approach groups are detailed in Table 2. The SINS score was slightly higher in the PPSF5.5-mm group than in the AC group, with no significant difference (13.0± 1.4 vs. 14.0± 1.8, p= 0.085). Additionally, there were no significant differences in the improvement of clinical parameters such as the McCormick scale, Nurick grade, and ECOG performance status from before surgery to 1 month after surgery in both groups (p-value, McCormick scale= 0.162, Nurick grade= 0.888, ECOG performance status= 0.640). However, for the neck/back VAS score, the difference between the values presurgery and one-month postsurgery was 3.7± 2.1 in the AC group and 6.3±2.1 in the PPSF5.5 group, indicating greater improvement (p= 0.003) in the PPSF5.5 group.
Radiological and surgical outcomes are presented in Table 3. A significant difference in segmental Cobb angle change was observed between the groups. The analysis revealed that a higher proportion of patients in the AC group experienced a segmental Cobb angle recovery of 10° or more postsurgery than the PPSF5.5 group (4 vs. 1, p= 0.045). However, the AC group also had more instances of index vertebral collapse, defined as a segmental Cobb angle change of negative 10° or more during follow-up than the PPSF5.5 group (6 vs. 1, p< 0.001). In the AC group, 2 cases required revision surgery due to instrumentation failure, whereas no such cases were reported in the PPSF5.5 group (p= 0.005). Revision surgery due to tumor regrowth was needed in one case in the AC group compared to 6 in the PPSF5.5 group, but this difference did not reach statistical significance (p= 0.203). As a minor complication, 1 case in each group required revision surgery due to infection (p= 0.665).
Multivariate linear regression analysis of the change in the segmental Cobb angle between postoperative and last follow-up indicated that the approach was the only significant variable (p < 0.001) associated with kyphotic progression after fusion surgery, whereas the use of a cage was not related (p = 0.736) (Table 4).
1. Cases of Revision Surgery due to Instrumentation Failure
A 62-year-old woman underwent a corpectomy at C6–7 for spinal metastasis from lung cancer (Fig. 4A). Six months later, she exhibited motor weakness in all 4 limbs. X-ray analysis revealed cage dislodgment, resulting in severe kyphotic deformities (Fig. 4B). Her compromised general health condition prevented further surgical intervention, and she later passed away due to pneumonia. In another case, a 56-year-old man underwent a corpectomy at T1 for spinal metastasis from cholangiocarcinoma (Fig. 4C). One-month postsurgery, he developed weakness in his lower extremities. A CT scan indicated cage displacement causing severe spinal cord compression (Fig. 4D). Although revision surgery was advised, the patient declined due to his limited life expectancy.
DISCUSSION
The CTJ is a transitional area characterized by significant changes in alignment and biomechanics where the lordotic and mobile cervical spine transitions into a kyphotic and rigid thoracic spine [1]. This junction, bridging the mobile cervical and rigid thoracic regions, is under increased mechanical stress, necessitating robust instrumentation at the CTJ to withstand these forces [2]. Stabilizing the CTJ, especially in patients with spinal metastases, remains a controversial topic regarding the optimal intervention strategy. Chakravarthy et al. [18] conducted a retrospective analysis of their posterior instrumentation techniques for CTJ metastasis, evaluating posterior instrumented fixation with a combined LMS and pedicle screw instrumentation along with tapered or fixed rods, optionally supplemented by anterior column reconstruction. Their findings suggest that a posterior approach at the CTJ yields sustainable improvements in patient-reported quality of life, with a complication rate of 18.8% and a 2-year hardware failure rate of 11%.
Hubertus et al. [19] carried out a multicentric study to define surgical techniques for metastases at the CTJ. They classified surgical procedures into 4 categories: (1) posterior decompression only, (2) posterior decompression and fusion, (3) AC and fusion, and (4) combined anterior and posterior approach and 360° fusion. Their results indicate that posterior decompression alone is feasible when spinal stability is intact, offering benefits due to shorter duration and comparatively low complication rates. They also observed that AC created a stable construct with lower hardware failure rates and nearly half the surgical complications compared to other techniques, challenging the widely held belief that the anterior approach has a higher instrumentation failure rate [1]. Siemionow et al. [20] reported a high incidence of complications associated with the combined anterior and posterior approaches. Furthermore, various methods have been developed to stabilize the CTJ. Boriani et al. [21] described an innovative method using preshaped carbon rods anchored to the laminae with sublaminar bands, whereas Obeidat et al. [22] introduced a novel technique involving LMS for posterior instrumentation in the subaxial cervical spine and cortical bone trajectory screws for the proximal thoracic spine.
As described above, there remains no consensus on the optimal CTJ fixation technique. Patients with spinal metastasis have a median survival of 21.6 months, surgical complications exceeding 5%, and a recurrence rate of 32.3%, with an average of 9.8 months until reoperation [23]. Given the traits of spine metastasis patients, the risks associated with major surgery and complications can critically affect their prognosis, necessitating a stable and simplified surgery. As noted in the study by Chakravarth et al., a posterior-only approach can provide effective and robust stabilization. In addition, the efficacy of pedicle screws in the subaxial cervical spine has been well-documented [8,10,24-26]. By applying this CPS for stabilizing the CTJ, the process became more robust, which our group has further advanced by using a 5.5-mm rod. We adopted a technique that combines pedicle screws with 5.5-mm rods commonly used in lumbar surgeries to minimize surgical extent, reduce operative time, and preserve kyphotic bending momentum. This technique has shown superior outcomes compared to the traditional 3.5-mm rods in prior studies [13]. In the current study, we specifically addressed CTJ metastases and compared the outcomes with the anterior approach.
In general, the pullout strength of screws in posterior fixation correlates with screw diameter and length [27], whereas vertebral collapse is influenced by the posterior rod [28]. Our technique uses screws with 5.5-mm rods. The thicker and longer screws enhance pullout strength and thicker rods reduce collapse (Fig. 1E), decreasing the risk of vertebral collapse compared to AC. To identify other factors affecting collapse, we conducted a logistic linear regression test using age, sex, tumor center, fusion level, use of a cage, and approach as independent variables. All factors except for the approach did not significantly affect the degree of collapse, particularly the presence of anterior support, such as AC with TMC insertion and posterior transpedicular approach with 3D-printed Genoss cage insertion, which did not influence vertebral collapse. Although AC initially provided better correction of the segmental Cobb angle postoperatively, the anterior approach group exhibited greater kyphosis progression over time due to vertebral collapse and showed a significantly higher incidence of surgical complications due to instrumentation failure. Previous studies, including subgroup analyses focusing on CTJ cases, consistently demonstrated greater stability when CPS was used with 5.5-mm rods than when CPS was used with 3.5-mm rods [13]. Moreover, 4 patients belonged to the CTJ metastasis 3.5-mm group in our cohort. The average fusion level for these 4 patients was 6.25, which was greater than the average of 3.47 in the 5.5-mm rods group. Integrating these findings with our current data, we confirm the PPSF5.5 approach provides superior stabilization and short-segment posterior fixation compared to both the AC approach and traditional posterior fixation methods.
Clinical outcomes also revealed the significance of the PPSF5.5 technique. Although not significantly different, the SINS score in the PPSF5.5 group was higher than that in the AC group (AC group: 13.0± 1.4 vs. PPSF5.5 group: 14.0± 1.8, p= 0.085). This finding suggests that there were more patients experiencing instability in the PPSF5.5 group, which also aligned with the higher preoperative VAS score in that group (AC group: 5.0± 2.5 vs. PPSF5.5 group: 7.9±1.7, p=0.001). However, both groups showed similar improvement in VAS scores postsurgery (AC group: 1.3± 1.4 vs. PPSF5.5 group: 1.5± 1.3, p= 0.646), which indicated that the posterior approach provided better instrument stability, leading to reduced pain.
On the other hand, patients in the AC group experienced fewer recurrences, although this difference was not significant. This outcome may relate to tumors localized to the anterior vertebral body, allowing AC to directly remove nearly all tumors and potentially reduce the likelihood of recurrence. The posterior approach was also able to remove the ventral tumor through the transpedicular approach, but the range of removal was limited compared to the anterior approach. However, when comparing survival periods, the PPSF5.5 group had a longer survival period (AC group: 12.0±14.0 months vs. PPSF5.5 group: 21.2± 17.5 months, p= 0.141). This suggests that the prognosis of spine metastasis is more directly influenced by the control of the primary tumor rather than the spine tumor itself. Of course, in cases where tumor growth is slow or the tumor is resistant to radiation therapy, and the primary cancer is managed to some extent, it may be worth considering a more aggressive approach to remove spine metastasis. From this perspective, the anterior approach could selectively offer advantages. Nevertheless, the anterior approach is typically limited to one or 2 levels and is generally more feasible for tumors positioned in the upper CTJ levels. Still, for patients with spine metastasis, selecting a less invasive surgical method that minimizes complications is essential. Based on our experiences and outcomes, we advocate for PPSF5.5 method as a viable alternative.
This study has several limitations. First, it was a retrospective analysis with a relatively small sample size, which constrains the scope of our analysis. Second, there were no cases with a tumor center at T2 in the AC group, whereas there were 13 such cases in the PPSF5.5 group. This difference in numbers suggests the possibility of selection bias. However, it should be noted that all the 13 cases in the PPSF5.5 group had tumor extensions involving T1 and C7. Consequently, they were biomechanically similar to patients with the tumor center at C7 and T1. Third, to demonstrate the stability and superiority of PPSF5.5, a direct comparison with PPSF3.5 (posterior pedicle screw fixation with 3.5-mm rods) was necessary. Unfortunately, our cohort contained only a few cases that could be included the 3.5-mm rods group for CTJ metastasis, making direct comparison challenging. This highlights an area that requires further research. Fourth, the effect of screw diameter on construct durability was not evaluated, which could be an important variable influencing the outcomes. Fifth, in our logistic linear regression analysis, TMC used in AC and 3D-printed cage used in the posterior approach were grouped as anterior support. To determine if anterior support influences vertebral collapse, further studies should separately examine these 2 cage types. Sixth, with a minimum follow-up period of 3 months, some cases lacked sufficient follow-up to fully assess instrumentation durability, likely due to the short life expectancy associated with the patients’ conditions. Nevertheless, our novel technique demonstrates promise in CTJ stabilization, and further research with a systematic design is essential to validate the effectiveness of this method.
CONCLUSION
In cases of CTJ metastatic tumor, our novel PPSF5.5 technique demonstrates superior resistance to forward bending and collapse and minimized instrumentation failure compared to AC technique. This method can also reduce operative time, fusion levels, and complication rates. AC may reduce the risk of tumor recurrence. However, this approach is recommended only when the tumor is confined to the vertebral body and situated at the upper level of CTJ. Based on our findings, we advocate for PPSF5.5 as a viable alternative for stabilizing CTJ metastasis.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
