INTRODUCTION
Spinal fusion surgery is an effective and well-established procedure for treating various conditions affecting the spine, namely low back pain due to degenerative disc/facet disease, neurogenic claudication, symptomatic spondylolisthesis, degenerative scoliosis, and radiculopathy due to foraminal stenosis [1]. Adjacent segment disease (ASD) is a potential complication (5%–30%) of spinal fusion procedures; proximal junctional kyphosis (PJK) is a specific radiographic manifestation of ASD and is diagnosed when the proximal junctional sagittal Cobb angle is ≥10° and at least 10° greater than the preoperative values [2,3]. The definition of PJK varies slightly across different studies. Risk factors for PJK include older age, osteoporosis, low bone mineral density (BMD), and poor global spinal alignment [4,5]. Long-term spinal interbody fusion alters the biomechanical stress on adjacent spinal segments, and such fusion may contribute to new-onset radiologic changes and physical symptoms requiring revision surgery. The most severe form of PJK is proximal junctional failure (PJF). PJF is a condition in which mechanical failure at the proximal junction causes symptomatic vertebral fracture, instrumentation failure, or spondylolisthesis, which typically results in revision surgery [6].
Research has suggested that proximal junctional tethering (PJT) mitigates the severity of and even prevents the development of PJK or PJF. Some retrospective and prospective cohort studies and case series have attempted to demonstrate the effectiveness of such tethers, reporting promising results [6-17]. However, few systematic reviews and meta-analyses have examined the efficacy of tethering, possibly owing to the lack of relevant randomized controlled trials. Vercoulen et al. [18] conducted a systematic review of spinal instrumentation techniques designed to reduce the postoperative rate of PJK; they observed a significant difference in PJK incidence between groups involving the use of various configurations of tethering and a group involving the use of pedicle screws. Zhao et al. [19] conducted a systematic review and meta-analysis; they suggested that ligament enhancement significantly reduced the risk of PJK, and they determined that ligament strengthening was superior to the use of laminar hooks. Moreover, Sursal et al. [20] concluded that although most early studies have indicated tethering for ligament augmentation could exert protective effects against PJK, a single technique is unlikely to solve this complex problem given the multifactorial etiology of PJK. However, the overall evidence regarding the efficacy of tethering remains scarce and weak. Accordingly, we conducted a meta-analysis to assess the effectiveness of PJT in reducing the odds of PJK or PJF in adults undergoing spinal interbody fusion surgery.
MATERIALS AND METHODS
1. Protocol and Registration
This review was conducted in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement [21]. The protocol of this meta-analysis was prospectively registered in PROSPERO on July 6, 2024 (ID: CRD42024562327).
2. Selection Criteria
Eligible studies were selected in accordance with the PICO (population, intervention, control, and outcome) framework. Specifically, studies were included if the population (P) comprised patients of both sexes, all ethnicities, and all nationalities receiving spinal fusion; the intervention (I) was PJT performed along with spinal fusion for reducing the rates of ASD; the control (C) comprised patients undergoing spinal fusion without tethering; and the outcome (O) was the odds of PJK or PJF at various follow-up intervals.
The inclusion criteria were as follows: (1) including patients receiving spinal fusion surgery, (2) applying prospective/retrospective comparative designs, (3) including only adult patients (≥18 years old), and (4) clearly reporting the number/percentage of patients who developed PJK or PJF in both arms. The exclusion criteria were as follows: (1) applying a single-arm design; (2) not being published in English; (3) being systematic reviews, case reports, or case series; (4) being cadaveric or purely biomechanical studies; (5) including pediatric patients; or (6) failing to generate data consistent with our definition of PJK or PJF.
3. Literature Search and Data Collection
A systematic literature search was conducted using PubMed, Embase, and Cochrane Library from their inception up to March 2025. The following keywords were used in the search: (adjacent segment disease OR proximal junctional kyphosis OR proximal junctional failure OR proximal junctional kyphosis OR proximal junctional tethering) AND (tethering OR strap OR ligamentous banding OR ligamentous augmentation OR ligament augmentation OR sublaminar OR banding OR tension band OR tension banding OR tape OR taping OR tendon augmentation). No restrictions were implemented in the search.
A comprehensive systematic review protocol was established, and appropriate articles for our meta-analysis were selected according to this protocol. Two researchers, YCH and HTL, independently screened the titles and abstracts of 312 articles to identify studies that met the inclusion or exclusion criteria. Discrepancies in study eligibility assessments between the reviewers were discussed and resolved. Thus, 12 studies were finally selected for inclusion in our meta-analysis.
4. Data Extraction and Study Quality Assessment
Two researchers, YCH and HTL, independently conducted data extraction by using a predefined data extraction form. Data on the following were extracted from the included studies: names of authors, year of publication, geographical location of the original study, design of the study, surgical indication, presence of crosslink, numbers of patients treated with tethering and control patients, types and level of tethers, follow-up time, and outcome of interest (rate of PJK or PJF development in patients treated with tethering and control patients).
The same 2 authors also independently performed a risk-ofbias assessment. The risk of bias in the 11 studies [6-15,17] was determined using the Newcastle-Ottawa quality assessment scale for cohort studies [22]. This scale consists of 3 key domains: selection, comparability, and ascertainment of outcome. Each domain contains 1–4 items to be rated, and high-quality choices are rated with a star, with the maximum possible score being 9. Two stars may be awarded for the comparability domain. Studies with a score of ≥7 are considered to be of high quality.
5. Statistical Analysis
Pooled OR and the corresponding 95% confidence intervals (CIs) were calculated using the inverse-variance weighting method. For studies with zero events in one or both arms, a continuity correction of 0.5 was added to all cells to enable calculation of the OR. The heterogeneity of the included studies was explored using the I2 statistic and is presented as the percentage of variation across the included studies. I2 values of <25%, 25%– 50%, and >50% were considered to indicate low, moderate, and high heterogeneity, respectively. If I2 exceeded 50%, a randomeffects model was used for the pooled analysis; if I2 was between 25% and 50%, a random-effects model was preferred; otherwise, a fixed (common)-effect model was applied. Egger test was used to detect any potential publication bias in analyses including more than 10 studies. A trim-and-fill adjusted funnel plot was used to explore asymmetry and estimate the potential influence of unpublished studies. In analyses with 10 or fewer studies, Egger test and the trim-and-fill method were not applied due to limited power; instead, a funnel plot was used to visually assess potential publication bias. All analyses were performed using R v. 4.4.1 (R Foundation for Statistical Computing, Austria).
RESULTS
1. Study Selection and Characteristics
The literature search conducted using PubMed, Embase, and Cochrane Library identified a total of 312 studies. After implementing the comprehensive review protocol, YCH and HTL independently screened the titles and abstracts of the identified studies. Potentially overlapping cohorts and duplicates were identified and omitted. Of the 312 studies, 20 were selected for fulltext evaluation. Of these 20 studies, 11 were found to be eligible after the application of the inclusion and exclusion criteria. These 11 studies comprised 8 retrospective cohort studies and 3 propensity-score-matched analyses; these studies comprised a total of 1,424 patients. The flowchart of study selection is demonstrated in Fig. 1.
Table 1 provides a summary of the characteristics of the 11 studies. An illustration of the various tethering techniques used in the studies is provided in Fig. 2. The year of publication ranged from 2008 to 2025. Among these 11 studies, adult spinal deformity was the major surgical indication, followed by lumbar spinal stenosis. The sample size ranged from 43 to 452. The mean follow-up period ranged from 1 year to 40 months. The main outcome of interest was the number of patients treated with tethering/control patients who subsequently developed PJK or PJF. Although the included studies had slightly different definitions of PJK, these variations were considered acceptable because the studies shared a common description of the degeneration occurring in the upper adjacent vertebral segment. PJF was defined as symptomatic PJK or any fracture/subluxation/failure of fixation and other similar complications requiring revision surgery.
2. Assessment of Study Quality
The results of the quality assessment are presented in Table 2. All studies were determined to be of high quality (score ≥7).
3. Statistical Analysis
The odds ratio (OR) and heterogeneity were calculated using R ver. 4.4.1. The rates of PJK and PJF in the studies were analyzed, and the results are presented in Figs. 3 and 4, respectively. Fig. 4 provides the details of 2 studies that only presented results regarding PJF. PJT was found to be associated with reduced odds of developing PJK or PJF.
The heterogeneity analysis results indicated moderate heterogeneity among the studies on PJK (I2=46.3%), as shown in Fig. 3. Accordingly, a random-effects model was applied for the pooled analysis. The results revealed a significant reduction (OR, 0.44; 95% CI, 0.27–0.71) in the odds of PJK in patients receiving PJT compared with control patients.
The heterogeneity analysis revealed moderate heterogeneity among the studies on PJF (I2=32.2%), as demonstrated in Fig. 4. Therefore, a random-effects model was applied for the pooled analysis. The results revealed a significant reduction (OR, 0.36; 95% CI, 0.19–0.69) in the odds of PJF in patients receiving PJT compared with control patients.
4. Subgroup Analysis
We divided the study groups into subgroups according to geographical location (North America or Japan), tethering with or without crosslinks, and tethering technique (interspinous or sublaminar).
Our subgroup analysis results revealed no significant difference in the odds of PJF (Fig. 5) between geographical locations. Only 1 study reported PJK in the North America arm; hence, PJK was not included in the analysis.
Our subgroup analysis results revealed no significant differ-ence in the odds of PJK (Fig. 6) and PJF (Fig. 7) between patients who received tethering with or without crosslinks. In 3 studies, patients treated with tether-only (TO) and tether with crosslink (TC) procedures were initially grouped into a single cohort, aligning with the goals of our analysis [6,8,12]. However, only 2 of these studies provided separate counts for patients in the TO and TC arms [8,12]. Consequently, the study that did not specify the distribution of TO and TC patients was excluded from our subgroup analysis [6].
Our subgroup analysis results demonstrated no significant difference in the odds of PJF (Fig. 8) between the interspinous and sublaminar techniques. Only one study reported PJK in the sublaminar technique arm; hence, PJK was not included in the analysis.
5. Publication Bias
Publication bias was assessed for both PJK and PJF. For PJK, formal statistical testing for publication bias was not performed due to the limited number of studies. Visual inspection of the funnel plot suggested potential asymmetry, though interpretation is limited by subjectivity and small sample size (Fig. 9). Therefore, the presence of publication bias cannot be ruled out. For PJF, Egger test indicated significant funnel plot asymmetry (p=0.003), as summarized in Table 3, suggesting potential publication bias. Following the application of the trim-and-fill method, the effect size of PJF became nonsignificant (OR, 0.66; 95% CI, 0.36–1.21), as presented in Table 4 and Fig. 10.
DISCUSSION
Spinal fusion surgery is a well-established procedure for treating various spinal disorders. Because of the widespread application of this procedure, the subsequent development of ASD, namely PJK and PJF, has been adequately documented. Many cases of PJK are associated with increased patient pain scores and Oswestry disability index values, rather than being just a radiologic phenomenon [23,24]. The pathophysiological mechanism of PJK and PJF has been hypothesized to be multifactorial.
PJK or PJF can be subdivided into 2 main categories on the basis of the underlying cause: ligament failure and osseous failure. Ligament failure accounts for >70% of PJK cases [25]. Ligament failure can be attributed to iatrogenic damage to the posterior supraspinous and interspinous ligaments, spinous processes, or paraspinal musculature [25]. Ligament failure can lead to increased angular displacement, likely predisposing patients to PJK or PJF [26,27]. Spinal fusion surgery for adult spinal deformity necessitates long-segment posterior instrumentation that inherently introduces considerable stress at the proximal junction, which can engender various effects, including PJK and PJF development [20].
Scholars have demonstrated several factors, including sagittal imbalance and low BMD, that play critical roles in spinal surgical outcomes [7,12,13,17,28,29]. For example, Rabinovich et al. [12] identified severe preoperative sagittal imbalance as a risk factor for PJK or PJF. Alluri et al. [7] noted that significant operative corrections in sagittal vertical axis (SVA) and pelvic incidence-lumbar lordosis (PI-LL) mismatch could increase the risk of complications. Nevertheless, Rodnoi et al. [13] observed that patients who underwent ligament augmentation had a higher baseline SVA, but these patients exhibited a lower rate of PJF. Yagi et al. [17] demonstrated that postoperative spinopelvic alignment, particularly in the C7 SVA and pelvic tilt, was associated with significantly greater improvements in the sublaminar tether group than in the control group over a 2-year follow-up period These results indicate the benefits of ligament augmentation [13,17]. However, studies have demonstrated that BMD significantly influences postoperative outcomes [28,29]. Chen et al. [28] conducted a systematic review and meta-analysis, and they revealed that patients who developed PJK had notably lower preoperative dual-energy x-ray absorptiometry T-scores and vertebral body Hounsfield units (HU) at the upper instrumented vertebra, indicating a mean difference of -0.69 in T-scores (p<0.001) and -32.35 HU (p<0.001). Additionally, Echt et al. [29] demonstrated that teriparatide treatment is the most effective intervention for increasing BMD and reducing the rates of PJF in osteoporotic patients undergoing surgery for treating adult spinal deformity. These findings demonstrate the importance of optimizing BMD before surgical interventions. Moreover, the findings suggest that proactive management of bone health may enhance surgical success and reduce the rate of complications in patients with spinal deformities undergoing spinal fusion surgery.
Various strategies have been proposed to reduce the development of ASD after spinal fusion surgery. Some surgeons perform spinal fusion along with the application of a dynamic hybrid stabilization device or interspinous process device; this is known as the “topping-off” technique [30]. Chiou et al. [31] conducted a network meta-analysis and demonstrated that the topping-off technique reduced the incidence of ASD, reoperation rate due to ASD, and intensity of back pain. However, a disadvantage of the topping-off technique is that it theoretically provides a relatively rigid, rather than flexible, construction. PJT, which provides semirigid, dynamic stabilization, aims to provide a smoother transition zone for reducing proximal junctional stress; thus, PJT is another surgical technique used to reduce the rates of PJK or PJF [32]. The results of this meta-analysis indicate that in adults undergoing spinal fusion surgery, PJT effectively reduces the odds of both PJK (OR, 0.44; 95% CI, 0.27–0.71) and PJF (OR, 0.36; 95% CI, 0.19–0.69). The included studies had moderate heterogeneity for both PJK and PJF. The included studies had moderate heterogeneity for both PJK and PJF. The quality of the included studies was acceptable. These findings underscore the robustness of PJT as a preventive surgical intervention against complications associated with spinal fusion surgery.
We conducted subgroup analyses to further explore the consistency of these findings across different parameters. The subgroup analyses did not reveal significant differences in the odds of ASD between geographical locations, between TO and TC use, or between tethering techniques. This finding suggests that PJT provides consistently protective effects against PJK and PJF, regardless of the geographical location, the presence or absence of crosslink reinforcement, or surgical tethering techniques (e.g., interspinous and sublaminar).
Table 1 describes whether the tether was passed through or around the spinous process or whether the sublaminar tether approach was applied in each study. Our subgroup analysis did not reveal any significant difference in the odds of ASDs between the interspinous and sublaminar tethering techniques. The strength of the lamina is considerably greater than that of the spinous process, especially in patients with osteopenia or osteoporosis [20]. In particular, spinous process banding may lead to clay-shoveler’s fractures in patients with osteopenia or osteoporosis [17]. Viswanathan et al. [33] conducted a biomechanical study and demonstrated that sublaminar banding creates a transition zone; this technique is more effective than passing a Mersilene tape through the spinous process, and it reduces stress at the proximal junction. Both Tachibana et al. [16] and Yagi et al. [17] have used the sublaminar banding technique with polyethylene Nespron tape, and the results revealed a decreased incidence of PJK or PJF. Alluri et al. [7] adopted a novel approach by utilizing a semitendinosus allograft as an alternative to traditional polyethylene tapes or other surgical sutures in sublaminar banding. Pham et al. [34] retrospectively reviewed 4 patients who underwent posterior spinal fusion for adult spinal deformity, with the allograft used to enhance the posterior ligamentous complex augmentation. Short-term follow-up yielded encouraging results, with none of the patients developing PJK after 5.5 months. However, scholars have reported concerns about retrolisthesis after sublaminar banding [11,16]. The sublaminar tethering passage occasionally leads to injury of the underlying spinal cord in patients with adolescent idiopathic scoliosis (AIS) [35]. A large retrospective series of AIS also reported that the rate of neural injury was 0.8%, despite a meticulous surgical technique [36]. The advantages and risks associated with the interspinous and sublaminar tether approaches mentioned above are illustrated in Table 5.
Currently, no consensus has been attained regarding the tension required in interspinous/sublaminar tethering techniques. Only 2 studies [7,11] included in our meta-analysis described the strength of their knotting processes. Doodkorte et al. [37] conducted a systematic review of human and animal cadaveric biomechanical studies, and they concluded that the flexion or extension range of motion (ROM) is generally reduced by applying sufficient pretension to the suture loops and tethers. These reduced flexion/extension ROMs are usually accompanied by decreased intradiscal pressure, indicating the reduced transfer of stress to the vertebral bodies. This may reduce the incidence of both PJK and PJF. Regarding the application of sublaminar tension bands with tethering, Cho et al. [38] observed that the flexion ROM decreased significantly by 42% and 57% when tension values of 250 and 350 N were applied using a mechanical jig and force sensor, respectively. In their study, the manually tightened suture loop technique did not significantly reduce the flexion ROM, and the variability was significantly greater than that achieved by the mechanical tethering technique. This may be attributed to intra- and intersurgeon variability when tightening the suture loops. In addition, the pretension values are unlikely to be manually achieved to the same extent as those achieved by a mechanical device (up to 250 N). The optimal tension for spinous process tethers is theoretically smaller than that for sublaminar tethers due to the greater strength of the lamina. Buell et al. [39] conducted a finite element analysis and determined that the optimal tension for spinous process tethering was 100 N. Mar et al. [40] executed a human cadaveric study and demonstrated that flexion ROM was linearly related to the extent of pretension applied. The tension force ranged from 0 to 88 N, which was applied through mechanical tethering. Applying pretensioning to 88 N in one specimen led to failure of the cortical bone in the spinous process. This result supports the hypothesis that lower strength is recommended when the spinous process approach is applied. These results demonstrate that mechanical tethering may be more reliable than manually tightening the suture loop; moreover, the optimal tension required in the interspinous/sublaminar techniques warrants further investigation.
Bess et al. [41] conducted a finite element analysis and demonstrated that creating posterior tethers resulted in a gradual transition of forces at the UIV; the forces were further dissipated as the levels of tethering increased. Buell et al. [39] also conducted a finite element analysis and concluded that UIV+2 with loop or weave tether configurations is the most effective method for reducing stress on the adjacent segment. However, increasing the levels of tethering leads to more aggressive soft tissue disruption, which may neutralize the protective effects of tethering [20]. Most of the included studies extended the tethering to UIV+1, which resulted in favorable outcomes in terms of decreasing the rates of both PJK and PJF.
Several limitations of this study merit discussion. First, the included studies were predominantly retrospective in nature, which may have introduced selection and confounding biases despite our efforts to account for study quality. Second, variations in surgical techniques may contribute to the heterogeneity of the studies and influence the generalizability of our findings. Third, the definition of PJK or PJF varied slightly among the studies, and this may have introduced heterogeneity in our outcome assessments, influenced our data analyses, and affected the clinical relevance and applicability of the findings. This observation highlights the lack of a universally accepted definition for ASD. Fourth, publication bias was identified in the PJF analysis based on Egger test. After applying the trim-and-fill method, the pooled effect estimate became nonsignificant, suggesting that the original effect may have been overestimated due to bias. This implies that the current body of literature may be skewed toward reporting favorable outcomes for PJF. For PJK, publication bias could not be formally assessed due to the small number of included studies. However, visual inspection of the funnel plot suggested potential asymmetry, and thus the presence of bias cannot be excluded. These limitations should be considered when interpreting the findings. Fifth, the relatively limited sample size may restrict the generalizability of our findings. Sixth, patient characteristics and follow-up durations varied and should be considered when assessing the validity of the study results. Finally, the specific tension required to cause neurologic injury or retrolisthesis remains unclear and warrants further investigation.
Prospective randomized controlled trials with standardized protocols should be conducted to elucidate the optimal techniques and long-term outcomes of PJT. Additionally, investigating the cost-effectiveness of this intervention compared with traditional approaches would provide valuable insights for shared decision-making in clinical settings.
CONCLUSION
PJT significantly reduces the odds of development of PJK (OR, 0.44; 95% CI, 0.27–0.71) and PJF (OR, 0.36; 95% CI, 0.19–0.69) in adults undergoing spinal fusion surgery. Subgroup analyses confirmed that these protective effects were consistent across geographical locations, tether types, and surgical techniques. Although optimal tensioning parameters remain undefined, our current evidence supports PJT as an effective intervention to mitigate proximal junctional complications and enhance postoperative outcomes.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Full text via PMC
Download Citation
Print
