Proximal junction uninstrumented bone fusion for proximal junctional kyphosis/failure prevention: a retrospective matched cohort study in Korea
Article information
Abstract
Study Design
Retrospective matched cohort study.
Purpose
This study aimed to determine the efficacy of proximal junction uninstrumented bone fusion (UBF) to prevent proximal junctional kyphosis/failure (PJK/F) in adult spinal deformity (ASD) surgeries.
Overview of Literature
PJK/F represents significant complications in ASD surgeries. Despite many preventive strategies being developed, which have been largely categorized as additive surgical procedures and optimal alignment correction, PJK/F still remains a prevalent complication.
Methods
This study retrospectively analyzed patients who underwent ASD corrective surgeries incorporating UBF at the proximal junction at a single academic institution. The patients with ASD were individually matched (1:2) for age, gender, osteoporosis status, and uppermost instrumented vertebrae. The demographic, operative, and radiographic parameters were compared, while the complication-free periods for PJK/F and any mechanical failure (MF) were separately assessed. A multivariable survival analysis identified the significant factors for mechanical complications.
Results
A total of 21 patients in the UBF group and 42 patients in the non-UBF group were enrolled. No significant differences in the demographic, operative, and radiographic parameters and the patient-reported outcome scores were observed. The median time to PJK/F was longer in the UBF group (11.0 months vs. 3.5 months, p=0.037). In the UBF group, six patients had successful bone mass fusion without any MF. The odds ratio of UBF was 0.55 (p=0.28) for PJK/F and 0.35 (p=0.056) for any MF. The results of the Kaplan-Meier analysis indicated a significantly extended MF-free period in the UBF group (p=0.028), but not for PJK/F (p=0.16).
Conclusions
Although proximal junction UBF in ASD surgeries did not show a statistically significant PJK/F reduction, its safety, technical simplicity, and the absence of complications in patients with successful bone mass fusion suggest its potential clinical value and warrant further research.
Introduction
Despite advancements in the field of adult spinal deformity (ASD), mechanical failure (MF) remains a significant challenge. Notably, proximal junctional kyphosis (PJK) and proximal junctional failure (PJF) are the most prevalent complications. While the incidence of PJK is broadly agreed upon to range between 20% and 40%, PJF varies significantly from 1.4% to 35%, partly due to the non-unified definitions [1–3].
The risk factors for PJK/F have been extensively studied along with surgical techniques and spinal alignment guidance for prevention. Preventive surgical techniques, either additive or avoidance strategies, have been proposed. They include careful posterior soft tissue dissection, the use of hooks or tethering, prophylactic vertebroplasty at the proximal junction, and avoidance of stiff rods and bicortical screw fixation at the uppermost instrumented vertebra (UIV) [4–7]. Ideal spinal alignment continues to be a subject of debate and advancement, with numerous concepts being investigated and validated [8–10]. However, despite these worldwide endeavors, the incidence of PJK/F remains consistent [11].
Based on the historical insights of uninstrumented bone fusion (UBF), originally described by Hibbs [12] in 1911 and widely used to achieve spinal fusion before the advent of modern instrumentation [13–15], we investigated its potential to make a gradual transition of stiffness for long-construct spinal fusion without any instrumentation at UIV+1. Since 2017, we have attempted UBF at UIV+1 by lamina decortication and cranially extending onlay bone grafts for ASD surgeries, hypothesizing that this proximal junction UBF can be a preventive measure for junctional problems. This study aimed to determine the efficacy of proximal junction UBF to prevent PJK and PJF in ASD surgeries. To our knowledge, this is the first study to present clinical data regarding proximal junction UBF for ASD corrective surgery.
Materials and Methods
Ethics statement
This study was conducted in compliance with the principles of the Declaration of Helsinki. The study protocol was reviewed and approved by the Institutional Review Board (IRB) of Seoul National University Bundang Hospital (IRB no., B-2307-842-102). The requirement for informed consent from individual patients was omitted because of the retrospective design of this study.
Study design
This is a retrospective matched cohort study. Initially, by reviewing our institution’s electronic medical records, we classified the patients (UBF group) who (1) underwent ASD corrective surgery with proximal junction UBF by a single surgeon who specialized in spinal deformities between 2017 and 2021, (2) were followed up for at least 2 years or had radiographic evidence of early MF, and (3) had preoperative and postoperative whole-spine standing radiographs. MF incorporated PJK, PJF, distal junctional failure, and implant loosening or fracture. These were categorized as either PJK/F or other MFs. PJK is defined as the proximal junctional sagittal Cobb angle between the lower endplate of UIV and the upper endplate of UIV+2, at least 10° greater than the preoperative measurement. On the other hand, PJF is defined as a vertebral fracture or subluxation at UIV or UIV+1, failure of proximal fixation, or neurological deficit [16]. Furthermore, the patients who had (1) congenital or neuromuscular spinal deformities, (2) adolescent idiopathic scoliosis, or (3) tumors or spondylitis as the main etiology of ASD were excluded.
Surgical technique
For the transition technique at the proximal junction to prevent PJK/F, our institution has occasionally used proximal junction UBF since 2017. Specifically, this technique involves decorticating the lamina at UIV+1 unilaterally or bilaterally, followed by the extension of the onlay bone graft onto the decorticated side. The bone graft comprised autobone mixed with allobone chips and biological materials to improve bone fusion. The biological materials included demineralized bone matrix and bone morphogenic protein. The midline structures at the proximal junction, including the facet joints and ligaments, were always preserved. All cases with proximal junction UBF had a pedicle screw-only construct at UIV, which was identical to the non-UBF group. Moreover, thoracic-lumbar-sacral orthosis was applied and maintained for 3 months to all patients with ASD who underwent corrective surgeries.
Matching
Twenty-one patients were selected for UBF. The patients were matched 1:2 based on age, gender, osteoporosis, and UIV level, all of which are widely accepted risk factors for PJK/F. Age was categorized into three groups (55–64, 65–74, and 75–84 years) and individually matched to identify patients within the same age group. Osteoporosis was diagnosed through conventional dual-energy X-ray absorptiometry, wherein a T-score of −2.5 or lower was indicative of osteoporosis. The level of UIV was categorized into four groups (i.e., upper thoracic [T1–T4], mid thoracic [T5–T8], lower thoracic [T9–T12], and lumbar [L1–L2]) and individually matched in the same manner. With consistent exclusion criteria, twice the number of matched patients were randomly selected from the patient pool, yielding a total of 42 patients.
Data collection and analysis
Data on the patients’ demographics, baseline medical conditions, follow-up duration (months), and operative characteristics, preoperative and postoperative Scoliosis Research Society (SRS)-22 scores, and spinopelvic/spinal balance parameters were obtained. As all patients underwent computed tomography (CT) spine imaging at least 1 year postoperatively, successful UBF bone mass fusion to the UIV+1 lamina was assessed using axial, coronal, and sagittal reconstructions. The evaluation was performed by a co-author (K.J.K.), who was initially blinded to the study hypothesis. The hip axis (HA) was defined as the midpoint of the center of the bilateral femoral heads from sagittal radiographs. The vertical line drawn from the center of the acoustic meatus to the ground was defined as the gravity line (GL). The offset between GL and HA, or GL-HA offset, was defined as the shortest sagittal plane distance between GL and HA, with positive values indicating an anterior positioning of the GL relative to the HA [17].
The statistical analyses were performed using R ver. 4.2.2 (open-source software; The R Foundation for Statistical Computing, Vienna, Austria; https://www.r-project.org/). The between-group differences were analyzed using the Wilcoxon rank-sum test for the quantitative data and Fisher’s exact test for the qualitative binary data. Statistical significance was set at p-values of 0.05. A sensitivity analysis was performed for PJK/F as the primary outcome. The continuous variables were presented as the median with interquartile range (25%–75%), while the categorical variables were presented as counts with percentages. The odds ratio (OR) and 95% confidence interval (CI) for PJK/F or MF were measured by defining the proximal junction UBF as a protective factor. The survival time was measured from the date of ASD corrective surgery to the date of PJK/F or MF, if any, or to the end of the last follow-up. The survival curves were estimated and plotted using the Kaplan–Meier method, and the two groups were compared using the log-rank test.
Results
Demographic data, operative characteristics, and patient-reported outcomes
Table 1 shows the baseline demographic data, operative characteristics, and SRS-22 scores for both groups. Beyond the individually matched factors, no significant differences in terms of overweight status (body mass index of ≥25 kg/m2 and <30 kg/m2), obesity (body mass index of ≥30 kg/m2), smoking status, and duration of follow-up were observed between the two groups. All patients, except one who had fusion from T12 to L4 without proximal junction UBF, underwent fusion to the sacrum/pelvis. No statistically significant differences in the proportions of patients who underwent three-column osteotomy, cement augmentation at UIV and UIV+1, proximal junction tethering, the use of cobalt–chrome rod, and the revisional status of the ASD surgery were observed between the groups. Transverse process hooks at UIV were not used. The preoperative and postoperative SRS-22 scores did not statistically differ between the groups.
Patient demographics, operative characteristics, and SRS-22 score
Radiographic parameters
Table 2 shows the improvements in the radiographic parameters through deformity correction. No significant differences in pelvic incidence (PI), preoperative and postoperative lumbar lordosis (LL), pelvic tilt (PT), PI–LL mismatch, and sagittal balance parameters such as C7–sagittal vertical axis (SVA) and GL-HA offset were observed between the groups. The postoperative proximal junctional angle also showed no significant difference.
Radiographic parameters
Mechanical complication analysis
Table 3 shows the incidence and time to postoperative MFs. In the UBF group, nine out of 21 patients (42.8%) had any MFs until the end of the follow-up, while the majority (eight out of 21; 38.0%) had PJK/F. In the non-UBF group, 27 out of 42 patients (64.2%) had MFs, with the majority (20 out of 42; 47.6%) being PJK/F. In the non-UBF group, three patients had PJK/F with rod fracture, distal junctional failure, or both. No statistically significant difference in the incidence of any type of MFs was observed between the groups. However, the non-UBF group had a significantly shorter time to PJK/F, although no difference was observed for time to any MFs. With proximal junction UBF as a protective factor, the measured OR for PJK/F was 0.55 (95% CI, 0.17–1.65; p=0.280) and 0.35 (95% CI, 0.11–1.02; p=0.056) for any MFs. Although the PJK/F-free survival did not show a significant difference (p=0.16), the difference in any MF-free survival was statistically significant (p=0.028) (Figs. 1, 2). Two-year PJK/F-free probability was 54.1% for the UBF group and 70.3% for the non-UBF group. Moreover, the 2-year MF-free probability was 41.0% for the UBF group and 70.3% for the non-UBF group. Age, the only continuous demographic variable in our analysis, was separately assessed by logistic regression for both PJK/F and MF. Age was not a significant predictor of either outcome (p=0.486 and p=0.245, respectively).
Mechanical failures
Proximal junctional kyphosis (PJK) and proximal junctional failure (PJF)-free survival function estimated by the Kaplan-Meier method for two groups of patients, one group that received proximal junction uninstrumented bone fusion (UBF) along with deformity correction surgery and another group that did not. The 95% confidence interval bands are also included. Censoring is represented by vertical marks. The graph at the bottom displays the number of patients at risk at various time points.
Any mechanical failure (MF)-free survival function estimated by the Kaplan-Meier method for two groups of patients, one group that received proximal junction uninstrumented bone fusion (UBF) along with deformity correction surgery and another group that did not. The 95% confidence interval bands are also included. Censoring is represented by vertical marks. The graph at the bottom displays the number of patients at risk at various time points.
Sensitivity analysis
As this study used the maximum available sample size based on the predefined inclusion and exclusion criteria (21 patients in the UBF group and 42 in the non-UBF group), a sensitivity analysis was conducted using Fisher’s exact test (two-sided α=0.05, 80% power) for the PJK/F incidence. Assuming a PJK/F incidence of 40% in the control group (non-UBF)—consistent with previous reports [1–3]—the study was powered to detect risk differences of ≤−32% or ≥38%, which exceed the observed PJK/F risk difference in our cohort (9.6%).
Discussion
Since the history of spinal deformity surgery began, PJK/F has imposed a significant burden on both patients and surgeons. Current understanding suggests that specific surgical decisions and techniques can partially help to improve these complications. The representative examples include the selection of the fusion level and alignment correction goal [1–3,18,19]. With regard to UIV selection, while the thoracolumbar junction is generally considered a surgical risk factor [20], it often cannot be avoided. Furthermore, fusion down to the sacrum is a well-known PJK/F risk factor, while stopping at L5 or above raises another concern for distal failure [21]. Ideal alignment goals continue to be an area of active research. Certain baseline risk factors, such as preoperative deformity severity, are non-modifiable. Moreover, numerous studies have highlighted that the greater correction of LL, PT, or C7–SVA is associated with PJK/F. However, these are mostly unmodifiable because adjusting the alignment goal to limit correction contradicts the principle of achieving ideal alignment [22]. Thus, most known PJK/F risk factors are difficult to avoid. This highlights the importance of incorporating additional surgical procedures to best prevent PJK/F in given clinical situations. Prophylactic vertebroplasty, for instance, has been extensively studied for PJK/F prevention, albeit with mixed results [23–25]. Transverse process hooks have shown a preventive effect for PJK, albeit with ongoing controversy for PJF prevention [25]. Tethering techniques using diverse tape configurations have also been assessed in relation to controversial benefits [25–27]. Considering these mixed findings, it is important to investigate various preventive measures.
UBF, a pioneering tool in spinal fusion, has been proven to be successful for short-segment fusion with a high success rate, even in the absence of interbody fusion or screw fixation [13–15]. In the era of instrumentation, UBF alone showed success in fusing spondylolisthesis segments or adolescent scoliosis deformity segments [28,29]. One study has shown the potential of UBF by reporting unintended posterolateral fusion extension due to the incidental migration of the bone graft [30]. These findings encouraged us to consider applying UBF at the proximal junction in ASD corrective surgery.
To achieve between-group homogeneity, individual 1:2 matching was applied. Unmatched demographic and operative factors, along with various spinopelvic and sagittal balance parameters, showed no significant differences between the groups (Tables 1, 2). Although statistically insignificant, the incidence of PJK/F was lower in the UBF group (38.0%) than in the non-UBF group (47.6%). Time to PJK/F was significantly longer in the UBF group (11.0 months) compared with the non-UBF group (3.5 months) (Table 3). However, this finding should be interpreted with caution, as multiple factors could influence the timing of PJK/F onset. It should be acknowledged that, although not statistically significant, the higher use of cobalt–chrome rods in the non-UBF group and the greater use of tethering and cement augmentation in the UBF group may have biased the PJK/F outcomes toward less favorable results in the non-UBF group. Of particular note, six patients in the UBF group (28.5%) achieved complete fusion between the UIV+1 bone mass and adjacent lamina/spinous processes, while none developed PJK/F during follow-up (Figs. 3, 4). Statistically, UBF was not a significant preventive factor for PJK/F (OR, 0.55; p=0.28) or for any MF (OR, 0.35; p=0.056), although both trends favored UBF. The results of the Kaplan-Meier analysis showed significantly longer MF-free survival in the UBF group (p=0.028), whereas PJK/F-free survival did not reach statistical significance (p=0.16) (Figs. 1, 2). A closer evaluation revealed that all MFs occurring within the first postoperative year were PJK/F events, and their early divergence—reflected by nonoverlapping 95% confidence intervals—was significant. Most PJK/Fs (22 out of 25; 88.0%) developed within 1 year postoperatively, which was likely because approximately 85% of the index cases had the UIV at the thoracolumbar junction, which predisposed to acute PJK/F, and because the majority of the subjects were elderly Asian women with a high prevalence of osteoporosis, making them susceptible to fracture-type PJF. No complications specifically attributable to the proximal junction UBF were identified, supporting its safety profile. All MFs other than PJK/F—mainly rod fractures—occurred after the first postoperative year, contributing to the between-group difference in the MF-free survival (Fig. 2). It is possible that the higher frequency of the use of cobalt–chrome rods in the non-UBF group, given their greater stiffness, may have predisposed them to these fractures, although this did not reach statistical significance.
A representative case of a 73-year-old female with successful bony fusion of left unilateral proximal junction uninstrumented bone fusion (UBF) without any mechanical failure. (A–D) Computed tomography showing left unilateral proximal junction UBF. (A, B) Two images on the left depicts the status 2 days postoperatively, while (C, D) two images on the right depicts the status 2 years postoperative. (E–G) Radiographs showing chronological change of sagittal spinal balance postoperatively. (E) shows preoperative C7–sagittal vertical axis (C7–SVA) of 278.0 mm and gravity line (GL)-hip axis (HA) offset of 248.5 mm. (F) shows postoperative 3 months C7–SVA of 2.0 mm and GL-HA offset of −49.3 mm. (G) shows postoperative 4 years C7–SVA of 45.4 mm and GL-HA offset of 15.1 mm, signifying maintained stable sagittal alignment.
A representative case of a 69-year-old female with successful bony fusion of bilateral proximal junction uninstrumented bone fusion (UBF) without any mechanical failure. (A–D) Computed tomography showing bilateral proximal junction UBF. (A, B) Two images on the left depicts the status 2 days postoperatively, while (C, D) two images on the right depicts the status 3 years postoperatively. (E–G) Radiographs showing chronological change of sagittal spinal balance postoperatively. (E) shows preoperative C7–sagittal vertical axis (C7–SVA) of 159.0 mm and gravity line (GL)-hip axis (HA) offset of 63.9 mm. (F) shows postoperative 2 months C7–SVA of 8.3 mm and GL-HA offset of −11.8 mm. (G) shows postoperative 3 years C7–SVA of 26.5 mm and GL-HA offset of −23.6 mm, signifying maintained stable sagittal alignment.
The principal advantage of proximal junction UBF is that it can be performed through minimal muscle dissection without violating the facet joint using a limited diagonal exposure of the UIV+1 lamina. Like UIV+1 vertebroplasty, it can be combined with other proximal junction–protective techniques. Its main limitation lies in the absence of an immediate preventive effect, as stabilization depends on bone fusion into UIV+1. Thus, early PJK/F may develop before osseous union occurs. Overall, the clinical value of UBF may lie in its safety and feasibility as a complementary, minimally invasive additive procedure.
However, this study has several limitations that should be addressed. First, the relatively small sample size compromised the statistical power. As demonstrated by the sensitivity analysis, the study was powered to detect only large differences in PJK/F incidence (32%–38%), which is substantially greater than the observed absolute risk difference of 9.6%. Accordingly, the lack of statistical significance should be interpreted as inconclusive rather than as evidence of no effect, and larger studies are required for validation. This limited power may also partly explain the lack of significant differences in postoperative SRS-22 scores between the groups, despite a higher trend toward MF in the non-UBF group that would be expected to adversely affect the clinical outcomes. Moreover, because proximal junction UBF was performed by a single surgeon at our institution, the generalizability of the findings should be further assessed in multicenter studies involving multiple surgeons. Second, our effort to ensure baseline homogeneity through individual matching could not control for every variable. For instance, only four patients in the UBF group fell within the 55–64 age category, resulting in a cohort skewed toward older patients. Third, the demographic characteristics of this Asian cohort differ from Western populations, with significantly lower rates of obesity (9.5%) and smoking (4.7%). The predominance of elderly Asian females contributed to a high prevalence of osteoporosis (33%), which may negatively affect the fusion rates. Moreover, reduced posterior musculature and multiple medical or musculoskeletal comorbidities that limit activity and impair postural control may collectively bias the results toward higher PJK/F and MF rates. Finally, the successful fusion of the proximal junction bone graft to UIV+1 may reflect overall surgical stability rather than a distinct benefit of UBF itself, potentially confounding the interpretation of its preventive effect.
Conclusions
Proximal junction UBF showed a trend toward reducing early PJK/F, suggesting a potential protective effect at the proximal junction. Notably, no PJK/F occurred in the patients with successful osseous fusion, supporting its stabilizing role. This technique is safe, minimally invasive, and easily applicable without additional instrumentation. These findings suggest that proximal junction UBF can serve as an effective adjunctive measure in ASD surgery. Further research is needed to confirm our findings.
Key Points
The prevention of proximal junctional kyphosis/failure (PJK/F) in adult spinal deformity surgery remains challenging.
Proximal junction uninstrumented bone fusion (UBF) is a safe, minimally invasive adjunctive pro-cedure.
No cases of PJK/F occurred among patients with successful proximal junction bone fusion.
UBF can potentially prevent PJK/F in adult spinal deformity surgery.
Notes
Conflict of Interest
No potential conflict of interest relevant to this article was reported.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Author Contributions
Conceptualization: HAS, SA. Methodology: SJH. Investigation: SJH, KJK. Formal analysis: KJK, SA. Writing–original draft: SJH, KJK. Writing–review & editing: HAS, SA. Final approval of the manuscript: all authors.