Secondary Logo

Journal Logo


Should Thoracolumbar Junction Be Always Avoided as Upper Instrumented Vertebra in Long Instrumented Fusion for Adult Spinal Deformity?

Risk Factor Analysis for Proximal Junctional Failure

Park, Se-Jun MD; Lee, Chong-Suh MD, PhD; Park, Jin-Sung MD; Lee, Kyung-Jun MD

Author Information
doi: 10.1097/BRS.0000000000003364
  • Free
  • Blog

Proximal junctional failure (PJF) following long instrumented spinal fusion for adult spinal deformity (ASD) is a well-recognized complication that negatively affects the clinical outcomes sometimes warranting revision surgery.1–3 Therefore, identifying the risk factors is important to prevent PJF after the surgery for ASD. Based on a literature review, several risk factors have been identified including older age, low bone density, presence of comorbidity, high body mass index, great preoperative sagittal imbalance, large correction of sagittal deformity, anterior-posterior spinal fusion, and uppermost instrumented vertebrae (UIV) at the thoracolumbar junction (TLJ).4–10

Proper selection of UIV is important in planning the long fusion surgery for ASD. The UIV level should be set with neutral and stable vertebrae to maintain overall balance and thereby reduce the need for revision.11 Generally, UIV at TLJ levels, especially between T11 and L1, is associated with the development of PJF.4,7,9,10,12–15 This is likely due to the transition from the less mobile thoracic spine to the highly mobile lumbar spine. Therefore, a few surgeons advocate extension UIV to upper thoracic level or at least above T10 to avoid PJF.4,14–16 However, it is also well known that extending the fusion more proximally may be associated with higher perioperative complications, pseudoarthrosis, and revision surgery.16

The primary goal of corrective surgery in ASD is to balance global spinal alignment with minimal fusion levels without increasing the risk of PJF development. AS the major deformity components are limited to the lumbar spine in most of degenerative cases, surgeons occasionally wonder whether or not to extend the fusion beyond TLJ levels to prevent future PJF, even though the disc space at the above levels appears non-diseased. In this study, the authors hypothesized that a certain group of patients may not experience PJF in the setting of the UIV at TLJ. Thus, in this study, the authors aimed to investigate the risk factors for PJF after long instrumented fusion stopping at TLJ in ASD and to determine which cases are suitable for TLJ stop without increasing the risk for PJF.


Study Design

This study is a retrospective case series, which were retrieved from ASD database in our institution. Of the 290 patients who underwent reconstructive surgery for adult spinal deformity between 2005 and 2016, cases with UIV level at TLJ were selected. The primary deformity was sagittal imbalance caused by degenerative flack back, and correcting this imbalance was the main objective of operation. For some patients with degenerative lumbar scoliosis, neutral vertebra in coronal plane was selected as UIV.17 To minimize the confounders, the distal fusion level was set to sacrum. The inclusion criteria were patients older than 50 years with ASD, and UIV at TLJ (T11, T12, or L1), with sacral fusion of the lowest instrumented vertebra (with or without iliac fixation), and a minimum follow-up duration of 24 months. Patients with ankylosing spondylitis or postoperative surgical infection were excluded from the study. All patients underwent all-pedicle-screw instrumentation and fusion without the use of a laminar or transverse process hook at the UIV.

The materials of the pedicle screws and rods were titanium alloy in all cases. The iliac fixation was determined according to the fusion level and degree of correction. However, since 2011 the iliac fixation has been performed routinely in all cases with greater than a four-level fusion to the sacrum to prevent nonunion at the L5-S1 levels.18 The patients were followed up routinely at weeks 2 and 6, and months 3 and 6 postoperatively, and thereafter every 6 months.

Radiographic Measurements

The following parameters were measured on the digitalized radiographs of the whole spine, preoperatively and at 2 weeks postoperatively (Figure 1): pelvic incidence (PI), sacral slope (SS), pelvic tilt (PT), lumbar lordosis (LL, L1-S1), upper LL (ULL, L12-horizontal line; geometrically same as LL minus SS), proximal junctional angle (PJA), thoracic kyphosis (TK, T5-T12), and sagittal vertical axis (SVA). Spinopelvic parameters and global spinal alignment, such as PI, SS, PT, LL, TK, and SVA, were measured using currently methods.19 PJA was measured from the caudal endplate of UIV to the cranial endplate of the two supra-adjacent vertebrae.20 Positive values of PJA and TK indicated kyphotic curvature, whereas positive values of LL and ULL mean lordotic curvature. The sagittal balance was defined as negative when the C7 plumb line fell behind the reference point.

Figure 1
Figure 1:
Plain radiographs of 73-year-old female who underwent corrective surgery for degenerative flat back. Preoperative proximal junctional angle was 3° (kyphotic). Lumbar lordosis was corrected from −20° to 33° after long fusion surgery with uppermost instrumented vertebra at T12. Two years after surgery proximal junctional failure was diagnosed with PJA of 34°. LL indicates lumbar lordosis; PI, pelvic incidence; PJA, proximal junctional angle.

Definitions of PJF

PJF was defined as postoperative PJA >20° at any time during follow-up (Figure 1), vertebral fracture at the UIV or UIV+1, failure of UIV fixation, myelopathy, or need for proximal extension of fusion. The cutoff value of PJA for diagnosis of proximal junctional kyphosis (PJK) or PJF has been reported as 10°, 15°, or 20°.4,7,20 In this study, we adopted 20° as the diagnostic criterion for PJF because PJK >20° has been closely associated with negative clinical results.4 Based on the current definition of PJF, all participants were divided into PJF group and non-PJF group.

Presumed Risk Factors

The presumed risk factors were evaluated under three categories which included the patient, surgical, and radiographic factors. Patient factors included age, sex, presence or absence of osteoporosis, body mass index (BMI, kg/m2), smoking, and the American Society Anesthesiologists physical status classification (hereinafter, ASA grade). The surgical factors included UIV level (T11, T12, or L1), previous spine surgery, surgical approach (posterior-alone vs. combined anterior-posterior approach), pedicle subtraction osteotomy (PSO), and iliac fixation. Radiographic factors included the preoperative values of PI, SS, PT, LL, ULL, PJA, TK, and SVA. All the parameters underlying the postoperative values and changes thereof were evaluated except for PJA because the authors assumed that the postoperative value of PJA was not altered directly by surgical procedures but responded to the shape of reconstruction. The value of PI-LL was calculated to determine the completeness of optimal correction and a value within ± 10° was considered for an optimal correction.21

Patient Outcomes

The scoliosis research society-22 (SRS-22) questionnaire was used for the preoperative evaluation of functional outcomes and at the last follow-up. For comparison between non-PJF and PJF groups, scores averaged with a total score of 5 for each domain. Higher score on the SRS indicates a better result and a lower score implies a worse result. The Oswestry Disability Index (ODI) was also used to assess the percentage of disability.

Statistical Analysis

Data were presented as frequencies with percentages for categorical variables and means with standard deviation for continuous variables. Fisher exact tests were used to compare categorical variables and Student t tests were used to assess the differences in means between non-PJF and PJF groups. The variables of the P values <0.05 on the univariate analyses were submitted to the multivariate analyses. Multivariate analyses were performed using logistic regression. For the risk factors from multivariate analysis, cutoff values of each parameter for PJF development were calculated using receiver-operating characteristic (ROC) curve as the point where the sensitivity and specificity were equal. Statistical analyses were performed using SPSS software (version 25.0.0; SPSS Inc., Chicago, IL). P values <0.05 were considered statistically significant.


A total of 63 patients met the inclusion criteria and comprised the study cohort. Mean age was 67.2 ± 6.3 years and there were 57 females. Follow-up duration was 51.7 ± 33.1 months. UIV level was at T11 in six, T12 in 12, and L1 in 45 patients.

PJF developed in 23 of 63 patients (36.5%) at a mean time point of 9.3 ± 14.1 months (range, 1.2–55.0 months) after surgery. Thus, non-PJF group comprised 40 patients and PJF group 23 patients. The PJF mode was cases with PJA >20° in nine patients, fracture at the UIV or UIV+1 in 13 patients, and a screw pullout in one patient. Six of 23 patients (26.1%) with PJF underwent fusion extension surgery up to T3 in two patients and T4 in four patients. Three patients who were recommended fusion extension surgery refused revision surgery.

Risk Factor Analysis

Univariate analysis of the patient's and surgical factors (Table 1) showed that older age (P = 0.038) and the presence of osteoporosis (P = 0.046) were identified as significant risk factors for PJF. No difference was found between non-PJF and PJF groups in terms of sex, BMI, smoking status, ASA grade, UIV level, previous surgery, surgical approach, PSO, or iliac fixation. Univariate analysis of radiographic factors revealed that greater preoperative PT (P = 0.026), preoperative PI-LL (P = 0.038), and preoperative PJA (P = 0.028) were significant risk factors for PJF (Table 2). The following tended to cause PJF development without reaching significance: change in SS (P = 0.098), change in PT (P = 0.061), and change in LL (P = 0.071). Optimal correction (PI-LL within ±10°) did not significantly affect the occurrence of PJF (P = 0.432). Multivariate analyses revealed that older age (odds ratio [OR] = 1.119, 95% confidence interval [CI]: 1.002–1.250, P = 0.046), osteoporosis (OR = 4.459, 95% CI: 1.128–17.616, P = 0.033), and greater preoperative PJA(°) (OR = 1.138, 95% CI: 1.033–1.253, P = 0.009) were significant risk factors for PJF development (Table 3). The cutoff value of age and PJA for PJF occurrence was calculated as 69.5 years (P = 0.017) and 0.3° (P = 0.012), respectively, under the ROC curve.

Univariate Analysis of Patient and Surgical Risk Factors
Univariate Analysis of Radiographic Risk Factors
Multivariate Analysis of Risk Factors

Accordingly, risk factors were summarized as age ≥70 years, presence of osteoporosis, and PJA ≥0°. In this study, no PJF developed in 14 patients lacking any risk factors (Figure 2, Table 4). However, presence of a single risk factor increased the probability of PJF development up to 55.6%. If patients had two risk factors, the probability increased up to 73.3%. Four of four patients (100%) carrying all the three risk factors experienced PJF.

Figure 2
Figure 2:
Plain radiographs of 66-year-old female who underwent corrective surgery for degenerative lumbar scoliosis. Preoperative lumbar T-score on bone mineral density was −0.9. Preoperative lumbar lordosis was 42° and proximal junctional angle was −4° (lordotic). Long fusion surgery was performed with uppermost instrumented vertebra at T11. Three years after surgery, proximal junctional failure did not occur. LL indicates lumbar lordosis; PI, pelvic incidence; PJA, proximal junctional angle.
PJF Development According to the Number of Risk Factors

Patient Outcomes

Forty-eight (76.2%) of 63 patients completed the questionnaire for the evaluation of clinical outcomes (Table 5). No significant differences existed between non-PJF and PJF groups in preoperative ODI scores and all domains of SRS-22 questionnaire. At the last follow-up, the ODI score was significantly higher in the PJF group than in the non-PJF group (P = 0.012). Scores of all SRS-22 domains were significantly lower in the PJF group than in the non-PJF group at the last follow-up (P = 0.002 for function/activity, P < 0.001 for pain, P = 0.001 for self-image/ appearance, P = 0.022 for mental health, and P = 0.007 for satisfaction, respectively).

Comparison of Clinical Outcome Between Non-PJF and PJF Groups


Proper selection of UIV in ASD surgery is crucial to completely restore the sagittal and coronal imbalance without causing the proximal junctional problems. Classical recommendations include UIV at neutral and stable vertebrae with healthy adjacent segments without degeneration or instability in any planes.2,15,22 However, making a decision on the UIV level is complicated by the deformity type, patient condition, and surgical goal.

The UIV at TLJ has attracted special attention because it has distinct anatomic characteristics to serve as the transition from the immobile thoracic spine to the mobile lumbar spine leading to potential biomechanical disadvantages.23 Due to this anatomic weakness, it is generally recognized that stopping at TLJ levels increases the likelihood of proximal junctional problems such as PJK or PJF.4,7,9,10,12–15 However, for a long time, the UIV level at TLJ in ASD surgery has been disputed. Suk et al15 has suggested that fusion should not stop at TLJ and instead be extended to T10 or above to prolong survival of the adjacent levels. Similarly, recent studies also demonstrated that the UIV at T8 or below levels increased the risk of PJK or PJF.4,12,14,15 However, contrasting studies reported the risk of PJK or PJF did not increase even with UIV at TLJ levels compared with UIV at above-TLJ levels.11,17,24,25 Kim et al11 reported that there was no significant radiographic or clinical outcome whether the UIV stopped at T9, T11, or L1. The Cho et al's study17 concluded that fusion to T11 or T12 was acceptable when UIV was above the upper end vertebra in adult degenerative scoliosis.

Fusion routinely extending above T10 or up to upper thoracic level merely to prevent potential PJK or PJF in ASD surgery should be reconsidered because it is known that increased number of fusions prolongs operative time, increases intraoperative blood loss, leading to greater perioperative morbidity and mortality, especially in elderly patients.16,26,27 Stopping at TLJ level in long spinal fusion is still associated with a risk of PJF, and therefore, it is necessary to identify the risk factors for PJF when TLJ is considered as the UIV level.

In the current study, we identified that age >70 years, osteoporosis, and preoperative PJA >0° (positive value of PJA means kyphosis) were significantly associated with PJF occurrence when the UIV stopped at TLJ. In the univariate analysis, the parameters indicating greater preoperative sagittal imbalance, such as preoperative PI-LL mismatch and preoperative PT, significantly increased the risk of PJF. Parameters representing the amount of correction, such as a change in SS, PT, and LL, also had a tendency to be associated with PJF, although these values did not reach statistical significance. These findings are consistent with previous studies reporting that greater correction of deformity as well as higher preoperative sagittal imbalance may contribute to the development of PJK or PJF.1,3,28,29 However, in multivariate analysis, no statistical significance was observed for radiographic parameters except for preoperative PJA.

In this study, the authors also found no correlation between optimal correction and the development of PJF in univariate analyses. It means PJF could not be prevented even if the deformity is corrected to an optimal range. Similar results were found in our previous study.10 Multifactorial analyses for risk factors for PJF demonstrated that the optimal correction of deformity did not prevent the development of PJF in ASD surgery.10

In this study, we found that no PJF developed the absence of any risk factors. Therefore, we suggest TLJ level as UIV can be considered selectively in patients younger than 70 years without osteoporosis, and with preoperative PJA <0° (lordotic curvature of proximal junctional level). However, notably, we also observed the odds of PJF development exceeded 50% even with one risk factor. Therefore, extending the fusion above TLJ level should be considered to prevent PJF in patients even with one risk factor.

The definition of PJF may be disputed because PJK ≥20° was included in this study. In early literatures, PJK was not related to significant clinical symptoms or structural failure.20,30,31 However, recent studies demonstrated PJK might be associated with negative clinical outcomes if the PJA exceed 20° in PJK.4,16 The current study demonstrated worse clinical outcome measures (ODI and SRS-22) at last follow-up in the PJF group compared with the non-PJF group. Although the clinical outcome assessment was not the primary endpoint, our results in the current study were clinically significant, suggesting PJF should be prevented because it represents not only radiographic failure but poor clinical results.

Several limitations in this study need be acknowledged. First, this study is retrospective in nature; therefore, the patients are subject to inconsistencies in surgical techniques and surgeons’ experience especially in selecting the UIV level. Second, our definitions of PJF might not reflect all the junctional complications, such as myelopathy and spondylolisthesis. However, in our experiences, pure myelopathy or spondylolisthesis without PJF were rare and the most commonly encountered junctional problems were kyphosis or fractures. Third, the relatively small size of the study cohort was another limitation potentially preventing the generalization of the results. However, this study was performed with narrow inclusion criteria including UIV at TLJ, fusion to sacrum, age greater than 50 years, and longer than 2-year follow-up. The relatively high statistical significance and consistency of the results might overcome this weakness.


Age >70 years, osteoporosis, and PJA >0° were identified as significant risk factors for PJF. Therefore, the TLJ level can be considered as UIV selectively for patients younger than 70 years without osteoporosis and with lordotic preoperative PJA.

Key Points

  • Age >70 years, osteoporosis, and PJA >0° were identified as significant risk factors for PJF in UIV at TLJ.
  • No PJF occurred in 14 patients lacking any risk factors.
  • The TLJ level can be considered as UIV selectively for patients younger than 70 years without osteoporosis and with lordotic preoperative PJA.
  • The odds of PJF development exceeded 50% even with one risk factor.
  • Therefore, extending the fusion above TLJ level should be considered to prevent PJF in patients even with one risk factor.


1. Kim HJ, Bridwell KH, Lenke LG, et al. Proximal junctional kyphosis results in inferior SRS pain subscores in adult deformity patients. Spine (Phila Pa 1976) 2013; 38:896–901.
2. Kim YJ, Bridwell KH, Lenke LG, et al. Proximal junctional kyphosis in adult spinal deformity after segmental posterior spinal instrumentation and fusion: minimum five-year follow-up. Spine (Phila Pa 1976) 2008; 33:2179–2184.
3. Yagi M, King AB, Boachie-Adjei O. Incidence, risk factors, and natural course of proximal junctional kyphosis: surgical outcomes review of adult idiopathic scoliosis. Minimum 5 years of follow-up. Spine (Phila Pa 1976) 2012; 37:1479–1489.
4. Bridwell KH, Lenke LG, Cho SK, et al. Proximal junctional kyphosis in primary adult deformity surgery: evaluation of 20 degrees as a critical angle. Neurosurgery 2013; 72:899–906.
5. Cho SK, Shin JI, Kim YJ. Proximal junctional kyphosis following adult spinal deformity surgery. Eur Spine J 2014; 23:2726–2736.
6. Diebo BG, Shah NV, Stroud SG, et al. Realignment surgery in adult spinal deformity: Prevalence and risk factors for proximal junctional kyphosis. Orthopade 2018; 47:301–309.
7. Hostin R, McCarthy I, O’Brien M, et al. Incidence, mode, and location of acute proximal junctional failures after surgical treatment of adult spinal deformity. Spine (Phila Pa 1976) 2013; 38:1008–1015.
8. Hyun SJ, Lee BH, Park JH, et al. Proximal junctional kyphosis and proximal junctional failure following adult spinal deformity surgery. Korean J Spine 2017; 14:126–132.
9. Kim DK, Kim JY, Kim DY, et al. Risk factors of proximal junctional kyphosis after multilevel fusion surgery: more than 2 years follow-up data. J Korean Neurosurg Soc 2017; 60:174–180.
10. Park SJ, Lee CS, Chung SS, et al. Different risk factors of proximal junctional kyphosis and proximal junctional failure following long instrumented fusion to the sacrum for adult spinal deformity: survivorship analysis of 160 patients. Neurosurgery 2017; 80:279–286.
11. Kim YJ, Bridwell KH, Lenke LG, et al. Is the T9, T11, or L1 the more reliable proximal level after adult lumbar or lumbosacral instrumented fusion to L5 or S1? Spine (Phila Pa 1976) 2007; 32:2653–2661.
12. Fujimori T, Inoue S, Le H, et al. Long fusion from sacrum to thoracic spine for adult spinal deformity with sagittal imbalance: upper versus lower thoracic spine as site of upper instrumented vertebra. Neurosurg Focus 2014; 36:E9.
13. Lau D, Clark AJ, Scheer JK, et al. Proximal junctional kyphosis and failure after spinal deformity surgery: a systematic review of the literature as a background to classification development. Spine (Phila Pa 1976) 2014; 39:2093–2102.
14. Scheer JK, Lafage V, Smith JS, et al. Maintenance of radiographic correction at 2 years following lumbar pedicle subtraction osteotomy is superior with upper thoracic compared with thoracolumbar junction upper instrumented vertebra. Eur Spine J 2015; 24: (Suppl 1): S121–S130.
15. Shufflebarger H, Suk SI, Mardjetko S. Debate: determining the upper instrumented vertebra in the management of adult degenerative scoliosis: stopping at T10 versus L1. Spine (Phila Pa 1976) 2006; 31:S185–S194.
16. O'Shaughnessy BA, Bridwell KH, Lenke LG, et al. Does a long-fusion “T3-sacrum” portend a worse outcome than a short-fusion “T10-sacrum” in primary surgery for adult scoliosis? Spine (Phila Pa 1976) 2012; 37:884–890.
17. Cho KJ, Suk SI, Park SR, et al. Selection of proximal fusion level for adult degenerative lumbar scoliosis. Eur Spine J 2013; 22:394–401.
18. Lee CS, Chung SS, Choi SW, et al. Critical length of fusion requiring additional fixation to prevent nonunion of the lumbosacral junction. Spine (Phila Pa 1976) 2010; 35:E206–E211.
19. Vaz G, Roussouly P, Berthonnaud E, et al. Sagittal morphology and equilibrium of pelvis and spine. Eur Spine J 2002; 11:80–87.
20. Glattes RC, Bridwell KH, Lenke LG, et al. Proximal junctional kyphosis in adult spinal deformity following long instrumented posterior spinal fusion: incidence, outcomes, and risk factor analysis. Spine (Phila Pa 1976) 2005; 30:1643–1649.
21. Schwab F, Patel A, Ungar B, et al. Adult spinal deformity-postoperative standing imbalance: how much can you tolerate? An overview of key parameters in assessing alignment and planning corrective surgery. Spine (Phila Pa 1976) 2010; 35:2224–2231.
22. Lee J, Park YS. Proximal Junctional Kyphosis: Diagnosis, Pathogenesis, and Treatment. Asian Spine J 2016; 10:593–600.
23. Chou WY, Hsu CJ, Chang WN, et al. Adjacent segment degeneration after lumbar spinal posterolateral fusion with instrumentation in elderly patients. Arch Orthop Trauma Surg 2002; 122:39–43.
24. Kim HJ, Boachie-Adjei O, Shaffrey CI, et al. Upper thoracic versus lower thoracic upper instrumented vertebrae endpoints have similar outcomes and complications in adult scoliosis. Spine (Phila Pa 1976) 2014; 39:E795–E799.
25. Ha Y, Maruo K, Racine L, et al. Proximal junctional kyphosis and clinical outcomes in adult spinal deformity surgery with fusion from the thoracic spine to the sacrum: a comparison of proximal and distal upper instrumented vertebrae. J Neurosurg Spine 2013; 19:360–369.
26. Fu X, Sun XL, Harris JA, et al. Long fusion correction of degenerative adult spinal deformity and the selection of the upper or lower thoracic region as the site of proximal instrumentation: a systematic review and meta-analysis. BMJ Open 2016; 6:e012103.
27. Silva FE, Lenke LG. Adult degenerative scoliosis: evaluation and management. Neurosurg Focus 2010; 28:E1.
28. Smith MW, Annis P, Lawrence BD, et al. Acute proximal junctional failure in patients with preoperative sagittal imbalance. Spine J 2015; 15:2142–2148.
29. Maruo K, Ha Y, Inoue S, et al. Predictive factors for proximal junctional kyphosis in long fusions to the sacrum in adult spinal deformity. Spine (Phila Pa 1976) 2013; 38:E1469–E1476.
30. Kim YJ, Bridwell KH, Lenke LG, et al. Proximal junctional kyphosis in adolescent idiopathic scoliosis following segmental posterior spinal instrumentation and fusion: minimum 5-year follow-up. Spine (Phila Pa 1976) 2005; 30:2045–2050.
31. Yagi M, Akilah KB, Boachie-Adjei O. Incidence, risk factors and classification of proximal junctional kyphosis: surgical outcomes review of adult idiopathic scoliosis. Spine (Phila Pa 1976) 2011; 36:E60–E68.

adult spinal deformity; long instrumented fusion; proximal junctional failure; thoracolumbar junction; uppermost instrumented vertebra

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.