ORIGINAL ARTICLE

Proximal Junctional Kyphosis above Fusions to the Thoracolumbar Spine for Adult Scoliosis: Rate and Predictive Factors

Rajiv K. Sethi *, Mital Patel *, Danielle Geula *, Vedat Deviren*, Shane Burch *, Bobby Tay *, Serena Hu *, David Bradford *, Sigurd Berven*.

*University of California at San Francisco, Department of Orthopaedic Surgery, San Francisco, CA, USA

Address for Correspondence:

Dr. Mital Patel (M.S.(ortho))
a/1, kapilakunj apptt. Karelibaugh, baroda.
Email : mital1982@gmail.com

Abstract:

Study Design: A retrospective study of adult deformity patients treated with instrumentation of the lumbar curve to the thoracolumbar spine.

Objectives: To determine the rate of radiographically apparent proximal junctional kyphosis (PJK), the rate of PJK requiring revision surgery, and factors that predict the occurrence of PJK in adults with long fusions and a cephalad level at the thoracolumbar junction.

Summary of Background Data: PJK is an important and significant sequela of long constructs from the lumbosacral spine to the thoracolumbar junction. The rate of PJK and the factors that may predict PJK after surgery is not well-defined. The role of fusion levels and radiographic parameters including lumbar lordosis, pelvic incidence, and sacral slope on the occurrence of PJK is unknown.

Methods: Retrospective study of patients with adult scoliosis treated with instrumentation and fusion from the lumbosacral spine to a cephalad level at the thoracolumbar junction (T10-L1). These patients were treated at a single institution between 1992-2006 were followed for a minimum of two years after surgery. PJK was defined as a progression of kyphosis above the instrumented fusion of >5 degrees.
Predictive variables identified include radiographic parameters of lumbar lordosis (L1- S1), pelvic incidence, sacral slope, change in lordosis and ratio of lordosis and pelvic incidence, and surgical variables including cephalad and caudad levels of instrumentation and fusion.

Results: 199 patients underwent instrumented surgery for adult degenerative scoliosis with cephalad instrumented level to T10-L1 between 1992-2006 at a single institution. Eighty-seven of the 199 patients were found to have complete pre-operative, post-operative, and most recent follow-up radiographs. The radiographic rate of PJK was 34% (30/87). 10% (9/87) of patients required revision surgery for PJK during the followup period. Patients fused to T10 had a 3 fold lower rate of PJK than those fused to T11, a 2 fold lower rate than those fused to T12, and a 1.5 fold lower rate than lose fused to L1. The effect of proximal level was not a statistically significant predictor of PJK. The patients requiring revision surgery for PJK had an average increase in lordosis by 12 degrees, and the patients not requiring PJK had an increase of 4 degrees compared with pre-operative lordosis. This difference was not statistically significant.

Conclusions: This study did not demonstrate an effect of cephalad or distal level of fusion on the rate of radiographically apparent PJK, or rate of revision surgery for PJK. There was however a trend seen with those patients fused to T10 having lower rates of PJK, and patients with a greater increase in lordosis having a higher increase in revision surgery for PJK.

Key Points:

This study did not demonstrate a significant effect of cephalad or distal level of fusion on the rate of radiographically apparent PJK, or rate of revision surgery for PJK. There was however a trend seen with those patients fused to T10 having lower rates of PJK, and patients with a greater increase in lordosis having a higher increase in revision surgery for PJK. There was no difference in pelvic incidence, sacral slope, or the ratio of lumbar lordosis to pelvic incidence between the cohorts studied

J.Orthopaedics 2011;8(4)e14

Keywords:

kyphosis: junctional: scoliosis: adult: revision

Introduction:
Selection of fusion levels in spine deformity patients is a complex decision and there is little evidence in the literature to guide informed decisions. There is significant variability in the choice of cephalad and caudad levels in patients with adult lumbar spine deformity. Junctional complications are common and significant in fusions to the lumbosacral and thoracolumbar regions of the spine. Segmental instrumentation is an effective technique for selective of thoracic and lumbar deformity, and the technique generates a significant gradient in mobility of the spine between instrumented and noninstrumented levels. Long fusions with a cephalad level at the thoracolumbar junction
may be compromised by progression of kyphosis above the fusion related to implant failure or adjacent segment kyphosis. Proximal junctional kyphosis is defined as a progression of segmental kyphosis of more than 5 degrees at the level above the instrumented fusion (9). The purpose of this paper is to define the rate of proximal junctional kyphosis in patients with fusions to the thoracolumbar spine, and to identify radiographic and surgical strategy factors that may be independent predictors of proximal junctional kyphosis.

The challenges of thoracolumbar and lumbosacral junctions are well-recognized by surgeons who care for patients with adult deformity (1-7). A goal of this study is toprovide surgeons with information that may guide informed-decision making in choosingfusion levels for the surgical management of adult spinal deformity.

With the introduction of modern segmental instrumentation and selective posterior fusions that permit selective fusions and limit the caudal and cephalad aspect of
fusion, junctional kyphosis has become a common radiographic finding at the transition
between fused and mobile segments (8) (Figure 2). Several papers (8-12) have reported the incidence and risk factors of development of PJK after limited thoracolumbar fusion with cephalad level ending at the thoracolumbar junction. The purpose of this paper is to determine the rate of PJK and the rate of PJK requiring revision surgery, and to review radiographic parameters to identify risk factors
for development of PJK.3

METHODS

One hundred and ninety-nine patients underwent instrumented surgery for adult degenerative scoliosis with cephalad instrumented level to T10-L1 between 1992-2006 at a single institution. Eighty-seven of the 199 patients were found to have completepreoperative,post-operative, and most recent follow-up radiographs due to conversion from a film-based system to an electronic file system. Inclusion criteria included primary diagnosis of adult degenerative scoliosis treated with posterior spinal fusion and instrumentation with or without anterior spinal fusion and instrumentation, age>30 years at time of surgery, and proximal level of fusion at T10-L1. The cohort included patients undergoing primary and revision surgery. Instrumentation consisted of fixation with bilateral pedicle screws in all cases. All operations were performed at a single institution by six fellowship trained spinal deformity surgeons. Radiographic follow-up was not obtained for many patients in the originally identified cohort due to the destruction of the radiographs taken prior to 2000 by the medical center when converting to digital films if the patients did not have followup within a certain time period. Therefore a sub-group (n=87) of the original study population (n=199), who underwent surgery between 2000 and 2006, was used for the final analysis. Complete pre-operative, post-operative, and most recent follow-up radiographs were obtained for this cohort. Radiographic parameters evaluated included lumbar lordosis (L1-S1), adjacent segment kyphosis, thoracolumbar kyphosis (T10-L2), sacral slope, proximal and distal extent of instrumentation, and pelvic incidence. The standard Cobb method was employed. All parameters were measured for pre-operative 4 baseline, immediate post-operative and most recent follow-up radiographs. PJK was defined as >5° change of adjacent segment kyphosis over successive radiographs (9). All radiographic measurements were made digitally by one investigator using a web-based (PACS) measuring system. Accuracy was assessed by random selection of
approximately ½ of the total radiographs which were subsequently re-measured by the original investigator who is a fellowship trained spine surgeon. These measurements were all within 5° of the original measurement and the radiographic dimensions were therefore judged to be accurate. A descriptive profile of the cohort was constructed and risk factors for the development of PJK after posterior spinal fusion were evaluated. Demographic data and surgical data collected included age, gender, length of time between original surgery and development of PJK, number of fused vertebrae, and the level of the upper-instrumented vertebra. The data were analyzed using SPSS version 12.0. Correlation analysis was performed using Spearman’s rho. P values for categorical values were made using Fisher’s exact test or Chi square test. P values of < 0.05 were considered significant. Logistic regression analyses were used to assess factors independently associated
with the development of PJK. The development of PJK as defined by radiographs was used as the dependent variable; and age, gender, upper instrumented vertebrae, lower instrumented vertebrae, and all radiographic parameters at all time points as well as the change in lumbar lordosis from pre-op to immediate post-op and change in kyphosis from immediate post-op to most recent follow-up or revision surgery, and ratio of change in lordosis to pelvic incidence were selected as independent variables.5

Results :
A group of 87 patients who underwent surgical correction of adult scoliosis between 2000-2006 was examined in detail. In this cohort, the rate of proximal junctional kyphosis (requiring surgery or not) was 44% (n=39). 10% (n=9) of patients required revision surgery for PJK. See Figure 1. The most common reason for revision surgery for PJK was proximal screw pull-out. The correlation between proximal extent of instrumentation and the development of PJK requiring surgery or not was not statistically significant by Spearman correlation, when examined overall (p=.329). However, a trend was noted, with those fused to T10 having a lower rate of PJK occurrence. Of those whose fusion stopped at T10 proximally, 29% (7/24) developed PJK. 8% (2/24) fused to T10 proximally required revision surgery for PJK. Of those fused to T11 proximally, 58% (11/19) developed PJK. 11% (2/19) whose fusion ended proximally at T11 required revision surgery for PJK. Of those fused to T12 proximally, 50% (15/30) developed PJK. 13% (4/30) fused proximally to T12 required revision surgery for PJK. Of those fused proximally to L1, 36% (6/14) developed PJK. 7% (1/14) fused to L1 required revision surgery for PJK.

See Table 1, 2, and 5.

Logistic regression was used to identify variables which had a significant effect on the risk of developing PJK. Two analyses were run, with the dependent variable being development of PJK (requiring surgery or not) and development of PJK requiring surgery, respectively. For the analysis in which cephalad fusion level was examined as the independent variable, the cohort was first examined to see if a statistically significant difference with respect to development of PJK, age, gender, pre-operative lordosis, and 6
pre-operative kyphosis between groups whose distal extent of fusion ended at a mobile vertebral segment (L3, L4, and L5) vs. non-mobile segment (S1 and Ilium). No significant difference was found (p=.717). Logistic regression revealed that the risk of developing PJK with a proximal extent of fusion at the level of T11 compared to T10 was nearly 3 fold. (odds ratio:2.92, p=0.09). The risk of developing PJK with a proximal extent of fusion at the level of T12 compared to T10 was approximately 2 fold (odds ratio: 2.12 , p=.181). The risk of developing PJK with a proximal extent of fusion at the level of L1 compared to T10 was approximately 1.5 fold (odds ratio: 1.41, p=.608). Caudal extent of instrumentation was not statistically related to the development of PJK (p=.424). There is no correlation between distal level of fusion and prevalence of revision surgery for PJK (p= .331). 12 patients had their caudal fixation ending at L5; 4 developed PJK as evidenced by radiographs, however, none required revision surgery for PJK. 58 patients had their caudal fixation ending at S1. 27 patients (47%) with distal extent of instrumentation extending to S1 developed PJK, 5 of whom (9%) required revision surgery. 9 patients had their caudal fixation ending at the ilium. 5 patients with caudal instrumentation ending at the ilium developed PJK. 3 (33%) required revision surgery for PJK. 8 patients had their fusion ending at L4 or above and were not included in this analysis.

No radiographic parameters measured were correlated with an increased rate of PJK development. Although there was a trend of increased lumbar lordosis in patients requiring revision surgery for PJK, with a 12 degree average change compared to a 4 degree change for those not developing PJK, this trend was not statistically significant (p= .215). There was no statistical difference in pelvic incidence (p=.203), sacral slope (p=.773), or the ratio of lumbar lordosis to pelvic incidence (p=.273) between the groups.

See Table 3 and Table 4.

Patient demographic characteristics were similar between all groups and were not related to the development of PJK. The average age of patients included in the analysis was 59 (28-81). The average age of patients undergoing revision for PJK was 62 (48-77). Age did not have an effect on the development of PJK (p=.381). The gender distribution was similar in all groups and was not statistically related to the development of PJK (p=.386). The cohort consisted of 70% (n=61) females and 30% (n=26) males.

Discussion :

The ability to end a scoliosis fusion proximally at the thoracolumbar junction that does not result in proximal junctional kyphosis remains a clinical challenge for surgeons and patients. This study did not demonstrate an effect of cephalad or distal level of fusion on the rate of PJK or rate of revision surgery for PJK. There was however a trend seen with those patients fused to T10 having lower rates of PJK. Regression analyses revealed a strong trend of increased odds of developing PJK at T11, T12, and L1 compared to T10. A trend was noted, with those fused to T10 having a 3 fold lower rate of PJK than those fused to T11, a 2 fold lower rate than those fused to T12, and a 1.5 fold lower rate than lose fused to L1. Although there was a trend of increased lumbar lordosis in patients requiring revision surgery for PJK, this trend was not statistically significant. There was no difference in pelvic incidence, sacral slope, or the ratio of lumbar lordosis to pelvic incidence between the cohorts studied.The problem of proximal junctional kyphosis has been studied by authors previously in adolescent idiopathic patients (8). Yang et al. have recommended extending the fusion into the structural thoracic spine if focal kyphosis above the fusion is greater than 10 degrees. Their paper also suggested that lumbar lordosis should be preserved carefully especially for patients whose preoperative lumbar lordosis is greater than 35 degrees. They also further concluded that the avoidance of a distraction maneuver in the lumbar region and proper positioning of the patients with the hip extended are important in preserving lumbar lordosis, and preventing the occurrence of proximal junctional kyphosis. The literature has several examples from the adolescent idiopathic population. McCance et al. reviewed 67 AIS patients with selective posterior thoracic Harrington fusions and found that radiographic distal junctional kyphosis
occurred in one patient with no cases of proximal junctional kyphosis (13). Lee et al. also looked at a population of 69 AIS patients and found that 46% of patients had proximal junctional kyphosis (5 degrees or greater) at 2 year followup (14). Rhee et al. studied a cohort of AIS patients and found PJK (10 degrees or greater) in 35% of their patients treated with posterior instrumented fusion (15). Outcome studies in adult deformity surgery have commonly presented data on process variables, including curve correction, fusion status, and complications (16-18). Previous work from our institution has revealed that radiographic parameters are not necessarily an accurate predictor of patient self-assessment of health status (19). However it is important to ascertain which radiographic parameters may be related to proximal junctional kyphosis and revision surgery for this important clinical problem in scoliosis surgery. We have attempted to provide an analysis of commonly considered variables such as pelvic incidence, sacral slope, and lumbar lordosis and their relationship to proximal junctional kyphosis and its relationship with the need for revision surgery. Glattes et al. have done the most comprehensive analysis about proximal junctional kyphosis in adult deformity patients, reviewing incidence, outcomes, and risk factor analysis. The authors questioned whether PJK occurs iatrogenically after fusion or if it is a consequence of natural age-related degeneration (9). It is well known that age related osteoporosis and disc degeneration leads to a kyphotic spine (20). Glattes et al. conclude from their study that the occurrence of PJK is likely multifactorial in nature. They analyzed the incidence of proximal junctional kyphosis to be 26% but stated that SRS-24 scores in these patients were not significantly affected. Their study did not comment on the proximal level of fusion at the thoracolumbar junction and its potential association with adjacent segment disease or kyphosis nor on specific radiographic parameters, including lumbar lordosis, sacral slope, pelvic incidence and their relationship to the development of PJK and PJK requiring revision surgery. Our study concurs with the assessment that the development of PJK is likely to be multifactorial but also suggests specific radiographic parameters the surgeon needs to consider in long instrumented constructs for adult degenerative scoliosis whose cephalad level ends at the thoracolumbar junction. Ten percent of the patients required revision surgery for PJK in our study. Of the patients that required revision surgery for PJK, there was a trend toward increased lumbar lordosis, although not statistically significant due to the small number of patients in this cohort. Surgeons typically strive to achieve physiological lumbar lordosis in their constructs and this maintains equilibrium of the pelvis and spine (21-22). However this study raises the possibility that overzealously increasing the patient’s baseline lordosis may predispose patients to PJK requiring revision surgery. Physiological lordosis must be obtained in order to provide sagittal homeostasis. Three dimensional regulation of spinal sagittal alignment has been shown to be closely correlated to pelvic incidence (23-24). Our study is the first to take pelvic incidence and the lumbar lordosis/pelvic incidence ratio into account with lack of correlation to proximal junctional kyphosis. A major weakness of this study is that we are unable to provide long standing C7 to sacrum plumb line measurements because of the transition from traditional scoliosis films to digital radiographs. Also because the fusions in the cohort stop at thethoracolumbar junction, many radiographs taken are not entire spine views that allow us to visualize C7. The fact that we were not able to capture a consecutive number of patients also is
a weakness of this study. This was due to the fact that a number of patients were lost to radiographic and clinical followup and therefore were unable to be included in the study. Another weakness in this study comes from the error associated with measurement. Carmen et al. have reported on the difficulty of measuring kyphosis in a pediatric population (25). Namely they talk about the difficulty of measuring cartilaginous endplates resulting in greater intraobserver and interobserver variability. No similar study exists in the adult deformity population but such variability could indeed exist. Our study supports the likelihood that the causes of PJK and PJK requiring revision surgery are multifactorial in nature. Future studies should consider the role of
osteoporosis, body mass index, and medical comorbidities as risk factors for this proximal decompensation. Taking these factors into account may help the surgeon decide on the best cephalad level for long fusions ending at the thoracolumbar junction. with PJK requiring surgery.

Cephalad Level (by group)

Table 1


Table 2

Table 3

Table 4

Table 5

 
Reference :

1. Polly DW, Hamill CL, Bridwell KH. Debate: to fuse or not to fuse to the sacrum, thefate of the L5-S1 disc. Spine. 2006 Sep 1;31(19 Suppl):S179-84.
2. 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. 2006 Sep 1;31(19 Suppl):S185-94.
3. Wigh RE: The thoracolumbar and lumbosacral transitional junctions.Spine. 1980 May-Jun;5(3):215-22.
4. Kwon BK, Elgafy H, Keynan O, Fisher CG, Boyd MC, Paquette SJ, Dvorak MF.Progressive junctional kyphosis at the caudal end of lumbar instrumented fusion:
   etiology, predictors, and treatment. Spine. 2006 Aug 1;31(17):1943-51.
5. Lowe TG, Lenke L, Betz R, Newton P, Clements D, Haher T, Crawford A, Letko L, Wilson LA. Distal junctional kyphosis of adolescent idiopathic thoracic curves following anterior or posterior instrumented fusion: incidence, risk factors, and prevention. Spine.
   2006 Feb 1;31(3):299-302.
6. Frez R, Cheng JC, Wong EM. Longitudinal changes in trunkal balance after selective fusion of King II curves in adolescent idiopathic scoliosis. Spine. 2000 Jun
   1;25(11):1352-9.
7. Swamy G, Berven SH, Bradford DS. The selection of L5 versus S1 in long fusions for adult idiopathic scoliosis. Neurosurg Clin N Am. 2007 Apr;18(2):281-8 * References (cited in order of appearance)
8. Yang SH, Chen PQ. Proximal kyphosis after short posterior fusion for thoracolumbar scoliosis. Clin Orthop Relat Res. 2003 Jun;(411):152-8.
9. Glattes RC, Bridwell KH, Lenke LG, Kim YJ, Rinella A, Edwards C 2nd. Proximal junctional kyphosis in adult spinal deformity following long instrumented posterior spinal fusion: incidence, outcomes, and risk factor analysis. Spine. 2005 Jul 15;30(14):1643-9.
10. Oskouian RJ Jr, Shaffrey CI. Degenerative lumbar scoliosis. Neurosurg Clin N Am. 2006 Jul;17(3):299-315.
11. Lippman CR, Spence CA, Youssef AS, Cahill DW. Correction of adult scoliosis via a posterior-only approach. Neurosurg Focus. 2003 Jan 15;14(1):e5.
12. Kumar MN, Baklanov A, Chopin D. Correlation between sagittal plane changes and adjacent segment degeneration following lumbar spine fusion. Eur Spine J. 2001Aug;10(4):314-9.
13. McCance SE, Denis F, Lonstein JE, et al. Coronal and sagittal balance in surgically treated adolescent idiopathic scoliosis with the King II curve pattern. A review of 67 consecutive cases having selective thoracic arthrodesis. Spine 1998;23:2063–73.
14. Lee GA, Betz RR, Clements DH III, et al. Proximal kyphosis after posterior spinal fusion in patients with idiopathic scoliosis. Spine 1999;24:795–9.
15. Rhee JM, Bridwell KH, Won D, et al. Sagittal plane analysis of adolescent idiopathic scoliosis: the effect of anterior versus posterior instrumentation. Spine 2002;27:2350–6.
16. Sponseller PD, Cohen MS, Nachemson AL, et al. Results of the surgical treatment of adults with idiopathic scoliosis. J Bone Joint Surg. 1987;69A: 667–675.
17. Kostuik JP, Israel J, Hall JE. Scoliosis Surgery in Adults. Clin Orthop. 1973;225–234.
18. Swank S, Lonstein JE, Moe JH, et al. Surgical Treatment of Adult Scoliosis. A Review of 222 Cases. J Bone Joint Surg. 1981;63A:268–287.
19. Berven SD, Deviren V, Demir-Deviren S, et al. Studies in the Modified Scoliosis Research Society Outcomes Instrument in adults; validation, reliability, and discriminatory capacity. Spine 2003;28:2164–9.
20. Hammerberg EM, Wood KB. Sagittal profile of the elderly. J Spinal Disorder Tech 2003;16:44–50
21. Kobayashi T, Atsuta Y, Matsuno T, Takeda N. A longitudinal study of congruent sagittal spinal alignment in an adult cohort. Spine. 2004 Mar 15;29(6):671-6.
22. Vaz G, Roussouly P, Berthonnaud E, et al. Sagittal morphology and equilibrium of pelvis and spine. Eur Spine J 2002;11:80–7.
23. Legaye J, Duval-Beaupe`re G, Hecquet J, et al. Pelvic incidence: a fundamental pelvic parameter for three-dimensional regulation of spinal sagittal curves. Eur Spine J 1998;7:99–103.
24. Jackson RP, Kanemura T, Kawakami N, et al. Lumbopelvic lordosis andpelvic balance on repeated standing lateral radiographs of adult volunteers and untreated subjects with constant low back pain. Spine 2000;25:575–86.
25. Carman DL, Browne RH, Birch JG. Measurement of scoliosis and kyphosis radiographs: intraobserver and interobserver variation. J Bone Joint Surg1990;72-A:328–33.

This is a peer reviewed paper 

Please cite as : Rajiv K Sethi,Proximal Junctional Kyphosis above Fusions to the Thoracolumbar Spine for Adult Scoliosis: Rate and Predictive Factors.

J.Orthopaedics 2011;8(4)e14

URL: http://www.jortho.org/2011/8/4/e14