The Spine Journal

-

(2014)

-

Clinical Study

Preoperative computer-based simulations for the correction of kyphotic deformities in ankylosing spondylitis patients Ye-Soo Park, MD, PhD*, Hong-Sik Kim, MD, Seung-Wook Baek, MD, Jeong-Han Oh, MD Department of Orthopaedic Surgery, Guri Hospital, Hanyang University College of Medicine, Yeongchunro153, Guri-Si, Gyeonggi-Do, 471-701, Korea Received 29 January 2013; revised 13 January 2014; accepted 3 February 2014

Abstract

BACKGROUND CONTEXT: A preoperative plan is important to obtain appropriate balance of the sagittal plane in patients with kyphotic deformity. Previous methods to calculate the correction angle are inconvenient and complicated, whereas the method using computer simulations may be very effective and much simpler than existing methods. PURPOSE: To evaluate the efficacy of preoperative measurements using a computer simulation for corrective osteotomy for the surgical treatment of kyphosis caused by ankylosing spondylitis (AS). STUDY DESIGN: Retrospective clinical data analysis. PATIENT SAMPLE: The sample comprises 18 AS patients with fixed kyphotic deformity who underwent corrective osteotomies at our hospital between October 2007 and January 2010. OUTCOME MEASURES: Thoracic kyphosis, lumbar lordosis, and the sagittal vertical axis (SVA) of the spine were evaluated by preoperative computer simulation and radiologic measurement. Clinical assessments were performed according to the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI), Bath Ankylosing Spondylitis Functional Index (BASFI), Short Form-36 (SF-36), and EuroQol-5 dimension (EQ-5D) before and after the surgery. METHODS: The coincidence between the preoperative computer simulation and postoperative radiologic parameters was evaluated. We also analyzed the changes derived from each clinical and radiologic measurement before and after the surgery. RESULTS: Mean thoracic kyphosis changed from 32.4 to 31.9 , mean lumbar lordosis was corrected from 11.5 to 26.9 , and the SVA was improved from 125.7 to 65.1 mm after surgery (p!.001). The correlation coefficients within groups between the computer simulations and radiologic parameters were 0.9, 0.6, and 0.7, showing significant congruency. Although BASDAI and BASFI did not significantly differ (p5.53 and p5.45, respectively), SF-36 and EQ-5D were significantly increased (p!.05 and p!.001, respectively). CONCLUSIONS: Comparisons of preoperative simulations and actual surgical outcomes showed significant coincidences; thus, evaluations through computer simulations before surgery are expected to help predict the level of correction possible after surgery and improve surgical planning. Ó 2014 Elsevier Inc. All rights reserved.

Keywords:

Ankylosing spondylitis; Kyphotic deformity; Corrective osteotomy; Preoperative measurement; Computer simulation; Sagittal vertical axis

FDA device/drug status: Not applicable. Author disclosures: Y-SP: Nothing to disclose. H-SK: Nothing to disclose. S-WB: Nothing to disclose. J-HO: Nothing to disclose. This study was not supported by any financial sources nor do the authors have any financial relationships to disclose. No funds were received in support of this work. No benefits in any form have been or will be received from a commercial party related directly or indirectly to the subject of this manuscript. Conflict of interest: The manuscript submitted does not contain 1529-9430/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.spinee.2014.02.022

information about medical device(s)/drug(s). Approved for the study by the institutional review board (IRB) on Human Subjects Research and Ethics Committees, Hanyang University Guri Hospital, Korea. * Corresponding author. Department of Orthopaedic Surgery, Guri Hospital, Hanyang University College of Medicine, Yeongchunro153, Guri-Si, Gyeonggi-Do, 471-701, Korea. Tel.: (82) 31-560-2316; fax: (82) 31-557-8781. E-mail address: [email protected] (Y.-S. Park)

2

Y.-S. Park et al. / The Spine Journal

-

(2014)

-

Introduction

Methods

Ankylosing spondylitis (AS) is a chronic inflammatory disease that primarily invades the spine and sacroiliac joint [1–3]. As the disease progresses, it moves to the upper part of the spine and the mobility of the joints disappears. In some patients, destructive lesions, such as Andersson lesions, are seen alongside fixed kyphosis because of reduced lumbar lordosis or increased thoracic kyphosis [4]. The prevalence of AS ranges from 0.1% to 1.1%, and spinal deformity may be reduced by the use of antitumor necrosis factor [5]. Because there are few compensatory mechanisms for kyphosis caused by AS, it causes many limitations in everyday life and may generate problems with walking and communication because it is impossible for patients to face forward. Therefore, surgical correction is necessary. Corrective osteotomy of fixed kyphosis is very difficult. It is thought to be very effective in patients with long segment kyphosis who lose compensation of sagittal imbalance. Pedicle subtraction osteotomy was first introduced by Thomasen [6], and today, it is widely used for the correction of fixed sagittal curvature. Through this technique, corrections up to 40 can be derived by segmental osteotomy. It is necessary to define a preoperative plan and calculate the correction angle to obtain appropriate balance of the sagittal plane [7]. Previous studies have addressed methods that use the chin-brow vertical angle [8], radiograph cutout [9], tangent [10], and pelvic parameters [11]; however, most of these methods are inconvenient and complicated. When designing the preoperative plan for an osteotomy, especially for the treatment of kyphotic deformity in patients with AS, computer simulations can be very effective and are much simpler than the existing methods. We evaluated the feasibility of preoperative computer measurements for kyphosis correction in patients with AS.

In this retrospective study, we reviewed the records of 18 patients with AS accompanied by fixed kyphotic deformity and who underwent pedicle subtraction osteotomy from October 2007 to January 2010. The average age of the patients was 36.5 years (30–53 years). Seventeen of the patients were male, whereas one was female. The enrolled subjects were followed-up for an average of 37.2 months (25.3–51.7 months). The correction angle was measured through a computer simulation before pedicle subtraction osteotomy for every patient. Inability to maintain standing posture or face forward while standing were the main indications for operation. Surgimap Spine software (Nemaris, Inc., New York, NY, USA) was used for computer simulations. The distances of thoracic kyphosis, lumbar lordosis, and the sagittal plane from the posterosuperior end plate of S1 were measured through long-cassette standing lateral spinal radiographs, and computer simulations were promoted by setting an appropriate angle at the lumbar spine (Fig. 1). To select the segment for corrective osteotomy, simulations of osteotomies in three different vertebral bodies, including the upper and lower vertebral bodies of the apex of kyphosis, were conducted to compare the sagittal vertical axis (SVA) in each segment. The segment that realized the best correction of the SVA was selected for surgery (Fig. 2). To determine the correction angle, osteotomies in the shapes of isosceles and right triangles at the selected segment were simulated. When conducting the isosceles triangle osteotomy simulation, the correction angle was the largest and the SVA were most improved; thus, isosceles triangle osteotomies were conducted (Fig. 3). The segment and angle determined via computer simulations were used in the actual surgeries. Long-cassette standing lateral spinal radiographs that were taken in the standing posture before and after surgery

Fig. 1. Surgimap Spine is a software tool that organizes simulated postoperative results via intuitive mouse clicks.

Y.-S. Park et al. / The Spine Journal

-

(2014)

-

3

Fig. 2. Osteotomy is simulated at the apex of the kyphosis and the adjacent level. The resulting sagittal balance is compared, and the appropriate osteotomy area is selected. preop., preoperative.

were used for the preoperative simulation, and the values of thoracic kyphosis, lumbar lordosis, and SVA before surgery and the computer-measured values before and after surgery were assessed. Each value was measured three times by three observers.

For clinical evaluation, the Short Form-36 and EuroQol5 dimension were used to determine the quality of life. The Bath Ankylosing Spondylitis Functional Index, composed of 10 questions for the evaluation of physical functions and the degree of ordinary life limitation, and the Bath

Fig. 3. The sagittal vertical axis is greatly improved by simulations using an isosceles triangle.

4

Y.-S. Park et al. / The Spine Journal

Ankylosing Spondylitis Disease Activity Index, composed of 6 questions related to major symptoms, were used to evaluate disease states. Values before and after the surgery were compared and analyzed. For statistical analysis, SPSS version 13.0 (SPSS, Inc., Chicago, IL, USA) was used. The coincidences in terms of intraobserver reproducibility and interobserver reliability and the scale of the sagittal plane before and after the simulation and surgery were evaluated using intraclass correlation coefficients introduced by Shrout and Fleiss [12], and the clinical results were evaluated using paired t tests. The intragroup correlation coefficients ranged from 0 (not related) to 1 (perfect match) and values greater than 0.5 were defined as significantly coincident. The significance level was set to be p!.05.

-

(2014)

-

summary after the operation was 67 points (48–84) and the average mental component summary was 65 points (48–78), showing statistically significant increases (p!.05). The EuroQol-5 dimension value showed a statistically significant improvement from an average of 5.27 points (4–7) before the operation to 7.23 points (5–12) after the operation (p!.001). The average Bath Ankylosing Spondylitis Disease Activity Index and Bath Ankylosing Spondylitis Functional Index before the operation were 5.9 points (4–7) and 6.0 points (2–6), respectively, and 4.8 points (3–6) and 5.0 points (4–6) after the operation; nonetheless, these results were not statistically significant (p5.53 and p5.45, respectively). There were no signs of surgically related complications, such as paralysis, vessel injury, or cerebrospinal fluid leakage in any of the patients.

Results Of the 18 patients, osteotomies were conducted at L2 in 2 patients and at L3 in 16 patients. The thoracic kyphosis was corrected from an average of 32.4 (15 –63 ) before the operation to an average of 31.9 (10 –57 ) after surgery. The lumbar lordosis was significantly improved from an average of 11.5 (1 –25 ) to an average of 26.9 (14 – 36 ), whereas the SVA was improved from an average of 125.7 mm (69–206 mm) to an average of 65.1 mm (25– 101 mm) (p!.001). Regarding the respective intraobserver reproducibility of the thoracic kyphosis, lumbar lordosis, and SVA, the values of the intragroup correlation coefficients ranged from 0.98 to 1.00; indicating significant coincidence. Similar coincidence values (between 0.98 and 1.00) were found for interobserver reliability. When comparing the preoperative computer simulation with the postsurgery results, thoracic kyphosis showed similar coincidence with 32.4 (15 –63 ) and 31.9 (10 –57 ) on average and an intragroup correlation coefficient of 0.9. The lumbar lordosis was 25.1 (2 –52 ) and 26.9 (14 –36 ) on average with an intragroup correlation coefficient of 0.6. The SVA showed significant coincidence (59.9 mm [3–145 mm] and 65.2 mm [25–101 mm], respectively) with an intragroup correlation coefficient of 0.7 (Table). The average physical component summary from Short Form-36 before the operation was 50 points (12–72) and the average mental component summary was 64 points (33–81). Meanwhile, the average physical component Table Agreement between simulated and postoperative values 

Mean thoracic kyphosis ( ) Mean lumbar lordosis (  ) Mean SVA (mm)

Simulated

Postoperative

ICC value

32.4 (15–63) 25.1 (2–52) 59.9 (3–145)

31.9 (10–57) 26.9 (14–36) 65.2 (25–101)

0.9 0.6 0.7

SVA, sagittal vertical axis; ICC, intraclass correlation coefficient (poor [!0.40], moderate [0.40–0.75], excellent [0.75–1.00]).

Discussion Osteotomy is widely used in patients with kyphotic deformities because of AS to recover sagittal plane and coronal plane balance [13]. Osteotomy aims to obtain such balance by correcting the spinal deformity, and balance in the sagittal and coronal planes can be obtained through various types of osteotomy depending on the shape of the deformity [14]. The degree of correction can be refined by using combinations of osteotomy techniques and different numbers of segments in osteotomy. Smith-Peterson osteotomy is a surgical method that is mainly used for round and nonfixed kyphosis in the long segments [7]. An angle correction of approximately 1 can be achieved for each 1 mm of osteotomy [15]. However, this method is limited in application if the anterior spine is fused. Pedicle subtraction osteotomy can be used for spinal deformities that need large corrections (up to about 40 ) [16]. It also may be used on patients with anterior spinal stiffness or kyphotic deformities in patients with AS. Sagittal plane balance is considered to be the most important aspect when conducting an osteotomy. Osteotomies are complicated and may cause postoperative complications. Despite such problems, favorable radiologic and clinical outcomes can be realized after osteotomy. Therefore, thorough analysis and planning are necessary to realize satisfactory outcomes [7]. Potential problems include proximal junctional kyphosis in cases of overcorrection [17– 19] or poor clinical outcomes including pain and poor function in cases of undercorrection [8]. There are various methods for planning a treatment of kyphotic deformity in patients with AS. The classical method includes a radiograph trace and the cutout method [9] measures the angle needed for correction by physically cutting X-rays; however, these methods are inconvenient and time-consuming. The method that uses pelvic parameters [11] measures the sacral slope and lumbar lordosis by applying the pelvic index, but fails to consider thoracic

Y.-S. Park et al. / The Spine Journal

kyphosis and individual differences in sacral slope and pelvic tilt. Many effective methods, including the tangent method [10], chin-brow vertical angle method [8], and whole body kyphosis angle method [20] have been introduced; however, all fail to produce as much improvement in angle correction as the film cutout method. In this study, pedicle subtraction osteotomy was conducted after preoperative simulations with Surgimap Spine software. Through the preoperative computer simulations of osteotomies conducted on L2, L3, and L4, degrees of SVA recovery were compared and used to select osteotomy segments and determine appropriate correction angles. Correction angles were planned on the radiograph of each entire spine that was entered into the program before surgery. Then, the distances between the perpendicular, which was automatically drawn using the radiograph, thoracic kyphosis, lumbar lordosis, the center of the C7 vertebra after the osteotomy, and the posterosuperior of the first sacral (S1) were calculated; thus, it was possible to easily compare the radiographs before and after the operation. Such comparisons enable verification by conducting an osteotomy directly to the spine of a patient and are much easier and less time consuming than the film cutout method. The results of computer simulation showed excellent intraobserver reproducibility and interobserver reliability with better than normal levels of sagittal plane index coincidence between simulation and postoperative results. We observed no complications, such as infection, paralysis, or cerebrospinal fluid leakage, as found in other studies [9,10,20]. Computer simulations improve preoperative planning for the treatment of complicated kyphosis and produce relatively accurate measurements. Thus, the use of simulations is considered to be effective for preoperative measurements and can reduce the incidence of complications such as overand undercorrection. Conclusion Preoperative computer simulations are simpler and easier than previously suggested methods for selecting correction segments and angles in patients with kyphotic deformities because of AS. The values of sagittal plane indexes between radiographs taken before and after surgery were similar. Computer simulations facilitate preoperative planning when correcting complicated cases of kyphosis and accurately predict outcomes. Patients may achieve increased levels of postoperative satisfaction because they can view postoperative posture changes before the actual operation.

-

(2014)

-

5

References [1] Kim TJ, Kim TH. Clinical spectrum of ankylosing spondylitis in Korea. Joint Bone Spine 2010;77:235–40. [2] McMaster MJ, Coventry MB. Spinal osteotomy in akylosing spondylitis. Technique, complications, and long-term results. Mayo Clin Proc 1973;48:476–86. [3] Pham T. Pathophysiology of ankylosing spondylitis: what’s new? Joint Bone Spine 2008;75:656–60. [4] Park YS, Kim JH, Ryu JA, et al. The Andersson lesion in ankylosing spondylitis: distinguishing between the inflammatory and traumatic subtypes. J Bone Joint Surg Br 2011;93:961–6. [5] Braun J, Sieper J. Therapy of ankylosing spondylitis and other spondyloarthritides: established medical treatment, anti-TNFalpha therapy and other novel approaches. Arthritis Res 2002;4: 307–21. [6] Thomasen E. Vertebral osteotomy for correction of kyphosis in ankylosing spondylitis. Clin Orthop Relat Res 1985;194:142–52. [7] Bridwell KH. Decision making regarding Smith-Petersen vs. pedicle subtraction osteotomy vs. vertebral column resection for spinal deformity. Spine 2006;31:S171–8. [8] Suk KS, Kim KT, Lee SH, et al. Significance of chin-brow vertical angle in correction of kyphotic deformity of ankylosing spondylitis patients. Spine 2003;28:2001–5. [9] Gertzbein SD, Harris MB. Wedge osteotomy for the correction of post-traumatic kyphosis. A new technique and a report of three cases. Spine 1992;17:374–9. [10] Ondra SL, Marzouk S, Koski T, et al. Mathematical calculation of pedicle subtraction osteotomy size to allow precision correction of fixed sagittal deformity. Spine 2006;31:E973–9. [11] Legaye J, Duval-Beaupere G. Sagittal plane alignment of the spine and gravity: a radiological and clinical evaluation. Acta Orthop Belg 2005;71:213–20. [12] Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull 1979;86:420–8. [13] Christodoulou A, Ploumis A, Terzidis J, et al. Surgical management of a congenital kyphotic deformity in an adolescent. Stud Health Technol Inform 2002;91:454–6. [14] Cho KJ, Bridwell KH, Lenke LG, et al. Comparison of SmithPetersen versus pedicle subtraction osteotomy for the correction of fixed sagittal imbalance. Spine 2005;30:2030–7. [15] Booth KC, Bridwell KH, Lenke LG, et al. Complications and predictive factors for the successful treatment of flatback deformity (fixed sagittal imbalance). Spine 1999;24:1712–20. [16] Murrey DB, Brigham CD, Kiebzak GM, et al. Transpedicular decompression and pedicle subtraction osteotomy (eggshell procedure): a retrospective review of 59 patients. Spine 2002;27:2338–45. [17] Yu WD, Bernstein RM, Watts HG. Autogenous tibial strut grafts used in anterior spinal fusion for severe kyphosis and kyphoscoliosis. Spine 2003;28:699–705. [18] Farcy JP, Weidenbaum M, Glassman SD. Sagittal index in management of thoracolumbar burst fractures. Spine 1990;15:958–65. [19] Hosman AJ, Langeloo DD, de Kleuver M, et al. Analysis of the sagittal plane after surgical management for Scheuermann’s disease: a view on overcorrection and the use of an anterior release. Spine 2002;27:167–75. [20] Min K, Hahn F, Leonardi M. Lumbar spinal osteotomy for kyphosis in ankylosing spondylitis: the significance of the whole body kyphosis angle. J Spinal Disord Tech 2007;20:149–53.