CC BY-NC-ND
0Journal of Orthopaedics, Trauma and Rehabilitation xxx (2014) 1—6
Contents lists available at ScienceDirect
Journal of Orthopaedics, Trauma and Rehabilitation
Journal homepages: www.e-jotr.com & www.ejotr.org
Original Article
Clinical Outcome and Complications of Transpedicular Closing-wedge Osteotomy for Correction of Deformity in Ankylosing Spondylitis in a Regional Hospital
Yim Wing-Ngai*, Cheung Ka-Kin, Chan Andrew Pak-Ho, Cheng Hung-On, Chow Yuk-Yin
Department of Orthopaedics and Traumatology, Tuen Mun Hospital, Tuen Mun, Hong Kong
article info
Article history: Accepted August 2013
Keywords:
ankylosing spondylitis clinical outcome complications kyphotic deformity
transpedicular closing wedge osteotomy
abstract
Study design: Prospective study of surgical correction of thoracolumbar kyphotic deformity caused by ankylosing spondylitis.
Objectives: To assess surgical outcomes and complications of thoracolumbar kyphotic deformity corrected with transpedicular closing wedge osteotomy performed in a regional hospital. Summary of background data: There have been several studies reporting on the results of surgical correction of deformity in ankylosing spondylitis all over the world. However, there has not been any local data published.
Method: From 2003 to 2011, we had performed 12 transpedicular closing wedge osteotomies in 9 patients with ankylosing spondylitis for correction of kyphotic and scoliotic deformity in thoracolumbar spine. Operative outcomes were assessed clinically by recording the Japanese Orthopaedic Association (JOA) scores, visual analogue scale (VAS) pain scores, Oswestry Disability Index (ODI) preoperatively and post-operatively and patient satisfaction postoperatively. Radiological outcome was assessed by measuring thoracic kyphosis, lumbar lordosis and sagittal plumb line preoperatively and postoperatively as well as the degree of surgical correction. Occurrence of complications was recorded by our standard audit protocol. Results: All patients had a single level of osteotomy done at a time. Most of the osteotomies were done at L2 or L3. The mean amount of correction was 21.6°. Complications included dural tear, pseudoarthrosis and transient radiculopathy. The extend of correction and incidence of complications improved with experience. Conclusions: Despite transpedicular closing wedge osteotomy being a major operation and is not without complications, most of our patients had good clinical results and subjective satisfaction.
* £ « rn
aiAifîiiflitteÊftfftt. (JOA) , ( vas ) ««f
» , ( ODI ), SftiAMiS, , ,
21.6Ä. »mfe-stmrnrnm, isfflfiittisi.
Conflicts of interest: All authors declare no conflicts of interest.
* Corresponding author. E-mail: acmondhk@yahoo.com.hk.
2210-4917/$ — see front matter Copyright © 2014, The Hong Kong Orthopaedic Association and Hong Kong College of Orthopaedic Surgeons. Published by Elsevier (Singapore) Pte Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jotr.2013.08.001
2 W.-N. Yim et al. / Journal of Orthopaedics, Trauma and Rehabilitation xxx (2014) 1—6
Introduction
Ankylosing spondylitis (AS) is a chronic inflammatory arthritis. It is often associated with severe postural deformity. There is progressive ossification of the ligaments of the spine and major joints resulting in ankylosis of the spine.1 AS is associated with a progressive loss in lumbar lordosis and an increase in thoracic kyphosis. The deformity often results in body imbalance and poor posture. The chief complaint of patients is difficulty in looking straight ahead due to the fixed deformity of the spine. The kyphotic deformity also restricts patients' daily activities such as driving a car, interpersonal communication, and personal hygiene. Some patients may experience back pain because of muscle imbalance or pseudoarthrosis caused by a fracture of the fused spine. In case of severe kyphotic deformity, a flexed posture may cause visceral compression.2
The goal of surgery in AS is to correct sagittal imbalance with extension osteotomy. Thoracolumbar transpedicular closing-wedge osteotomy is one of the widely accepted operations to correct this deformity.2—5 Currently, there is no local study reporting the clinical outcome of this operation in Hong Kong. Therefore, our study aims to report the outcome and complications of this surgery performed in a regional hospital in Hong Kong.
Materials and methods
In the period from 2003 to 2011, nine patients (7 men and 2 women) with a known history of AS underwent closing-wedge osteotomy in Tuen Mun Hospital (Tuen Mun, Hong Kong). In total, 12 osteotomies were performed. A single surgical team carried out all the operations. Patients who had kyphotic deformity together with back pain, pseudoarthrosis, and subjective functional impairment were indicated to have surgery. All the patients who underwent transpedicular closing-wedge osteotomy were included in our study. Surgery was not performed in one patient because of small pedicles, as measured by computed tomography (CT) scan, through which screw insertion was impossible. The average follow-up period was 54 months, ranging from 7 months to 105 months. Three patients had a follow-up of less than 2 years. Surgical outcomes were measured by Japanese Orthopaedic Association (JOA) scores, visual analogue scale (VAS) pain scores, and Oswestry Disability Index (ODI), both pre- and postoperatively, together with subjective satisfaction since 2005.
Data on the following parameters were also recorded: age, sex, comorbidity, operative time, blood loss, operated level, surgical technique, length of stay, use of neural monitoring, and complications.
Figure 1. Position of patient on Jackson frame.
Assessment of deformity
The authors use the standard Cobb's method on standing lateral radiograph of the whole spine to measure the thoracic kyphosis and lumbar lordosis pre- and postoperatively. The degree of operative correction was measured by the Cobb's method at the upper end plate of the first fusion level and the lower end plate of the last fusion level. Sagittal alignment of the spine was measured by a sagittal plumb line, which measured the distance from T1 to S1, pre- and postoperatively.
Surgical planning and technique
Preoperative CT scan and magnetic resonance imaging were performed in all patients to assess the structural anatomy, presence of silent pseudoarthrosis, and narrowing of neural canal. The patients were placed in the prone position on a Jackson frame, with the chest and pelvic being supported and the abdomen allowed to hang free (Figure 1). A midline longitudinal skin incision was made to allow the exposure of the laminae and facet joints. Pedicle screws were inserted usually at three levels above and two levels below the osteotomy segment, under the control of an image intensifier. Most of the osteotomies were performed at L2 or L3 level. If a patient had concomitant pseudoarthrosis, fusion would be extended to two levels above the pseudoarthrosis segment.
The spinous processes of the osteotomy segment were excised and used for bone graft. Laminectomy was performed using a highspeed air drill and a Kerrison rongeur to expose the dura and nerve roots. Transverse processes were cut at the base. Bone cuts were made transpedicularly up to the anterior cortex of the vertebral body. The dura and roots were dissected free and protected during the whole procedure. Closure of the wedge was performed under the control of an image intensifier and the direct visualisation of the dural sac and nerve roots, with the anterior cortex acting as a hinge. Precontoured rods were applied and tightened, with nerve roots confirmed free from compression after correction of deformity. Cross-linked rods were applied. Bone grafts were placed posteriorly after decortication of the transverse processes. The wound was closed in layers, with a drain being inserted. Somatosensory- and motor-evoked potentials were monitored intraoperatively; this service was introduced to our centre in 2009.
All patients were monitored closely for their postoperative neurological status. Thoracolumbosacral orthosis was given to all patients for 3 months.
Results
In nine patients, transpedicular closing-wedge osteotomy was performed for correction of kyphotic deformity. Three of them had a concomitant pseudoarthrosis. All patients had single-level osteotomy (Figure 2) done at a time. Three patients underwent a second-level osteotomy 3—5 years after the first operation: one was performed for remedy of complications, one was performed for inadequate correction, and one was a planned two-stage operation performed to correct both sagittal and coronal malalignments. Osteotomies were performed at T12 in one case, L1 in one case, L2 in four cases, L3 in five cases, and L4 in one case.
Three patients had fixed flexion contractures of the hips. Two of them underwent bilateral total hip replacement prior to spinal osteotomy. The other one did not undergo hip replacement because the flexion contracture was mild and he did not have any symptoms related to hip contracture.
The average operative time was 428 minutes (Figure 3). Three patients had preoperative pseudoarthrosis, which required more levels of instrumentation, resulting in longer operative times in
W.-N. Yim et al. / Journal of Orthopaedics, Trauma and Rehabilitation xxx (2014) 1—6
Figure 2. Lateral view of the vertebral resection (left) and the X-ray of the corrected spine with instrumentation (right).
these patients. The average length of stay in hospital was 17.3 days (6—39 days). Since 2009, when we incorporated this service in our hospital, six patients underwent somatosensory- and motor-evoked potential monitoring during the operation. The average blood loss during the operation was 3175 mL.
Complications
Complications (Table 1) were of three types: intraoperative, early postoperative, and late postoperative.
Regarding the intraoperative complications, three patients (25%) had dural tears during the operation. One of them was repaired successfully; the other two were beyond repair due to dural ectasia. All the dural tears were healed without subsequent complications. One patient (8.3%) had severe blood loss (15,000 mL) due to extensive fusion from T7—L5 and improper positioning prior to the operation. Massive blood transfusion was required. None of the patients had any major vascular injuries.
Four patients (33%) developed transient L2 or L3 radiculopathy as the early postoperative complications. All of them had complete recovery. No superficial or deep wound infection was reported in these patients. None of them developed deep vein thrombosis and postoperative ileus.
Regarding the late complications, one patient (8.3%) had nonunion at the osteotomy segment due to a faulty technique that resulted in pseudoarthrosis. Revision surgery was performed 4 years after the initial operation. One patient (8.3%) had junctional pseudoarthrosis; the involved segment healed spontaneously during subsequent follow-ups. One patient lost the initial
correction with progressive kyphosis and required another level of osteotomy 4 years after the first osteotomy. None of the patients experienced implant loosening or failure.
Outcomes
The average preoperative deformity was 44.2° (33.9—48.9°), measured from T2—SI. The average correction achieved was 21.6° (9—38.5°; Figures 4 and 5). The average pre- and postoperative T1 — S1 distance was 118 mm (65—180 mm) and 43 mm (5—88 mm), respectively. The average improvement in T1—S1 distance after operation was 75 mm (38—119 mm).
Since 2005, the authors had pre- and postoperative charting for the JOA score, ODI, VAS pain score, and subjective improvement (Table 2). Out of the 12 transpedicular closing-wedge osteotomies, the above clinical charting was recorded for six. The data were analysed by Wilcoxon signed rank test, and p < 0.05 was considered statistically significant. The average preoperative JOA lumbar score was 17.5/29. At 6 months after the operation, the average JOA lumbar score was 23.4/29 (p = 0.043). The average JOA lumbar score at the latest follow-up (12—36 months) improved to 25.4/29 (p = 0.043).
For the pain assessment, all patients demonstrated improvement immediately after the osteotomies. VAS scores were charted for back and leg symptoms. The average VAS score for back and leg improved from 6.5/10 to 2.5/10 (p = 0.102) and from 4.5/10 to 3.5/ 10 (p = 0.705), respectively, at 6 months after the operation. The VAS back and leg scores at the latest follow-up (12—36 months) were 4.1/10 (p = 0.416) and 2.1/10 (p = 0.465), respectively. However, the results were not statistically significant.
Functional outcome was assessed using ODI scores. There was a significant improvement of ODI from 41.6% preoperatively to 21.6% (p = 0.028) at 6 months and 22.2% (p = 0.046) at the latest follow-up.
In general, patients were satisfied with the results. Subjective improvement quoted from patients was 80% at 6 months after the operation, and the satisfaction remained until the latest follow-up. All patients would like to undergo the same operation if they were allowed to make a decision again.
Discussion
AS-related kyphotic deformity poses significant restrictions on patients' physical and psychosocial functions. Patients often have difficulties in daily activities as simple as changing clothes and taking care of personal hygiene. More importantly, they have low self-esteem due to their stooped posture, which affects their social life. Camargo et al6 reported that an improvement of appearance is of psychological importance to patients, and results in better mental attitude and ability to participate socially. Therefore, correction of kyphosis is one of the goals of our operation. Another goal of our operation is to relieve patients' back symptoms. People used to believe that back pain in AS is related to the inflammatory nature of the disease. However, most patients had back pain because of muscle imbalance, pseudoarthrosis, or couching gait
Table 1
Complications of spinal correction in ankylosing spondylitis
Operative time —Linear (Operative time) Figure 3. Operative time.
Dural tear 3 25
Massive blood loss 1 8.3
Neurological complication 4 33
Wound infection 0 0
Nonunion 1 8.3
Implant failure 0 0
W.-N. Yim et al. / Journal of Orthopaedics, Trauma and Rehabilitation xxx (2014) 1—6
Figure 4. Photograph showing the posture of our patient preoperatively and postoperatively.
Figure 5. X-ray showing the alignment of lumar spine preoperatively and postoperatively.
secondary to lower limb weakness. These can be improved by surgical correction of the deformity.
Different surgical techniques have been introduced to correct the kyphotic deformity in AS; of these, transpedicular closing-wedge, open-wedge, and polysegmental osteotomies are the
Table 2
Charting of clinical outcomes
Preoperative 6 mo after P Latest P
operation follow-up
Average JOA score 17.5/29 23.4/29 0.043 25.4 0.043
Average ODI 41.6 21.6 0.028 22.2 0.046
Average VAS score (back) 6.5 2.5 0.102 4.1 0.416
Average VAS score (leg) 4.5 3.5 0.705 2.1 0.465
JOA = Japanese Orthopaedic Association; ODI = Oswestry Disability Index; VAS = visual analogue scale.
three most popular techniques used nowadays. '9 In our hospital, transpedicular closing-wedge osteotomy is preferred due to a lesser chance of major vascular injury and higher union rate as compared to open-wedge osteotomy.10—13 Kiaer and Gehrchen7 and Kim et al16 reported that thoracolumbar closing-wedge pedicular resection osteotomy is effective to correct kyphotic deformity in AS and with good clinical results. In our study also, patients showed good clinical results and high subjective satisfaction. We could achieve an average angle of correction of 21.6°. Patients had significant functional improvement, as demonstrated by the improvement of JOA scores from 17.5 to 25.4 and ODI values from 41.6% to 22.2%. The VAS pain scores also improved in 6 months' postoperative time. All patients would like to have the same operation if they were asked to make the decision again.
Despite the encouraging clinical outcomes and satisfaction associated with transpedicular closing-wedge osteotomy, it is still a
W.-N. Yim et al. / Journal of Orthopaedics, Trauma and Rehabilitation xxx (2014) 1—6
major operation and not without complications. AS is an autoimmune disease, which is also associated with significant comorbidities. Patients usually have a limited chest expansion, resulting in high anaesthetic risk. Restrictive lung also increases the chance of postoperative chest infection. The operative time is expected to be long. The average operative time in our hospital is 428 minutes. Therefore, adequate padding to pressure points is important to prevent pressure sore. Prolonged prone positioning also increases the risk of ischemic optic neuropathy.14 No ischaemic optic neuropathy was seen in our cases. Significant blood loss is another major concern of this operation (Table 3 and Figure 6), which may be caused by several reasons. First of all, a large wound is required for the fusion of long segments; bleeding is expected during soft tissue dissection and surgical exposure. Second, resection of the posterior spinal column and part of the vertebral body is required for closing of the wedge; significant bleeding from the bone and epidural venous plexus is also expected. The average blood loss in our cases was 3175 mL. In one extreme case, severe blood loss (15,000 mL) occurred, which required 16 units of pack cell transfusion as well as massive fluid replacement. This was a case of pseudoarthrosis at T9, and osteotomy was performed at L2 level; hence, extensive fusion from T7 to L5 was required. The case was positioned improperly preoperatively, not allowing the abdomen to hang free. Therefore, venous return was disturbed, resulting in massive blood loss. As significant blood loss is usually expected in this type of operation, pack cells should be prepared prior to the operation. Recently, we introduced a cell saver to reduce the use of pack cells, which may reduce the amount of blood transfusion, thereby reducing the complications related to transfusion.
Comparing the operative times and blood loss of the first-time osteostomy cases, we reduced our operative time from 440 minutes to 340 minutes in the recent two cases. The blood loss remained almost static. We believe that with experience, the operative time can be further reduced.
Infection
Several factors may attribute to the chance of wound infection, including the autoimmune character of the disease, long operative time, massive blood loss, haematoma, and immunosuppressive effect of blood transfusion. No wound infection was reported in our patients, which may be related to the small number of cases.
Duval tear
Dural ectasia is common in AS patients due to the inflammatory nature of the disease7; this makes dissection more difficult and the risk of dural tear is high. Three out of 12 cases (25%) had dural tears during transpedicular closing-wedge osteotomy, the incidence of which is higher than that during other spine surgeries in our hospital (10%). For the two patients who had dural tears beyond repair,
Table 3
Blood loss and operative time in first-time osteotomy
Age (y) Gender Operative Level of Levels of Blood Operative
year osteotomy fusion loss (mL) time (min)
41 M 2003 L2 T12—L4 2600 420
41 M 2005 L3 T11—L5 2000 405
28 M 2005 L2 T11—L4 2100 363
34 M 2006 L3 T10—L5 2000 450
50 M 2007 L3 T12—L5 2000 430
32 F 2009 L3 T11—L5 1500 580
49 M 2010 L3 T12—S1 1300 355
32 M 2011 L2 T11—L4 2700 331
54 F 2010 L2 T7—L5 15,000 470
16000 14000 12000 10000 (mL) 8000 6000 4000 2000 —m 0
2003 2005 2005 2006 2007 2009 2010 2010 2011 Operative year ^^"blood loss [mL)
Figure 6. Total blood loss during operation.
we kept the drain for a longer time to observe for any CSF leak. We also advised the patients to have bed rest for a longer time if there was CSF leakage.
Neurological deficit
During closing-wedge osteotomy, neural structures are at risk.15 Nerve entrapment may occur during the closure of the wedge. Intraoperative assessment of any nerve root entrapment after closing the wedge is important. The use of somatosensory- and motor-evoked potential monitoring may further decrease the risk of nerve root entrapment. Despite the use of the above measures, we still observed transient radiculopathy in four out of 12 cases (33%). All of them have complete recovery within 3 months, without further surgical exploration. In two of them, no somato-sensory- and motor-evoked potential monitoring was performed, as we incorporated this service only in 2009. Two of them had the above monitoring, but did not show any abnormality from the monitor during operation.
Pseudoarthrosis
Among our patients, one had junctional pseudoarthrosis. We believed that in this case, the fusion was just extended to the apex of thoracic kyphosis level; the increase in stress at that level led to pseudoarthrosis. The residual kyphosis above the level of fusion was also biomechanically unfavourable to the implants, increasing the chance of pull-out. Although pseudoarthrosis finally healed spontaneously, it is recommended that fusion should not end at the apex of deformity. We also recommended a standard alignment X-ray of the spine for all patients to assess the sagittal alignment of the whole spine. Try to avoid the apex of deformity by extending the level of fusion or adjust the level of osteotomy accordingly.
Nonunion
One patient had nonunion due to technical error, which required revision. Osteotomy was made faultily through endplate into the intervertebral disc. There was inadequate bone contact, which resulted in nonunion.
Implant failure
Another concern of osteotomy in AS patients is the osteoporotic bone quality, which may increase the risk of implant pull-out or loosening. We did not have any complications related to loosening and pull-out of screws. This may be related to the higher number of fusion levels. We fused at least three levels above and two levels below, which reduced the stress at each level. Moreover, our patients were relatively young and their bone quality was expected to be better. The design of implants also accounts for the low pull-out
W.-N. Yim et al. / Journal of Orthopaedics, Trauma and Rehabilitation xxx (2014) 1—6
rate. Larger-size pedicle screws and screw heads with variable angles were used, which increased the stability of fixation and decreased the stress on implants. Furthermore, most of our patients achieved an acceptable sagittal alignment after correction. This can reduce the stress on implants as well.
Conclusion
Kyphotic deformity in AS affects patients' physical and psychosocial functions. Transpedicular closing-wedge osteotomy is an operation that can correct the malalignment. It has a good clinical outcome, with high patient satisfaction. However, it is also a major operation and is not without complications. It has a steep learning curve. The authors believe that with more experience the operative time, blood loss, and associated complications can be reduced.
References
1. Fox MW, Onofrio BM, Kilgore JE. Neurological complications of ankylosing spondylitis. J Neurosurg 1993;78:871—8.
2. Kim KT, Lee SE, Kim YW, et al. Lumbar posterior wedge osteotomy of kyphotic deformity in ankylosing spondylitis. J Korean Orthop Assoc 1997;32:1756—65.
3. Emneus H. Wedge osteotomy of spine in ankylosing spondylitis. Acta Orthop Scand 1968;39:321—6.
4. Van Royen BJ, Slot GH. Closing-wedge posterior osteotomy for ankylosing spondylitis. Clin Orthop 1985;194:142—51.
5. Jaffray D, Becker V, Eisenstein S. Closing wedge osteotomy with transpedicular fixation in ankylosing spondylitis. Clin Orthop 1992;279:122—6.
6. Camargo FP, Cordeiro EN, Napoli MM. Corrective osteotomy of the spine in ankylosing spondylitis: experience with 66 cases. Clin Orthop 1986;208: 157—67.
7. Kiaer T, Gehrchen M. Transpedicular closed wedge osteotomy in ankylosing spondylitis: results of surgical treatment and prospective outcome analysis. Eur Spine 2010;19:57—64.
8. Thomasen E. Vertebral osteotomy for correction of kyphosis in ankylosing spondylitis. Clin Orthop 1985;194:142—51.
9. PuchelJ, Zielke K. Corrective surgery for kyphosis in bekhterev's disease -indication, technique, results [Korrektur operation bei Bechterew-Kyphose. Indikation, Technik, Ergebnisse]. Z Orthop Ibre Grenzgeb 1982;120:338—42 [In German].
10. Adams JC. Technique, dangers and safeguards in osteotomy of the spine. J Bone Joint Surg Br 1952;34:226—32.
11. Weatherley C, Jaffary D, Terry A. Vascular complications associated with osteotomy in ankylosing spondylitis: a report of two cases. Spine 1988;13: 43—6.
12. Lichtblau PO, Wilson PD. Possible mechanism of aorta rupture in orthopaedic correction of rheumatoid spondylitis. J Bone Joint Surg Am 1956;38:123—7.
13. Chang KW, Chen YY, Lin CC, et al. Closing wedge osteotomy versus opening wedge osteotomy in ankylosing spondylitis with thoracolumbar kyphotic deformity. Spine 2005;30:1584—93.
14. WarmanJR, Crawford J, Evans R, et al. Dependency-induced intraocular pressure: a contributing factor for postoperative visual loss after spinal surgery. Seattle, WA: Poster presentation at the Scoliosis Research Society; September 2002.
15. Willems KF, Slot GH, Anderson PG, et al. Spinal osteotomy in patients with ankylosing spondylitis: complications during first postoperative year. Spine 2004;30:101—7.
16. Kim KT, Suk SK, Cho JY, et al. Clinical outcome results of pedicle subtraction osteotomy in ankylosing spondylitis with kyphotic deformity. Spine 2002;27: 612—8.
