AIS is the most common type of spinal deformity treated by surgical fixation. Its onset is usually insidious and its complications may be deadly   . Instrumental correction of scoliotic deformities has progressed from Harrington rods, to Cotrel-Dubousset (CD) instrumentation, to the more recent advent of thoracic pedicle screw fixation  -  . Harrington rod application improved coronal deformity yet, sagittal deformity was almost associated flat-back syndrome    . The use of CD instrumentation improved outcomes, but correction of axial rotation remained questionable     . Recently, the application of pedicle screws-used in all patients of this study-has improved upon many of the limitations from the preceding constructs and provided good maintenance of correction in the short-term  -  .
Suk et al.  reported outcomes in 78 patients treated with surgical fixation by pedicle screws and reported improved correction rate at 2 years follow up when compared to hybrid constructs. Other studies showed pedicle screws to provide best immediate postoperative correction of coronal curves, planar translation, and rotation    . Good maintenance at 2 years has been reported in both fused and compensatory curves, as well as a reduction in the number of segments fused and a lower revision rate associated with pedicle screw constructs   .
Although long-term radiological outcomes have been reported for earlier spinal constructs, the radiological evaluation of spinal instrumentation with pedicle screws were mainly based on short follow-up studies with scarce longer term results      . Kim et al.  compared hybrid and screw constructs with 5-year minimum follow-up in proximal junctional kyphosis mainly. This study included a large consecutive number of patients (100) to evaluate the long term radiographic outcomes of the recent spinal instrumentation trend in the management of AIS with correlation to the possible clinical scenarios.
2. Materials and Methods
This study was conducted in Sultan Bin Abdulaziz Humanitarian City (SBAHC), Riyadh, KSA, where 100 AIS patients with spinal pedicle screws fixation were included, after obtaining the informed consents and approval of the SBAHC ethical board. Cases were tracked through the SBAHC Health Information System (HIS) database. Basic demographic data of all patients was retrieved from medical record files. Patients were followed for 5 years (from October 2012 to October 2017) with average duration of follow-up was 3.7 years for all cases.
Radiological examinations included total spine x-rays, CT scan and MRI at regular 2 years and 5 years intervals. CT scans with 3D reconstruction were performed on a 64-slice CT scanner (Ingenuity; Philips Medical Systems). Parameters used were: Helical CT protocol, Iterative reconstruction: low dose, iDose Level 3 (range, 1 - 5), slice thickness 0.5 mm, slice interval 0.5 mm, matrix 512 × 512, collimation 64 × 0.625 mm, OMAR software, bone window (YD kernel, level 800 HU, width 2000 HU) and soft window (Soft kernel B). The degree of metallic artifacts and clarity of the scanned vertebrae of the 3D reconstruction images using OMAR were evaluated to determine their diagnostic acceptance, only those of satisfactory subjective quality were included.
Radiological evaluation was done by measuring the Cobb angle on the anteroposterior radiographs usually performed immediately postoperative. Preoperative Cobb angle was then compared to the postoperative Cobb angle to calculate the percentage correction rate. This process was repeated and recorded with each follow up visit. Clinical scenarios and complications were divided into groups including good response and maintenance, infection, surgical revision, pseudarthrosis, neurological deficits, instrumentation-related as screw loosening or complete hardware failure. All complications were managed successfully by surgical revisions, physiotherapy and medications.
Image quality evaluation was conducted on the source bone and soft tissue windows as well as the 3D reconstruction images. A blinded consultant radiologist, with more than 15 years of experience in radiological diagnosis, conducted the subjective evaluations, including severity of the spinal metallic artifacts, density of the related paraspinal muscles, clarity of the inter-muscular fat planes, and definition of the scanned vertebral elements using a 5-point scale, Table 1  . Images scoring 3 or more were considered radiologically satisfactory.
Statistical analysis was performed using SPSS 12.0.2® (IBM, Armonk, NY) software using Student t tests, Chi square tests and Fisher’s exact tests. Results
Table 1. Radiological evaluation of OMAR reconstructed image quality.
were reported as average and percentage as number of study group was 100. A “p” value of 0.05 was considered statistically significant.
A total of 100 patients matched the inclusion criteria, 24 (24%) were females and 76 (76%) were males. The average age at operation was 16.5 years and the average duration of follow-up was 3.7 years for all cases. Sixty seven (67%) patients showed good correction and maintenance of the spinal curves over the follow up periods at 2 and 5 year intervals, where the mean Cobb angle was 68˚ preoperatively and 21˚ postoperatively (Figure 1). The calculated correction rate was 74%
(a) (b) (c) (d)
Figure 1. (a)-(d) Maintained correction over 5-year period with no complications. (a) Preoperative whole spine PXR AP view showing marked smooth left dorso-lumbar scoliosis. Cobb angle = 45˚; (b) immediate post-operative scan with good correction and satisfactory alignment. Cobb angle = 9˚; (c) 2-year follow up scan showing maintained correction with no significant change in the Cobb angle measurement; (d) axial CT scan (bone window with OMAR software) at 5-year interval showing optimum screw positioning with no evidence of loosening or pull-out.
in this group. Thirty three (33%) patients showed complications including postoperative infection (13 cases) (Figure 2), surgical revision (4 cases), pseudarthrosis (3 cases), neurological deficits (5 cases) (Figure 3, Figure 4), screw loosening, malpositioning and pull-out (7 cases) (Figures 5-7) and complete
Figure 2. Post-operative infection with screw loosening. Axial CT scan (bone window with OMAR software) showing marked osteolysis due to severe post-operative infection with marked loosening of both fixing screws.
Figure 3. Neural injury. Axial CT scan (bone window with OMAR software) showing the left malpositioned screw markedly encroaching upon the spinal canal.
Figure 4. Screw malpositioning with direct cord and nerve root injury. Axial CT scan (bone window with OMAR software) showing the right malpositioned screw at the level of the exit foramen with direct neural injury.
Figure 5. Screw pull-out. PXR AP view of the dorsal spine, 1-year post-operative showing pull-out and loosening of the right upper fixing screws.
hardware failure (1 case) (Figure 8); in whom the preoperative mean Cobb angle was 56˚ compared to the patients with no complications (p = 0.288). In males, there were postoperative complications in 27 out of 76 (35.5%) and in females 5 out of 24 (21%), (p = 0.027) (Table 2).
Complications took place before the 2-year follow up evaluation in all patients and were managed effectively, resulting in no significant changes between the 2-year and 5-year assessments. Postoperative infection represented almost half of this group and was caused by gram positive organisms in all cases, 12 cases
Table 2. Patient information.
Figure 6. Screw loosening. Sagittal CT scan (bone window with OMAR software) showing complete and partial loosening of the proximal right 5th and 6th row screws respectively.
were treated successfully over a mean period of 9 months and one case progressed to complete vertebral osteolysis with large epidural abscess and intraspinal extension, treated by antibiotics and requiring drainage of a deep infection (Figure 2). One patient had malpositioned screw with high risk of aortic injury (Figure 7), which required revision operation. Four patients complained of post- op radicular symptoms and sensory deficits that did not require surgical revision. One patient had a neurological deficit in the form of weak right knee
Figure 7. Vascular injury with screw malpositioning. Axial CT scan (bone window with OMAR software) showing direct injury of the descending thoracic aorta by a malpositioned screw.
Figure 8. (a) (b) Hardware failure. (a) PXR AP view 17 weeks post-operative showing rod and screw dissociation with failed fixation; (b) Axial CT scan (bone window with OMAR software) showing separation of the metallic components.
flexion postoperatively. CT scan was done for this patient and found to be a misplaced screw encroaching upon the corresponding nerve root at the exit foramen (Figure 4), which was subsequently revised. This patient had almost complete neurological recovery at the 5-years follow up visit. One patient had multiple screw pull outs (Figure 7) and one had complete hardware failure and the fixation was removed with replacement (Figure 8). The other cases were treated by physiotherapy with medications (Figure 9).
Figure 9. Patient outcome.
Although the treatment of AIS by spinal instrumentation using pedicle screws is well known in many countries all over the world, this is the first study of its kind in the kingdom of Saudi Arabia, with this large cohort and long term follow up evaluation. Many previous studies comparing earlier instrumentation systems have included the radiological long term outcome evaluation in AIS while the more recent pedicle screws had fewer data to review   . Other case studies reported slight increases or progression of radiographic parameters, but most have displayed evidence of correction maintenance with increased follow up periods  . Suk et al.  reported their outcomes on a study group of 203 patients having pedicle screw fixation with at least 5-year follow up, and noted a loss of 3% correction of the main corrected curve over time. Other studies have also reported similar results, but had limited study groups like Hwang et al.  who assessed pedicle screw fixation in 57 patients with 5-year follow up and noted minimal loss of correction from the immediate post-op (2%), but did not comment on changes took place at the 2-year follow up. Di Silvestre et al.  compared hybrid and pedicle screw constructs with a mean follow up of 6.7 years and found good maintenance of correction with screws.
Our calculated correction rate was 74% in this study and this comes in agreement with other similar yet smaller cohort studies  . We observed that there was no correlation between percentage correction and postoperative clinical scenarios (Table 2). About third of patients showed complications in whom the preoperative mean Cobb angle was 56˚ compared to the patients with no complications (p = 0.288). Therefore, those with more correction didn’t necessarily have better postoperative outcomes. This was the same conclusion by Akil and Riaz  , who stated that the surgeon should only correct to a point that is safe to do so, and correction does not yield better results and may predispose the young patients to postoperative complications, especially neurological injury.
The clinical application of OMAR for post-spinal instrumentation was performed on all patient population coming for regular postoperative evaluations in SBAHC, one of the most prestigious rehabilitation centers in the Kingdom of Saudi Arabia with two international accreditations of CARF and JCI and one national accreditation of CBAHI. All cases were scored 3 or more and were accepted by the radiologist (Table 1), which confirmed the feasibility of OMAR in spinal instrumentation imaging. The degree of metallic artifacts and the definition of the adjacent vertebrae were all evaluated 4 and 5 points, impressive of little metallic artifacts with no obvious distortion. With the resulting clear 3D spinal reconstructions, the spinal surgeons could be sure about the relationship between vertebrae and the fixing hardware (e.g., the proper placing and depth of transpedicular screws, no injury of the related nerve roots, no loosening or malplacement), evaluate the status of vertebrae surrounding the implant for osteoporosis or infection, and follow up the healing process of fractures. Thus, this technique would help accurate postoperative assessment of spinal fixation of different reasons. To have a complete evaluation of this OMAR technique, it would be of interest to perform a quantitative image analysis of CT numbers as a complement to the qualitative analysis carried out in this study.
Five-year follow up of 100 Saudi patients with AIS treated with spinal instrumentation showed good improvement and maintenance of the corrected spinal curves in two thirds of patients with the regular interval follow up visits. One third of patients in this study showed variable minor and major complications with good overall treatment response at the 5-year interval. Low dose CT scan with the use of orthopedic metal artifact reduction software provided an effective modality for evaluation and follow up.
The author would like to thank the SBAHC Biomedical Statistics Department for performing the statistical analysis in this work.
This study was approved by the SBAHC Ethical Board.
Conflicts of Interest
The author declares that he has no conflicts of interest.
Amir MA selected the subject, reviewed the literature, wrote the paper, selected the images and edited the whole work.