POSTOPERATIVE RADIOGRAPHIC CONTROL OF INSTRUMENTATION WITH THORACIC PEDICLE SCREWS IN ADOLESCENT IDIOPATHIC SCOLIOSIS

Objective: To evaluate the efficiency of intraoperative and postoperative radiographs of the dorsal pedicle screw placement and its relationship with computed tomography (CT). Methods: Thirty six patients with adolescent idiopathic scoliosis (AIS) were evaluated. Three different observers classified the placement of screws through radiographs. The responses were compared with CT scans. Results: A total of 280 pedicle screws were evaluated. The interobserver assessment was 0.56 (moderate match). The assessment of intraobserver matching was 0.79 (good match). No vascular or neurological lesions have been observed. Conclusions: According to the results of intra and inter-observers, the reproducibility of postoperative radiograph is not very effective to diagnose malposition of spinal pedicle screws. Computed tomography showed very useful to determine the position of the spinal pedicle screws .


INTRODUCTION
Transpedicular instrumentation, initially combined with plates, was originally used in the 1970s for spinal fixation of lumbar fractures. 1From 1986, it began to be used for surgical corrections and stabilization of deformities, with hybrid constructions, and from 1994, it was used as a complete construction of pedicular screws. 2,3[6][7][8][9][10] Thoracic instrumentation has potential dangers, due to the risk of neurovascular complications due to malpositioned screws, therefore, it is very important to have a diagnostic method that recognizes this risk.Radiographies and tomographies are the diagnostic methods currently used for its control.
The objective was to evaluate the efficacy of intra-and postoperative radiography to assess the position of thoracic transpedicle screws in patients with idiopathic scoliosis.

MATERIAL AND METHODS
The study included a total of 36 patients with adolescent idiopathic scoliosis (AIS) treated with posterior transpedicular instrumentation.The pattern of the scoliotic curve was classified according to the study published by Lawrence Lenke in 2001. 11e analyzed a total of 280 thoracic pedicle screws (TPS), inserted free hand by the technician, 12 with a radioscopic guide in the anteroposterior and lateral planes.Polyaxial titanium screws of 4.5 and 5.5 mm in diameter were used, with lengths ranging, depending on the vertebral level, from 30 to 40 mm.In all the surgeries, continuous intraoperative electrophysiological monitoring was performed, using motor evoked potentials (MEP) and somatosensory evoked potentials (SSEP).
In the second postoperative week, on an outpatient basis, the following images were taken: 1) X-ray of the spine-anterior-posterior and lateral views; and 2) computed tomography of thoracic spine (in bone window, with thin 2 mm slices taken at the widest axis of each screw, with multi-plane reconstruction).
The X-rays of the spine (immediate postoperative and at 2 weeks) were shown to three orthopedists of the spine, at three separate times, one month apart, to analyze the intraobserver variable.The orthopedists were asked to evaluate the position of the thoracic pedicle screws as either inside, outside, or uncertain.
Tomographic classification: The radiographic data were compared with the tomographic images of each screw.The position of the screw was classified based on the work of Bride Upendra of the All India Institute of Medical Sciences (AIIMS) group, 13 which makes the following divisions, for evaluating the position of thoracic screws: Type I: position acceptable (Figure 1): A -the screw is completely within the pedicle; B -the screw is perforating some pedicular cortex, but part of it is still within the pedicle; Cthe screw is penetrating less than 2 mm through the pedicular cortex; D -the screw is located in the costovertebral unit (screw insideoutside-inside).Type II: placement unacceptable; those that do not correspond to Type I and that do not show signs of neurovascular compromise.(Figure 2) Type III: screws that are causing neurovascular injury.(Figure 3)

Statistical Analysis
The data collected were analyzed using the kappa method, according to the interpretation by Landis and Koch, who propose margins for assessing the number of agreement based on the kappa method.The evaluations were divided into six groups.(Table 1)

RESULTS
A total of 36 patients with idiopathic scoliosis were analyzed: thirty-one women and five men.The average age at the time of surgery was 16.7 years (range 11-19 years).The distribution of Table 1.The data collected were analyzed using the method of Kappa, with interpretation by Landis and Koch.scoliotic curve patterns, according to Lenke, was: 1-12; 2-1; 3-13; 4-1; 5-3 and 6-6.The average Cobb angle of the curves was 58.3°.The total number of thoracic screws analyzed was 280.Radiological evaluation: In the evaluation of the orthopedists (Observers 1, 2 and 3), for screws inside the pedicle was an overall average of 163.3 (58.5%).For the screws outside pedicle, the orthopedists found an overall average of 65 screws (23.2%).And finally, for those whose position was uncertain, they found an overall average of 51 (18.3%).(Table 2) Tomographic evaluation: The results of the tomographic evaluation are shown in Table 3.

Agreement
The interobserver evaluation was 0.59 (moderate agreement) and the intraobserver evaluation was 0.79 (good).(Figure 4) Complications.No vascular or neurological complications were recorded.There were two cases of pull out in high thoracic screws which coincided in the type 2 pattern.

DISCUSSION
.15 Study of pedicle instrumentation has been focused on its effective means of placement, which occurs when the screw is located completely within the limits of the pedicular cortex.][18] Some works in the literature have reported the safety of this method in frontal deformities, 3,5 -7,19 beginning with Suk et al. who report 47 patients treated for idiopathic scoliosis with pedicle screws and hybrid constructions, although they do not show the absolute number of thoracic screws, and a total of 3.0% incidence of poorly-positioned screws in the X-ray, without neurological complications . 10The work of Liljenqvist, 20 based on postoperative evaluation with tomography, evaluates 120 thoracic screws in patients with idiopathic scoliosis.Of these 8.3 % penetrated the medial cortex by an average distance of 1.5 mm, without neurological complications.The study of Belmont et al. 21reports 279 thoracic screws, with 57% of the screws completely contained in the pedicle, with both medial cortical (14%) and lateral (29%) penetration, and no neurological complications.It is important to recognize that in these publications, minor infringements of the thoracic pedicle screws, although not ideal, are considered perforations within an acceptable range.

CONCLUSION
Based on the intra-and interobserver results, the use of intraoperative and postoperative radiography did not demonstrate great effectiveness in the diagnosis of misplaced thoracic pedicle screws.Tomography showed great usefulness for determining the position of the screws.
All authors declare no potential conflict of interest concerning this article.

Figure 2 .
Figure 2. Type II placement unacceptable, without injury to vital structure.A, 2 mm medial perforation.B, perforation of the anterior cortex.

Figure 3 .D
Figure 3. Type III placement unacceptable, without injury to vital structure.A, spinal cord injury.B, Screw impinging the aorta.

Table 2 .
Results of the radiological evaluation.

Table 3 .
Results of the tomographic evaluation.