POSITIONING OF PEDICLE SCREWS IN ADOLESCENT IDIOPATHIC SCOLIOSIS USING ELECTROMYOGRAPHY

Objective: To analyze the occurrence of poor positioning of pedicle screws inserted with the aid of intraoperative electromyographic stimulation in the treatment of Adolescent Idiopathic Scoliosis (AIS). Methods: This is a prospective observational study including all patients undergoing surgical treatment for AIS, between March and December 2013 at a single institution. All procedures were monitored by electromyography of the inserted pedicle screws. The position of the screws was evaluated by assessment of postoperative CT and classified according to the specific AIS classification system. Results: Sixteen patients were included in the study, totalizing 281 instrumented pedicles (17.5 per patient). No patient had any neurological deficit or complaint after surgery. In the axial plane, 195 screws were found in ideal position (69.4%) while in the sagittal plane, 226 screws were found in ideal position (80.4%). Considering both the axial and the sagittal planes, it was observed that 59.1% (166/281) of the screws did not violate any cortical wall. Conclusion: The use of pedicle screws proved to be a safe technique without causing neurological damage in AIS surgeries, even with the occurrence of poor positioning of some implants.


INTRODUCTION
The use of the pedicle screw is becoming increasingly frequent for rigid fixation in spine fusion, and is especially common in the surgical treatment of adolescent idiopathic scoliosis (AIS).4][5][6][7] Segmental instrumentation with pedicle screws enables better correction of deformities in the coronal, sagittal and rotational planes, less loss of reduction, shorter constructions, and improvement in lung function, without any increase in neurological complications, when compared with instrumentation with hooks, or hybrid instrumentation (proximal hooks and distal pedicle screws). 8,9ue to the unique anatomical, vascular and neurological characteristics around the vertebral canal, care is needed when inserting the pedicle screws, ensuring precise insertion and confirming the intraosseous position of the screws, with the aim of improving the safety of the procedure. 10][13] For this technique, motor evoked potential (MEP), somatosensory evoked potential (SSEP), free-running electromyography (EMG-FR) and stimulated electromyography (EMG-ST) are used, the latter being performed directly on the inserted implants. 14,15n patients with scoliosis, due to the rotational deformity, the risk of violation of the medial cortical or lateral wall of the pedicle is even greater, but is not always easy to spot during the surgical procedure.Intraoperative evaluation of the position of the screws with conventional fluoroscopy or radiography will assist in the detection of poorly positioned implants, although the accuracy obtained with computed tomography (CT) will be greater for this determination. 16n this aspect, EMG-ST provides additional information, establishing a link between the implants and the neural elements.
The objective of this study was to analyze the position of pedicle screws implanted during surgical treatment of AIE with the aid of intraoperative EMG-ST, with postoperative computed tomography exam.

METHODS
This is a prospective study in patients submitted to surgery by the posterior approach for correction of adult idiopathic scoliosis.After approval of the study by the Institutional Review Board (IRB) of the service responsible for the study (opinion number 533.892), patients were included after they, or their guardians, had signed the Informed Consent Form (ICF) and who had undergone a postoperative CT scan of the vertebral spine.Those with secondary scoliosis by another cause were excluded, as were patients submitted to revision surgery, and those who had not undergone a postoperative CT scan.
All the surgical procedures were performed at the same service, and by the same team, between March and December 2013.In all the surgeries, the instrumentation was exclusively by pedicle screws (all the implants were from the same supplier), inserted using the Free Hands technique. 10ll the surgeries were performed under ION monitoring, by the same team, using the same technique and the same equipment.The anesthesia used was totally intravenous, in order to interfere as little as possible in the responses of the neurophysiological tests. 17fter placement of the screws in their respective pedicles, the EMG-ST stage was performed directly on the implants.(Figure 1) For this stimulus, a monopolar electrode (cathode) and a straight subdermal needle injected inserted into the paravertebral musculature (anode) were used.The technical stimulation parameters that we used were: frequency of 3Hz, duration of 0.1ms, with stimulation intensity (milliamperes) increased to provoke the emergence of a response in the EMG. 14,15,18The maximum stimulation value used was 30 milliamperes (mA).
After surgery, the patients underwent a computed tomography exam.The positioning of the screws was evaluated and classified according to the classification proposed by Abul-Kasim et al. 19 which considers and grades the medial cortical perforation (MCP), lateral cortical perforation (LCP), and foraminal perforation (FP).(Figure 2)

RESULTS
The study included 16 patients: 12 female (75%) and four male (25%), with ages ranging from 11 to 26 years (average of 16.6 years).In all, 281 pedicles were instrumented (average of 17.5 per patient) with a minimum of 12 and a maximum of 26.The minimum follow-up time after surgery was 3 months; no patient presented any neurological complaint or alteration in the physical exam; and there was no complaint of irradiated pain in dermatomal region that would lead to a suspicion of radicular compression or lesion.
According to Table 1, in relation to the axial plane of the pedicle, 195 screws were in the ideal position (69.4%).41 screws violated the lateral cortical bone, of which 25 (8.9%) were classified as LCP1 and 16 (5.7%)as LCP2.45 violated the medial cortical bone, of which 27 (9.6%)were classified as MCP1, and 18 (6.4%)as MCP2.As illustrated in Table 2, in relation to the sagittal plane, 226 screws (80.4%) were in the ideal position, of which 48 (17.1%) violated the inferior foramen (FP1 INF) and 7 (2.5%) violated the superior foramen (FP1 SUP).Considering both the axial and sagittal planes, it was observed that 59.1% (166/281) of the screws did not violate any cortical bone (lateral, medial and upper and lower foramens), while 40.9% (115/281) presented at least a minimal degree of cortical violation in some or more than one plane.(Table 3)

DISCUSSION
The use of pedicle fixation systems in the surgical treatment of thoracolumbar spine has become increasingly common.Biomechanical studies demonstrate the superiority of pedicle fixation, in terms of the fusion rate, power of correction and early mobilization, over  fixation systems by hooks, or mixed systems. 20,21The neurological complications reported in the literature, resulting from inadequate positioning of the pedicle screws, range from 0% to 0.9%. 22,23In the present study, no complications were observed during the entire patient follow-up.
2][13] The persistently electrified pedicle stimulation instrument technique, EMG-ST, enables monitoring and identification of perforations of the pedicle wall, before a lesion of the neural root occurs. 11,24Obtaining responses to EMG-ST with thresholds lower than 4 or 5 mA is strongly suggestive of perforation of the cortical bone by transpedicular instrumentation, while a response with thresholds higher than 15 mA is indicative of adequate positioning. 25n relation to pedicle instrumentation by the Freehand technique, there is a fear of neurological damage, especially when the cortical wall of the medial pedicle is violated. 10However, small violations are common, and are generally asymptomatic.Numerous publications demonstrate that violations ≤2mm are harmless, 7,10,21 hence the term "questionable safe zone" is attributed to violations between 4 and 8 mm. 10 The occurrence of poorly positioned screws in the pedicle varies between 3% and 44%. 22In Brazil, De Marco et al. 26 studied the position of pedicle screws inserted without the use of EMG in 24 patients, but 183 pedicles -that study did not determine a specific pathology as in our study, with all the patents having AIE -and observed that 36.06% of the screws present a lesion of the cortical wall of the pedicle.In our study, even with the aid of ECG, misplaced positioning was observed, with at least one pedicle cortical wall being violated, in 40.9% of the screws.However, we considered only patients submitted to surgery for correction of AIS, in the presence of vertebrae with rotational and structural deformity.
In the present study, there was perforation of the lateral cortical wall of the pedicle in 14.6% of the screws.These values agree with what is described in the literature. 2,10,26,27There was no case of vascular, neurological or visceral damage.In relation to the medial cortical wall of the pedicles, values of between 1.4% and 14% of inadequate positioning are described, 2,10,26,27 reaching up to 28% according to Farber et al. 28 In the present study, there was perforation of the medical cortical in 16% of the screws.
Table 1.Description of the positions of the screws in the axial plan for each patient, and in the total pedicles evaluated.Table 2. Description of the positions of the screws in the sagittal plane for each patient, and in the total pedicles evaluated.

Figure 1 .
Figure 1.Electromyography stimulated through the electrode positioned directly in the screw implanted in the pedicles.

Table 3 .
Description of the screws in the ideal position in the sagittal and axial planes (FP0 and LCP/MCP0) and in inadequate position in some of the planes (FP ≠0 or LCP/MCP ≠0).