Comparison of robot versus fluoroscopy-assisted pedicle screw instrumentation in adolescent idiopathic scoliosis surgery: a Meta-analysis Authors: Yifan Kang, Lindong Wei, Jianrong Liu Published: 2024 ABSTRACT AbstractAimTo investigate the safety and accuracy of robot-assisted pedicle screw placement in adolescent idiopathic scoliosis (AIS) surgery.MethodsPubMed, Medline, Embase, Cochrane Library, Web of Science, CNKI, and EMCC databases were searched for studies utilizing the clinical data of adolescent idiopathic scoliosis patients (robot-assisted and fluoroscopy-assisted group) who underwent posterior pedicle screw fixation to collect preoperative, postoperative and follow-up indexes and imaging parameters, including screw implantation accuracy, the number of intraoperative fluoroscopy times, amount of bleeding, operation time, postoperative VAS score and postoperative Cobb angle.ResultsSix studies were included in this meta-analysis, including 154 patients in the intervention group and 171 patients in the control group. Our results showed that screw implantation accuracy (WMD 1.07, 95% CI (1.01,1.14), P = 0.03) was higher than that in the control group. The number of intraoperative fluoroscopy times (WMD -6.11, 95% CI (-9.60, -2.62), P = 0.0006) and operation time (WMD 31.52, 95% CI (5.70,57.35), P = 0.02) in the robot group were less, and the differences were statistically significant. Intraoperative blood (WMD -59.05, 95% CI (-212.81,94.70), P = 0.45), postoperative VAS score (WMD -0.07, 95% CI (-0.24,0.11), P = 0.46), preoperative Cobb angle (WMD 0.79, 95% CI (-1.12,2.69), P = 0.42) (Fig. 7), postoperative Cobb angle (WMD -0.25, 95% CI (-1.12,0.62), P = 0.57) (Fig. 8) had no statistical significance in the robot group compared with control group.ConclusionThe robot navigation system can improve screw implantation accuracy and reduce the operation time and radiation exposure during the operation.Keywords: Adolescent idiopathic scoliosis, Robot, Pedicle screw, Meta-analysis Introduction In 2003, the Spine Assist robot system developed by Mazor Corporation of Israel was the first to realize the clinical application of spine surgery. Subsequently, a number of robots have been developed to continuously improve the technology and realize visual navigation. Although the use of orthopedic robots in China started late, it has developed rapidly in recent years. The application of the navigation system of the surgical robot for accurate positioning and accurate screw placement in the three-dimensional space has shown great potential in the correction of spinal deformity [1]. Adolescent idiopathic scoliosis (AIS) is the most severe and common spinal deformity, which refers to the complex three-dimensional deformity of one or more segments of the spine bending to the lateral side accompanied by vertebral rotation [2], and the prevalence of childhood population is 0.5%-5.2% [3, 4]. Pedicle screw is the main instrument used in AIS correction surgery. Robot-assisted spinal surgery seems to be more accurate than free-hand technique in pedicle screw placement [5], less soft tissue injury [6], and simple operation [7], reducing the risk of secondary surgery [8], so it has been gradually widely used. Croissant et al. [9] used a robot-assisted system for spinal correction surgery, and the results showed that the robot-assisted device had high accuracy and postoperative safety. However, screw malposition has been frequently reported, with an incidence of 20% to 30% [10, 11], and the risks of malposition include nerve injury, visceral injury, and the potential need for revision surgery [12, 13]. A large study of 181 AIS patients found that after an average follow-up of 8.5 years (range 3–18 years), 51 patients underwent 54 revision surgeries, with the most common reason for revision being implant misalignment/implant failure (37%), followed by poor wound healing/infection (23%) [14]. For patients with severe scoliosis and vertebral rotation or small pedicle, the traditional method of screw placement is difficult and complicated, and it is easy to cause screw misplacement [15]. Conservative treatment for AIS also plays a certain role, but it cannot fundamentally change the Angle of spinal curvature [16]. Therefore, how to reduce and prevent the serious incidence of pedicle screw misplacement and neurovascular complications is still a challenge for spine surgeons. This study aims to evaluate the accuracy and safety of robot-assisted pedicle screw placement in adolescent idiopathic scoliosis (AIS) surgery. The purpose of this study is to reduce the risk of surgery, the occurrence of postoperative complications and promote the rapid recovery after surgery. Materials and methods A research protocol was registered through PROSPERO: International Prospective Register of Systematic Reviews (CRD42023445711) and completed conforming to the Preferred Reporting Items for Reviews and Meta-Analyses (PRISMA) guidelines for systematic review. Literature search Studies were identified through a systematic literature search on online databases: PubMed, Medline, Embase, Cochrane Library, Web of Science, CNKI, and EMCC. We performed an electronic database search for full-text articles and published abstracts from the inception of each database to April 2023. We did not limit the search by language, geographic origin, date of publication, or study type. For database searches, the following main keywords were the following text words: “Adolescent Idiopathic Scoliosis” OR “AIS” AND “Robot” AND “Pedicle Screw”. Inclusion and exclusion criteria. All available studies included AIS patients aged 10–19 years who underwent corrective surgery. Inclusion criteria: (1) User posterior correction and internal fixation; (2) patients with a main curve Cobb angle more than 40°that needed posterior correction surgery Surgical treatment can be tolerated; (3) Divided into Robot group and control group; (4) Sufficient data. Exclusion criteria: (1) Patients with prior spinal surgery; (2) Other types of scoliosis, such as congenital scoliosis, neuromuscular scoliosis and syndrome scoliosis; (3) Patients with severe cardiovascular and cerebrovascular diseases; (4) Combined with severe spinal cord injury or nerve injury. Data extraction Data collected consisted of basic demographic information including age, sex, and Lenke types. Indexes and imaging parameters analysis was performed screw implantation accuracy, the number of intraoperative fluoroscopy times, operation time, intraoperative blood, postoperative VAS score, preoperative and postoperative Cobb angle. VAS score is a commonly used method to assess pain intensity. Its operation method is simple, usually draw a straight line on a piece of white paper, one end of the line is marked with the number 0, representing that the patient does not have any pain; The other end is marked with the number 10, which means excruciating pain, unbearable. Depending on how much pain they feel, the patient chooses a point on this line to indicate their level of pain. Cobb Angle is a standard measure used to assess and track the severity of scoliosis. The Cobb Angle is usually measured on an X-ray by identifying the upper and lower end vertebrae of the lateral bend and drawing the corresponding vertical line, the intersection of the two lines is the Cobb Angle. Quality assessment and statistical analysis All meta-analyses were performed using Review Manager 5.3 (Cochrane Collaboration, Oxford, UK). Continuous and dichotomous variables were analyzed using weighted mean differences (WMDs) and risk ration (ORs) with 95% confidence intervals (CIs) respectively. The statistical heterogeneity was quantified using the I2. The random-effects model was used if there was heterogeneity between studies (I2 > 50%); otherwise, the fixed-effects model was used (I2 < 50%). Results Study selection and data extractions All studies met the inclusion criteria and were subsequently reviewed and analyzed. The detailed study selection process is documented in the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flowchart. The search strategy is illustrated in Fig. 1. From the literature search, 78 abstracts of studies were retrieved and independently screened for inclusion. The information extracted included study general information (title, author and year), study characteristics (type of study design), and number of cases (Table 1.). We excluded 64 articles by reading the abstracts for any one of the following reasons: same articles or articles with unavailable data and incomplete outcome indicator. 6 full-text articles were reviewed for inclusion. 7 papers were ineligible for the following reasons: 4 papers without same trials, 1 paper with inappropriate subjects, 1 paper had small cases and 1 paper did not provide complete data. Ultimately, 6 studies met the screening criteria and were included in this meta-analysis (Fig. 1). Literature screening flow chart and results Table 1General characteristics of the included studiesStudyCountryresearch typecasesagemale/femaleLenke types1AN1A+3CN1BN5CNRAFARAFARAFARAFARAFARAFARAFARAFAXin 2023 [17]Chinacase-control182215.17±1.8215.59±1.973/155/1771011223257Hou 2023 [18]ChinaRCT455614.69±1.9314.49±2.0132/1338/18----------Linden 2022 [19]Switzerlandcohort303015.0±2.0115.3±1.96/176/17----------Cao 2021 [20]ChinaRCT151516.08±2.3116.16±2.269/610/51011------54Chen 2021 [21]Chinacase-control222413.914.14/185/197922232262Shao 2017 [22]ChinaRCT242414.7±3.415.3±3.04/203/21---------- General characteristics of the included studies The authors independently implemented the Newcastle–Ottawa Scale Assessment Scale (NOSSA) and Cochrane Risk of Bias Assessment Tool to assess for the biases Tables 2, 3and 4. Consequently, the quality of evidence for this study is high. Table 2Results of Bias risk assessment in included case–control studiesStudySelectionComparabilityExposureScoresAdequate definition of casesRepresentativeness of casesSelection of controlsDefinition of controlsControl for important factorAscertainment of exposureSame methods of ascertainment for cases and controlsNon-response rateXin 2023 [17]111121119Shao 2017 [22]111121119 Results of Bias risk assessment in included case–control studies Table 3Results of Bias risk assessment in included cohort studyStudySelectionComparabilityExposureScoresAdequate definition of casesRepresentativeness of casesSelection of controlsDefinition of controlsControl for important factorAscertainment of exposureSame methods of ascertainment for cases and controlsNon-response rateLinden2022 [19]111121108 Results of Bias risk assessment in included cohort study Table 4Results of bias risk assessment included in randomized controlled studiesStudyRandom sequence generationAllocation concealmentBlinding of participants and personnelBlinding of outcome assessmentIncomplete outcome dataSelective reportingOther biasHou 2023 [18]Low riskUnclear riskUnclear riskLow riskLow riskLow riskUnclear riskCao 2021 [20]Low riskUnclear riskUnclear riskLow riskLow riskLow riskUnclear riskShao 2017 [22]Low riskUnclear riskUnclear riskLow riskLow riskLow riskUnclear risk Results of bias risk assessment included in randomized controlled studies Outcome comparison Six studies were included in this meta-analysis, including 154 patients in the intervention group and 171 patients in the control group. Our results showed that screw implantation accuracy (WMD 1.07, 95% CI (1.01,1.14), P = 0.03) (Fig. 2) was higher than that in the control group. The number of intraoperative fluoroscopy times (WMD -6.11, 95% CI (-9.60, -2.62), P = 0.0006) (Fig. 3) and operation time (WMD 31.52, 95% CI (5.70,57.35), P = 0.02) (Fig. 4) in the robot group were less, and the differences were statistically significant. Intraoperative blood (WMD -59.05, 95% CI (-212.81,94.70), P = 0.45) (Fig. 5), postoperative VAS score (WMD -0.07, 95% CI (-0.24,0.11), P = 0.46) (Fig. 6), preoperative Cobb angle (WMD 0.79, 95% CI (-1.12,2.69), P = 0.42) (Fig. 7), postoperative Cobb angle (WMD -0.25, 95% CI (-1.12,0.62), P = 0.57) (Fig. 8) had no statistical significance in the robot group compared with control group. Forest plot of the accuracy of screw placement between RA group and FA group Forest plot of number of perspective between RA group and FA group Forest plot of operation time between RA group and FA group Forest plot of amount of bleeding between RA group and FA group Forest plot of postoperative VAS score between RA group and FA group Forest plot of preoperative Cobb angle between RA group and FA group Forest plot of postoperative Cobb angle between RA group and FA group Sensitivity analyses and publication bias Sensitivity was performed by excluding and calculating each study from the meta-analysis one at a time and counting the pooled effect size to determine the influence of each study. Publication bias was screened using funnel plots. A P < 0.05 was considered statistically significant. An example is indicated by sensitivity analysis showing the funnel plot of the accuracy of screw placement reported in this meta-analysis for RA group and control group (Fig. 9). After the heterogeneity test, any study could be omitted without any significant impact on the overall statistical significance, indicating that the results of this meta-analysis were stable. In addition, the shape of the funnel plot was symmetrical, indicating no publication bias in our study. Risk of publication bias in the included literature Discussion Surgical treatment should be considered for patients with progressive development or severe scoliosis (Cobb Angle ≥ 40°) of AIS. At present, posterior pedicle screw fixation has gradually become a common surgical fixation method, which can provide patients with good fixation effect and satisfactory coronal correction [23]. Although free-hand screw placement can be completed, free-hand screw placement mainly depends on the pedicle and the anatomic structures such as the zigzag ridge, transverse process and articular process to determine the entry point of screw. In AIS, the pedicle diameter is small, and severe variation of spinal anatomical structure often occurs. For example, asymmetric development of spinal canal and pedicle, relative stenosis of pedicle, vertebral rotation, pedicle deformation, spinal cord deviation, etc. [15, 24, 25], not only increase the difficulty of determining the correct entry point through anatomical landmarks, but also easily lead to screw malposition, which affects the accuracy of screw placement and affects the orthopedic effect. Even the screw breaks through the bone cortex, causing compression and injury of the nerve, spinal cord and aorta [26]. Therefore, it is difficult to place the screws accurately by hand. At the same time, the tapered pedicle on the concave side increases the risk of fracture and screw pullout during screw placement on the concave side [27]. Researchers have found that the incidence of pedicle screw malposition in patients with thoracic scoliosis is as high as 6.0%-54.7 [28], and due to the abundant blood vessels and nerves on the thoracic side, vascular and nerve injuries are more likely to occur. In addition, scoliosis surgery often takes a long time, and surgeons are prone to fatigue and lack of concentration, which increases the number of fluoroscopy exposures. Robot-assisted surgery, an emerging surgical technique in spine surgery in recent years, has been one of the most robust methods to improve the accuracy of screw placement and reduce the risk of potential neurological complications [29, 30]. At present, the robot navigation system has been widely used, which can show more details and three-dimensional complex structures during the operation, obtain real-time information comprehensively, adjust the direction and depth operation of the inserted screws at any time [31], and reduce the incidence of screw misplacement, neurovascular injury, thoracic cavity damage and other serious complications [32]. It is easy to operate and greatly reduces the operation time. This study showed that robot-assisted AIS surgery can improve the accuracy of pedicle screw placement. MOLLIQAJ et al. [33] also found that robot-assisted spinal surgery is safer and more accurate than traditional spinal surgery, and the accuracy of robot screw placement can reach 92%-98% [34]. The results suggest that the surgical effect of robot navigation is significantly better than that of conventional operation. AIS patients have a unique pedicle structure, including pedicle dysplasia, vertebral rotation [35, 36], thin pedicle, and spinal cord deviation [15], which are prone to screw misplacement. Therefore, accurate positioning during operation cannot be ignored, thus reflecting the superiority of three-dimensional conditions. The results of Blas et al. [37] and Abul-Kasim et al. [38] and Khan et al. [38] also support the conclusion of this study. The results indicate that with the application and proficiency of robot-assisted pedicle screw placement in orthopedics, robot-assisted pedicle screw placement will be safer and more accurate than traditional manual pedicle screw placement. It can reduce the radiation exposure during the operation and avoid the failure of screw insertion caused by fatigue or physiological tremor. The pedicle diameter on the concave side of the apical vertebral region is small, and the vertebral body rotation is severe, which significantly increases the risk of screw insertion [39]. Zhang et al. [40] reported that the accuracy of safe screw placement in the apical vertebral region in the navigation group was significantly higher than that in the free-hand group, and the number of screws placed in the apical vertebral region was significantly increased. The number of screws implanted is very important for vertebral derotation in the apical region [41], and more anchors can achieve good curvature correction and greater derotation rate [42]. This study shows that the advantages of intraoperative navigation-assisted screw placement are even more prominent in the face of complex deformities and severe anatomical variations. This study also found that the number of fluoroscopy and operation time under robot navigation were significantly reduced, which greatly reduced the radiation exposure of patients during operation. If the operation time was too long, there would be metabolic acidosis, decreased coagulation function, and the risk of thrombosis. Prolonged intubation leads to ischemia and even injury and necrosis of the tracheal mucosa. The incidence of these risks is significantly reduced with robotic assistance. The small number of studies that observed the number of screw adjustments among the included studies prevented a comprehensive evaluation, but intraoperative blood loss was described. If The Times of screw adjustment are too many, the amount of bleeding cannot be ignored. Among them, Chen's study pointed out that the number of adjustments to the implant was reduced in the robot group, so the amount of intraoperative blood loss was significantly less than that in the manual group. However, there was no significant difference in intraoperative blood loss between the two groups, which may be related to the different times of screw adjustment and the different experience of surgeons. Cobb Angle is an important index to evaluate scoliosis. Meta analysis of the combined effect size of 4 literatures showed that the surgical methods of the two groups were no significantly different. In addition, there was no significant difference in VAS score between the robot group and the freehand group. Given the lack of long-term follow-up, the number of included outcomes is insufficient to fully address the advantages of robotic assistance. Most of the studies included in this paper mainly used "Tianji" robot system and Renaissance robots, and the observed benefits of robot assistance are mainly reflected in the changes of various intraoperative indicators. Compared with the high cost, expensive and complex operation of robots, which require regular repair and maintenance, this study has not clearly shown that robot-assisted treatment is indeed better, but in terms of cost performance, it may not have advantages over ordinary radiation therapy. However, for young orthopedic surgeons, such a difficult operation of the spine may not be learned all at once, and experience in more than 46 surgeries is required to acquire the surgical proficiency of trained traditional open scoliosis surgeons [43]. And the robot has a learning curve, the more times it is used, the more accurate the instrument becomes. Using robots will increase the success rate of surgery, which seems to be worthwhile for the patient. In short, the advantages of robotic surgery have been initially demonstrated, but the cost is high, and it is still a challenge to evaluate the advantages and disadvantages and choose a better surgical method. Surgeons need to carry out professional preoperative evaluation to provide patients with practical and beneficial treatment plans and improve their life happiness. Limitations There are some potential limitations that need to be addressed. The included studies did not distinguish the division of pyramidal regions, did not included the revision rates, neurological outcomes, costs and infection rates, and did not conduct quality of life scale evaluation, such as SRS-22 patient questionnaire. The follow-up time is short, and there are no reports on long-term prognosis and correction effect. Conclusion In conclusion, robot-assisted pedicle screw placement in idiopathic scoliosis can improve the accuracy of pedicle screw placement, reduce intraoperative fluoroscopy times and operation time, which is a more accurate and safe assisted pedicle screw placement method. When facing the complex deformity and severe anatomical variation of AIS patients, the advantages of navigation system assisted pedicle screw placement are more prominent. Surgeons should consider the severity of the patient's condition and economic situation, carefully select the surgical method according to the anatomical structure and characteristics of the spine, and evaluate whether the radiation exposure is reasonable. Acknowledgements We are thankful to all the authors and all the study participants in this study. Authors’ contributions The authors' contributions are as follows: YiFan Kang and LinDong Wei collected and reviewed the literature and evaluated its quality, evaluated the data, and wrote and interpreted the paper. JianRong Liu reviewed papers, corrected and provided financial support. Funding The research was funded by the Central Government Guide Local Science and Technology Development Fund Project (YDZJSX2022A069). Data availability The datasets used and/or analysed during the current study available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate No ethical approval was required for this article. Consent for publication This unit has given its consent for publication. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Yifan Kang and Lindong Wei are co-first author. References Associated Data Data Availability Statement The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.