Spine surgery in a state-of-the-art hybrid operating room: an experience of 1745 implanted pedicle screws in the thoracolumbar spine

Hybrid-operating rooms are more and more commonly used in different fields of surgery like cardiothoracic surgery, vascular surgery, neurosurgery and orthopaedic trauma surgery especially at the spine and pelvis [1‐5]. The improvement of visualization and intraoperative imaging, the 3D-Scan capability and possible computer assisted 3D navigation are clear advantages of a hybrid operating room which allows safe and minimal invasive surgery. Compared to a traditional operating room setting the imaging device is directly linked to operating table and navigation system and all devices are directly controlled by the attending surgeon. The combination of high-resolution images and easy to use 3D navigation if needed might allow the surgeon to increase the accuracy like in this study for pedicle screw placement even further. A recent systematic review and meta-analysis of 51.161 screws still found misplacement of less than 4 mm reaching from 5.0 to 18.9% depending on the used technique [6]. The authors of this study used since 2015 a newly build hybrid operating room for 5 years on a daily bases for all types of thoracolumbar fractures (Fig. 1). A spinal fracture algorithm was developed using fluoroscopy for minimal invasive pedicle screw placement under intraoperative control with 3D imaging and the possibility of 3D navigation if needed (Fig. 2). The aim of this study was to prove that an hybrid operating room can improve the accuracy of minimal invasive pedicle screw placement in fractures of the thoracolumbar spine.

Institutional and prior ethical committee approval for the use of data in this study was obtained. Between June 2015 and December 2020, all patients with fractures of the spine were treated in newly build hybrid operating room (Fig. 1).

The special designed operating room had a room capacity of 55 square meters and consisted of a fixed robotic 3D flat-panel detector (Artis zeego, Siemens Healthineers, Germany), which is linked to the operating table (Trumpf, Germany) and a navigation system (BrainLab Curve, BrainLab, Germany). 326 patients with fractures of the thoracolumbar spine were included in the study. Figure 2 shows the developed and used algorithm of the authors. All included patients were treated in prone position and with a minimally invasive approach. Draping around the patient and for both branches allowed sterile free movement of the C-Arm (Fig. 2a). After initial fluoroscopy (Fig. 2b) if the pedicle was not clearly seen in anterior posterior (a.p.) views, a reference base was fixed to the processus spinosus near the fracture and intraoperative 3D scan was performed and automatically sent to the navigation system (Fig. 2d). In all cases, a Jamshidi needle was advanced under fluoroscopic control or with assistance of the navigation through the pedicle. After implanting a 1.45 mm guide wire (Fig. 2c) for all planned screws a 3D-Scan was performed with K-wires fixed flat using a sterile draping and a clamp. The 3D scan was then evaluated through the attending surgeon and the screw diameter and length were measured and planned (Fig. 2e). If one of the K-Wires was misplaced, it was corrected and the 3D-Scan was repeated. If all K-Wires showed perfect positioning the Screw-Rod-System (Viper2, DePuy Spine, 325 Paramount drive, Raynham, USA) was implanted (Fig. 2f).

For retrospective evaluation, clinical records including patient charts and radiographic images were reviewed. Patient-related factors like age, gender, body mass index and ASA classification were retrieved. Preoperative computed tomography images were used to classify the fractures applying the AO Spine thoracolumbar classification system [7]. Postoperative computed tomography images were used to assess screw placement. Screw perforation was divided in 4 grades according to Gertzbein et al. [8] (grade 0 perforation: no perforation, grade 1: 0–2 mm, grade 2: 2–4 mm, grade 3: > 4 mm).

Data analysis was performed with IBM SPSS Statistics (V21.0) and Microsoft Excel (V16.3). Demographic characteristics are described as mean and standard deviation. For the primary outcome measures, ordinal regression was performed considering all variables related to the grade of screw perforation. For the secondary outcome measures, logistic regression was performed considering all variables related to the used radiation.

This study showed that utilizing all the features of a hybrid operating room can improve the accuracy of pedicle screw placement for fractures on all thoracolumbar levels from T1 to L5. Compared to the systematic review of Perdomo-Pantoja et al. that reported an accuracy rate of 95.5% with CT-navigation and 91.5% in fluoroscopy assisted technique, we achieved a considerably lower screw perforation rate. Using the described algorithm in Fig. 2, perforation of more than 2 mm only occurred in 0.9% of the cases. This is one of the lowest evaluated rates of perforation compared to literature [6, 9]. After navigated or non-navigated K-Wire placement, a 3D-Scan was performed and misplaced K-Wires were corrected. This was necessary in 19.3% of the cases and significantly more often in navigated procedures. The replacement in these cases prevented perforation rates like seen in comparable studies. The higher rate on perforation in navigated procedures as well as the higher rate of replaced K-wire might be explained by smaller pedicles and more complex trajectories. Even with 3D navigation there is always a small percentage of misplacement. Compared to the study of Tjasic et al., who also used 3D-navigation, our study had more minor breaches [10]. While Tjasic et al. reported 1.6% perforation over 4 mm, only perforation under 4 mm were evaluated in this study which also can be explained by the K-wire correction. Liu et al. compared pedicle screw placement for fractures in the thoracic spine either using fluoroscopy or 3D-Scan via O-arm to control K-wire position. A significantly higher rate of satisfactory positioned pedicle screws (grade 0 and I) was observed in the O-arm control group [11]. The rate of perforations with less than 2 mm was with 96.6% comparable to the 98.3% in our study underlining the importance of a K-wire control after fluoroscopy guided pedicle screw placement. Perforation over 4 mm could be prevented in both studies. The slightly higher accuracy in our study might be explained by the use of 3D-navigation in complex cases with insufficient imaging. Comparable results for perforations less than 2 mm were seen by Shin et al. with 97.1% in the navigated group using O-arm navigation and 94.1% in the non-navigated group [12]. Still in both groups 1 vs 4 pedicle screws with perforations over 4 mm were found. Due to the K-wire correction after placement in both groups we could achieve a slightly higher accuracy and prevent perforation over 4 mm. In 16.9% of the cases 3D-navigation was used mainly in the thoracic spine because of insufficient fluoroscopy. In our study, navigated procedures showed 81.2% of the screws completely inside the pedicle and 15.3% showed perforation less than 2 mm. In comparison, Scarone et al. showed a higher rate of accuracy for O-arm and iCT-Airo navigated pedicle screw placement with 92.1% [13]. The rate of pedicle screws completely inside the pedicle also was higher in the study of Shin et al. with 93.3% and in the study of Tjasic et al. with 92.8% for 3D-C-Arm navigated and 98.9% for O-Arm navigated [12, 14]. The higher rates might be explained by the selection bias of our study that was not concepted as a comparison between navigated and non-navigated cases. Therefore the majority of navigated case were thoracic cases from T1-T10 with often poor imaging quality and small pedicle diameters, while Scarone et al. and Tajsic et al. placed more navigated screws in the lumbar spine [13, 14].

Main disadvantages of a hybrid operating room are extensive costs for the initial outlay of about 1.2 million dollars and the anticipated high radiation dose. While the costs for the initial outlay are high the operating time might be lower due to better imaging quality and through direct control of the C-Arm by the surgeon. Bronsard et al. reported an operating time of 83.5 min for mainly lumbar fractures with 4 percutaneously implanted screws using only fluoroscopy [15]. In comparison, mean operating time in the present study for 4 lumbar screws was 56 min. For fluoroscopic thoracic pedicle screw placement mean time was 99 min which is considerably faster than the study of Liu et al. with an average of 195 min. Navigated surgery showed significantly longer operating time which is in line with literature [9, 16]. Shin et al. demonstrated placing the screws with help of navigation is comparable time consuming than under fluoroscopic control but uses much more time for preparation [12]what might explain the longer operating time. Compared to O-Arm navigation with about 200 min operating time in different studies [9, 13] the operating time with 3D-C-Arm with navigation in this study was noticeable shorter with a mean time of 135 min. Also, with a comparable technique Tjasic et al. needed about 200 min. As described by Ryang et al. navigated procedures come with a steep learning curve and constant improvement for in radiation and operating time [17]. The authors are using 3D navigation since 2015 for pelvic and spinal surgeries and accumulated a lot of experience and cases what might explain the faster operating times.

As expected, thoracic cases needed more radiation dosage than thoracolumbar and lumbar cases. Furthermore, navigated cases used more radiation than non-navigated cases. Schuetze et al. showed that navigation reduces the radiation exposure for the operating personnel because it can leave the room during 3D-Scans [18, 19]. Furthermore, with the help of the ALARA principles radiation can be further reduced. The study of Bronsard et al. and Tajsic et al. had an effective dose of about 1.5 mSv only using fluoroscopy [14, 15]. Mean dosage for non-navigated cases was 11.6 mSv considerably higher but can be explained by the control 3D-Scan to verify K-Wire position. Furthermore, the 3D-Scan was even repeated if a K-Wire needed revision. Compared to the mentioned studies the accuracy of the pedicle screws was as mentioned above higher which in our opinion justifies the higher radiation dosage. For navigated procedures we measure a mean radiation dosage of 20.3 mSv. This is in line with a comparable study using iCT-Airo and O-Arm navigation [13].

This study is the first study to evaluate the daily use of a state-of-the-art hybrid operating room in the treatment of spinal fractures from T1 to L5. Due to the high-resolution imaging, 3D-Scan controlled K-wire placement and 3D-navigation when fluoroscopy was not sufficient an accuracy of 99% for perforations less than 2 mm could be achieved. Still, costs for the initial outlay and radiation must be taken into account as a disadvantage.