In this study two robots were used to collaborate during thoracolumbar or sacral fracture surgical fixation. The first an imaging robotic 3D C-arm (ArtisZeego) that allows accurate intraoperative imaging and the second the Renaissance robotic navigation platform (Mazor Robotics) that assists in proper pedicle screw placement based on intraoperative image acquisition.
The study period implementation of the combined usage of these technologies was prior to the development of the newer robotic navigation platforms, when these technologies had yet to be combined as such. We found that the dialog created between these two robotic technologies increases patient safety and improves surgical results. The ability to have repeated real time intraoperative imagining in traumatic spine cases allowed us to increase the spectrum of cases that can be treated percutaneously, with improved robotic guidance for screw placement and validated instrumentation positioning.
The accuracy of percutaneous pedicle screw placement has been extensively studied and compared between the various technologies. Using the freehand technique the reported malposition rate is 10–15% of cases [4, 15]. Indeed, compared to freehand, the use of computer assisted navigation has achieved improved accuracy by 11% as first reported in a large meta-analysis, [4] additional meta-analyses further confirmed the superior accuracy of the robotic navigation technology [3, 6, 16].
Prior to combing these two technologies, in our previous robotic assisted surgery cases the percutaneous pedicle screw malposition rate was 5% (data collected in our institution). Others reports using the Renaissance (Mazor Robotics) reported similar rates [12, 17–21]. Keric et al reported in the placement of 2067 screws in 406 patients an accuracy rate of 96.9% [17]. A growing body of evidence in the literature reports on high rates of pedicle screw positioning accuracy associated with low complication rates with the newer systems as 3D fluoroscopy and intraoperative CT integrated with navigational systems and robotic assisted navigation [12, 18]. Hyun et al. reported superiority in accuracy of placement of 130 screws in the 30 patient Robotic assisted group without a single misplaced screw when compared to the free hand technique with only 1.44% misplaced screws [18]. Similarly, Laudato et al. reported in 84 patients placement of 569 screws, no significant difference and very low rates of screw misplacement, between Robotic assisted (1.56%), O-arm (2.62%) and free hand technique (2.55%), which they attribute to the surgeon’s experience [12].
While in this series, we have shown an increased rate to 100% - achieving ultimate patient safety and accuracy. Although trajectory drilling and hardware placement were executed manually, these surgeries are another step towards surgical automation.
These findings have been since validated in studies using the newer Mazor X robotic guidance platform, which has an option of combination with intraoperative CT- based spinal navigation as reported by Khan et al. (2018) in 50 patients with degenerative disc disease who underwent robot guided placement of 190 pedicle screws achieving an accuracy rate of 99.5% [22]. Recently, initial results from a high-power multi-center prospective study from the U.S. MIS-ReFRESH compared 374 Mazor robotic-guided to 111 fluoroscopic guided minimally invasive spinal fusions and found Mazor robotic-guided had 5.2 fold lower risk of surgical complications and 8.8 fold lower risk for revision surgery [23]. While a similar multi-center prospective comparative European Robotic Spinal Instrumentation study (EUROSPIN) is still ongoing [24].
Operating room time is a further consideration regarding this method. Several authors described the surgeons quick learning curve of the new spinal navigation techniques for robotic or imaged guided pedicle screw placement and thus decreased surgical time per screw [10, 19, 25, 26]. In our study all cases were performed by surgeons experienced with this technique, thus allowing quick screw placement. The learning curve was mainly for proper patient positioning in the hybrid OR and for proper draping and bridge placement to allow the Artis Zeego fixed imaging device to perform a clean spin of the patient's spine. Once this challenge was overcome, case time dropped with cases taking as short as 75 minutes. Richter et.al. presented similar challenges using an Artis Zeego hybrid OR in combination with a robotic navigation, yet concluding no considerable time was added to each procedure [27].
In hybrid ORs the use of a fixed imaging device (as the 3D robotic C-arm used in our study) are associated with risk of radiation exposure and pose a significant occupational hazard for surgical staff. A recent report recommended leaving the OR as no radiation during a 3D scan was measured behind closed doors of the OR [28]. Correspondingly, according to our Artis Zeego imaging protocol all OR staff are in a protected room while the imaging takes place, thus decreasing radiation exposure to a minimum. Furthermore, due to modifications in the Artis Zeego software, which automatically adapts the radiation dosage to the anatomic region, while assuring good constant image quality, the reduced dosage protocol in the thoracolumbar spine further reduces patient radiation exposure, as compared to a conventional post- operative CT scan [14]. As we have also previously demonstrated, robotic surgery has been shown to reduce radiation exposure [29].
The limitations of this study are its lack of randomization. However, this new merging of technologies is another step towards surgery automation decreasing malposition and increasing patient safety.