Robotic-Assisted Navigation Guided Kyphotic Deformity Correction Surgery
Harvinder S Chhabra1, Jitesh Manghwani2
1 Chief of Spine Service and Medical Director, Indian Spinal Injuries Centre, New Delhi, India
2 Consultant Spine Surgeon, Department of Spine Service, Indian Spinal Injuries Centre, New Delhi, India
Correspondence Address:
Harvinder S Chhabra
Chief of Spine Service and Medical Director, Indian Spinal Injuries Centre, New Delhi
India
Source of Support: None, Conflict of Interest: None
CheckDOI: 10.4103/0028-3886.360927
Pedicle screws offer the advantage of 3-column fixation and hence are the most common modality of instrumentation in the spine. However, pedicle screw fixation is a challenge in cases of complex pediatric deformity, especially in those with destroyed or anomalous anatomy.[1],[2],[11],[12] Misplacement of the implant is thus a common complication in such cases.[3] In addition, radiation exposure poses a risk to both the patient and the surgeon.
To address this limitation, technology has been constantly evolving to increase the efficacy and reduce complications. The most recent has been the robotic arm trajectory coupled with real-time navigation technology. There is a paucity of literature on this combined technology, especially in pediatric spine deformities.[13],[14] With this case, we present the use of robot-assisted navigation in a complex post-tubercular pediatric kyphotic deformity.
Video link: https://youtu.be/CTnvYCEVM9s
QR Code:
Objective
This is a case of a 16-year-old female patient with post-tubercular kyphoscoliotic deformity (00:00–00:12) [Figure 1]. After histopathological diagnosis of Pott's spine 3 years ago, the patient was conservatively managed (elsewhere) with antitubercular chemotherapy (drug sensitive) for 9 months. The patient presented with a painless progressive deformity causing difficulty in lying supine and walking (00:13–00:24). The neurology was intact, and the lumbar spine kyphosis was more apparent on forward bending. X-ray scans and computed tomography (CT) revealed a destroyed L2 vertebrae with residual body in the posterolateral quadrant (00:34–00:45) [Figure 2], [Figure 3], [Figure 4]. The 52° kyphosis was correctable to 48° on push-prone X-rays. The patient had associated compensatory thoracic scoliosis (Cobb's 47°). The thoracic scoliosis corrected very well on bending films. As it was a non-structural curve secondary to lumbar deformity, it was not planned for correction.
Figure 1: Preoperative standing whole spine AP X-ray showing associated scoliotic deformityFigure 2: Preoperative standing lateral X-ray lumbar spine showing kyphotic deformity of 51.81°Figure 3: Preoperative CT scan showing residual destroyed L2 vertebrae and the associated deformityFigure 4: Preoperative CT scan showing residual destroyed L2 vertebrae in the posterolateral quadrant and the associated deformityMRI revealed no cord anomalies or signal intensity changes (00:48–00:53) [Figure 5]. In view of progressive deformity and kyphoscoliotic lumbar spine, the patient was planned for deformity correction with robot assistance and real-time navigation after obtaining written informed consent. The plan was for T12–L5 pedicle screw instrumentation and posterior closing wedge osteotomy after L2 vertebral column resection.
Figure 5: Preoperative MRI scan showing no cord anomalies or signal intensity changes ProcedureAfter induction of general anesthesia, the patient was positioned on an Allen radiolucent table with the MazorX robot mounted on the side of the table and the stealth navigation camera unit on the head side of the patient (00:54–01:06). This enabled seamless cone beam CT scan by using an O-arm (01:07–01:14). Motor evoked potential (MEP) and somato-sensory evoked potentials (SSEP) were then set up for multimodal intra operative neuro monitoring (01:15–01:26). T12–L5, the area of surgical interest, was exposed by subperiosteal dissection (01:27–01:33). The robot was fixed on the patient by using a PSIS pin (01:34–01:46). The robot and patient then act as one unit. Alternatively, specific spinous process clamps can also be used.
The robot then defined a safe area for its movement, capturing the three-dimensional spatial orientation by using optical cameras mounted on the robotic arm (01:54–02:03). Snapshot tracker (02:04–02:09) was used for robot-navigation registration, following which navigated probe was used to define region of interest (02:10–02:16). Star tracker was used for intraoperative cone beam CT or O-arm scan (02:17–02:26). Once images were transferred, intraoperative planning of screw trajectory in the desired trajectory was done (scan and plan). Once the CT was done and the images were transferred, there was confirmation of star marker being found on the robotic system (02:27–02:35). We then marked the area of the vertebral column of surgical interest in AP and lateral views of the imaging taken by the O-arm on the robotic system (02:36–02:48). There was auto-segmentation of vertebrae by the robotic system, which was improvised manually (02:49–02:54). These lines were registered in the independent disc spaces (02:55–03:00). The vertebrae were marked and leveled. Manual correction of the axis is important, especially in cases of deformity in the axial, sagittal, and coronal planes (03:01–03:13). Once the axis was corrected, we proceeded with the planning of the trajectory of the pedicle screw (03:14–03:19). The system allows choosing different type and size of implants and even normal or reduction screws (03:20–03:28). Planning should be such that any skiving potential (slipping of drill causing error in the trajectory) is avoided (03:29–03:36). Once the screws were planned in the desired trajectory, we stacked and saw the images across small submillimetric CT cuts to confirm any error in the planning of the screw (03:43–03:55). It is recommended that the trajectory in axial, sagittal, and coronal planes is checked (03:56–04:01). The planning of L2 level screw (around the apex of the deformity) the size of the pedicle screw on the right side was increased from 5.5 mm to 6.5 mm as this pedicle was found to be accommodating a 6.5-mm screw. We stacked and played to reconfirm any breach in the trajectory of the screw (04:27–04:32).
In the planning of the L5 screw, the exclamation mark denotes high soft tissue pressure, and a warning is given for the same by the robotic system (05:05–05:11). We reduced the medialization and smoothened the bony surface at the entry points of the screws to prevent any possible skiving in the trajectory (05:12–05:23). The robot was then directed to the left T12 pedicle screw (05:29–05:34). The cannula was inserted through the robotic arm (05:35–05:38). The navigation probe confirmed the direction of the pedicle screw (05:39–05:42). The inner sleeve was inserted (05:43–05:45). It is recommended to use the mallet so that the sleeve engages and holds on to the bone properly to prevent skiving (05:50–05:58). Saline should be added within the sleeve before drilling inside to prevent any thermal injuries (05:59–06:05). The high-speed drill is also navigated to obtain a live feed to confirm the trajectory in which drilling is performed (06:06–06:14). This real-time navigation confirms that we are in the proper trajectory of the pedicle screws, and once the pedicle is crossed, the drill should be withdrawn (06:15–06:27). It is advised to sound and check the integrity of the walls of the pedicle as patient safety comes first (06:28–06:37). Once confirmed, we can either tap or directly insert the self-tapping screw in the desired trajectory created by the drill (06:38–06:47).
In this case, we could appreciate the cannula getting skived and hence we withdrew it instead of mounting it on the bone directly (07:31–07:40). These dynamic decisions are important to prevent misdirected trajectory (07:43–07:49). Once all the pedicle screws were inserted, we proceeded with the laminectomy (08:07–08:12). A temporary rod was inserted on one side (08:20–08:23). L2 corpectomy was done. The navigated burr can also be used to mark the predecided angle of the osteotomy or the edges of the corpectomy (09:55–10:06). Compression was done bilaterally to correct the kyphosis (10:49–10:54). We were able to achieve a bone-on-bone contact in this case and hence only bone graft was added and no cage was required (11:01–11:09). The bone bed was prepared by shingling for better fusion (11:19–11:22). Surgical wound closure was done in layers over a negative suction drain (11:23–11:28).
There were no intraoperative complications. The kyphosis was corrected from 52° to 20° of lordosis [Figure 6] and [Figure 7]. The patient was mobilized on postoperative day 1. On follow up at the end of 1 year, the patient was doing well with no complications such as implant loosening and maintained correction (11:38–11:46). The postoperative images validate the decision of not including thoracic curve in the correction as the thoracic curve got corrected substantially. There was good coronal alignment with a residual curve of 20°.
Figure 6: Postoperative standing AP X-ray showing corrected deformity and instrumentation from T12-L5Figure 7: Postoperative standing lateral X-ray showing corrected deformity and instrumentation in T12–L5 with bone-on-bone contact and lordosis of 19.08°Pearls and Pitfalls
Spinous process clamps or PSIS pins depend on the type of surgery, MIS or Open, and on the level being operated upon.
The initial marking with navigated probe should be as close to the area of interest as possible.
It is also recommended to smoothen the bony surface at the entry points of the screws to prevent any possible skiving in the trajectory.
While planning, if the robotic system shows an indication of alert of high soft tissue pressure (indicated by an exclamation mark), the trajectory of screws may be changed a bit (decrease the amount of medialization) to prevent errors.
It is advised to sound and check the integrity of the walls of the pedicle before tapping or inserting the screws.
The authors emphasize that the principles of kyphotic deformity correction, including appropriate neuromonitoring, need to be meticulously followed.
DiscussionThere are two major limitations of spine surgery: visual and manipulative. Insufficient vision and forced pressures can cause errors in spine surgery, leading to malpositioned screws. The complications of malpositioned pedicle screws are multifold. It can lead to neurological deficits, loss of correction on follow-up, non-union, and may require revision surgery for correcting the screw trajectory.[4],[15],[16],[17],[18] Traditional techniques also result in radiation exposure, both to the patient as well as the surgeon. The advantages of robotics as published in the literature include improved accuracy and decreased radiation exposure.[5],[6] However, there is paucity of literature on the usage of robotics in pediatric deformities, although published literature suggests that the use is safe and efficacious.[7],[19],[20],[21]
One of the disadvantages of robotics was not having a real-time feed. This has evolved over time by the integration of navigation with the robotic system. This helps in the planning of screws, using a robotic arm for the trajectory and navigation, real-time feed, and confirmation of three-dimensional anatomy.[8] The optical camera of MazorX system helps to perform a 3D assessment of work for self-reference of the location with the surroundings, thus avoiding any form of collision intraoperatively. Once articulated with the patient, the robot and the patient act as a unit, and any forms of movement intraoperatively can be readily picked up by the robotic system.[9],[22]
The registration can be done either by a preoperative CT scan with intraoperative fluoroscopic images or by obtaining an intraoperative CT scan by using an O-arm (scan and plan). Patients' anatomy can pose a challenge. However, this can be overcome to a certain extent with the MazorX system's feature of segmenting each vertebral body separately.
Robotic technology helps beyond pedicle screw placement. Preoperative planning and intraoperative registration help to visualize the invisible. By using the software tool “X-Align,” deformity correction can be accurately predicted. It meticulously helps in planning the exact position of the screw to give the best possible correction. Along with this, patient-specific rods are emerging with robotic technology. It obviates the need for intraoperative rod bending, which may allow for better correction, predicting correction, lowering rod breakage, and preventing under correction of deformities.[10] The future applications will also involve performing complex procedures with reproducibility, safety and accuracy.
Limitations
The limitations of robotic system involve the associated cost and availability. More than the cost to the patient, it is a major investment for the hospital and hence availability is not common. However, if the cost–benefit ratio is worked out, it may reflect an advantage as the technology aids in preventing complications and adds to precision. Spine surgical robotics is widely believed to have a great potential for future application, but it needs to solve the disadvantages of not having a wide application range.
ConclusionsRobotic technology coupled with navigation and integrated with intraoperative CT scans allows for precise instrumentation, reduced complications, lower radiation exposure, and better patient outcomes, especially in complex deformity cases. While these technologies are making a lasting impact, they are still evolving, and their full potential will be gradually tapped.
Patient's consent
Full and detailed consent from the patient/guardian has been taken. The patient's identity has been adequately anonymized. If anything related to the patient's identity is shown, adequate consent has been taken from the patient/relative/guardian. The journal will not be responsible for any medicolegal issues arising out of issues related to the patient's identity or any other issues arising from the public display of the video. The journal reserves the right to withdraw or pull out the video at any point of time without providing any reason whatsoever.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
References
留言 (0)