In this retrospective cohort study, we examined a consecutive series of patients diagnosed with AAD-HVA who underwent PRFI. The diagnostic criterion for high span is CTA coronal view, based on the fact that the highest span of the vertebral artery at C2 is located below the horizontal line of the exit of the transverse process foramen and within the transverse process foramen. This study evaluated and compared the perioperative outcomes between patients undergoing RA-PRFI and FH-PRFI performed by the same group of surgeons. The surgical team consists of three doctors, who are chief physician, deputy chief physician, and resident physician. This three people medical team only performs upper cervical spine surgery. This research was approved by the hospital institutional review board and conducted between July 2018 and January 2022 (approval number: ID 20180101). The FH group comprised 14 patients who received FH -PRFI, while the other group consisted of 11 patients treated with RA-PRFI. Detailed patient information, including basic demographics, medical history, and relevant scores, can be found in Table 1.

Following administration of anesthetic agents, patients were placed in the prone position with routine sterile sheeting. The PRFI procedure involved a median incision, ensuring the careful protection of vessels while exposing the posterior arch entry point of cervical 1 and the entry point of cervical 2. The Harms technique was employed for fixation, involving the meticulous exposure of the posterior part of the pedicle. This was followed by the use of a grinding drill for opening, manual drilling of holes, and implantation of C1 lateral mass screws and C2 pedicle screws under fluoroscopic guidance. Intraoperative confirmation of the precise placement of the internal fixation was performed using the C-arm. Two appropriately sized rods were used to connect C1-2, and a large amount of autologous iliac bone is used for posterior bone grafting.

RH-PRFI was performed using the Beijing Tinavi robot (Fig. 1) as the surgical tool. Following general anesthesia, the patient was positioned in the prone position, and the head was securely fixed to the head frame. Meanwhile, a tracer was attached to the head frame to maintain the alignment between the head, head frame, and tracer throughout the procedure. The posterior resection fusion internal fixation was then performed through a median incision, with careful attention given to protecting the vessels and exposing the posterior arch entry point of cervical 1 and the entry point of cervical 2. Import preoperative CT data into the robot for intraoperative data matching, the data was uploaded to the robot’s central processing unit for optimal screw sizing and determination of the screw entry point (Fig. 2). To assess accuracy, three reference points were identified within the surgical field and validated using the robotic system. Intraoperatively, a guide wire was inserted through a trocar on the robotic arm, guided by fluoroscopy, and C1 lateral mass screws and C2 pedicle screws were then implanted. Finally, an intraoperative O-arm evaluation was performed to determine the exact position of the internal fixation. Two appropriately sized rods were used to connect C1-2, and a large amount of autologous iliac bone is used for posterior bone grafting.

Schematic diagram of robot surgery operation

Robot intraoperative planning diagram

Before closing the surgical wound, use 20 ml 2 g tranexamic acid (First Pharmaceutical Co., Ltd.) to routinely rinse the wound. Placement of drainage tube for measuring postoperative drainage volume. The radiation dose will be recorded by C-ARM, and the data will be exported by a specialized intraoperative radiologist. The intraoperative blood loss is estimated as the total amount of suction minus the amount of flushing fluid plus the amount of blood on the gauze.

Postoperative assessments included the evaluation and comparison of various factors between the two study groups, including intraoperative hemorrhage, overall operative time (from initial incision to application of surgical dressing), length of hospital stay following surgery, radiation exposure, and fusion rate.

Two days after the surgical procedure, CT scans were performed to determine the position of the screws. Radiologists and chief surgeons reviewed the CT images to determine the number of screws in a competent position, and the Gertzbein and Robbins grading system [13] was used for grading. The grading system consisted of four grades: grade 0 indicated that the screw was entirely within the bone, grade 1 indicated that the screw penetrated the cortex by < 2 mm, grade 2 indicated penetration of ≥ 2 but < 4 mm, and grade 3 indicated penetration of ≥ 4 mm. Screws graded as 0 or I were considered clinically acceptable, while screws graded as 2 or 3 were deemed unacceptable and could potentially lead to serious complications. Patients with misaligned screws underwent immediate magnetic resonance angiography or CT angiography to assess potential vascular injury.

Patient evaluations were conducted at 1, 4, 12, 24, and 36 weeks postoperatively, according to the Japanese Orthopaedic Association (JOA) score and visual analog scale (VAS). Additionally, the Neck Disability Index (NDI) score was measured at 3, 6, and 12 months post-surgery to evaluate the overall outcome.

Statistical tests were carried out with SPSS v20.0. Continuous data are reported as mean ± standard deviation, while categorical variables are presented as numbers (%). Data normality was examined using the Shapiro‒Wilk test. Follow-up and preoperative parameters, such as JOA and VAS scores, were analyzed with paired t-tests. Independent t-tests were utilized to analyze surgical outcomes and postoperative imaging data between the two groups. Descriptive statistics were applied to enumeration data based on group characteristics. Statistical significance was defined as p < 0.05.