2026 年 13 巻 p. 355-360
C2-C3 disc herniation is an extremely rare condition, accounting for <1% of all cervical disc herniations. Discectomy using the posterior transdural or anterior transretropharyngeal approach has been used for its treatment. However, there are still concerns regarding the invasiveness and potential complications of these techniques. Herein, we report a case of C2-C3 disc herniation treated with transvertebral discectomy, assisted by an intraoperative navigation system. An 80-year-old woman presented with worsening numbness and pain in the left upper extremity, as well as gait instability. Magnetic resonance imaging revealed a left-dominant C2-C3 disc herniation compressing the spinal cord. As symptoms showed minimal improvement with conservative treatment, surgical intervention was performed. Through a 4-cm transverse incision on the left side of the neck and a 6-mm bone window, a transvertebral discectomy was performed to remove the C2-C3 disc herniation. An intraoperative navigation system was used to enhance surgical accuracy. The herniated disc fragment was successfully removed via the transvertebral approach without intraoperative complications. Intraoperative navigation helped ensure a safe and accurate surgical procedure. The patient's symptoms improved immediately, and she was discharged to home 1 week later. Transvertebral discectomy enables minimally invasive removal of C2-C3 disc herniation. Combined preoperative magnetic resonance imaging and intraoperative computed tomography using navigation system supports a safer and more comprehensive removal.
C2-C3 disc herniations account for <1% of all cervical disc herniations. Regarding the pathomechanism of upper cervical disc herniation, age-related degenerative changes and segmental fusion in the lower cervical spine increase the relative mobility and mechanical stress on the upper cervical levels, eventually leading to disc herniation.1,2) C2-C3 disc herniations were large, and were centrally located.3) Due to its clinical characteristics, the posterior approach is associated with a high risk of spinal cord compression or traction, potentially resulting in spinal cord injury. Posterolateral approaches require resection of the C2 transverse process and mobilization of the vertebral artery, thereby making the procedure technically demanding and increasing the risk of vascular injury.4)
Anterior cervical discectomy and fusion (ACDF) is the standard approach for cervical disc herniation.5) Securing an adequate surgical field is essential because the anterior cervical region has several critical soft-tissue, vascular, and neural structures, including the pharynx, retropharyngeal structures, carotid artery, and cranial nerves. However, as the mandible limits surgical exposure, ACDF at the C2-C3 level is challenging to perform.6) Excessive retraction of the pharyngeal and airway structures is often required, which increases the risk of postoperative dysphagia and vocal cord palsy.
Written informed consent was obtained from the patient, and institutional review board approval was waived in accordance with our institutional policy.
Case presentationAn 80-year-old woman with a history of posterior cervical laminoplasty presented with worsening numbness and pain in her left upper extremity. She had undergone 6 months of conservative treatment consisting of medication and rest at another hospital; however, her symptoms showed little improvement and gradually worsened, with gait instability becoming increasingly apparent, leading to her referral to our institution. Neurological examination revealed numbness and pain extending from the left upper arm to the radial aspect of the forearm. The visual analog scale (VAS) score (0-10) was 8.5. Grip strength was 19 kg on the right and 16 kg on the left, with no obvious weakness in upper extremity muscle groups. The Romberg test demonstrated postural instability. Although the patient was able to walk independently, her gait was unsteady. The Japanese Orthopaedic Association (JOA) score at the time of presentation was 11. Magnetic resonance imaging (MRI) revealed a left-dominant C2-C3 disc herniation causing spinal cord compression. At the C4-C5 level, left foraminal stenosis was also identified (Figure 1A-C). Based on these findings, the gait disturbance was considered to be due to proprioceptive impairment caused by the C2-C3 disc herniation, and the numbness and pain in the left upper extremity were attributed to C5 radiculopathy associated with foraminal stenosis at the C4-C5 level. Therefore, we planned a minimally invasive surgical procedure that allowed simultaneous treatment of the two lesions. A single 4-cm transverse incision was made in the left neck. Transvertebral discectomy was performed for the C2-C3 disc herniation (Figure 1D), followed by transvertebral foraminotomy for the left C4-C5 foraminal stenosis.

(A) Sagittal T2W MRI showed a disc herniation at the C2-C3 level. (B) Axial T2W MRI (C2-C3 level) demonstrated a left-sided disc herniation compressing the spinal cord. (C) Axial T2W MRI (C4-C5 level) demonstrated a left-sided foraminal stenosis. (D) Schematic illustration of transvertebral discectomy for C2-C3 disc herniation.
MRI: magnetic resonance imaging; T2W: T2-weighted imaging
Preoperative imaging, including dynamic radiography, computed tomography (CT) scan, and MRI, was performed to evaluate spinal motion and lesion location and to determine the entry point. Intraoperative navigation system was prepared to accurately determine the entry point and prevent intraoperative disorientation.
OperationThe head was fixed using a Mayfield head holder (Integra LifeSciences, Princeton, NJ, USA), and the neck was extended. The intraoperative navigation system (StealthStation S8; Medtronic, Dublin, Ireland) was positioned on the cranial side (Figure 2A). Under fluoroscopic guidance, the entry point was confirmed (Figure 2B), and a 4-cm skin incision was made along a natural skin crease (Figure 2C). The anterior aspect of the C3 and C4 vertebral body was accessed via the interval between the sternocleidomastoid and omohyoid muscles. Retractors (TrimLine, Medtronic) were used to adequately secure surgical fields and to effectively protect the soft tissues (Figure 3A). Intraoperative CT scan was conducted with O-arm (Medtronic) and fused with preoperative MRI to facilitate accurate trajectory planning and navigation (Figure 3B and C). A bone window with a diameter of approximately 6 mm was created at the anterior aspect of the C3 vertebral body (Figure 3D and E). The C3 vertebral body was drilled toward the herniated disc with a navigation-guided drill (Stealth-Midas, Medtronic). After reaching the disc space, the herniated disc was cautiously removed with meticulous attention to protect the spinal cord (Figure 3F-I). Navigation was used as needed to confirm anatomical orientation. After removal of the herniated disc, a 6-mm bone window was similarly created on the anterior aspect of the C4 vertebral body, and a transvertebral foraminotomy was performed. The final intraoperative CT scan was carried out to validate the extent of bone removal and the depth of the surgical trajectory, thereby preventing incomplete decompression. Hemostasis was achieved, a drainage tube was placed, and the wound was closed in the standard fashion. The operative time was 181 minutes, and estimated blood loss was minimal.

(A) The head was fixed using a Mayfield head holder, and the neck was placed in extension. The navigation system was positioned on the cranial side. (B) The entry point and trajectory were determined preoperatively under fluoroscopic guidance. (C) A 4-cm skin incision was designed along the natural skin crease.

(A) The anterior aspect of the C3 vertebral body was exposed, and the surgical field was secured using retractors. (B) The preoperative MRI and intraoperative CT were fused. (C) Under navigation guidance, the entry point and trajectory were confirmed. (D) Drilling was performed from the anterior aspect of the C3 vertebral body toward the herniated disc. (E) A bone window was approximately 6 mm in size. (F) Herniated disc was observed. Blue arrowhead, herniated disc. (G and H) Herniated disc was removed. Blue arrowhead, herniated disc. (I) The herniated disc was completely removed. Yellow arrowhead, dura mater.
CT: computed tomography; MRI: magnetic resonance imaging
The symptoms improved immediately after surgery. Postoperative CT scan demonstrated bone removal at the intended location (Figure 4A-C), and postoperative MRI confirmed adequate removal of the herniated disc (Figure 4D and E). The patient was discharged to home 1 week after surgery, without complications (Figure 4F). The postoperative JOA and VAS scores improved to 14 and 2, respectively, and further improved to 15.5 and 1, respectively, at 3 months postoperatively. At 6 months postoperatively, cervical spine radiographs showed no evidence of dynamic instability (Figure 4G). The patient remained symptom-free, and follow-up will be continued.

(A) Postoperative sagittal CT. (B) Postoperative axial CT at the C2-C3 level. (C) Postoperative three-dimensional CT. Black arrowhead, bone window for C2-C3 discectomy; blue arrowhead, bone window for C4-C5 foraminotomy. (D) Postoperative sagittal T2W MRI showed the disc herniation at the C2-C3 level was removed. (E) Postoperative axial T2W MRI showed the spinal cord compression was also improved. (F) Postoperative skin incision. (G) Dynamic X-ray images at 6 months after surgery.
CT: computed tomography; MRI: magnetic resonance imaging; T2W: T2-weighted imaging
Our approach is minimally invasive, allowing the procedure to be performed via a 4-cm incision. This technique enables access to multiple lesions via the same skin incision, as demonstrated in this case. Another advantage is the reduced risk of infection, as no implants are required. At our institution, an intraoperative navigation system was used to confirm anatomical orientation. However, anteroposterior and lateral fluoroscopy could be an alternative option. Although this approach requires removal of a small volume of the vertebral body, raising concerns regarding postoperative instability, a previous report using similar approaches has demonstrated preservation of both biomechanical and radiographic spinal stability.7) In this case, radiographic spinal instability has not been observed at 6 months postoperatively; however, careful follow-up will be continued.
We performed this procedure through a 6-mm keyhole for two reasons. First, if the keyhole is smaller than this, instruments tend to interfere with each other during insertion and removal, resulting in reduced maneuverability. Second, minimizing bone removal may reduce the risk of postoperative cervical instability. As the amount of bone removal increases, the risk of postoperative instability is expected to rise; therefore, we avoided enlarging the keyhole unnecessarily and used what we consider the minimum adequate size of 6 mm. Depending on the lesion, a slight increase of a few millimeters may be acceptable; however, we do not have precise data to define the upper limit of an acceptable keyhole size.
Compared with previously reported transvertebral approaches, which have mainly been applied to the lower cervical spine under fluoroscopic guidance,7) this technique has several distinguishing features. First, it was applied to the C2-C3 level, where anterior exposure is technically challenging due to anatomical constraints such as the mandible and a narrow operative corridor. Second, an intraoperative navigation system was used to determine the entry point and trajectory with high accuracy, which may reduce the risk of malposition and intraoperative disorientation. Third, this technique enabled simultaneous treatment of multiple lesions through a single small incision, highlighting its potential utility as a minimally invasive strategy.
Recently, favorable surgical outcomes of endoscopic-assisted ACDF for C2-C3 disc herniation have been reported.6) Although an endoscope provides excellent visualization, performing the procedure safely within the narrow operative space demands advanced surgical skills. Moreover, its steep learning curve has limited widespread adoption,8) and the number of institutions capable of offering this technique remains small. In contrast, our technique can be performed using a standard microscopic equipment and by incorporating the conventional ACDF steps, allowing for a safer and more reproducible minimally invasive procedure. The operative time for endoscopic-assisted ACDF for C2-C3 disc herniation has been reported to be approximately 3 hours.6) In this case, our procedure, including both C2-C3 discectomy and C4-C5 foraminotomy, was completed in 181 minutes. Although a direct comparison is limited by differences in surgical techniques, our approach appears to be sufficiently minimally invasive.
This approach is indicated for C2-C3 disc herniations that clearly compress the spinal cord from the anterior aspect and are predominantly localized to either the right or the left side. Additionally, patients with a thin chest wall and a long neck, facilitating anterior exposure of the vertebral bodies, are considered good candidates for this procedure. In contrast, this approach may not be suitable for patients with obesity, a short neck, or a thick chest wall. In such cases, ACDF has been reported to involve a deeper operative corridor, difficulty in tissue exposure and retraction, and interference between the chest wall and the surgeon's hands, all of which can compromise operative maneuverability.9,10) In our approach, adequate exposure of the C3 vertebral body is required, and the surgical trajectory is directed cranially; therefore, in these patients, inadequate exposure and partially blind manipulation are anticipated, and this approach is not recommended. Although the exoscope with a steeper oblique viewing angle may help overcome these limitations, our experience remains limited.
Similar to conventional ACDF, this technique carries risks of postoperative dysphagia, hoarseness, and neck hematoma. In addition, it has its own procedure-specific risks. An inaccurate entry point or trajectory may lead to insufficient decompression, spinal cord injury, or postoperative instability. In addition, the narrow working corridor through the keyhole can limit instrument maneuverability and increase the risk of intraoperative complications. These challenges may be more pronounced at the C2-C3 level due to anatomical constraints such as the mandible and limited operative space. Therefore, careful patient selection, meticulous preoperative planning, and the use of an intraoperative navigation system may help minimize these risks and improve surgical safety.
All authors have no conflict of interest.