Two patients presented with distinct symptoms. A 38-year-old male experienced severe upper back pain with right arm pain (visual analogue scale [VAS] 8), tingling, and numbness in the right third, fourth, and fifth fingers. A 28-year-old male reported right scapular pain radiating to the arm (VAS 8), fifth finger numbness, and hand weakness (fifth finger abduction and grasp III).

Both patients exhibited right-sided paracentral T1–2 disc herniations with slight upward migration on magnetic resonance imaging (MRI). The largest size of disc herniation on axial MRI image measured 11 mm × 6.5 mm, while the other was 10.5 mm × 6 mm. Computed tomography (CT) revealed soft disc herniations without bony spurs, annular calcification, or ossification of the posterior longitudinal ligament (Figure 1).

Despite conservative treatments (rest, oral medication, pain block), both patients continued to experience severe pain and neurological deficits. Given the success of UBE discectomies in the lower cervical area (C6–7–T1), we elected to perform UBE discectomy at the T1–2 level.

The procedure utilized general endoscopic surgical instruments, including an endoscopic camera and display monitor (Stryker, Kalamazoo, MI, USA), a 0° scope (Techcord, Daejeon, Korea), a high-speed drill (Primado2, NSK, Kanuma, Japan), a radiofrequency coagulator (VAPR, DePuy Synthes, Raynham, MA, USA), 1, 2-mm Kerrison punch, and angled curette, and other common endoscopic tools.

Under general endotracheal anesthesia, the patient was positioned prone (Concorde position). Patient’s head was placed on soft sponge cubic, without using rigid head fixator. The neck was maintained in a neutral position to prevent hyperextension or hyperflexion, and potential cord compression from the herniated disc. Both shoulders were retracted caudally and stabilized.

Under C-arm fluoroscopy, the medial margins of the T1 and T2 pedicles were confirmed in the true anteroposterior view (Figure 2A). The working portal (upper endplate of the T2 vertebra) and scope portal (1.5–2.0 cm caudal to the working portal) were marked. Lateral C-arm imaging was limited due to shoulder obstruction of the T1 and T2 vertebrae (Figure 2B). Using the visible C5–6–7 endplates as a reference, we roughly assessed the appropriateness of the sagittal working angle for surgery. Returning to the anteroposterior view, incisions were made, and 2 portals were established. After securing the scope at the T1 lamino-facet junction, the surgical level was confirmed again with C-arm oblique imaging.

After denuding muscles and soft tissues from the lamina, the T1 and T2 lamino-facet junctions were identified (Supplementary video clip 1). The T2 upper lamina and proximal superior articular process (SAP) were also located. The lower T1 lamina and medial inferior articular process were drilled, extending cranially to the ligamentum flavum insertion and laterally to approximately one-third of the facet joint (2- to 3-mm width). Subsequently, the medial T2 SAP and upper T2 lamina were drilled. The remaining bone and ligamentum flavum were removed using 1 and 2 mm Kerrison punches.

Following ligamentum flavum removal, the neural structures were carefully separated from the epidural soft tissues. Epidural vessels were coagulated, and fat tissues were cleared to expose the axillary region comprising the T1 nerve root and thecal sac. An upward-migrated disc fragment was identified within the axillary region and cautiously extracted with a non–ball-tipped nerve probe and pituitary forcep. After confirming reduced tension of nerve root and dura, a ball-tipped nerve probe was used to explore the ventral epidural space for any residual disc fragments, and to compress the annulus to expel subannular disc material.

Meticulous hemostasis was achieved using radiofrequency ablator, bone wax, and a hemostatic agent. A drainage catheter was inserted, followed by subcutaneous would closure.

Written informed consent for publication was obtained from both patients. They agreed to provide their medical information for the purpose of this study, with the understanding that their identities would remain confidential.

UBE has demonstrated outcomes comparable to other surgical options (e.g., microscopy, PELD) for lumbar discectomy [1-3] and cervical discectomy [6-8]. Based on the results reported in the literature, we proceeded with UBE T1–2 discectomy, which we believe to be the first report of its kind. Our patient presented with acute aggravated symptoms which implies an acute disc rupture, not chronic. Preoperative imaging revealed a paracentral location of soft disc herniation without a hard component (such as a calcified disc or severe bony spur) on CT scans. These factors led us to anticipate favorable outcomes.

The T1–2 UBE discectomy procedure is influenced by several unique features of the thoracic spine. Its intermediate position between the cervical and lumbar regions allows for leveraging specific characteristics during surgery. Compared to the confined cervical spine, the thoracic spine provides a larger operative space, accommodating instruments like drills, curettes, and Kerrison punches more easily. However, due to the presence of spinal cord, it produces more surgical stresses compared to lumbar area which contains only nerve roots. Several steps are required to avoid cord injury during UBE procedure. A solid working portal should be established to allow smooth in and out passage of surgical instruments. This is also crucial for maintaining constant saline outflow to prevent increased intracranial pressure [9]. During surgery, it is crucial to ensure that surgical instruments do not press on neural structures. When performing a laminectomy, it is advisable to use a burr for most of the bone work. Only small-sized Kerrison punches (1–2 mm) or curettes should be used sparingly to prevent nerve compression.

Tanaka et al. [10] reported increased distance between exiting nerve roots and disc spaces as one moves caudally in cervical spine. This trend continued in our procedure on T1–2 spine, providing larger safety margins for the exiting root during T1–2 discectomy. In contrast to cervical discectomies, which sometimes require partial pediculotomy due to limited space, T1–2 area provided enough space between the exiting root and index disc without the need for pediculotomy. Nevertheless, caution must have been taken when dissecting around the neural structures. Especially in the ventral epidural space, the ball-tipped probe was used only after adequate space had been secured.

Level confirmation during lower cervical (C6–7–T1) and upper thoracic spine surgeries can be challenging due to patient anatomy and positioning. Short, thick necks or well-developed shoulders often obscure lateral views, while the index vertebrae and disc space may be completely invisible in these procedures. Even when visible on preoperative x-rays, C-arm imaging under prone positioning can render these structures indistinct.

To address these challenges, we relied on anterior-posterior (AP) and oblique views for level confirmation. On AP views, the C7 and T1 transverse processes, along with the first rib, served as reference points. The presence of a cervical rib [11] can occasionally complicate level identification, necessitating preoperative CT scans. On oblique views, the prominent C2 lamina serves as a reliable landmark for counting levels and ensuring that the scope is docked at the correct level (Figure 4).

During posterior cervical surgery, UBE provides a highly magnified and clear surgical view, which is essential for observing tiny structures such as the nerve root, perineural tissues, and epidural vessels. Additionally, the gentle pressure created by continuous saline irrigation allows the thecal sac to subtly collapse, creating a larger operating space. For these reasons, many UBE surgeons find it more comfortable compared to open techniques. These advantages were instrumental during our T1–2 discectomy, enabling efficient and safe surgical procedures. The enhanced surgical view provided by the endoscope aided surgeons in distinguishing neural structures from other perineural elements. The subtly collapsed thecal sac facilitated the safe use of surgical instruments, and it was also possible to avoid unnecessary facet resection. Intraoperative bleeding and muscular damage were minimized. With further surgical trials and accumulated data, the UBE technique is anticipated to become a notable surgical option for T1–2 discectomy.

However, thoracic spine surgery, especially at the spinal cord level, carries significant risks, including cord injury. As the UBE procedure has a steep learning curve [12], acquiring substantial experience in UBE procedures in the lumbar and cervical regions is strongly recommended before attempting thoracic procedures.