Laminectomy without reconstruction has been the standard surgical treatment method for intradural and posterior epidural spinal lesions7). Although this surgical technique provides an adequate operative field, it may lead to hematoma in the spinal canal, invasion of scar tissue after surgery, or postoperative instability and scoliosis45). As microscopic surgical techniques have evolved, new surgical methods such as laminoplasty with the use of a T-saw for reconstruction36) or laminoplasty with a 90° or 60° change in the lamina with en bloc removal14) have been introduced. Some authors have reported that certain spinal cord tumors can be removed through hemilaminectomy and a minimally invasive technique that involves the use of a tubular retractor system, which enables maximum preservation of the posterior elements210). Here, we report a spinous process-splitting hemilaminoplasty technique that maintains the shape and integrity of the posterior elements and retains a barrier that prevents hematoma in the spinal canal and scar tissue damage.

The age of the patients ranged from 28 to 76 years (mean, 60.2 years). Four patients had radiculopathy and five exhibited motor weakness. Eight had an intradural extramedullary spinal tumor and one had ossification of the ligament flavum. The surgical sites were cervical in six patients, thoracic in one, and lumbar in two (Table 1). Spinous process-splitting hemilaminoplasty was performed on levels 1-3 with a mean operation time of 237.8±47.0 min (range, 160-315 min) and a mean EBL of 116.7±61.2 mL (range, 50-200 mL). Post-surgical diagnoses were schwannoma in five patients, meningioma in two, neurofibroma in one, and ossification of the ligament flavum in one. We performed total resection in the schwannoma, neurofibroma, and ossified ligament flavum cases, and Simpson grade II resection in the patients with meningioma (Table 2).

The postoperative follow-up period was 6 to 31 months, and all patients were able to walk 1 day after surgery. All patients recovered without neurologic deficits or pain aggravation. CT scans were obtained for each patient 6 months after surgery, and the findings confirmed that laminoplasty was achieved on the bone fusion without displacement.

Although laminectomy has previously been used to treat intradural lesions, the surgical outcomes have often been compromised by hematoma, scar tissue invasion, and/or postoperative instability. With the development of microsurgical techniques, various alternative methods have been introduced1249). Alternative methods include rotation laminoplasty, hemilaminectomy, and en bloc laminoplasty.

Asazuma et al.1) performed 90 degree-rotation laminoplasty on spinal cord and cauda equina tumors in seven patients and reconstructed the vertebral arch. The authors reported satisfactory operative exposure, a maintained range of motion, and an improved Japanese Orthopedic Association score. Hida et al.4) introduced a 60 degree-rotation laminoplasty with titanium mini-plates. Of the eight patients who underwent surgery, six had an intradural tumor as well as ossification of the ligament flava and spontaneous spinal cord herniation.

These methods involve rotating the lamina because of bone loss due to cutting. This can only be conducted on the thoracic and lumbar spine where the spinous process is long, but not on the cervical spine where the spinous process is short. Therefore, these techniques can only be used to treat thoracic and lumbar lesions. Furthermore, because bone fusion does not mirror the original state, this technique cannot be considered true reconstructive laminoplasty. In contrast, the method that we introduced here involves minimizing bone loss by cutting the spinous process with a T-saw or oscillating saw to obtain a wide epidural space, enabling preservation of the pre-surgical appearance. This technique is not only applicable to the thoracic and lumbar spine regions, but also to cervical lesions. In addition, because it does not involve the use of titanium mini-plates, there are no artifacts, and therefore, imaging evaluations are more accurate, especially in cases of tumors that could be recurrent.

Hemilaminectomy has been introduced as a surgical treatment for intradural lesions using a unilateral approach28). It preserves the contralateral muscles and ligaments, which prevents secondary instability and scoliosis after laminectomy. However, in hemilaminectomy, the lamina is removed via a unilateral approach, and thus, the operative field is restricted. In contrast, our technique involves splitting the spinous process, thereby allowing the midline of the dura to be seen (Fig. 3). Undercutting of the spinous process enables the surgeon to see around the midline of the dura and dura margins, and therefore, both sides become visible.

Hara et al.3) performed en bloc laminoplasty on 16 patients using a T-saw, and similar to the results obtained using our technique, the bone shape remained similar before and after surgery. The authors reported that by cutting the spinous process around the surgical level and placing the sutures, they were able to preserve posterior supporting elements such as the supraspinous and infraspinous ligaments, preventing kyphotic deformity after laminoplasty. However, our method involves a smaller incision because the spinous process around the surgical area is not removed. Moreover, the midline and contralateral posterior elements remain intact, and thus, the physiological integrity of the spinous ligament is maintained while postoperative kyphotic deformity is prevented. It should also be noted that splitting the spinous process where there is a large amount of cancellous bone, rather than the laminofacet junction, results in better bone fusion.

Spinous process-splitting hemilaminoplasty and other surgical methods were compared with respect to operative time and blood loss12348). The operation time for 60-degree rotation laminoplasty3) was 233.75±63.46 min (170-380) and blood loss was 218.75±176.33 mL (45-600), while for 90-degree rotation laminoplasty1), the operation time was 338.29±144.35 min (205-445) and blood loss was 578.43±176.33 mL (73-2038). For heminlaminectomy8), operative time was 246.67±61.54 min (180-320) and blood loss during surgery was 55.83±12.01 mL. For en-bloc laminoplasty3), surgery time was 350.63±81.20 min (242-485) and blood loss was 589.31±282.89 mL (90-1209). These results are similar to those obtained using our surgical technique. And the minimal operative time and bleeding on our surgical method were similar for other surgical methods.

The majority of spinal tumors are either intradural extramedullary (40%) or extradural (55%) tumors, and the intradural lesions often lean toward one side8). In our cases, during same period, we treated a total of 11 cases of spinal cord tumors. Except for two patients with intradural intramedullary tumors and one patient with previous posterior spine surgery, spinous process-splitting hemilaminoplasty was performed in all cases where conventional techniques were applicable. Complete tumor removal was achieved in all cases.

Our novel surgical technique has advantages over previous laminectomy, laminoplasty, and hemilaminectomy techniques. Primarily, this new procedure enables adequate visualization during surgery and prevents scar tissue invasion and hematoma. Moreover, it does not cause injury to the muscles and ligaments on the contralateral side of the lesion, and thus prevents postoperative instability and subluxation. Finally, it should be noted that no imaging artifacts are produced as a result of the surgery, and therefore, imaging evaluations related to tumor recurrence are more accurate.

Spinous process-splitting hemilaminoplasty for intradural and posterior epidural lesions preserves the contralateral posterior elements and maintains stability compared to previous reconstructive laminoplasty techniques. Visualization is also improved compared to previous hemilaminectomy techniques, and laminoplasty following this procedure helps maintain physiological integrity.