In this study, we compared the outcomes of 76 patients who underwent laminoplasty combined with C2DL or C2OL for treating multilevel cervical OPLL involving C2. C2DL was comparable to C2OL in terms of the functional outcomes evaluated by the NDI, SF-36 score, JOA score, neurological RR. However, the volume of blood loss, postoperative VASSNP and incidence rate of neck axial pain in the C2OL group were higher than those in the C2DL group. Additionally, C2DL was superior to C2OL in terms of maintaining the C2–C7 Cobb angle and cervical ROM postoperatively.
Nowadays, there are controversies regarding the surgical intervention indications for treating OPLL involving C2[16-18]. Liu et al.[5] considered that the spinal canal at the C2 segment was larger than the spinal canal in the subaxial spine, which may tolerate slight neural elements compression, and only patients with the spinal cord severely compressed at C2 or obvious kyphosis at the C2–C3 segment were surgically indicated. However, the authors only focused on the local region of C2, ignoring whether the C2 segment would affect the overall decompression effect after surgery. Hirabayashi et al.[19] introduced two mechanisms of neural decompression because of laminoplasty: dorsal shift of the spinal cord and local decompression effect. Therefore, Matsuzaki et al.[12] proposed the following surgical indications for OPLL involving C2. First, patients with severe stenosis of the spinal canal at C2 due to OPLL must undergo decompression surgery. Second, patients with a relatively stenotic spinal canal at C2 and obviously narrow spinal canal from C3 downward, as well as a narrow spinal canal from C3 downward combined with obvious cervical lordosis, were recommended for surgical interventions because C3 downward decompression alone may affect the spinal cord posterior shift. Anyway, any elements precluding spinal cord posterior shift will compromise the decompression effect. In the current study, the mean SAC at the C2 spinal canal was about 10 mm, and the OPLL extended upward to C2 with spinal canal stenosis at C2 caudal side. We predicted that C3 downward decompression alone would compromise the spinal cord posterior shift; so C2 segment decompression was performed in all patients, and satisfactory spinal cord posterior shift was achieved.
Nevertheless, there is no consensus over optimal surgical strategies for the treatment of cervical OPLL involving C2 because C2 occupies a pivotal position for maintaining the cervical spine alignment and stability[12]. There are several extensors attached to the C2 lamina and spinous process, including the semispinalis cervicis, obliquus capitis inferior, and rectus capitis posterior major. Biomechanical experiments have demonstrated that these extensors and C2 combined with ligaments act as the main dynamic stabilizers that are beneficial for maintaining static and dynamic equilibria of the cervical spine and play a major role in maintaining the lordosis and alignment of the cervical spine[20, 21]. Inoue et al.[22]noted that a kyphotic deformity in the upper cervical spine and compensatory lordosis of the lower cervical spine developed in patients who underwent C2 lamina removal. In traditional cervical laminoplasty, many patients have suffered from neck axial pain and neck stiffness due to destruction of the posterior cervical deep extensors[23]. Therefore, modified laminoplasty that preserves the deep extensors has been developed, such as C4–C7 laminoplasty combined with C3 laminectomy, or laminoplasty with preservation of the C2 and C7 spinous processes achieved better clinical outcomes. The authors considered that protection of the anatomical structures around the cervical spine such as the extensors and ligaments resulted in better cervical ROM and alleviation of neck axial pain[8, 24, 25].
There are various surgical strategies for cervical OPLL involving C2 with individual advantages and disadvantages. Liu et al.[5] preferred C2–C7 open-door laminoplasty due to the satisfactory functional outcomes, but the incidence of neck axial pain and volume of blood loss were significantly greater than those in patients who underwent C3–C7 laminoplasty alone. The clinical results were consistent with our study; the higher incidence of postoperative neck axial pain and increased volume of blood loss in the C2OL group might be attributable to extended destruction and additional injury of significant musculature attachments at C2[26]. Takeshita et al[27]. applied French-door laminoplasty for C2 decompression. Although the musculature attachments at C2 were reconstructed intraoperatively, the C2–C7 Cobb angle decreased significantly. Wang et al.[28] applied C2–C7 open-door laminoplasty combined with instrumented fusion to prevent cervical alignment loss, but cervical spine fusion resulted in cervical ROM reduction. When OPLL involves the C2 segment, adequate posterior decompression and preservation of the musculature attachments at C2 seems to be contradictory, and achieving both is technically demanding for clinicians. However, Matsuzaki et al.[12] invented open-door laminoplasty combined with C2 dome-like decompression for cervical OPLL involving C2 that fully preserved the C2 dorsal structure intact and can be performed safely. During the procedures, a dome-like cavity was created on the ventral cortex of the C2 lamina using a high-speed burr; moving the burr in a semicircular motion to prepare a well-proportioned groove according to the configuration of the dura mater was crucial to generate sufficient space available for spinal cord posterior shift. The authors applied this novel technique in 33 patients, and improvements in clinical symptoms were observed in all patients without catastrophic complications[12]. Sun et al.[29] applied C2DL for the treatment of cervical spinal stenosis in 28 patients. The neurological function improved significantly, mean cervical ROM decreased slightly, and incidence of neck axis pain was 14.29%. In our study, we applied open-door laminoplasty combined with C2 dome-like decompression in the C2DL group with satisfactory functional outcomes. The incidence of neck axis pain in the C2DL group was 15.0%, which was similar to that reported by Sun et al.[29] Compared with those in the C2OL group, the lower incidence of neck axial pain and volume blood loss in the C2DL group may be attributable to preservation of the musculature attachments at C2[5].
Although the original intention of laminoplasty was to preserve the cervical ROM, It has been reported that 30%–70% of patients experienced loss of cervical ROM during the follow-up[30, 31]. According to study by Hyun et al. [31], the overall cervical ROM decreased by 38.5% compared to the baseline after laminoplasty, 77.7% of patients with OPLL developed autofusion, and the mean cervical ROM decreased by 52.8% five years postoperatively. In the current study, the ROM of the cervical spine in the C2OL and the C2DL groups decreased by 57.2% and 40.0% at the final follow-up, respectively. We suggest that the overall reduced cervical ROM is attributed to alterations in tissue elasticity after posterior neck muscle dissection. Laminoplasty including the C2 lamina also resulted in worsening of cervical alignment. An analysis conducted by Kawaguchi et al.[32] showed that the alignment decreased by 7.3° after en bloc laminoplasty in 16 patients. In the study by Takeshita et al.,[27] the mean C2–C7 Cobb angle decreased in the French-door laminoplasty group and dome-like expansive laminoplasty group were 8.3° and 5.2°, respectively, which were similar to that in our study. The authors insisted that laminoplasty including the C2 lamina resulted in worsening of cervical alignment compared with cases in which the C2 lamina was retained. In our study, the mean C2–C7 Cobb angle decreases in the C2OL and C2DL groups was 9.2°and 6.9°, respectively. Based on the radiographical outcomes, the patients who underwent C2DL achieved better cervical alignment and ROM than those who underwent C2OL. We suggest that the cervical deep extensors attached to C2 spinous process and lamina play an important role. Therefore, we can conclude that preserving the attachment of the extensors at C2 facilitates the maintenance of cervical ROM and cervical lordosis preliminarily.
Although satisfactory outcomes were obtained in both groups and there were no severe complications in our study, complications associated with C2DL should be noted. Shimizu et al.[33] reported a case of fracture of the C2 lamina that occurred after dome-like expansive laminoplasty and caused neural compression due to excessive lamina grinding. Therefore, preoperative feasibility and safety evaluation of C2DL based on CT is required. Additionally, care should be taken to avoid excessive lamina removal, and the possibility of axial fracture occurring after the procedure should be considered. In our study, we implemented the following strategies to preclude insufficient decompression or excessive C2 lamina removal. First, we measured the thickness of the C2 lamina accurately on CT scans and predesigned the decompression area preoperatively. Second, it was necessary to explore the residual thickness of C2 the lamina during grinding. The principle of decompression is to ensure that there is no obvious pressure on the dura mater and to avoid an excessively wide scope of bilateral decompression and inordinate lamina removal.
Our study has some limitations. First, this was a single-institution retrospective study with a small sample size, this limits the external validity and conclusions that can be drawn. Second, although we reviewed cases using the same inclusion criteria, the selection bias with retrospective studies may have an impact on the conclusions. Third, the factor of OPLL progression was not considered, potentially affecting the final functional outcomes. Therefore, further prospective multi-institutional studies with larger sample size and long-term follow-up should be conducted to achieve a more reliable conclusion.