Surgical Treatment for Ossification of the Posterior Longitudinal Ligament (OPLL) at the Thoracic Spine: Usefulness of the Posterior Approach

Various methods via anterior or posterior approach with or without spinal stabilization have been performed in accordance with the level and configuration of ossification of the posterior longitudinal ligament (OPLL) as the decompression surgery for thoracic myelopathy due to OPLL. Among them, anterior decompression at the middle thoracic level (T4/T5-T7/T8) is especially difficult to perform because of the special anatomical structures, where the spinal alignment is kyphotic and the thoracic cage containing circulatory-respiratory organs exist nearby. Of the anterior decompression procedures at this level, the posterior approach has various advantages compared to the anterior one. In the anterior approach, the procedure is complicated and the effect of decompression of the spinal cord can be obtained only by direct resection or anterior floating of the OPLL. However, complications such as spinal cord injury and dural tear are most likely to occur at that time. On the contrary, in the posterior approach, the procedure is simple, and various options to obtain decompression can be selected from, these are, laminectomy, laminoplasty, dekyphosis surgery, staged decompression surgery (Tsuzuki's method), circumferential decompression via posterior approach alone (Ohtsuka's method), and circumferential decompression via combined posterior and anterior approaches (Tomita's method). Among them, in laminectomy, laminoplasty, and dekyphosis surgery, anterior decompression can be obtained to some extent without performing direct procedure on the OPLL. In Ohtsuka's method, complete decompression can be obtained via posterior approach alone, although it is somewhat technically demanding. It is preferable to drop the shaved down and separated OPLL anteriorly instead of trying to remove it completely to avoid complications, especially in patients with severe adhesion between the dura mater and OPLL.