Laboratory examinations did not show any evidence of a coagulopathy. Cervical T2-weighted magnetic resonance imaging (MRI) showed foraminal stenosis due to a herniated disc at the C5-6 level on the right side. He underwent anterior cervical discectomy and fusion (ACDF) at the C5-6 level. During the surgery, active bleeding from an epidural vein was noticed at the C5-6 foramen, which was controlled with hemostatic materials. Anterior cervical fusion was performed with a polyetheretherketone cage filled with cancellous bone graft. The operating time was 120 minutes and blood loss was 200 mL.

A postoperative SEH extending to non-decompressed, adjacent levels has been seldom reported after cervical spine surgery. Hans et al.²⁾ performed a second-look surgery via the anterior route under the suspicion of a loculated postoperative SEH for a patient presenting with respiratory distress and flaccid tetraplegia after ACDF at the C6-7 level. However, the patient's condition did not improve after the revision surgery. Then they performed follow-up MRI, which revealed a postoperative SEH extending from C3 to T3. They performed a posterior cervical laminectomy with evacuation of the hematoma thereafter and the patient completely recovered after the posterior decompressive surgery. In the present study, the patient with postoperative SEH complained of severe pain only without any neurological deterioration. Based on the clinical symptom of the patient, we considered the loculated SEH at the C5-6 level as main lesion, though follow-up MRI showed significant compression of cervical spinal cord due to cervicothoracic SEH. The patient's symptoms totally relieved after anterior removal of loculated SEH at the C5-6 level without removing the whole cervicothoracic SEH. Ikuta et al.³⁾ reported a 33 percent overall incidence of postoperative SEH after microendoscopic posterior decompression for lumbar stenosis, and only three developed symptoms of radicular pain. MRI demonstrated spontaneous regression of the SEH in all patients at 3 months after surgery³⁾. However, they observed less expansion of the dural sac after 1 year despite sufficient widening of the osseous spinal canal in patients with postoperative SEH. Postoperative SEH also caused a delay in the patient's recovery and led to a poor clinical improvement. They assum-ed that transient compression of the nerve roots and poor expansion of the dural sac might prevent the initial recovery of the nerve damage; fibrosis induced by SEH permanently prevents the expansion of the dural sac and might cause nerve root irritation. Therefore, it was suggested that prevention of postoperative SEH might be required to prevent not only neurological deterioration but also a delay in the patient's recovery³⁾. In the present case, remnant SEH at non-decompressed, adjacent levels would have resolved spontaneously. However, further follow-up evaluation seems necessary since postoperative SEH in the cervicothoracic spine might cause a delay in patient's recovery.

An age of more than 60 years old, multi-level procedure, and/or preoperative coagulopathy have been suggested as risk factors for postoperative SEH^5,9,11). One important factor for a postoperative SEH is the completeness of the hemostasis during the initial surgery^2,5,11). Bleeding form the fusion bed might be one cause of postoperative SEH after ACDF. The authors try not to prepare the fusion bed vigorously, because it can cause troublesome bleeding and furthermore delayed subsidence of cage or graft. If there is significant bleeding from the fusion bed, the authors usually control bleeding with a small amount of hemostatic material such as, bone wax. Another important cause of postoperative SEH is bleeding from the epidural vein at the foramen. When it is not feasible to coagulate epidural vein directly, the authors use commercially available hemostatoc products, such as thrombin-soaked gelfoam, surgicel, and/or avitene. Though effective in control of epidural bleeding, these products can embolize, induce a localized inflammatory response, be easily washed away, and result in rebleeding⁸⁾. In the present case, hemostasis was considered adequate during the initial surgery since there was no evidence of active bleeding before closure. Rather, coughing during wake-up, which is associated with sudden elevation of the venous pressure, probably caused rebleeding from the epidural vein at the foramen. Straining, sneezing, and lifting have been previously suggested as probable predisposing factors for a spontaneous spinal epidural hematoma, all of which are also related to a sudden elevation of the venous pressure⁷⁾. Therefore, it may be helpful to induce elevation of the venous pressure using a Valsalva maneuver before closure. When rebleeding happens with Valsalva maneuver, additional hemosatsis of epidural vein is necessary before closure. Care also should be taken to prevent severe coughing during wake-up from general anesthesia, since it can cause rebleeding of epidural vein as with our case.