'Spring-back' closure associated with open-door cervical laminoplasty

Abstract

BACKGROUND CONTEXT: Spring-back complication after open-door laminoplasty as described by Hirabayashi is a well-known risk, but its definition, incidence, and associated neurologic outcome remain unclear. OBJECTIVE: To investigate the incidence and the neurologic consequence of spring-back closure after open-door laminoplasty. STUDY DESIGN: A retrospective radiographic and clinical review. OUTCOME MEASURES: Lateral cervical spine X-rays were evaluated. Anteroposterior diameters (APD) of the vertebral canal of C3-C7 were measured. Spring-back was defined as loss of APD on follow-up in comparison to immediate postoperative canal expansion. The loss of the end-on lamina silhouette with consequent reappearance of the lateral profile of the spinous processes was also assessed to verify the presence of spring-back. Spring-back closure was classified based on whether the collapse was total or partial, and whether all the operated levels or only a subset had collapsed (ie, complete vs. partial closure, segmental closure vs. total-construct closure). Neurologic status was documented using the Japanese Orthopaedic Association (JOA) score. METHODS: Thirty consecutive patients who underwent open-door laminoplasty from 1995 to 2005 at a single institution with a minimum follow-up of 2 years were assessed. They were all operated on using the classic Hirabayashi technique. Radiographic outcomes were assessed independently by two individuals. RESULTS: Sixteen men and 14 women with an average follow-up of 5 years (range, 2-12 years) were included. Of these patients, 24 had cervical spondylotic myelopathy and six had ossification of the posterior longitudinal ligament. Spring-back closure was found in three patients (10%) and 7 of 117 laminae (6%) within 6 months of the operation, which was further confirmed by computed tomography and magnetic resonance imaging. All spring-back closures were partial segmental closures. Gender and age were not significant factors related to spring back (p>.05). The mean JOA score on follow-up was 12.5, with a recovery rate of 40%. All patients with spring back and available JOA data exhibited postoperative neurologic deterioration. Of the three patients with spring back, two patients underwent revision surgery, whereas one declined. CONCLUSIONS: Spring-back closure occurred in 10% of our patients at or before 6 months after surgery. The incidence of spring-back by level (ie, 117 laminae) was 6%, mainly occurring at the lower cervical spine. All spring-back closures were partial segmental closures, most commonly involving C5 and C6. Postoperative neurologic deficit was associated with spring-back closure; therefore, surgeons should adopt preemptive surgical measures to prevent the occurrence of such a complication.