Treatment of multilevel cervical fusion with cages

BACKGROUND Multilevel cervical discectomy usually requires plate and screw fixation for maintaining the spinal curvature, and increasing the graft fusion rate. However, the use of plate and screw fixation may cause a few complications, such as screw breakage, screw pullout, esophagus perforation, and cord or nerve root injury. In this study, we try to use cages to replace plate function in multilevel cervical fusion. METHODS From January 1997 to June 2001, there were 180 consecutive cases of multilevel cervical degenerative disease. We randomized them into three groups: Group A (60 patients) underwent anterior discectomy and polyetheretherketone (PEEK) fusion, Group B (50 patients) underwent anterior discectomy, autogenous iliac crest graft (AICG) fusion and plate fixation, and Group C (70 patients) underwent anterior discectomy and AICG only. X-ray of cervical spine was taken every 3 months until fusion was complete. Spinal curvature was measured by lateral view of X-ray. The functional and working status were evaluated by Prolo scale. Blood loss and operation time were recorded, respectively. RESULTS The total complication rates were 3.3%, 16%, and 54.3% in Groups A, B, and C respectively. The graft complications were evaluated by radiographic findings (graft collapse, nonunion, or dislodged graft). However, only 37.1% of patients (13/35) with graft complications had clinical symptoms (severe neck pain, radicular pain, or neurologic deficits). The fusion rate was better, and the time to fusion was sooner in Groups A and B than Group C, p < 0.001 (X-2 test). PEEK cage is statistically better than plating group in total complications, p < 0.05. Graft collapse and nonunion were the major graft complications in Group C (AICG without plating). Screw pullout, and screw breakage were the main causes of plating complication. Blood loss was minimum in Group A, p < 0.05. Spinal lordosis increased by a mean of 4.61 +/- 2.93 mm and 1.68 +/- 5.02 mm in Groups A and B, respectively, but spinal kyphosis increased by a mean of -2.09 +/- 4.77 mm in Group C. Group A had a statistically better Prolo scale than Group C, p < 0.0001. CONCLUSIONS Both PEEK cage without plating and AICG with plating are good methods for interbody fusion in multilevel cervical degenerative diseases. They increase spinal lordosis and graft fusion rate, and cause fewer surgical complications. However, PEEK cage is preferred in our study for multilevel fusion, because it has the fewest complication rates and the least amount of blood loss. (C) 2004 Elsevier Inc.