Prospective Controlled Study of Rectangular Titanium Cage Fusion Compared With Iliac Crest Autograft Fusion in Anterior Cervical Discectomy

Dimitris Zevgaridis, M.D., Claudius Thomé, M.D., Joachim K. Krauss, M.D., Department of Neurosurgery, University Hospital Mannheim, Mannheim, Germany.

Neurosurg Focus. 2002;12(1) 

In This Article


Outcome of motor and sensory radiculopathy was equally successful in both groups (p = 0.943 and 0.940, respectively). Patients with myelopathy fared somewhat less well, but there was significant improvement (postcompared with preoperative: p = 0.031 [iliac crest auto-graft group] and p = 0.016 [RTC group]) with no differences in outcome between groups (p = 0.867) (Fig. 3). The evaluation of arm pain before and after surgery, according to the VAS, showed highly significant improvement in both groups (post-compared with preoperative: p 0.001 [iliac crest group]; p <0.001; [RTC group]; p = 0.702 [both groups]). The relief of neck pain was less pronounced, but again highly significant and demonstrated a similar degree in both groups (post-compared with preoperative: p = <0.001 [iliac crest group]; p = 0.005 [RTC group]; p = 0.692 [both groups]) (Fig. 4).

Bar graph depicting the outcome of ACD for sensory deficit, motor deficit, and myelopathy. Note that the vast majority of patients improved in all three categories without statistically significant differences between the iliac crest autograft and RTC groups. ns = not significant.

Bar graphs showing the highly significant reduction of arm pain (left) and neck pain (right) in the iliac crest autograft and RTC groups assessed pre-and postoperatively using the VAS. There is no intergroup statistically significant difference.

According to Odom criteria, 83% of patients in both groups experienced good to excellent functional recovery (p = 0.824). According to the PSI, 93% of patients in both groups were satisfied and would undergo the same operation for the same results (p = 0.987) (Fig. 5).

Bar graph demonstrating functional outcome evaluated according to Odom criteria and PSI in the iliac crest autograft and RTC groups. Both implants yielded a similar outcome.

Three patients suffered severe pain (>3 of 10 on the VAS lasting >6 months) at the site of iliac crest harvesting. One patient in the iliac crest autograft group developed a pseudarthrosis. The patient underwent reoperation and fusion was achieved with autologous bone and a plate system. One case of transient recurrent nerve palsy was documented in the RTC group. There were no implant-related complications after placement of the RTCs (complications in four of 18 patients in the iliac crest autograft group and zero of 18 patients in the RTC group [p = 0.045]).

Graft collapse was demonstrated in three out of 22 levels in the iliac crest autograft group. One iliac crest auto-graft demonstrated slight extrusion with resulting pseudarthrosis (Fig. 6). Subsidence was seen in eight of 23 implants in the RTC group. Subsidence in the RTC group did not correspond with any clinical symptoms. It was more frequent early in the series (in four of the first five patients), when partial removal of the cortical endplates was performed in an attempt to promote rapid osseous growth and fusion. As a result of this observation, thereafter only the disc material was thoroughly removed, which subsequently reduced the rate of subsidence. Radiologically, the degree of subsidence was unchanged in all affected levels at 12 months compared with 3 months postoperatively, indicating no progression of subsidence over time.

With regard to fusion, RTCs proved to be equivalent to iliac crest autografts (p = 1.000). Ninety-one percent of the surgically treated segments (89% of patients) in the iliac crest autograft group and 87% of those (83% of patients) in the RTC group were considered stable and showed no signs of motion in flexion-extension on lateral radiographs (Table 2).