Cervical spine fusion with bioabsorbable cages

Object. Although it improves rates of fusion in surgical arthrodesis, conventional spinal instrumentation is associated with several risks, including hardware extrusion that injures adjacent anatomical structures, and disuse osteopenia as a result of stress shielding. The long-term effects of indwelling spinal instrumentation, although incompletely documented, may be detrimental. One way to avoid such problems would be to use bioabsorbable implantation devices. In this pilot study, performed in 1998 and 1999, the authors evaluated the use of a bioabsorbable interbody fusion device in a goat cervical spine model. Methods. Forty-two goats underwent two-level anterior cervical discectomy and fusion: eight received iliac crest autograft; 16 received a cage implant composed of 70:30 poly(L-lactide-co-D,L-lactide)/polyglycolic acid (70:30 PLDLLA/PGA)filled with either autograft or recombinant human bone morphogenetic protein-2 (rhBMP-2); and 18 received a cage implant composed of 85:15 PLDLLA/PGA filled with either autograft or rhBMP-2. Animals were killed at 3, 6, and 12 months postsurgery, and their cervical spines were evaluated histologically, radiographically, and physically for fusion. A primarily fibrous union was demonstrated in all animals killed at 3 months. At 6 months, bone trabeculae had become more prominent and the fibrous response less so in all cohorts. This occurred most frequently in the animals that underwent fusion with the 70:30 PLDLLA/PGA cages filled with rhBMP-2, in which 63 percent attained a histologically confirmed union that contained quantitatively more bone and less fibrous tissue than in the other cohorts; 63 percent of the unions in the aforementioned cohort were graded quantitatively as being stable. Conclusions. In this study the authors have established that the 85:15 PLDLLA/PGA cages are absorbed too quickly to be functionally useful in this model. The 70:30 PLDLLA/PGA cages worked as well as tricortical autograft when filled with cancellous autograft, and better when filled with rhBMP-2. At 6- and 12-month follow-up review, the 70:30 PLDLLA/PGA cages had not yet begun to be absorbed. There was little if any inflammatory response to these cages at 6 months. Future studies should include biomechanical and microradiographic testing, and a longer follow-up period is necessary in this model to determine when the 70:30 PLDLLA/PGA cages are absorbed.