The purpose of the study was to determine the effectiveness of rhBMP-2 when applied to an absorbable collagen sponge for interbody fusion with allograft spacers.

Seventy-seven consecutive patients requiring inter-body fusion with one hundred and eighteen levels were included in the study. In thirty-six patients cervical or lumbar interbody fusions were performed using allograft and rhBMP-2. In the remaing allograft was used with demineralised bone matrix. In patients undergoing anterior cervical decompression and fusion (ALIF) machined allograft spacers were supplemented with anterior locking plates. In those scheduled for anterior lumbar inter-body fusion (ALIF) or transforaminal lumbar interbody fusion (TLIF), machined allograft spacers were supplemented with posterior pedicle screw constructs. There were no stand alone ALIF procedures. All patients undergoing surgery were assessed preoperatively and at two weeks, six weeks, three, six, twelve, twenty-four months following surgery.Their Oswestry disability index (ODI) questionaires, Visual analogue scale (VAS) scores and a pain diagram were recorded. Radiographic measurements were made on the electronic public access computer system (EPACS).

All patients with allograft plus rhBMP-2 showed radiographic evidence of fusion by six months following surgery. The allograft with demineralised bone matrix group took considerably longer to achieve the same radiographic end plate appearance. Subsidence was obvious on radiographs in greater than 50% of cases with allograft and rhBMP-2 at three months. It was noted to occur between six weeks and three months and there was no significant subsidence after six months. This was statistically significant with a p< 0.0001 (fisher exact test). The average subsidence for the entire rhBMP-2 group at twelve months was 16.5%(SE 2.5% and range 0–58%) and for the allograft and bone matrix group was 4.6%(SE.74% range 0–15%)with a p< 0.0001 (independant t test).

Although high rates of fusion can be accomplished with allograft and rhBMP-2 (100%), significant subsidence occurs in greater than 50% of lumbar spine cases and 30% of cervical spine cases. This is possibly a result of early bone turnover with loss of structural support of the allograft spacer and end plate resorption.