Detection of bone graft failure in lumbar spondylodesis: spatial resolution with high-resolution peripheral quantitative CT.

OBJECTIVE

In spinal surgery, anterior spondylodesis is often combined with bone grafting, and graft integration is assessed with CT. High-resolution peripheral quantitative CT offers a resolution of 82 mum. The aim of this study was to compare the outcome of anterior spondylodesis as assessed with three radiologic procedures.

MATERIALS AND METHODS

Monosegmental lumbar spondylodesis with autologous iliac crest graft or solvent-preserved bovine cancellous bone was performed on seven sheep. The fused spinal segments were explanted after 24 weeks and examined with clinical 64-MDCT, high-resolution peripheral quantitative CT, and contact radiography. In 2D views, the area of the disk space bridged by bone was assessed, and the grafts were examined for fractures.

RESULTS

In three of seven sheep, clinical CT erroneously showed stable consolidation, whereas contact radiography revealed a clearly visible graft fracture, as did high-resolution peripheral quantitative CT. There was a statistically significant difference (p = 0.038) between bone volume assessed with clinical CT and that assessed with contact radiography. There was an almost significant difference (p = 0.053) between volumes assessed with high-resolution peripheral quantitative CT and clinical MDCT.

CONCLUSION

High-resolution peripheral quantitative CT, a technique approved for clinical use, has higher resolution in imaging of bone structure than does 64-MDCT. Our results show that high-resolution peripheral quantitative CT is superior to 64-MDCT in assessing osseous implant integration after anterior spondylodesis. The specimen size limit, however, prohibits in vivo use of this method in evaluation of the human spine. Our results suggest that in clinical practice, persisting symptoms despite radiologic findings of consolidated spondylodesis may be related to graft failure, which cannot be detected with clinically available methods.