Spinal Fusion Endpoint Adjudication

Challenge

Determining whether a spinal arthrodesis has achieved solid fusion is among the most subjective assessments in musculoskeletal imaging, yet it was the primary endpoint of this medical device trial. Site investigators' fusion determinations diverged substantially from one another and from imaging evidence, threatening the endpoint's credibility. The sponsor needed an independent, reproducible fusion assessment combining thin-slice CT morphology with functional flexion-extension radiographs.

Our Solution

Intrinsic Imaging established a dual-reader adjudication model in which two fellowship-trained musculoskeletal radiologists independently assessed bridging bone on thin-slice CT and angular motion on flexion-extension radiographs, applying pre-specified quantitative thresholds. Discordant determinations proceeded to a third senior reader whose assessment was documented with full rationale. Standardized acquisition requirements ensured that CT slice thickness and radiographic positioning supported reliable measurement at every site.

Results

• Reader concordance of 96% before adjudication
• Fusion determinations upheld at FDA panel review
• Endpoint variability reduced by half versus investigator assessment