Abstract
Background: There is a lack of knowledge about the biological process of intercorporal bone graft remodelling towards successful lumbar spine fusion with bridging vital bone between two adjacent vertebrae. Hounsfield Units (HU) highly correlate with Bone Mineral Density (BMD) and changes in HU may be valuable as proxy measure to monitor ongoing fusion. The aims of this study were to explore the feasibility to quantify BMD changes in the intercorporal bone graft after spinal fusion on consecutive CT-scans, and to explore whether trends in the individual changes in HU over time may serve as a proxy measure for successful fusion or non-fusion.Methods: A retrospective case series was conducted using available one-year and two-year CT-data from patients after posterior lumbar interbody fusion surgery. First, a standardized measuring procedure was developed, incorporating internal phantomless calibration, to establish the HU values of the grafted region of interest using consecutive CT-slices. Subsequently, changes in the participants’ bone graft HU over time between their first and second year after surgery were explored.Results: Between one and two years after surgery, seven out of nine (78%) participants showed an increase in their bone grafts’ HU, ranging from 3% to 41%. In two participants HU-values decreased 14% and 32%, respectively. The intraobserver reliability of the HU measuring procedure was excellent (ICC 0.93 (95% confidence interval 0.91 to 0.95)).Conclusion: The majority (78%) of participants in the current study showed increases in their bone graft HU, which suggests ongoing bone graft remodelling towards lumbar fusion. This result corresponds with the established percentages of achieved rigid bony fusion after lumbar spinal fusion surgery reported in the literature. Thus, prospective follow-up of CT-based HU measurements may have potential as diagnostic measure to monitor bone graft remodelling in time and trends may predict successful fusion or non-fusion. The measurement procedure developed in this study, using internal phantomless calibration, showed excellent intraobserver reliability and merits further study in larger patient numbers.