Femoral Neck Hounsfield Units as an Adjunct for Bone Mineral Density after Combat-Related Lower Extremity Amputation

Objectives: To correlate femoral neck Hounsfield units (HUs) measured on a computed tomography (CT) scan to dual-energy x-ray absorptiometry (DEXA) T-scores allowing evaluation of bone mineral density (BMD) over time after lower extremity trauma–related amputation. Design: Retrospective cohort study. Setting: United States military trauma referral center. Patients: Military combat-related lower extremity amputees with both DEXA and CT scans within 6 months of each other. Intervention: None. Main Outcome Measures: Correlation between femoral neck comprehensive mean HUs and BMD and HUs threshold for low BMD. Results: Regression model correlation (r) between CT HU and DEXA T-score was r = 0.84 [95% confidence interval (CI) 0.52–0.94] and r = 0.81 (95% CI 0.57–0.92) when CT imaging was separated from DEXA by less than 4 and 5 months, respectively. Beyond 5 months separation, correlation decreased to r = 0.60 (95% CI 0.29–0.80). Using a receiver operator characteristic curve for mean comprehensive HUs to determine low BMD with 4-month cut-off, a threshold of 151 HUs was 97% sensitive and 84% specific to identify low BMD, whereas 98 HUs was 100% sensitive and 100% specific to identify osteoporosis. Conclusion: Using opportunistic CT, clinicians can reliably estimate BMD in trauma-related amputees. This information will inform providers making decisions regarding weightbearing and bisphosphonate therapy to limit further loss. Future phases of this study will aim to use this correlation to study the effects of weightbearing advancement timing, bisphosphonate therapy, and interventions on the natural history of bone density after amputation.