Hounsfield Unit-Calculated Bone Mineral Density Loss Following Combat-Related Lower Extremity Amputations

Abstract

Update This article was updated on December 19, 2023, because of a previous error. Specifically, an extra footnote was inadvertently included at the bottom of Table III. That extra footnote has now been removed. As such, the text that had previously read “*HU = Hounsfield units. †The values are given as the mean and the standard deviation.” now reads simply “*HU = Hounsfield units.” Background: After combat-related lower extremity amputations, patients rapidly lose bone mineral density (BMD). As serial dual x-ray absorptiometry (DXA) scans are rarely performed in this setting, it is difficult to determine the timeline for bone loss and recovery or the role of interventions. However, a strong correlation has been demonstrated between DXA BMD and computed tomography (CT) signal attenuation. We sought to leverage multiple CT scans obtained after trauma to develop a predictive model for BMD after combat-related lower extremity amputations. Methods: We reviewed amputations performed within the United States military between 2003 and 2016 in patients with multiple CT scans. We collected pertinent clinical information, including amputation level(s), complications, and time to weight-bearing. The primary outcome measure was the development of low BMD, estimated in Hounsfield units (HU) from CT scans with use of a previously validated method. One hundred and twenty-eight patients with 613 femoral neck CT scans were available for analysis. A least absolute shrinkage and selection operator (LASSO) multiple logistic regression analysis was applied to determine the effects of modifiable and non-modifiable variables on BMD. A random-effects model was applied to determine which factors were most predictive of low BMD and to quantify their effects. Results: Both amputated and non-amputated extremities demonstrated substantial BMD loss, which stabilized approximately 3 years after the injury. Loss of BMD followed a logarithmic pattern, stabilizing after 1,000 days. On average, amputated limbs lost approximately 100 HU of BMD after 1,000 days. Other factors identified by the mixed-effects model included nonambulatory status (−33.5 HU), age at injury (−3.4 HU per year), surgical complications delaying weight-bearing (−21.3 HU), transtibial amputation (20.9 HU), and active vitamin-D treatment (−19.7 HU). Conclusions: Patients with combat-related lower extremity amputations experience an initially rapid decline in BMD in both intact and amputated limbs as a result of both modifiable and non-modifiable influences, including time to walking, amputation level, surgical complications, and age. The paradoxical association of vitamin-D supplementation with lower HU likely reflects this treatment being assigned to patients with low BMD. This model may assist with clinical decision-making prior to performing lower extremity amputation and also may assist providers with postoperative decision-making to optimize management for prophylaxis against osteoporosis. Level of Evidence: Therapeutic Level III. See Instructions for Authors for a complete description of levels of evidence.