Multi-parametric muscle and fat correlation of computed tomography parameters to outcomes in a total hip arthroplasty population

Abstract

Background: Cross-sectional imaging is not currently used in planning Total Hip Arthroplasty (THA). The aim of our study is to determine correlations between CT parameters and outcomes following THA. Methods: A prospective registry of patients who underwent total joint arthroplasty was reviewed for patients who: (1) underwent THA, (2) had a CT between 1 year before and 6 months after surgery, and (3) completed perioperative WOMAC and Harris Hip Score (HHS) questionnaires. Two readers measured CT parameters, yielding mean Hounsfield Units, area, average diameter, and perimeter of the psoas major, gluteus medius and minimus muscles. A segmentation algorithm determined visceral and subcutaneous fat area, and waist circumference. ICC was calculated for each measurement to examine inter-reader agreement. Regression analyses were performed to select measurements with most impact on outcome scores. Results: Twenty-eight patients met inclusion criteria (17 female, 11 male), having mean (+/- standard deviation) age of 54.4 +/- 14.8 years and BMI 29.0 +/- 6.3 kg/m2. Correlations were found between HHS and age (0.650, p = 0.018), height (-1.263, p = 0.009), visceral-to-subcutaneous fat area ratio at the psoas level (0.511, p = 0.018), and waist circumference at the psoas level (1.759, p = 0.002). Inter-reader analysis showed ICC > 0.850 for all measurements. Conclusion: Age and height, as well as CT-derived visceral-to-subcutaneous fat area ratio and waist circumference significantly correlate with postsurgical HHS scores following THA. Our study suggests that parameters derived from cross-sectional CT imaging can be useful additional preoperative planning tool for THA.