Measurement of Periarticular Subcutaneous Fat on CT Images and Adverse Outcomes Following Total Knee Arthroplasty

Background

Obesity is associated with a higher rate of wound complications following primary total knee arthroplasty (TKA). With readily available computer tomography (CT) images from robotic-assisted TKA, we analyzed measurement of fat content on preoperative CT images as a possible predictor of wound complications following primary TKA.

Methods

Patients who underwent robotic-assisted TKA at one institution in 2018 were included in this retrospective cohort study. Two independent reviewers measured three SCF areas at different axial CT cuts and normalized them by dividing the area of the distal femur. These areas were distributed into 4 groups. Any wound complication that required clinical or surgical intervention was reviewed and analyzed. For further comparison, prepatellar SCF thickness ratio measured on CT scan and BMI were grouped and analyzed similarly for wound complications. We also analyzed any association of SCF measurement with secondary outcomes such as operative time, length of stay, readmission, and reoperation.

Results

One hundred fifty patients with diagnosis of osteoarthritis, mean age of 64 years and BMI of 34.3 kg/m² were included in this study. Ninety-one patients (61%) were female. Normalized SCF measurements at 2 cm above the patella, mid-patella, and tibial tubercle had excellent intraclass correlation coefficient at 0.987, 0.989, and 0.989, respectively. When SCF at 2 cm above patella was analyzed, Group 1 (smallest amount of SCF) had a significantly higher wound complication rate compared with Groups 2 and 3 combined (18.9 vs 5.3%, p=0.036). Group 4 (largest amount of SCF) also had a significantly higher wound complication rate compared with Groups 2 and 3 combined (18.9 vs 5.3%, p=0.036).

Conclusions

Accurate and consistent measurement of periarticular fat around the knee based on axial CT images demonstrated that moderate amount of fat is associated with better clinical outcomes following primary TKA. Our study did not find any clinical significance of gender difference in fat distribution. Therefore, more studies should be undertaken to evaluate for any clinical association of gender-specific fat distribution and to confirm our finding that a certain amount of fatty tissue is necessary for improved outcomes following TKA.