Is Pre-Laparotomy CT-Derived Body Composition Associated with Surgical Site Infections after Trauma? An Exploratory Analysis

Authors:
William Rieger, Onur Sahin, Nicole Noto, Stephanie Martinez Ugarte, Renee Green, Anne Jeckovich, Parker Towns, Ronald Bilow, Rafael Bravo Santos, Julie Holihan, Lillian Kao

Body of Abstract:
Background: Sarcopenia and obesity have been well correlated to surgical and infectious outcomes in oncologic surgery, but it is unclear whether this relationship exists after surgery for trauma. We aimed to determine if pre-operative computed tomography (CT)-derived body composition measures are associated with surgical site infection (SSI) after trauma laparotomy. We hypothesized that higher skeletal muscle and lower visceral fat are associated with fewer SSIs.

Methods: We conducted a retrospective review of adult (≥16 years) patients who underwent pre-operative CT and subsequent laparotomy for trauma at a single center from 3/2021-7/2023. Patient details were obtained from the medical record and SSI status was determined per the Centers for Disease Control definition. Body composition data was obtained from Picture Archiving and Communication Systems as a single de-identified image at the 3rd lumbar vertebral level. Scans were analyzed using CoreSlicer 1.0 to quantify subcutaneous and visceral fat, abdominal wall muscle, and bilateral psoas muscle areas (Figure). Measures were analyzed independently, as ratios between muscle and fat areas, and as calculated indices per height: Skeletal Muscle Index (SMI), Visceral Adiposity Tissue Index (VATI), and Subcutaneous Adiposity Tissue Index (SATI). Measures and indices were analyzed both continuously and binomially via sex and age-adjusted values. Univariate and multivariable statistics were performed with a priori covariates restricted to those available pre-operatively.

Results: Of 655 patients, 428 patients who had pre-operative CT scans were included. Of those, the median age was 34 years (IQR 23-45), most were male (n= 316, 74%), nearly half suffered blunt injury (n= 221, 52%), and most were severely injured (median injury severity score 24, IQR 14-34). The majority of patients were non-obese (n= 305, 72%), and non-sarcopenic (n= 336, 78.5%), with a median skeletal muscle area of 176 cm2 (IQR 145-207), visceral adipose tissue area of 100 cm2 (IQR 37-163), and subcutaneous adipose tissue area of 169 cm2 (IQR 72-266). Median muscle to fat ratio was 0.60 (IQR 0.20-1.0), with 1.0 (SD 0.25) as expected mean for all-aged, non-obese males. Fifty-one patients (12%) had a documented SSI. After adjustment, SMI (OR 1.00, CI 0.99-1.01), VATI (OR 1.00, CI 0.99-1.01), and SATI (OR 1.02, CI 0.99-1.04) were not significantly associated with SSIs, but a higher total muscle to total fat ratio was associated with a decrease in SSIs (OR 0.65, CI 0.45-0.94).

Conclusions: While traditional sarcopenic definitions and body composition indices did not apply well to our generally young trauma population, a higher muscle to fat ratio correlated with decreased SSI. Though patients who were not stable enough to undergo pre-laparotomy CT were excluded, earlier CT-based information and SSI risk stratification might better inform clinicians and aid in delivery of personalized infection prevention.