Myosteatosis evaluation using erector spinae and psoas muscles to predict adverse events during adjuvant chemotherapy for breast cancer.

Abstract

Myosteatosis (intramuscular adiposity) is predictive of chemotherapy toxicity in women undergoing adjuvant chemotherapy for breast cancer (BC). We evaluated a novel, user-friendly and cost-effective technique utilizing a Picture Archiving and Communication Systems (PACS) tool that is readily available in the electronic medical record (EMR), using skeletal muscle density (SMD) to detect myosteatosis and then compared PACS results with those derived from widely used body composition software (SliceOMatic, QC, Canada). Using retrospective data from a sample of women with early BC (Stage I-III) who had CT scan and received chemotherapy. Pearson correlation coefficients were used to compare SliceOMatic with PACS results. Associations of PACS results with chemotherapy-related adverse events were evaluated using multivariable (MV) log-binomial models adjusted for age, race, BMI, anthracycline-based therapy, and number of comorbidities. In 338 patients, mean age was 51, 32% were non-white, and 40% had obesity (BMI ≥ 30kg/m2). Correlation of SMD using SliceOMatic whole muscle measurements with PACS psoas muscle was 0.76 (p < .0001) and with PACS erector spinae muscle 0.91 (p < .0001). Using PACS psoas muscle, myosteatosis was associated with any adverse event [RR 1.66, CI 1.22-2.26 (p < .0001)], dose reduction [RR 1.63, CI 1.01-2.65 (p = .05)], and early treatment discontinuation [RR 2.14, CI 1.10-4.14 (p = 0.03)]. Using PACS erector spinae muscle, myosteatosis was associated any adverse event [RR 1.59, CI 1.11-2.27 (p = 0.01)] and dose reduction [RR 1.91, CI 1.07-3.42 (p = .03)]. Skeletal muscle density measures using PACS correlated strongly with SliceOMatic results and both are similarly predictive of chemotherapy-related adverse events.