Body weight and composition changes, primary tumor side, and resection as survival predictors in metastatic colorectal cancer.

Abstract

850 Background: Common symptoms of metastatic colorectal cancer (mCRC) are sarcopenia and weight loss. We investigated whether weight loss, sarcopenia and primary tumor resection can predict overall survival among patients (pts) with mCRC Methods: Consecutive pts newly diagnosed with mCRC in a community hospital between 2012-2014 we reviewed. We gathered initial and 6-month follow up data on weight, standardized muscle mass measure (estimated psoas muscle area/m2 height, sEPA) and subcutaneous fat (SQ). Results: There were 109 pts (age 66.9±14.5, range 37-93 years). Chemotherapy recipients were younger (n = 59, 61.2±13.3 years) and survived longer (22.3 versus 5.3 months, p < 0.001) as compared to best supportive care patients (n = 50, 73.6±13.0 years, p < 0.001). There were no baseline and 6-month differences between pts with resected versus in situ primary tumor in age (p = 0.074), baseline weight (p = 0.728) or percent weight loss (p = 0.404), albumin (p = 0.322), hemoglobin (p = 0.301), creatinine (p = 0.791), initial standardized EPA (p = 0.866), percent of sEPA loss (p = 0.952), and percent subcutaneous thickness loss (p = 0.477). Cohort was further dichotomized by median anthropometric changes at 6 months: -7.1% for weight loss, -6.2% sEPA, -3.3% SQ fat. Cox proportional hazard models demonstrated that anthropometric measures and receipt of systemic chemotherapy were the strongest predictors of survival with their predictive strength surpassing traditional predictors as age, tumor sidedness, primary tumor resection, initial BMI and serum albumin level. Conclusions: Weight loss and anthropometric changes are strongly associated with shorter survival. Prognostic characteristics of loss of weight, muscle and fat should be investigated further using more robust datasets. [Table: see text]