Assessment of muscle mass with computerised tomography in patients with incurable gastrointestinal cancer. A prospective single centre study

Abstract

Body composition is often affected in patients with incurable cancer, but the prevalence of sarcopenia is unknown. Our aim was to evaluate sarcopenia as predictor of overall survival in a cohort of patients with incurable gastrointestinal cancer and furthermore to evaluate if this cohort had different characteristics than patients, from an identical cohort who accepted participation in a RCT. Design and methods In this single centre study, we prospectively included a cohort of patients with incurable gastrointestinal cancer nutritionally at risk (NRS 2002≥2). Patients were screened but refused participation in an RCT testing supplemental HPN. To assess sarcopenia, data on skeletal muscle mass (SMM) from the cross-sectional area of L3 were assessed using computerized tomography scan (CT scan). SMM evaluation was included if a CT scan was available within 60 days from the inclusion date. Differences in survival were tested according to sarcopenia and modified Glasgow Prognostic Score (mGPS). Survival was compared between the patients who refused to participate in the RCT and patients who actually did participate. Results Eligible for inclusion were 187 patients, and 165 had a CT scan available for analysis. Most prevalent diagnosis was pancreatic cancer (52%), median age was 70.5 (41.2-89.4), median BMI 22.3 (14.4-36.8) and 99% were receiving chemotherapy. Sarcopenia was present in 78% of the overall cohort, more women (88%) than men (70%) were sarcopenic at inclusion. There was a positive correlation between BMI and SMM, but SMM accounted for only 8% of the variance in BMI. Conclusions Prevalence of sarcopenia was high in this cohort of patient with incurable gastrointestinal cancer; SMM did positively correlate to BMI, but only accounted for minor variations. mGPS was in the multivariate cox regression model predictive of survival and sarcopenia did not add to this elevated risk.