Risk Assessment of Anticancer Treatments Beyond Performance Status: A Prospective Study in 277 Cancer Patients

9620 Background: Alterations of performance status (PS) and of nutritional status are predictors of acute toxicity following anticancer treatment. Resting energy expenditure (REE) is often increased in cancer patients and may be a major determinant for the alteration of nutritional status and the development of precachexia and cachexia. We investigated whether abnormal metabolism could predict early acute toxicity (EAT). Methods: In this prospective observational monocentric study, REE was measured by indirect calorimetry before anticancer treatment initiation. C-Reactive Protein (CRP), albumin, transthyretin and PS were collected. Measured REE was compared with predicted REE as defined by the Harris-Benedict formula. Patients were classified as hypometabolic (REE < 90%), normometabolic (90-110%) or hypermetabolic ( > 110%). Toxicity was assessed after the first cycle of treatment. An EAT was defined as any event leading to unplanned hospital admission, dose reduction, treatment delay ( > 7 days) or discontinuation. Results: A total of 277 patients (pts) with solid tumors (gastro-intestinal 25%, genitourinary 23%, thoracic 22%) were included. Sex ratio: 1.25; median age: 63 years (20-91). Most of the pts (211; 76%) had locally advanced or metastatic disease and were treated with chemotherapy (246; 89%) or tyrosine-kinase inhibitor (30; 11%). Calorimetry revealed pts with either normo- (29%), hyper- (51%) or hypometabolism (20%). A subset of 59 pts (21%) experienced an EAT. The occurrence of EAT was associated with poor PS (2-3 vs 0-1: OR = 2.04 [1.12-3.73], p = 0.029), low albumin ( < 35 vs ≥ 35g/l: OR = 2.39 [1.03-5.54], p = 0.048), inflammation (CRP ≥ 10 vs < 10mg/l: OR = 2.43 [1.35-4.37], p = 0.004) and abnormal metabolism (abnormal vs normal: OR = 2.36 [1.13-4.95], p = 0.023). In multivariate analysis, elevated CRP was an independent predictor of EAT (p = 0.047). The REE was associated with higher sensitivity (83%) than CRP (55%), PS (41%) and albumin (17%) to predict EAT. Conclusions: Pts with abnormal metabolism experience more early acute toxicity. The measurement of resting energy expenditure might improve the detection of patients at risk for toxicity and missed with the standard clinical evaluation.