ABSTRACT
BACKGROUND: Malnutrition is prevalent in cancer patients and is associated with inferior outcomes. We examined the association between malnutrition, as measured by the Subjective Global Assessment (SGA), and chemotherapy dose reduction in patients with gastrointestinal malignancies. We hypothesised that malnutrition, defined by a patient's baseline SGA, would be associated with a greater degree of chemotherapy dose-reduction, with the implication of greater chemotherapy related toxicity. DESIGN: We reviewed chemotherapy dosing and treatment related toxicity for patients enrolled in a prospective Gastrointestinal Cancer Registry over their first 8 weeks of treatment. We compared results between well-nourished and malnourished patients. RESULTS: Malnourished patients were more likely than well-nourished patients to have their starting chemotherapy dose reduced from standard published dosing (67% versus 35%, p=0.0001). Despite attenuated initial dosing, malnourished patients received a smaller fraction of planned chemotherapy (mean 80±23% versus 90±15% of cycle 1, p=0.005), primarily due to toxicity-related dose reductions. After controlling for age, gender, Eastern Cooperative Oncology Group performance status (ECOG), albumin, smoking status, body habitus, and weight loss, malnutrition remained the strongest independent predictor of the magnitude of chemotherapy dose reduction (estimate -10.3%, 95% confidence interval -19.0 to -0.1.6%, p=0.020). CONCLUSIONS: Malnutrition is an independent predictor of chemotherapy dose-reduction for toxicity. This study highlights the practical significance of malnutrition in gastrointestinal malignancies and provides a baseline for future nutrition intervention studies to improve chemotherapy tolerability in malnourished patients.