Abstract
INTRODUCTION: Cancer-associated malnutrition is a pervasive and under-recognized complication that profoundly impacts treatment tolerance, clinical outcomes, and quality of life. Despite the availability of multiple nutritional screening and assessment tools, these instruments differ widely in sensitivity, specificity, and ease of integration into clinical workflows, and no universally accepted standard exists. This review critically examines the current landscape of malnutrition assessment in oncology, summarizes tool performance across populations and cancer types, and proposes strategies-such as artificial intelligence-enabled models and internationally harmonized protocols-to improve diagnosis, treatment planning, and overall patient outcomes. METHODS: A comprehensive literature search was conducted across PubMed, Web of Science, Embase, and Elsevier databases, covering studies published up to 13 March 2025. Medical Subject Headings (MeSH) were used to identify terms including "malnutrition," "cachexia," "cancer," "nutritional status assessment," "nutritional screening," and "nutritional screening tool." Boolean operators refined the strategy, and a two-stage screening excluded studies with irrelevant populations, outcomes, or designs, as well as non-peer-reviewed sources. RESULTS: Significant heterogeneity was found in tool performance and applicability across cancer types, clinical settings, and demographic subgroups. General instruments such as the Malnutrition Universal Screening Tool (MUST) and Nutritional Risk Screening 2002 (NRS-2002) demonstrated strong predictive validity in broad clinical use, whereas condition-specific tools like Patient-Generated Subjective Global Assessment (PG-SGA) offered superior sensitivity in high-risk populations, including patients with gastric or head and neck cancers. However, variability in thresholds, assessment frequency, and validation approaches highlights the urgent need for standardization. DISCUSSION: Current assessment strategies are limited by subjectivity, static single-point evaluations, and inconsistent implementation. Future innovations should integrate artificial intelligence, dynamic longitudinal monitoring, and multimodal data analytics to develop objective and personalized evaluation systems. Establishing globally harmonized standards will be crucial to improving nutritional care, reducing malnutrition-related morbidity, and enhancing survival and quality of life for patients with cancer.