AI-enhanced oncology MDT 2.0: from multi-modal data synergy to value-based care reconstruction - a systematic review of clinical efficacy and socioeconomic benefits

Multidisciplinary team (MDT) meetings are the cornerstone of modern oncology, however, meantime face persistent challenges including variability in decision-making, time constraints, and unequal access to subspecialty expertise. The emergence of large language models (LLMs) has catalyzed a new paradigm-AI-Enhanced Oncology MDT 2.0. This systematic review synthesizes evidence from 29 peer-reviewed studies (2020-2026) to evaluate the clinical efficacy and socioeconomic implications of integrating AI into oncology MDT decision-making. Key findings demonstrate that AI systems achieve concordance rates with human tumor boards of 62-76% across multiple cancer types (excluding molecular tumor board studies), with substantial agreement in guideline-driven decisions but limitations in complex, individualized cases. AI demonstrates notable strengths in standardizing guideline-adherent recommendations and supporting molecular target identification, but still struggles with nuanced clinical judgment and patient-specific factors. From a socioeconomic perspective, AI offers potential for value-based care reconstruction through reduced MDT preparation time and democratized access to sub-specialty expertise, though rigorous cost-effectiveness evidence remains limited. The review concludes that AI-Enhanced Oncology MDT 2.0 represents strategic augmentation rather than replacement of human judgment.