Simulating a virtual tumor board with large language models: a pilot study in NSCLC patients receiving immunotherapy

Multidisciplinary teams (MDTs) are fundamental to cancer care but face increasing burdens. This pilot study evaluates a large language model (LLM) simulating a tumor board for a clinically complex cohort of non-small cell lung cancer (NSCLC) patients, using a full-context guideline injection methodology to ground its reasoning in authoritative standards. Ten real-world NSCLC cases were presented to Google’s Gemini 2.5 Pro using prompt engineering. The model was primed by providing the complete National Cancer Institute guidelines as in-context source data in a structured JSON file. AI-generated recommendations were scored against the institutional human MDT. The LLM demonstrated high performance, achieving mean scores of 4.9/5.0 for content accuracy, 5.0/5.0 for internal consistency, and 4.4/5.0 for clinical applicability. Importantly, no safety concerns were identified in the AI’s recommendations. However, the model did not generate any novel insights beyond those considered by the human MDT. An LLM primed with comprehensive guidelines can accurately and safely replicate MDT recommendations for complex NSCLC cases. The combination of guideline injection and meticulous prompt engineering is a critical strategy for ensuring LLM reliability. This positions these models as powerful decision-support tools to augment, not replace, expert clinical workflow. Cancer care often involves a team of specialists, called a multidisciplinary team (MDT), who meet to decide on the best treatment for each patient. Recently, artificial intelligence (AI) tools, especially large language models, have shown potential to help in such decision-making processes. In this pilot study, we tested whether an advanced AI system (Google Gemini 2.5 Pro) could simulate the discussions of an MDT for patients with non-small cell lung cancer receiving immunotherapy. To make sure the AI’s advice was safe and evidence-based, we gave it the full U.S. National Cancer Institute treatment guidelines in a structured format before asking it to evaluate 10 real patient cases. The AI’s recommendations were then compared with the actual decisions made by our hospital’s MDT. The AI performed very well, with high accuracy, logical consistency, and good clinical relevance. Importantly, none of its suggestions posed risks to patient safety. Our findings show that AI, when guided by trusted medical standards, can safely support cancer doctors in their decision-making. It has the potential to make MDT discussions more efficient, but it should be seen as a support tool – not a replacement for expert judgment.