Multimodal Artificial Intelligence System for Risk-Adapted Cancer Survivorship Surveillance: A Multicenter Target Trial Emulation.

PURPOSE: The growing population of cancer survivors faces immense monitoring burdens because of rigid follow-up guidelines, such as the intensive surveillance schedules recommended by the National Comprehensive Cancer Network. To address this issue, we engineered a multimodal artificial intelligence (AI)-based decision support system that integrates biological domain data (magnetic resonance imaging) and physical treatment domain data (radiation therapy dose maps) to guide individualized care.

METHODS AND MATERIALS: Using stage II nasopharyngeal carcinoma (N = 2148 across 5 centers) as a model, we first implemented a target trial emulation framework to confirm the safety of treatment deintensification and establish a baseline for streamlined surveillance. We then trained a transformer architecture to predict individualized treatment failure timing and translated these predictions into a risk-adapted surveillance strategy.

RESULTS

In the target trial emulation, omitting concurrent chemotherapy demonstrated comparable survival outcomes with concurrent chemoradiation therapy across all cohorts, establishing a safely deintensified clinical baseline. Subsequently, the AI system achieved high-fidelity predictions, with an area under the curve of 0.991 internally and 0.986 in the multi-institutional external validation cohort. This AI-guided strategy substantially reduced the need for follow-up visits for over 90% of failure-free patients, while recommending a maximum of only 6 visits for high-risk individuals over a 5-year period, demonstrating a high sensitivity for detecting true failures.

CONCLUSIONS

This generalizable AI framework can seamlessly complement the current National Comprehensive Cancer Network guidelines, offering a transformative, data-driven solution that reduces the global monitoring burden of cancer survivorship care.