HistoMet: A Pan-Cancer Deep Learning Framework for Prognostic Prediction of Metastatic Progression and Site Tropism from Primary Tumor Histopathology

2026-02-07

AI Summary

Accurately predicting cancer metastasis and its specific sites directly from primary tumor histopathology images remains challenging. To address this, this work proposes the first multiple instance learning framework that explicitly models the clinical two-stage decision process: first assessing metastatic risk and then predicting metastatic sites. The method integrates language-defined and data-adaptive metastasis-related concepts through a pretrained pathology vision–language model and concept-guided representation learning, achieving both decision awareness and concept alignment. Evaluated on a pan-cancer cohort of 6,504 patients, the approach significantly reduces downstream workload at 95% sensitivity. Among metastatic cases, it achieves a macro F1 score of 74.6 and a macro one-vs-rest AUC of 92.1, demonstrating high accuracy and interpretability.

Technology Category

• Machine Learning: Multi-instance/Multi-view Learning
• Search and Optimization: Metareasoning and Metaheuristics
• Computer Vision: Language and Vision

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Metastatic Progression remains the leading cause of cancer-related mortality, yet predicting whether a primary tumor will metastasize and where it will disseminate directly from histopathology remains a fundamental challenge. Although whole-slide images (WSIs) provide rich morphological information, prior computational pathology approaches typically address metastatic status or site prediction as isolated tasks, and do not explicitly model the clinically sequential decision process of metastatic risk assessment followed by downstream site-specific evaluation. To address this research gap, we present a decision-aware, concept-aligned MIL framework, HistoMet, for prognostic metastatic outcome prediction from primary tumor WSIs. Our proposed framework adopts a two-module prediction pipeline in which the likelihood of metastatic progression from the primary tumor is first estimated, followed by conditional prediction of metastatic site for high-risk cases. To guide representation learning and improve clinical interpretability, our framework integrates linguistically defined and data-adaptive metastatic concepts through a pretrained pathology vision-language model. We evaluate HistoMet on a multi-institutional pan-cancer cohort of 6504 patients with metastasis follow-up and site annotations. Under clinically relevant high-sensitivity screening settings (95 percent sensitivity), HistoMet significantly reduces downstream workload while maintaining high metastatic risk recall. Conditional on metastatic cases, HistoMet achieves a macro F1 of 74.6 with a standard deviation of 1.3 and a macro one-vs-rest AUC of 92.1. These results demonstrate that explicitly modeling clinical decision structure enables robust and deployable prognostic prediction of metastatic progression and site tropism directly from primary tumor histopathology.

Problem

• Metastatic Progression
• Site Tropism
• Histopathology
• Prognostic Prediction
• Pan-Cancer

Innovation

• decision-aware MIL
• concept-aligned representation
• metastatic site tropism
• vision-language model
• prognostic prediction