Federal Grant Award Summary

The National Cancer Institute (NCI) awarded MD Anderson Cancer Center a $500,384 Project Grant under the Cancer Cause and Prevention Research program (CFDA 93.393) on July 15, 2026, with a completion date of June 30, 2031. This award funds the development of advanced statistical methodologies for transfer learning in clinical decision-making that integrate diverse data sources including established risk prediction models, surrogate outcomes, and family medical history. The project will deliver three primary methodological innovations: rank-based transfer learning approaches that leverage relative risk rankings from existing models while accommodating new biomarkers and population heterogeneity; systematic frameworks for incorporating surrogate data to improve primary endpoint analyses without requiring complex joint modeling; and empirical likelihood approaches with bias-correction methods that integrate family health history data while accounting for recall bias and measurement error.

To ensure broad dissemination and clinical utility, the research team will develop and provide open-access, user-friendly software tools implementing these methodologies. The proposed innovations will be validated through application to motivating clinical studies to demonstrate improvements in statistical efficiency and clinical decision-making outcomes. This research addresses critical gaps in existing transfer learning approaches that often rely on restrictive assumptions incompatible with real-world heterogeneous data sources, particularly in cancer prevention and risk assessment contexts aligned with NCI's strategic priorities.