Individualized risk prediction for malignant transformation in intraductal papillary mucinous neoplasms of the pancreas

• Applicant: Dr. Viola Maria Pleines
• Subject Area: General and Visceral Surgery
• Term: since 2026
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 575952590

Due to the widespread availability of high-resolution imaging techniques, cystic lesions of the pancreas are increasingly detected incidentally. Intraductal papillary mucinous neoplasms (IPMN) are a subgroup of these lesions that are precursors to pancreatic cancer and vary greatly in terms of their biological behavior with regard to malignant potential. Pancreatic surgery can be associated with considerable risks. Simultaneously delaying surgery can mean overlooking a potentially curable carcinoma. Clinical management of IPMN therefore poses a considerable challenge. Previous guidelines have been based primarily on radiological criteria and overestimate the actual risk of malignancy, resulting in overtreatment in up to 30% of cases. There is a lack of reliable tools for personalized risk assessment. The planned project aims to develop a decision-making model that systematically links individual malignancy and surgical risk with patients' personal treatment preferences (Cancer Risk Threshold, CRT). The work program comprises four steps: In the first step, a prognostic risk model is created based on a comprehensive cohort from New York University (≈150 operated, ≈10,000 non-operated IPMN cases) using multivariate regression and machine learning. This combines clinical and radiological parameters for the individual prediction of high-grade dysplasia or invasive carcinoma. Secondly, discrete choice experiments are used to record the personal risk thresholds (CRT) of 350 patients, i.e., the threshold at which the benefits of surgery outweigh the risks. Latent class analysis is used to identify typical profiles of individual risk perception and preferences. In the third step, these risk profiles are integrated into a decision model that uses quality-adjusted life years to individually assess whether surgery or surveillance is preferable. This will be clinically piloted on a select group of patients. Finally, the model will be expanded to include molecular diagnostics (liquid biopsy) to evaluate the additive diagnostic benefit of molecular markers in risk stratification. For the first time, IPMN cases that have undergone surgery and those that have been conservatively monitored will be analyzed together, thereby reducing systematic biases. In addition, innovative molecular markers and individual patient preferences will be integrated to significantly improve the quality of clinical decision-making. In the long term, the resulting tool could help to identify high-risk cases at an early stage, avoid unnecessary surgery, and thus improve the quality of life of those affected.

• DFG Programme: WBP Fellowship
• International Connection: USA
• Host: Professor Dr. Greg Sacks