Related Experiment Video
Updated: May 12, 2026
An R-Based Landscape Validation of a Competing Risk Model
Published on: September 16, 2022
Upgrade risk in intraductal papillomas: A retrospective analysis of real-world data and predictive model development
Jenni Kotola1, Anselm Tamminen2
1Department of Clinical Medicine, University of Turku, Finland.
Background:
In current practice, the traditional strategy of excising all IDPs has been replaced by more selective management. However, criteria for selecting patients for surveillance remain unclear, and no widely accepted predictive model exists.
Methods:
We retrospectively analyzed real-world data from 325 cases of IDPs diagnosed via core needle biopsy (CNB) at a tertiary teaching hospital between 2010 and 2023. We assessed upgrade rates to malignancy and evaluated potential predictive factors. Two previously published models were applied to our cohort, and a new model was developed based on our data.
Results:
Overall, 17% (55/325) of IDPs were upgraded to malignancy. Among lesions without atypia on CNB (n = 215), the upgrade rate was 8.8% (19/215), compared to 40% (23/58) in those with atypia (p < 0.001). Previously suggested models yielded modest results when applied to our study population. First model would have spared 11% (24/215) of patients from surgery, while the second model would have spared 17% (36/215), with one missed upgrade. Our model identified all upgraded cases and would have spared 33% (72/215) of non-atypical IDPs from surgery.
Conclusions:
Atypia on CNB is a strong predictor of upgrade to malignancy. Existing models showed limited utility in reducing unnecessary surgeries. Our proposed model demonstrated improved performance and may support more individualized management of IDPs.
