A recent study from Washington University School of Medicine in St. Louis reveals a major advancement in the early detection and risk prediction of breast cancer. Extending from prior work on regular full field digital mammograms, they extend risk prediction to images from the now more common imaging with breast tomosynthesis. The research describes the development and validation of a novel breast cancer risk prediction model using synthetic digital breast tomosynthesis (DBT) images.

The study is published March 11 in Cancer Prevention Research.

The study compared data from women who had routine mammography screening in St. Louis to data from the Emory Breast Imaging Dataset (EMBED), validating risk-prediction methods developed by the research team in past studies.

We developed a model that estimates breast cancer risk over the next five-year period, calibrated to national risk strata, says senior author Graham Colditz, MD, DrPH, associate director of prevention and control at Siteman Cancer Center and chief of the Division of Public Health Sciences at WashU Medicine. It’s imperative that this model accurately identifies women at higher risk to facilitate better screening and preventive strategies, or risk reduction.

One of the strengths of this research lies in its focus on a racially and ethnically diverse population, addressing a significant gap in prior studies that often overlooked minority groups. The WashU Medicine cohort included over 10,000 women, with 27% being non-Hispanic Black women, while the validation cohort included 46% of the same demographic, highlighting the model’s adaptability and potential for wide-reaching impact.

Importantly, the researchers observed that the calibration of their model showed good agreement between predicted outcomes and actual cases, identifying 6% of the population at high risk over the next five years. This risk group accounted for 20% of total diagnosed cases, emphasizing the model’s predictive accuracy and its potential to streamline current clinical practices in breast cancer management.

Follow-up of cohorts is crucial for the development and validation of risk models, says co-author Shu (Joy) Jiang, PhD, a statistician, data scientist and associate professor of surgery in the Division of Public Health Sciences.

By only utilizing synthetic DBT, the model optimizes the use of existing screening tools while commencing follow-up of women from their first DBT exam. This potentially reduces unnecessary follow-ups and allows for timely interventions.

Colditz and Jiang partnered with Debbie Bennett, MD, a professor of radiology and chief of breast imaging for the Mallinckrodt Institute of Radiology at WashU Medicine, to conduct this and previous, related studies.

This study represents a critical step forward in breast cancer risk management, says Colditz, who is the Niess-Gain Professor of Surgery and deputy director of the Institute for Public Health at WashU Medicine. With further validation, this model could support more equitable and efficient breast cancer screening programs, especially in community settings.

Future research will aim to expand the inclusion of long-term follow-ups, offering more extensive evidence on the utility of DBT in routine care. There remain opportunities to improve the model by incorporating even more diverse racial and demographic groups, which will be essential for maximizing the model’s applicability across the general population.

The findings presented in this study are pivotal for the field of breast cancer research, promising to enhance early detection, personalize patient care, and ultimately, improve clinical outcomes for women at risk of breast cancer.

Source: https://surgery.wustl.edu/new-risk-prediction-model-shows-promise-for-early-breast-cancer-detection/