How a Novel Prediction Model May Help to Reduce False-Positive Recalls in Breast Cancer Screening
Posted: Friday, October 22, 2021
A novel prediction model, developed by Bianca M. den Dekker, MD, of the University Medical Center Utrecht, the Netherlands, and colleagues, may potentially help to reduce the false-positive rate in supplemental dense-tissue MRI breast cancer screening. According to the study published in Radiology, the prediction model is able to combine MRI findings with clinical characteristics such as family history, age, and body mass index.
“The reduction of the false-positive recall rate is an important issue when considering the use of breast MRI as a screening tool,” commented Dr. den Dekker in a press release from the Radiological Society of North America. “Our prediction models may identify a substantial number of false-positives after first-round supplemental MRI screenings, reducing false-positive recalls and benign biopsies without missing any cancers.”
Study data were prospectively collected from the randomized Dense Tissue and Early Breast Neoplasm Screening trial. Of 454 women who received positive MRI results in a first supplemental screening round, 375 had false-positive results. Using the full prediction model based on all collected clinical characteristics, 45.5% of false-positive recalls and 21.3% of benign biopsies could have been prevented without missing any cancers (AUC = 0.88; 95% confidence interval = 0.84–0.92). Solely focusing on MRI findings and age, the model could have prevented 35.5% of false-positive MRI results and 13.0% of benign biopsies (AUC = 0.84; 95% confidence interval = 0.79–0.88; P = .15).
It was noted that the false-positive rate in the study group dropped from 79.8 per 1,000 screenings in the first round to 26.3 per 1,000 in the second round. Dr. den Dekker explained that this could be partially due to the availability of prior MRI exams, which permit comparison for interval change. The researchers also added that subsequent incident screening rounds and validation studies using data from different populations are warranted.
Disclosure: For full disclosures of the study authors, visit rsna.org.