However, other models incorporating DWI or quantitative MRI features offered lower AUC ranges for malignancy detection (71 to 75 percent) and detection of invasive or high-grade DCIS (78 to 90 percent) in comparison to models utilizing DCE-MRI.
“While lesion visibility on DWI and quantitative MRI features derived from high-temporal resolution acquisition were associated with both malignancy and invasive/high-grade DCIS, these features did not show clear added value over standard qualitative DCE MRI for differentiating suspicious calcifications,” wrote lead study author Vivian Y. Park, M.D., who is affiliated with the Department of Radiology at the University of Washington in Seattle, and colleagues.
Three Key Takeaways
- Superior diagnostic performance with DCE-MRI. Dynamic contrast-enhanced MRI demonstrated up to a 30 percent higher detection rate for invasive breast cancer or high-grade DCIS in women with suspicious mammographic calcifications compared to a combination of mammography and clinical variables.
- Limited added value from DWI or quantitative MRI. Diffusion-weighted imaging and quantitative MRI features did not provide significant diagnostic improvement beyond standard qualitative DCE-MRI in differentiating malignant from benign calcifications.
- Potential to reduce unnecessary biopsies. In patients with BI-RADS 4A/4B calcifications, use of standard DCE-MRI could have avoided approximately 70–75% of unnecessary biopsies, particularly in cases of non-enhancing calcifications when invasive or high-grade DCIS was absent.
For patients with mammographic BI-RADS 4A and 4B calcifications, the researchers noted significantly reduced malignancy rates for non-enhancing calcifications and an absence of invasive cancer or high-grade DCIS detection in these cases. They also noted that three missed malignancies in this subgroup had intermediate-grade DCIS.
“ … Standard DCE-MRI could have reduced the number of planned biopsies in mammographic BI-RADS category 4a/4b calcifications by 77.8 percent (49/63) in our study cohort or reduced the total number of unnecessary biopsies by 73.0 percent (46/63) when accounting for MRI-prompted additional benign biopsies,” maintained Park and colleagues.
(Editor’s note: For related content, see “What New Research Reveals About Preoperative Breast MRI,” “Post-NAC Breast MRI Without Calcifications Associated with 65 Percent Higher Likelihood of Pathologic Complete Response” and “The Reading Room Podcast: Current and Emerging Insights on Abbreviated Breast MRI, Part 2.”)
In regard to study limitations, the authors acknowledged that small sample size prevented analysis of individual mammography features and thwarted assessment of MRI for detecting cases of low-grade DCIS alone. The researchers also conceded relatively low numbers of visible lesions on DWI MRI and DCE MRI.