Abbreviated breast MRI scans demonstrated comparable diagnostic accuracy to full multiparametric protocol MRI scans for women with extremely dense breasts, according to findings from the DENSE trial published in Radiology.
MRI scans have greater diagnostic accuracy than mammograms for women with dense breast tissue in identifying breast cancer; however, they are more costly and time-consuming. The investigators sought to determine a minimal MRI screening protocol that reduced the cost and time while still maintaining high diagnostic accuracy.
“The shortest abbreviated protocol demonstrated good performance, comparable to that of the full multiparametric protocol, while being up to four times faster to acquire and up to two times faster to read,” stated study author Wouter B. Veldhuis, MD, PhD, Associate Professor, Department of Radiology, University Medical Center, Utrecht University, the Netherlands.
Study Methods and Results
Seven radiologists analyzed a subset of patients from the DENSE trial, performing incremental readings of first-round screening MRIs among women with extremely dense breasts and who did not have cancer at the time of mammography.
Different sequences of abbreviated MRI scans were tested in four incremental steps. First, the radiologists tested both high-temporal low-spatial and low-temporal high-spatial dynamic T1-weighted series, up to 2 minutes after contrast was injected into the patient. The final step included all full-protocol sequences.
Each of the seven radiologists assessed all MRI scans and determined a BI-RADS score. Pooled sensitivity and specificity were calculated for all of their readings as well as reading time per step.
A total of 518 MRI scans from 518 women were included in the first-round screening. Of these patients, 83 had breast cancer, including 68 invasive cancers and 15 ductal carcinomas in situ. The radiologists completed a total of 2,072 reads.
Pooled sensitivity for the abbreviated protocol was 84.3% (95% confidence interval [CI] = 77.7%–89.2%) compared with 85.9% (95% CI = 80.0%–90.3%) for the full protocol (P = .68). No specificity difference was noted between the abbreviated (73.9%; 95% CI = 70.7%–76.9%) and the full protocols (75.8%; 95% CI = 72.8%–78.5%; P = .39).
The abbreviated MRI protocol had a pooled reading time of 49.7 seconds (95% CI = 48.5–50.9 seconds) vs 96.4 seconds with the full MRI protocol (95% CI = 94.3–98.5 seconds; P < .001). Scanning times were 70% to 80% shorter depending on the hospital and scanner vendor.
“Reducing the examination time and noise levels potentially improves patient experience,” Dr. Veldhuis stated. “Moreover, shorter reading and scan times may allow implementation of MRI in national screening programs, making it available for all women with extremely dense breasts.”
Disclosure: For full disclosures of the study authors, visit pubs.rsna.org.
