BONN, Germany -- MRI outperformed mammography in detecting ductal carcinoma in situ, especially high-grade lesions, researchers here found.
BONN, Germany, Aug. 10 -- MRI outperformed mammography in detecting ductal carcinoma in situ, researchers reported.
It was especially effective in detecting the high-grade lesions likelier to progress to invasive carcinoma, Cristiane K. Kuhl, M.D., of the University of Bonn, and colleagues, reported in the Aug. 11 issue of The Lancet.
Of 167 women with a diagnosis of ductal carcinoma in situ (DCIS), 92% were diagnosed before surgery by MRI compared with only 56% detected by mammography, the researchers said.
MRI showed even better performance on high-grade lesions, picking up 98% of them; 48% were missed by mammography.
Although commonly held to be a direct precursor of invasive breast cancer, ductal carcinoma in situ is a heterogeneous disease. High-grade, or pure, lesions are more likely to progress faster to invasive disease than the low-grade forms, the researchers wrote.
Until now, MRI has been considered less sensitive than mammography in picking up high-grade lesions, but recent changes in diagnostic criteria have made MRI more feasible. Some clinicians have found it more effective than mammography.
To test the two techniques, the researchers studied 7,319 women who were referred over a five-year period to the breast center at an academic tertiary care facility. All women had an MRI in addition to mammography.
Ductal carcinoma in situ was diagnosed in 167 of the women who had undergone both imaging tests preoperatively, 93 (56%) of these cases were diagnosed by mammography and 153 (92%) by MRI (P
The higher sensitivity of MRI did not translate into an unduly high number of false positives, the researchers said. The positive predictive values of both methods were comparable.
Stratified by imaging method, mammography produced 421 true-positive diagnoses and 344 false-positives, for a positive predictive value of 55%. With MRI there were 599 true-positives and 413 false-positives, providing a positive predictive value of 59%
MRI could help improve the ability to diagnose ductal carcinoma in situ, especially high nuclear-grade lesions, the researchers said, but added that recommendations for the use of MRI screening are not appropriate at this point.
These observations came from a single center that offers a high level of interpretation for both types of studies, they said, and since breast MRI is currently used only rarely in clinical practice, few radiologists can offer the necessary expertise compared with those skilled in reading mammograms.
A systematic multi-institutional screening trial will be necessary to further investigate the clinical role of MRI for diagnosing ductal carcinoma, and randomized trials are required to determine the effect on recurrence rates or mortality, Dr. Kuhl said.
However, the investigators said, their findings seem important, since the presence or absence of necroses are increasingly used to predict outcomes and thus guide decision-making.
"Current concepts with regard to the prognostic importance of necroses could conceivably be different if the non-calcifying, mammography-occult, MRI-only [lesions] are also taken into consideration," they concluded.
In an accompanying comment Carla Boetes, M.D., and Ritse M. Mann, M.D., of Nijmegen Medical Center in The Netherlands, wrote that these findings "can only lead to the conclusion that MRI outperforms mammography in tumor detection and diagnosis."
Therefore, they said, MRI should no longer be regarded as an adjunct to mammography but as a distinct method to detect early breast cancer.
A large multi-center breast-screening trial with MRI in the general population is essential, they said.
This study was supported by Frderverein fr Radiologie an der
Universitt Bonn, the Ricomagus Moesche Donation, and
The editorial writers reported no conflict of interest.