Diagnostic Imaging
December 2002

X-RAY VISION

MRI screening plan puts mammography in crosshairs

"You deserve better protection than a mammogram." A line from The New York Times' series criticizing the state of mammography in the U.S.? A comment by European or Canadian researchers questioning the efficacy of screening mammography?

Nope. It's a headline from a half-page newspaper ad promoting a screening center's MRI-based breast screening program. It has many in radiology in an uproar, and it's an indication of the type of things we may begin to see as consumer-based screening moves forward.

The ad, developed for radiologist Dr. Craig Bittner's AmeriScan screening clinics in California and Arizona, outraged experts here in the San Francisco Bay Area and has caused a stir in other quarters within radiology as well. While MR has a growing role in breast imaging, no research supports its use in screening asymptomatic, low-risk women. That the ad capitalizes on some of the studies that have unfairly maligned x-ray mammography as a breast cancer screening tool is certain to rile mammographers even more.

MRI does hold some potential for screening, but that potential has not been examined and is not well understood. It can be 100% sensitive for cancer, but its specificity falls into the 60% range in experienced hands and probably less among those new to the practice. False positives can be caused by a number of factors (fibroadenoma, lobular carcinoma in situ, atypical duct hyperplasia, fibrocystic changes, proliferative changes, papilloma, sclerosing adenosis, duct hyperplasia).

These issues have left MRI in a backup position to mammography in general screening settings: it can be used to stage and localize cancer or find suspected cancer and to screen high-risk women such as those with a family history of breast cancer or those who carry the BRCA1 or BRCA2 genes.

Further arguing against its use is the cost: AmeriScan charges between $1700 and $2200 for MRI breast screening, compared with $50 to $150 for an x-ray mammogram.

AmeriScan offered a spirited defense of its program in a report on its Web site (http://www.mribreastscreen.org/), but we suspect that some of the authors quoted would not recommend MRI as a screening strategy for a general population. A key problem with AmeriScan's arguments is that data from MRI breast imaging studies on high-risk women are being interpreted as supporting MRI screening for a general population. We can't assume that's true: MRI's ability to spot cancers in a general population may not duplicate its success in high-risk populations; false positives may turn out to be higher than with traditional x-ray mammography.

But as angry as this may make many mammographers and other radiologists, there's not much than can be done about it, and the likelihood is that more situations like this will occur. When you have entrepreneurial medicine like consumer-based screening, you will have practitioners who take positions well outside of the mainstream and make a lot of noise in doing so. Sometimes they promote these positions through negative comparisons with accepted and reasonably efficacious procedures like mammography. Best practices as defined by the ACR or other professional organizations don't have much clout in this setting. In fact, entrepreneurs may get a lot of mileage by simply positioning themselves as against the "establishment."

Compounding this is a willingness by medical consumers to pay more attention to the discussion and the debate and to seek out their own answers when support for a technique seems equivocal. Consumer screening has introduced significant new dynamics in radiology. Its effects will be felt for a long time to come.

What are your thoughts on this topic? Please e-mail me at jhayes@cmp.com.