Diagnostic Imaging
September 2003

BREAST IMAGING

Clamor for breast MRI could be boon or boondoggle

Consumer buzz drives expansion, posing political and clinical issues for breast centers

Three years ago, Barbara Regis had her first and last screening mammogram. Never again, said the 44-year-old graphic artist in California's Silicon Valley. As for risking undetected breast cancer, she thought she would take her chances.

These days, Regis is having second thoughts. Attracted by a local imaging center's advertisement for breast MRI that promises accurate cancer detection without painful compression, she is considering paying the $1200 screening fee out of pocket just for the peace of mind.

Some might say Regis is misinformed. Others would claim that she is a patient empowered. In either case, she isn't alone.

Breast MR has a buzz factor among consumers that is unmatched by any other imaging test. The number of breast MR scans performed in the U.S. is increasing by 40% annually, according to IMV Medical Information Division of Des Plaines, IL.

In recent years, the technique has rapidly diffused from academic sites to community hospitals and independent breast care centers. The average center performs one to three breast MRI exams daily. Many of these sites are grappling with strategies for managing walk-ins who want a screening breast MR.

But the growing popularity of breast MRI isn't due solely to self-referred patients. In competent hands and in select patient populations, the technique's sensitivity approaches 100% for detecting invasive breast cancer, a persuasive statistic for breast imagers and surgeons alike.

"It's got marketing legs," said Gerald Kolb, president of Breast Health Management, a consulting firm in Bend, OR. "When you have breast MRI at one center, it becomes a huge marketing tool, and the rest of the market seems to buy into it. People are looking at MRI as an indicator of quality for a breast center."

Breast MRI also has the potential to prop up plummeting profit margins at breast centers that have suffered through years of mammography reimbursement reductions. Like other breast interventions, breast MRI makes money, said Dr. William Poller, director of breast imaging at Allegheny General Hospital Breast Center in Pittsburgh.

"It makes money whether you are globally billing or professionally billing. It's the interventional procedures that make money in the breast center, and this one, because of the technical fee, is pretty good," Poller said. "But we're not doing this to round out our wallets. MR is a good study. After ultrasound, it's the most important thing we've come upon."

SLEUTHING TOOL

Most sites providing breast MRI rely on clinical indications shaped in equal parts by evolving research findings and payer policies. The most common indicators include local staging of breast cancer, monitoring response to chemotherapy, and implant imaging. Another indicator is also gaining acceptance: screening high-risk patients such as breast cancer gene carriers and those with prior lymphoma, and the contralateral breast in patients diagnosed with cancer.

Within these select subsets, every breast imager using MRI today can tell a dramatic tale of how the technique enabled him or her to discover an otherwise hidden cancer.

"Last week, I performed two biopsies on a woman who had an MRI because of a high-risk lesion near the site of a previous biopsy," said Dr. Edward B. Cronin, director of breast imaging at Hartford Hospital in Hartford, CT. "On the MR, two masses enhanced in her opposite breast that couldn't be seen on the mammogram."

At biopsy, both lesions were diagnosed as high-grade infiltrating carcinomas. The MRI also found residual enhancement at the margins of the preexisting lumpectomy site. A third cancer was found just beyond the margin.

"She wound up having an early diagnosis of three infiltrating breast cancers," Cronin said. "All different, and all separate. They would never have been found at this stage were it not for MRI."

Such cases are not rare, Cronin said. Unanticipated lesions that turn out to be cancerous are found in about 15% of the select population of high-risk patients who undergo breast MRI at Hartford.

"People are sitting up and taking notice because this is all disease that was missed before," he said. "Breast MRI might very well contribute to lowering recurrence rates because we are finding things that might not have come to the forefront for several years."

Looking for subtle clues is the name of the game in breast imaging, said Dr. Kamilia Kozlowski, medical director at Knoxville Comprehensive Breast Center in Tennessee. MRI adds confidence to differential diagnoses and decisions about whether a patient should proceed with a lumpectomy versus a mastectomy (Figure 1).

"Often, the mammogram is negative and the ultrasound is equivocal," Kozlowski said. "The biopsy comes back benign, but because of the ultrasound findings you wonder if you have a sampling problem. With MRI, if there's no uptake, I don't worry."

The clinical pedigree of breast MRI has a few splotches. As it is still being standardized, its accuracy varies considerably with technique. Typical sensitivity ranges between 95% and 98%, while specificity varies widely from 38% to 97%. The exam's negative predictive value is about 95%, while its positive predictive value falls between 65% and 75%. Common false positives include fibroadenoma, lobular carcinoma in situ, and atypical duct hyperplasia, among others.

Thus, developing a strict indications list and sticking to it is critical, said Dr. Bruce Porter, medical director of First Hill Diagnostic Imaging in Seattle.

"Breast MR is a problem-solving tool," Porter said. "A thorough workup with mammography and ultrasound should be performed first. You shouldn't turn to MR at the first sign of something suspicious."

VARIABLE TECHNIQUE

The majority of the sites performing breast MRI rely on general-use high-field systems, employ a dedicated breast coil, and administer intravenous contrast. A handful of sites are using the 0.5T dedicated breast MRI system developed by Aurora Imaging Technology. The exam takes 30 to 45 minutes. The complexity and time-consuming nature of breast MRI demand appropriate patient selection, Porter said.

Clinical indicators can be a moving target in breast MRI. An informal poll of members of the National Consortium of Breast Centers (NCBC) conducted by Diagnostic Imaging found that many sites providing the service are confused about when breast MRI is appropriate.

"The way breast MRI is being utilized around the country, there is no clear consensus," Kolb said. "Do only patients who are scheduled for a lumpectomy get it? Or should all patients who are going on to some kind of surgical intervention get it? If so, then we've got a pretty big pot. Someone-either the American College of Radiology or the Society of Breast Imaging-needs to draw the line."

The International Working Group for Breast MRI is developing standardized technical specs, clinical indications, and reporting lexicon, but its findings are not widely distributed.

The lack of consensus has led to spotty reimbursement for some applications. Use of breast MRI to screen high-risk patients is one example. Payers in the Northwest reimburse for the exam, but those in Tennessee don't. Despite recent research from the U.S. and Europe that appears to support the application, its acceptance is not universal.

In a study published in Radiology in July (226:773-778), researchers at the University of Pennsylvania demonstrated the feasibility of using breast MRI in a screening role to evaluate the contralateral breast in patients with newly diagnosed breast cancer. Fifteen patients out of 182 had suspicious lesions on MRI; of these, seven had malignant results and eight findings were benign.

Despite the results, screening all women at high risk may not be economically feasible, said lead author Dr. Steven G. Lee. Identifying patients at highest risk for synchronous bilateral breast cancer, as well as those in that group who would benefit most, is the first step.

A trio of other studies based on work at Memorial Sloan-Kettering Cancer Center, the University of Bonn in Germany, and Erasmus Medical Center in Rotterdam, the Netherlands, revealed similar findings. Each uncovered high sensitivity but troubling shortcomings when it came to specificity and false positives.

After a presentation of results from the German and Dutch MRI studies at the American Society of Clinical Oncology meeting in June, moderator Dr. William Gradishar noted that the use of MRI for high-risk women is "appropriate and promising" in identifying tumors, but its ultimate value requires more study.

Yet the promise of MRI in high-risk patients has proved persuasive to the American Cancer Society, which recommends use of ultrasound and breast MRI in its 2003 guidelines for screening women known to be at increased risk for cancer.

LAUNCHING THE SERVICE

Experts caution that performing breast MRI demands a high level of clinical expertise. Familiarity with breast pathology is critical. Most say it is easier for a breast imager specializing in mammography to learn MRI than it is for an MRI expert to successfully read breast scans.

"You are interacting with the patient much more than in other areas of radiology, and you are also working with the pathology," Porter said. "As a result, you have to be able to adapt. Compared with heads or knees or spines, these are complicated studies."

At First Hill Diagnostic Imaging, Porter employs what he calls a 4D approach, beginning with a series of T1-weighted images using fat saturation, followed by a bolus of gadolinium and a second series of images acquired over time. Digital subtraction produces a high-contrast image that allows creation of a series of maximum intensity projections. The technique tracks the image enhancement through time, allowing assessment of lesion morphology as well as biology (Figure 2).

In addition to clear clinical indications and good technique, other important aspects of breast MR include timeliness, the ability to follow up on findings, and rapport with local surgeons.

"Timeliness-the ability to move quickly when an MRI is indicated-is an important aspect of providing this service," said Dr. Alan Semine, chief of breast imaging at Newton-Wellesley Hospital in Newton, MA. "Evaluating a patient with MRI as soon as possible after she has been diagnosed with breast cancer is important clinically as well as for the patient's peace of mind."

For centers with a dedicated breast MRI unit onsite, quick access is not an issue. But not all facilities can afford the $1.2 million investment in a dedicated breast MRI unit. For most sites, which must schedule breast exams among other MRI studies, timeliness can be challenging.

"In some areas of the country, breast is treated like a second-class citizen on the magnets, as compared with neuro and other cases," Poller said.

FOLLOW-UP PROTOCOL

Developing a protocol for following up on suspicious breast MRI findings is another must. Not all agree on what the most appropriate follow-up should be. Options include MRI-guided localization, ultrasound-guided biopsy, and MRI-guided biopsy using a customized coil.

Some breast imagers believe that sites offering breast MRI should also offer MRI-guided biopsies; others say ultrasound-guided biopsies are faster and more cost-effective. But it is clear that tracking suspicious findings, not just identifying them, is a critical component of the service.

"You need to be thinking ahead to biopsy and be prepared to biopsy those lesions that you can't see except with MRI," said Dr. Frederic Kelcz, a breast imaging specialist at the University of Wisconsin, Madison.

It's common for the center to see referrals from sites up to 70 miles away that are not equipped to follow up on breast MRI findings.

Getting local surgeons involved while the breast MRI service is still in the planning stages is another consideration, said Dr. Justin P. Smith, a breast imaging specialist at First Hill Diagnostic Imaging.

"Reach out to the key breast surgeons in the community, let them know you are thinking of providing breast MR, and identify the key indications," Smith said. "You're looking for buy-in. Make the surgeon your ally."

SCREENING CONTROVERSY

While much variety exists in technique, equipment, clinical indications, payment policies, and follow-up strategies at breast MRI sites around the country, there is general agreement on one issue: Breast MRI is no match for mammography when it comes to general breast cancer screening, and its use as a general screening tool is wrong. Opposition to the practice is passionate.

Consultant Kolb refers to one national chain that promotes breast MRI for general screening as "Ameri-scam" and suggests the company should be put out of business.

Suggesting that breast MRI is a better screening tool than mammography does a public disservice, according to Dr. Mark Novick, medical director of Manhattan East Breast Imaging in New York City.

"It's a matter of overselling your technology," he said.

Moreover, he said, from a public policy perspective, it's unrealistic to expect that breast centers could survive on bargain-basement screening rates, which would result from any shift in public policy regarding screening benefits for MRI instead of mammography. Kolb agrees.

"The reason we now have universal screening in this country is that the price could be held low," Kolb said. "To think that we can substitute a $1000 procedure for an $80 one is at best wishful thinking. My concern is that we could get a more expensive test okayed as a screening tool and then have screening benefits pulled away because they are too expensive."

Breast imagers also worry that distribution of MRI will never match that of mammography, even with its increasing presence among community centers. They believe that for reasons of cost, access, and general adequacy, mammography should continue to be the gold standard screening tool.

"Even though MRI is more sensitive than mammography, it is not going to be embraced any time soon as a screening tool," Novick said. "When it comes to general screening, perhaps the greatest sensitivity is not what society is looking for. It may be looking for greater specificity at lower cost. That's why mammography will remain entrenched."

A vocal dissenter is Dr. Craig Bittner, founder and medical director of AmeriScan, a national chain that promotes breast cancer screening with MRI at its seven imaging centers across the U.S. Twelve additional sites are scheduled to open soon. Bittner believes patients are getting what they pay for with mammography screening: a low rate for what he calls an outdated technology.

"This has become a political and economic issue that ignores the facts of medical science," said Bittner, who studied with breast MRI pioneer Dr. Werner Kaiser at the University of Bonn to develop his clinical expertise. "Women deserve better than mammography."

AmeriScan centers equipped with MRI perform an average of 100 breast MRI screens monthly. Many women pay for the exam themselves, although some insurers reimburse for the test as a wellness exam, albeit on a sliding scale that ranges from 20% to 100%.

"Insurance companies are being forced to pay because the public is aware now that mammography is an extremely inaccurate technology," he said. "That is certainly the case in our practice."

Bittner predicts it will be only a few years before breast MRI becomes the recognized mainstay for breast cancer screening and detection.

"That day will come, not because radiologists will look to medical science and make good decisions, but because the opponents in radiology will have no choice," he said. "The public is going to demand it, and thank goodness we live in a society where there is a free flow of information. Because when medicine is controlled by insurance companies and self-interested physicians, medical science doesn't stand a chance."

Porter, who has spent 12 years developing breast MRI technique and applications, finds the idea of general screening both "inappropriate and disturbing."

"We still have a lot of work to do to get this established as a clinical problem-solving tool and to get radiologists to the point where they are comfortable and competent using it," he said. "If we are going to start saying that the key indication is having money and breasts, then we are going to have real problems."

And when it comes to creating peace of mind for women like Regis, the lack of correlation between an MRI screen and other breast studies could lead to confusing findings, provoking more anxiety than it relieves.

"General screening is going to give breast MRI a bad rap," said Kozlowski of Knoxville Comprehensive Breast Center. "It's not a good thing for the modality at all."