Genetic testing taps at-risk women for MRI monitoring

Some women who carry a strong genetic susceptibility to cancer have consented to the surgical removal of their breasts to avoid likely cancer. The combination of reliable genetic testing, however, and close surveillance using breast MRI, mammography, and ultrasound now offers an alternative to such excision or the acceptance of a strong possibility of cancer.

Public health officials have learned to appreciate a synergistic relationship between BRCA1/BRCA2 genetic testing for identifying breast cancer susceptibility and breast MRI screening for diagnosis of the disease. The American Cancer Society announced in March guidelines recommending breast MRI and annual mammograms beginning at age 30 for women at increased risk of cancer (CA Cancer J Clin 2007;57(2):90-104.). The ACS recommendations listed candidates for screening MR:

• women with a BRCA1 or BRCA2 mutation;

• women with a parent, sibling, or child with a BRCA1 or BRCA2 mutation but not themselves tested;

• women harboring a 20% to 25% or more lifetime risk of breast cancer established by standard risk assessment tools; and

• women with a history of radiation treatment to the chest between ages 10 and 30, such as for Hodgkin's lymphoma.

BRCA1 and BRCA2 were first identified in 1994 and 1995, respectively. Estimates vary, but women with the two mutations may have a 60% to 70% lifetime risk of breast cancer and a 20% risk of ovarian cancer. In comparison, women from the general population carry a 12% risk of breast cancer and 2% lifetime susceptibility to ovarian cancer.

On a parallel track, radiologists have been refining protocols for clinical breast MR applications for more than a decade. They developed special pulse sequences, bilateral breast coils, dedicated high-field scanners, and MR-guided breast biopsy devices, all geared toward the optimum application of MRI for diagnosing breast cancer. Clinical trials found that breast MRI is highly sensitive to the disease but inappropriate for general screening because of high costs and a high false-positive rate.

These findings led researchers to seek out specific patient groups for whom breast MRI would be well suited. Then Dutch consortium reported in 2004 that MRI was more than twice as sensitive as mammography, though slightly less specific, for detecting invasive breast cancer in women with a familial or genetic predisposition.

Medical literature shows that MRI screening combined with conventional screening mammography can double the number of diagnosed breast cancers, according to the expert panel that wrote the ACS guidelines. Up to 1.5 million women possibly fall into the high-risk group.

Within days after the ACS guidelines announcement, The New England Journal of Medicine published evidence favoring breast MRI of the contralateral breast in women with newly diagnosed cancer in one breast. MRI identified mammographically occult invasive cancers in the contralateral breasts of 30 of 969 women with recently diagnosed unilateral breast cancers. Follow-up analysis underwritten by the American College of Radiology Imaging Network and conducted by Dr. Constance Lehman, a professor of radiology at the University of Washington, suggests that about 200,000 women diagnosed with breast cancer annually could benefit from an MR exam.

The demand for contralateral breast MR after a definitive diagnosis in one breast resonates in the experience of imaging services. Radiology Associates of New Mexico completed an audit of 400 consecutive procedures performed in 2004 in Albuquerque, Santa Fe, and Alamogordo. It found that 40% to 50% of the scans were performed on the contralateral breast, said Dr. Michael Linver, director of mammography.

MR results altered case management in nearly half of the practice's breast cancer cases. For 7.5% of the confirmed cases, cancer was also found in the contralateral breast. Patients were spared considerable grief, and the healthcare system was saved hundreds of thousands of dollars, he said.

"They would have gone through this $150,000 treatment for cancer on one side, only to discover cancer on the other side three, four, or five years later," Linver said. "The woman would then have had to have another $150,000 treatment and face the risk of a more advanced cancer that might not have been curable."

Screening MR buys time for BRCA1/BRCA2-positive women who will eventually undergo prophylactic mastectomy and oophorectomy, said Dr. Bruce Porter, MRI director at Swedish Medical Center in Seattle. Many women learn about their high-risk status in their late twenties and early thirties. Delaying surgery means a lot for a woman's self-confidence in attracting a mate, bearing children, and nursing them, he said.

No good screening test exists for ovarian cancer, said Robert Resta, a genetic counselor at Swedish Medical Center. Serial screening with CA-125 blood tests and transvaginal ultrasound are prescribed, but these are prone to false positives, and cancer often progresses to stage III status before detection. Prophylactic oophorectomy is recommended after the end of child-bearing, ideally between ages 35 and 40. An oophorectomy before age 40 also cuts the woman's breast cancer risk in half.

TEAM APPROACH KEY

A few medical centers and large group practices have implemented programs that bring together genetic testing, counseling, and imaging. ACS officials and others acknowledge that many more clinics will be needed to address the demand created by the new guidelines and published research. Medical personnel at Memorial Sloan-Kettering Cancer Center in New York City, Swedish Medical Center, and Radiology Associates of New Mexico all stress the importance of teamwork among genetics counselors, oncologists, and radiologists.

The geneticist/counselor plays a key role, according to Dr. Elizabeth A. Morris, director of breast MRI at MSKCC. Separating women who are truly at high risk from those who only think they are susceptible is an important starting point. An expert is required to document family histories and order testing to identify suspected mutations.

Failing to hire a geneticist to handle initial patient contacts and serve as team leader to channel patients to the appropriate specialists will assure the demise of an otherwise well-intentioned program, Resta said. At least one representative as well from radiology, oncology, and surgery should be encouraged to develop a special interest in BRCA1/BRCA2 testing, monitoring, and treatment. A monthly meeting bringing the various practitioners together improves communication and problem solving, and a team approach can coordinate details.

Local circumstances, however, also play a role. In Albuquerque, Radiology Associates of New Mexico was drawn into breast MR service after Dr. Bernadette Redd, a fellowship-trained breast MR specialist, expressed interest in joining the group so she could live in the Land of Enchantment.

Dr. Paul R. Duncan, a clinical oncologist with Hematology Oncology Associates in Albuquerque, supplies referrals. He first became involved in genetic testing in the early 1990s after taking an interest in familial cancer. Research on the regional frequency of Lynch syndrome, a genetically influenced form of colon cancer, led him attend a two-week minifellowship in Chicago on genetic testing and counseling sponsored by the American Society of Clinical Oncologists. Duncan applied the training personally and has run seminars throughout New Mexico to teach physicians how to recognize and respond to genetically linked breast and other cancers.

Duncan meets with 10 families a month, building on case files of 200 families evaluated over seven years. He orders BRCA1/BRCA2 genetic tests when appropriate and advises patients about the results, sometimes referring them back to their personal physicians. Other patients may have already consulted a surgeon. He continues to monitor some patients. Those who opt for surveillance are referred to Radiology Associates of New Mexico for breast MR and mammographic screening. The service also receives referrals from physicians whom Duncan has trained over the years.

Cancer predisposition accounts for 90% of the 800 families evaluated annually by Dr. Mark Robson, clinic director of the clinical genetic service at MSKCC. He draws out family and medical histories and informs patients about the meaning of genetic alterations, especially the probability of carrying a mutation, and how that information can be used. He identifies slightly more than 100 BRCA1/BRCA2 carriers annually.

At Swedish Medical Center, Resta evaluated slightly more than 400 patients in 2006. The volume rose rapidly after the publication of the NEJM paper and ACS guidelines. The center's daily volume record of nine patients was broken repeatedly in April and peaked at 14 cases on April 27.

Associated Radiologists of New Mexico performed 950 to 1200 MR exams last year, and a majority were for contralateral breasts after a definitive unilateral diagnosis. About 10% to 15% of the exams are screening tests on high-risk women.

The criteria for MR screening are broader at MSKCC than detection of BRCA mutations alone, Morris said. The program also accepts some patients who have a personal history of breast cancer, lobular carcinoma in situ, or atypia hyperplasia or have first-degree relatives who have had breast or ovarian cancer. Her group has published findings that show no difference in the detection rate between screening patients with atypia and patients who were not screened, but the cancers diagnosed with MR were found at an earlier stage.

The MSKCC screening regimen reflects an emerging de facto standard for the field. Morris alternates digital screening mammography with MRI every six months. No data exist to support a specific approach, she said. Mammography is recommended annually anyway, however, so performing breast MRI annually six months after mammography seems to make sense.

Many facilities still perform screen-film mammography. Swedish Medical Center performs screening mammography and ultrasound conjointly. Porter found that cancers of high-risk women lack a clearly malignant morphology, but their kinetics will be clearly abnormal and detectible with computer-assisted diagnosis software for MRI. Porter uses CADStream, a firm in which he has a financial interest. An alternative is DynaCAD, made by Invivo.

Radiologists and imaging technologists who have a passion for breast MR add to the likelihood of success. Performing breast MR a few times a month does not lead to competence, Porter said.

Excellence comes from performing examinations in volume. Radiologists at MSKCC must perform at least 100 supervised MR breast scans before being considered competent enough to read them independently, Morris said.

Repetition also helps radiologists master the art of MR-guided biopsy. A breast MR service is not comprehensive unless its radiologist can competently perform MR-guided interventions. MR-guided biopsies are difficult because needle insertion must coincide with gadolinium injection and wash-out. Lesion conspicuity depends on the enhancement features of the lesion.

Redd uses cheaper and easier ultrasound guidance instead of MR-guided techniques when she can, and she places a clip in the lesion so its position can be confirmed with MR if the results are negative. Still, tumors must be seen to be biopsied, and there are times when MRI is the only modality that will do the job.

Appropriate imaging technology must be in place. Breast MRI hardware and software have improved markedly since 2003 when the first MR-capable biopsy device was introduced, Morris said. Parallel imaging emerged that year as well, boosting the spatial resolution that breast MR requires. Morris notes that 1.5T MR scanners have been the workhorses of breast scanning. MR breast imaging is impossible without a dedicated bilateral breast coil, and it improves greatly when the scanner is equipped with high-powered gradients to optimize spatial resolution. Because of faster gradients, breast MR exams that previously took an hour are now completed in 30 minutes.

Porter complains about the diversity of techniques and equipment used for breast MR. While some radiologists still acquire 3 to 4-mm resolution, Porter works with a standard of 0.8 mm for his low-resolution dynamic scans. His high-resolution images are 0.65 mm thick and have an in-plane resolution of 0.5 mm.

Thinner slices help Porter do a better job imaging ductal carcinoma in situ. With the help of thin-slice imaging, he can cut five slices through a 3-mm DCIS lesion, for example. Radiologists who still acquire 3-mm slices would be lucky to get just one, he said.

American College of Radiology accreditation for breast MR may eliminate some inconsistencies. A voluntary program may be announced this year, Morris said.

FOCUSED RESULTS

The experiences of women whose cancers were detected because of screening highlight the success of these three programs. From among the Hispanic women that oncologist Duncan would identify as Sephardic Jews (see sidebar, page 49), Redd remembers a 27-year-old woman who underwent a breast MR exam in 2004 on the basis of her family history. The exam revealed a 5-cm breast tumor. Subsequent genetic testing found that she was a BRCA1 carrier. She was treated with lumpectomy and neoadjuvant chemotherapy. Routine screening after treatment has shown that she has stayed disease-free for three years.

In Seattle, Resta recalls the surprising results from breast MR on nonidentical twins. One sister had breast cancer at age 37. Though a mammogram indicated that her sister was normal, both twins were tested and deemed positive for BRCA1/BRCA2 mutations. Breast MRI for both sisters found what screening mammography had missed. The second sister was also diagnosed with cancer.

"This isn't just about an individual patient," Porter said. "It's about your sisters, your daughters, and all their subsequent children."

Genetic testing and counseling produce definitive answers about suspected familial links to breast cancer. Accelerated screening with breast MR and other modalities serves as a safety net to diagnose breast cancers that lurk in the future for many susceptible women.

Mr. Brice is Senior Editor of Diagnostic Imaging.