ARRS 2009: Diffusion-Weighted Imaging May Improve Accuracy of Breast MRI

Alice McCarthy

May 04, 2009

May 4, 2009 (Boston, Massachusetts) — Diffusion-weighted imaging (DWI) used to distinguish malignant vs benign breast lesions during magnetic resonance imaging (MRI) is helpful and may reduce the number of unnecessary breast biopsies, according to new research.

DWI is a technique involving the exchange of water molecules (diffusion) between breast tissue compartments. Diffusion rates vary between normal and pathologic tissue. With DWI MRI, the MRI machine is set to detect small restrictions in the free movement of water within the breast. Restricted diffusion areas show as hot spots on the MRI.

"The main aim of our work is to help save women from unnecessary breast biopsies," said lead investigator R. El-Khouli, MD, who lead the study when she was at Johns Hopkins University School of Medicine in Baltimore, Maryland, but is now with the National Institutes of Health in Bethesda, Maryland. She presented study results here last week at the American Roentgen Ray Society (ARRS) 2009 Annual Meeting.

Dr. El-Khouli explained that although MRI detects most breast cancer lesions, specificity rates are approximately 70% to 85%. Techniques like DWI have been introduced to improve specificity and provide functional information about breast tissue. "We are hoping to improve the specificity of MRI by adding new techniques like DWI and spectroscopy," she told Medscape Radiology.

DWI is quantified by a calculation called apparent diffusion coefficient (ADC) mapping, a calculated measure of water diffusion through the breast tissue. Previous studies have confirmed that ADC values do vary between malignant and benign breast masses, but overlap has been reported because benign breast changes can mimic malignancies. Further complicating ADC evaluation, ADC values of the human breast are notoriously complicated to normalize. These values are affected by the hormonal status of the female body. Among the different phases of the menstrual cycle, ADC values can vary 5.5%.

Compromises and Tradeoffs

"This technique currently is a compromise," Dr. El-Khouli said. "You don't want to miss a single cancer. At the same time you don't want to get women into the trouble of having biopsies and introducing unnecessary concern. It is a very challenging task."

In her study of 81 patients with 85 lesions, DWI was added to the standard MRI. A total of 60 lesions were known to be malignant. With DWI MRI, the technique correctly retrospectively correctly diagnosed 50 of these (83%). A total of 23 (92%) of 25 benign lesions were diagnosed correctly.

The investigators also calculated ratios of lesion ADC to glandular tissue ADC (L/GT). They found no difference in benign and malignant ADC and L/GT ratio between pre- and postmenopausal women.

"The ADC value of benign lesions ranged from 0.7 to 3.3 × 10−3 mm2/sec (mean, 2 ± 0.76), while for malignant lesions it ranged from 0.4 to 1.9 × 10−3 mm2/sec (mean, 1.1 ± 0.37) (p<0.05). For L/GT ratio, benign lesions ranged from 0.5 to 1.7 (mean, 1.1 ± 0.39) and malignant lesions from 0.28 to 0.98 (mean, 0.53 ± 0.15) (p<0.05). The area under the [receiver operating characteristic] curve for the ADC values was good (0.84), while it was excellent for L/GT ratio (0.92)," Dr. El-Khouli and colleagues write in their abstract.

Helpful Technique, But Needs Work

"I think DWI will be a helpful technique but there are a lot of things that need to be hashed out," commented session moderator Lonie Salkowski, MD, associate professor of breast imaging at the University of Wisconsin School of Medicine and Public Health in Madison.

"There is no tool to foster ease of use like what we have with other techniques, like computer-aided detection. It is one of its limitations because everyone makes their own ad hoc system of how to do it. I think in certain cases you can use it if you are really suspicious of the lesion and want to gain more information without committing to biopsy. It has some potential, but it needs more work in terms of refining the technique as well as in interpreting the results," Dr. Salkowski pointed out.

At Johns Hopkins University, where the study was conducted, DWI MRI breast screenings have been routine for all breast MRI patients for the past 18 months. When asked if this practice has reduced the number of breast biopsies, Dr. El-Khouli said, "We are trying to integrate the diffusion imaging to see what effect it has had on our biopsy numbers. But until we have more solid information we can't really apply what we see in research to the clinic."

To further this goal, Dr. El-Khouli and colleagues are developing techniques for automated processing of normalized ADC values for the entire breast.

Dr. Salkowski said that her team also does DWI for all of their breast MRI patients. "We are using it solely for research now to see where it might fit into our whole scheme and whether or not we need to improve it," she said. "Now, DWI breast MRI is very ad hoc [in terms of] how we interpret the curves and ADC values. We need a better system."

The study was supported by the National Institutes of Health. Study coinvestigator Dr. Peter Barker is a consultant for Philip Healthcare. Dr. Salkowski has no relevant financial disclosures.

American Roentgen Ray Society (ARRS) 2009 Annual Meeting: Abstract 129. Presented April 29, 2009.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: