Could MRI Be a Better Breast Cancer Screening Tool Than Mammography?


Key Points

  • In 443 patients with breast cancer who had negative screening mammograms, additional screening by MRI identified 11 cancers not revealed by mammography.
  • Screening was accomplished via an abridged MRI protocol that was quickly performed and read.
  • MRI may identify prognostically “relevant” cancers that mammography typically misses.

German investigators reported at the 2013 Breast Cancer Symposium in San Francisco that an abridged magnetic resonance imaging (MRI) protocol can accurately detect cancers among women whose mammographic screenings were negative (Abstract 1). MRI, therefore, may reveal the type of tumor that mammography typically misses—and can do so in a time-efficient fashion, thus making MRI feasible for breast cancer screening, said Christiane K. Kuhl, MD, of RWTH Aachen University in Aachen, Germany.

“We already know that we find more cancers with MRI screening than with mammographic screening, including more invasive cancers and more ductal carcinoma in situ (DCIS),” she said.

Improving the Detection of Prognostically Relevant Cancers

These cancers may be more “prognostically relevant” than those detected by mammography, which tend to be “irrelevant,” she suggested. “The aim of new, nonmammographic screening strategies is not necessarily to detect more cancers but to improve the detection of cancers that are prognostically relevant.”

Current MRI protocols used for screening are identical to those used for diagnostic purposes and as such are time-consuming to acquire and to read. Dr. Kuhl and colleagues, therefore, designed an abridged breast MRI protocol that would be more suitable for screening purposes. The MRI consists only of the postcontrast subtracted (FAST) images and their maximum intensity projection (MIP). The MIP images allow readers to quickly detect enhancement while the FAST images provide further categorization of these enhancements, she explained.

Prospective Study of Abridged MRI Modality

The researchers evaluated the screening MRI protocol in an observational study of 443 asymptomatic women at intermediate or slightly increased lifetime risk of breast cancer. All subjects had no abnormalities on digital mammography.

Experienced breast MRI radiologists rated the MIP images as positive or negative depending on the presence or absence of significant enhancement. They read the FAST images to provide BIRADS (Breast Imaging Reporting and Data System) scores and the full diagnostic MRI protocol to compare the abridged modality for diagnostic yield and accuracy. 

Breast Cancers Detected in Women With Negative Mammograms

In 606 screening rounds, 11 breast cancers were detected, including 4 DCIS and 7 invasive cancers, for a cancer yield of 18.3 per 1,000. All lesions were Tis or T1, N0, and M0; virtually all were grade 2–3; and the median tumor size was 8.4 mm. No interval cancers were diagnosed.

The positive and negative predictive values of the abridged protocol were comparable to those of the full diagnostic protocol. The negative predictive value was 98.9%, Dr. Kuhl reported.

The MIP/FAST screening also compared favorably with mammographic screening with regard to the time needed to acquire or review images. The abridged protocol required 3 minutes of acquisition time (vs 21 minutes for the full protocol) and about 30 seconds of reading time.

The abridged MRI screening modality “allowed a substantial additional yield of biologically relevant invasive cancers and DCIS in this cohort of women at moderately or slightly increased risk of breast cancer,” Dr. Kulh concluded. “As long as so many women die of breast cancer, the search for improved screening strategies must continue. True screening breast MRI may be such a strategy.”

Randomized Trial Needed

Monica Morrow, MD, the Anne Burnett Windfohr Chair of Clinical Oncology and Chief of the Breast Service, Department of Surgery, at Memorial Sloan-Kettering Cancer Center, New York, congratulated Dr. Kuhl and colleagues for “an innovative approach to making screening MRI more accessible.”

But Dr. Morrow cautioned against using low-grade DCIS as a surrogate for breast cancers that are “not important.” She said, “We need to keep in mind that every single prospective randomized trial of DCIS radiotherapy vs no radiotherapy showed that the risk of progression to invasive cancer was equal, regardless of the grade of DCIS. I’m not sure that this, in and of itself, is a good surrogate.”

Dr. Morrow also suggested that if the two modalities detect different types of cancers, a randomized trial would be necessary to prove MRI superior to mammography, perhaps with an endpoint of breast cancer–specific survival.

Dr. Kuhl agreed on the need for additional data if MRI is going to be used for screening. “But this study shows proof of principle,” she said. “Can we think of using MRI for mass screening? I think the answer is ‘yes, maybe.’”

She further explained, “I don’t think MRI will necessarily identify different cancers, but I think MRI is specifically ‘blind’ to low-grade DCIS, or detects cancers that don’t have the proteomic tools required for growth. Being blind for these types of DCIS may be a virtue rather than a disadvantage.”

The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.