COMMENTARY

Does Elevated Breast Cancer Risk Require MRI?

Kathy D. Miller, MD

Disclosures

April 06, 2012

This feature requires the newest version of Flash. You can download it here.

Hi. This is Dr. Kathy Miller with another Medscape Oncology video blog. This time, I am talking to you about some of the recent studies looking at various aspects of breast imaging.

I am particularly struck by the results of the ACRIN6666 Trial.[1] ACRIN is the American College of Radiology Imaging Network, the research network of our radiology colleagues. This trial specifically looked at whether adding regular ultrasound screening to annual mammography would find more cancers in women at higher risk of developing breast cancer. After 3 years of annual mammogram and ultrasound screening, women were invited to have an MRI as a single-screening event to see whether that would find even more cancers.

Although these women were at somewhat higher risk (they typically would have been classified as intermediate risk), they represent a group of women for which we currently don't have enough data to recommend MRI as part of their screening, and for whom the risks of overcalling findings on imaging, and obtaining biopsies that find only a few additional cancers, are quite real.

So here are the numbers. They screened about 2600 women over this 3-year period and 110 (4%) of those women were diagnosed with cancer during that time. Of those cancers:

  • 33 were seen only on mammogram;

  • 32 were seen only on ultrasound;

  • 26 were seen on both mammogram and ultrasound;

  • 9 were seen only on MRI; and

  • 11 were not seen on any imaging.

These 11 interval cancers were found by a clinical evaluation abnormality that the patient or their physician found between annual mammography screening.

The rest of the numbers can get fairly technical, so let me remind you that sensitivity tells us how many real cancers were found by a certain imaging modality. Specificity is the flipside. Specificity looks at the false positives. Specificity looks at how many additional noncancers were seen that then required additional evaluation. You can boil that down to the positive predictive value (PPV). PPV is, very simply, when the radiologist looks at the films, calls you, and says, "I think Mrs. Jones needs a biopsy" -- what is the likelihood that that biopsy is going to find cancer?

For mammography alone, that PPV was 38%. If you added ultrasound, you found a few additional cancers but you did a lot more biopsies. In fact, in the first year of screening, 10% of women underwent a biopsy, and although that number fell with subsequent screening in years 2 and 3, still about 7% of women underwent a biopsy, so the PPV was cut more than half. Only 16% of the biopsies found cancer.

At the end of the 3 years, when they offered women a mammography free of charge, only 58% of women said, "Thank you. I would like to have that additional screening test." Although MRI found a tiny number of extra cancers, it didn't help the PPV much because a lot more biopsies were done, so the PPV stayed at 19%.

What I find most striking, and what I think deserves praise for this group of investigators, is that they concluded that it is unclear that the added cost and reduced tolerability of screening MRI were justified in this population of women at intermediate risk for the development of breast cancer. Recall that half of the women in the study were considered at high risk because of a personal history of breast cancer. This is a group of women who, in many community sites, are routinely recommended to have MRI screening, and yet a group of radiologists suggest that maybe it is not as worthwhile in this group. We owe them our thanks and our gratitude for the honesty of their conclusions and for conducting this study to give us these numbers so we can have those intelligent and sometimes difficult conversations with our patients about what screening is helpful and what screening might not be worthwhile for them.

I would welcome your thoughts on this article, and I will see you again soon.

Comments

3090D553-9492-4563-8681-AD288FA52ACE
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:

processing....