An Expert Discusses Breast Density Notification Laws: Are They Ahead of the Science?

John C. Hayes; Debra I. Monticciolo, MD

Disclosures

December 03, 2012

In This Article

A Connecticut Experience

Medscape: I'm sure you are aware that Radiology just published a study from one breast imaging practice that is based on a breast density notification law adopted in Connecticut in 2009.[1] That study found an increase of 3.2 additional cancers per 1000 women screened. Would you expect to see the same sort of results in Texas, and why or why not?

Dr. Monticciolo: I believe the article you are referring to is by Hooley and associates.[1]

Medscape: That is correct.

Dr. Monticciolo: The cancer detection rate in that study was similar to that seen in the ACRIN trials and other reports on screening ultrasonography. So I think that you could expect a similar study done elsewhere to produce similar results regarding the number of additional cancers detected. But there are additional issues to consider when you look at the data from that study and other studies very carefully. You will notice that these authors had a very high false-positive rate and a low positive predictive value for their biopsies.

This is the main problem with screening ultrasonography. In the Connecticut study, about 5% of patients were recommended for biopsy on the basis of their ultrasonography results, which is 4 to 5 times the rate for biopsies that we get from doing just standard mammography.

The cancer rate in the Connecticut study was similar to that seen in the ACRIN trials.

And when a biopsy was done for their ultrasonography-detected lesions, only 6% were positive -- which is good for those patients, but it's a very small percentage. It's not good for the 94% for whom you did a biopsy and it was negative. With mammography, about 30%-40% of recommended biopsies will be positive, which is a far, far better outcome.

So additional cancer is one part of that study, but there is a downside, and that is balancing those benefits and risks. It's really part of this issue.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....