We have learned a great many things about cancer prevention and control over the past three decades. One of the most definitive messages we have been able to deliver is: Routine mammograms can detect breast cancers at a treatable stage and save lives.

As digital mammography machines matured and entered the clinical realm, questions arose about whether this breast cancer screening method could or should supplant film. Digital mammography can supply images with greater spatial resolution and be manipulated to more clearly define a potentially lethal abnormality in breast tissue. So there was good reason to put digital mammography to the test.

This week's lead story describes the NCI-sponsored DMIST trial, which provides strong evidence that digital mammography is as good as film in the average population and that there is a significant subset of women who benefit from screening with it. Nearly two of three DMIST participants fell into one of the groups in which digital mammography proved more sensitive than standard film mammography, and a large portion of the malignancies caught by digital were of the type that can prove lethal if not detected early.

But there is much more to this story. Beyond its results, DMIST is significant for several other reasons.

The roots of this advance date back to the late 1980s. Digital x-ray machines were being developed for several applications, but because of the technical difficulty of developing devices that had the resolution needed for mammograms, that application was languishing. So in 1991, following the recommendations of an expert consensus conference, NCI issued a Request for Applications (RFA) to speed the development of digital mammography.

The RFA led to the formation of the International Digital Mammography Development Group, an effort that advanced digital mammography to the point where a trial such as DMIST was feasible.

It is particularly noteworthy that NCI and ACRIN secured the involvement of four companies that develop and manufacture digital mammography machines. These four companies are competitors, and their participation was a risk for each in that their equipment might have been shown to be inferior to the others. As it turned out, there was no significant difference demonstrated among the four machines used in DMIST.

DMIST also once again demonstrates NCI's commitment to working with other federal health agencies. Both FDA and the Centers for Medicare and Medicaid Services helped develop the DMIST protocol. Thus, we could ensure that the trial would generate data that would be useful in helping those agencies make regulatory decisions related to digital mammography, which could affect its continued development and availability.

And DMIST is another important addition to the "electronic health care" movement being championed by NCI through efforts such as the cancer Biomedical Informatics Grid. Digital mammogram results will be just one more component of a system in which researchers and clinicians are increasingly becoming virtually connected, sharing their expertise and data to advance cancer prevention, diagnosis, and treatment.

Of course, this trial would not have been possible without the participation of nearly 50,000 women and the dedicated staff at ACRIN and in the NCI Cancer Imaging Program.

Overall, DMIST offers a model case study of how NCI can be an agent of change, pursuing new approaches to research, partnering with the private and public sectors, and fueling the development of technologies to achieve an important payoff.

Dr. Andrew C. von Eschenbach
Director, National Cancer Institute