Digital Mammography

Summary by Cécile bour, MD

April, 26, 2021

In the Journal of The National Institute Volume 113, Number 1, January 2021, is published an article of 2020 on an important meta-analysis. It discusses the contribution of digital mammography in breast cancer screening, a publication that we had relayed here:

Digital mammography, approved in 2000 by the American FDA, is also widely used in France. 

Studies suggested that the new technology is equivalent to the old analog film technology for cancer detection.

A short technical description

Analog mammography produces an image printed directly on silver film. 

Digital sensor radiography (CR), which has now completely replaced the analog one, is an indirect mammography technique that captures the image on a reusable plate. This receptor of image contains a photostimulated luminophore, the X-rays cause an excitation of luminescent molecules which convert the X-rays to light. An analog-to-digital converter then produces a digital image that can be archived.

Direct Digital Radiography (DR) does not use reusable plates. Sensors convert the X-ray stream directly into an electrical signal which is then digitized and transferred to the screen. The image is visible and can be analyzed on the screen directly by the radiologist. The image can then be enhanced if it is over or underexposed for a better visualization. In this case too, the images can be stored as digital files in a computerized archiving system.

The article

In the article published in the National Institute's journal, authors Otis W Brawley (oncologist and epidemiologist in Baltimore, USA) and Channing J Paller (oncologist and urologist in Baltimore, USA) first recall the results of this important meta-analysis.

Digital mammography is certainly justified by easier storage and handling of images. There is also a possibility of computer-assisted diagnosis and better performance for the exploration of dense breasts. Digital mammography also has a lower radiation exposure than film mammography, provided that the number of images is not increased unreasonably... More on this later.

For effective screening, the authors point out, three objectives must be met: more localized tumors found at the same time as a decrease in interval cancers (tumors diagnosed between two screenings), and a decrease in the incidence of advanced cancers.

Regrettably, the Australian meta-analysis by Faber et al. (Sydney School of Public Health, Australia) confirms the problem of increased overdiagnosis with this method, which allows more detection of small lesions, particularly carcinomas in situ, the vast majority of which do not affect the life of the woman diagnosed, but there is no difference in the detection rate of invasive cancers. 

The study suggests that 11% of cancers detected by digital mammography are overdiagnosed. However, other analyses cited in the article are much more pessimistic about the rate of overdiagnosis attributable to the digital system [1].

The digital technique also has no effect on interval cancers, which are not reduced.

In the United States, the age-adjusted incidence rate (new diagnoses) of breast cancer increased by more than 30% from 1975 to 2000, while the incidence of advanced breast cancer at the time of diagnosis was stable for the 25 years, instead of decreasing as expected.[2]

Most importantly, the recall rate is significantly increased with digital technology because of the increase in false positives, which means that the claim of less radiation with digital technology is questionable, since these women recalled because of false suspicion of cancer will undergo, among other examinations, new X-rays.

Highlights from the article

Two interesting points are made by Brawley and Channing in relation to the results that can be extracted from the Australian meta-analysis:

1. The true measure of the value of effective screening at the population level is a reduction in cancer mortality and unnecessary treatment. 

Screening should not justify its apparent success on the discovery of more and more cancers, many of which are unnecessary detections, but rather on the detection of more cancers that are important to find because they are clinically dangerous to the women who develop them. But digital mammography is not more discriminating for these forms and overdetects a large number of cancers that would not have caused any consequence.

2. It is human nature to think that the new technology is always the best, and many experts felt that digital mammography would lead to better health outcomes. 

Sometimes we have to face the fact that the truth is different from what experts have put forward. These results demonstrate, according to the authors, the importance of post-marketing evaluation and open-mindedness. 

These results also show the biological variations of breast cancer, and confirm a non-linear natural history of the pathology: some cancers are useless to find because they regress or do not progress, others are immediately aggressive and develop between two screening mammograms without any possibility of stopping their occurrence.

Our commentary

We invite our readers to read the work of Bernard Junod and Dr Bernard Duperray on overdiagnosis. The natural history of cancer is of paramount importance, without this knowledge we will never be able to understand the problem of overdiagnosis and interval cancers.


[1] "One-third to one-half of breast cancers detected by mammography would not have been Clinical over a lifetime (overdiagnosis)."

[2]Welch HG, Gorski DH, Albertsen PC. Trends in metastatic breast and prostate cancer–lessons in cancer dynamics. N Engl J Med. 2015;373(18):1685–1687.

Cancer Rose est un collectif de professionnels de la santé, rassemblés en association. Cancer Rose fonctionne sans publicité, sans conflit d’intérêt, sans subvention. Merci de soutenir notre action sur HelloAsso.

Cancer Rose is a French non-profit organization of health care professionals. Cancer Rose performs its activity without advertising, conflict of interest, subsidies. Thank you to support our activity on HelloAsso.