Digital mammography not more effective in reducing the most serious cancers, according to Australian meta-analysis

by Cécile Bour, MD, June 28, 2020

This is a meta-analysis of 24 studies of more than 16 million screening mammograms (10,968,843 film and 5,614,900 digital mammograms) by Australian researchers, which reported the following results :

With digital mammography is observed :

  • A statistically significant increase in recalls and false positives; bearing in mind that false positives can generate considerable anxiety in women who are misidentified as having breast cancer. After the transition to digital mammography, which began around the year 2000, a recall increase rate of 6.96 per 1,000 screens was  reported.
  •  No impact on the rates of interval cancers, these cancers not being anticipated by screening because they are often very rapid and progressive and appear between two screening mammograms, despite a previous negative mammogram. Their rate is not lowered by digital technology.
  •  A modest but statistically significant increase in the cancer detection rate, however this improved detection rate is largely due to a higher detection of carcinoma in situ, the great majority of which are of unnecessary detection, with little difference in the detection of invasive cancers.

 Precisions and some reminders

1° Transition from analog to digital

Initially, mammography was analogical, i.e. using films that had to be developed. The radiological signal, to make it simple, is  transformed into a ‘visual’ signal, which involves technical uncertainties that can alter the quality of the film to be interpreted (film artifacts, quality and fragility of the films, problems with developing baths, etc.). With the digital process, used in screening since 2008, sensors retrieve the image, store it in memory, there is no transformation of the signal and the image is projected in real time on the computer screen (or console). The diffusion of the signal is limited to the maximum (‘photon counting’ technique) to keep only the useful information. This technology has been acclaimed for its better detection rate compared to the analog system, especially in case of dense breasts, and for its lower irradiation.

This last aspect is certainly not negligible, but should be relativized, since radio-biologists have made us understand the mechanisms of radio-toxicity, which is much less dose-dependent than subject-dependent. See:

2° The impact of digital in terms of benefits

 In 2010, questions were already being raised about the real impact for women in terms of mortality reduction, because the interest of screening is to detect the cancers that endanger the most women’s lives, and thus reduce mortality, a fact that is difficult to prove with breast cancer screening (reduction in breast cancer mortality since the 1990s, thanks to therapeutic advances, a reduction well before the advent of screening campaigns).

However, an increase in detection rate ranging from 20 to 28% with the digital technique (in fact according to the Australian study, there is 25% more detection for carcinoma in situ and only 4% more for invasive carcinoma), this represents only one additional cancer detected thanks to this new technique per 1000 women, and it does not mean that one more life is saved.

Screening performance is measured by mortality and not by the detection rate. The problem is that it preferentially detects a cancer that will be little or not life-threatening, and that would have been detected by analog mammography one year later anyway, or even in the absence of any screening by the appearance of a clinical symptom, without changing the therapeutic result. More cancers that would never have manifested themselves are also detected. The problem of overdetection by preferential detection of cancers with little or no evolution thus causing overdiagnosis, has already been raised.

3° The problem of over-detection of carcinoma in situ (CIS)

Below is a reminder of what we explained about in situ carcinoma of the breast:

In situ carcinoma of the breast is defined by the proliferation of cancer cells within a galactophoric duct without the cells protruding beyond the wall of the duct to invade the rest of the breast.

It is essentially a mammographic finding, as 90% of women diagnosed with DCIS had microcalcifications on mammography.

The vast majority of these lesions do not threaten women’s lives if they are not detected, their prognosis is very good, the 10-year survival rate, a parameter widely used by health authorities, is over 95%. There is the ductal form and the lobular form considered rather as a risk factor for breast cancer.

CIS contributes to overdiagnosis. Tests and studies show that the increasing detection of CISs has not contributed to the reduction of breast cancer mortality. Before the introduction of screening, CIS accounted for less than 5% of all breast cancers, rising to 15-20% in all countries where screening campaigns exist. They are not counted in the incidence figures (rate of new cases) given by the National Cancer Institute, as they are considered separately and not as “true” cancers.

In addition, there is a lack of real consensus among pathologists for the classification of these lesions in the analysis of the biopsies they receive, with a tendency to outclass them in poorer prognostic categories, for fear of underestimating a “disease”.

Most CIS are considered as non-mandatory precursor lesions to invasive cancer; paradoxically, the dramatic increase in their detection followed by their surgical ablation has not been followed by proportional decreases in the incidence of invasive cancers.

The major problem is that these particular breast cancer entities are treated with the same severity as breast cancer.

In November 2016, a study carried out by University of Toronto reported  the following results:

– Their treatment makes no difference in women’s survival.
– Women with CIS are heavily treated (sometimes by bilateral mastectomy) and have the same likelihood of dying from breast cancer as women in the general population.
– Treating CIS does not decrease their recurrence.
– Preventing recurrence by radiation therapy or mastectomy would also not reduce the risk of mortality from breast cancer.

Similarly, our study of mastectomies in France showed a steady increase in surgical procedures, our first hypothesis being the overtreatment of lesions that are not invasive cancers, but so-called pre-cancerous lesions and CIS [3] [4].

The long-term consequences of overtreatment can be life-threatening. For example, radiation therapy on these lesions appears to be unable to reduce the risk of death from breast cancer, but is associated with a dose-dependent increase (10-100% over 20 years) in the rate of major coronary events. [5]

In several countries, clinical trials are being conducted to test a simple active surveillance, especially for low-grade CIS, rather than aggressive treatment:


For Philippe Autier [6], from the International Prevention Research Institute (IPRI), the problem is undoubtedly inherent to routine mammography, in particular digital mammography, which is too efficient for the detection of small calcifications; these are the most frequent radiological sign of these forms, and mammogaphy has an excellent sensitivity for the detection of these microcalcifications.

You will find in our media library several clinical cases of carcinomas in situ, abusively called carcinomas.

Conclusion of authors of Australian meta-analysis

The increase in cancer detection after the switch to digital mammography did not result in a reduction in interval cancer rates. Recall rates were increased.

These results suggest that the transition from film to digital mammography, primarily motivated by reasons of better technological performance, did not result in health benefits for screened women.

This analysis reinforces the need to carefully evaluate the effects of future technological changes, such as tomosynthesis [1] [2], to ensure that the new technologies lead to better health outcomes, beyond the gains in technical performance alone.

[1] Read :[2] Read also  :

🛈 Nous sommes un collectif de professionnels de la santé, rassemblés en association. Nous agissons et fonctionnons sans publicité, sans conflit d’intérêt, sans subvention. Merci de soutenir notre action sur HelloAsso.
🛈 We are an French non-profit organization of health care professionals. We act our activity without advertising, conflict of interest, subsidies. Thank you to support our activity on HelloAsso.
Scroll to Top