Mammography screening: a major issue in medicine

Authors: Philippe Autier, Matthieu Boniol *

Abstract, C.Bour, MD

January 3, 2018

*Strathclyde University Institute for Global Public Health at IPRI, European Prevention Research Institute, European Area, Building G, Allée Claude Debussy, 69130 Lyon Ecully, France

*International Prevention Research Institute (iPRI), 95 Cours Lafayette, 69006 Lyon, France

In this publication to be published in the February issue of the European Journal of Cancer, the authors provide a comprehensive review of mammography screening from several perspectives: effectiveness, criteria for evaluating its effectiveness, economic aspect, over-diagnosis, over-treatment, criticism of the methodology of the studies that founded the first campaigns, natural history of the disease, screening, high-risk women, etc... and they make a comparison with the others screenings that, according to them, are effective in terms of reducing mortality, namely cervical and colon cancer, in order to analyze why screening for breast cancer has ended in the failure that we know today.

A comparison is also presented with the case of neuroblastoma screening, which finished in a similar deadlock to that of breast cancer screening, due to similarities in the epidemiological features of the screening for this cancer.

The highlights

After 20–30 years of mammography screening, the incidence rates of advanced and metastatic breast cancer have remained stable.

Breast cancer mortality rates have not decreased more rapidly in areas where mammography is in place since the late 1980s.

One third to one half of mammography-detected breast cancers would not have been clinical during lifetime (overdiagnosis).

Breast screening randomised trials have adopted distinctive methods that led to exaggerating the efficacy of screening.

The influence mammography screening may have on mortality decreases with the increasing efficiency of cancer therapies.

Two indicators for screening effectiveness

Methods for evaluating the effectiveness of cancer screening are based on :

- monitoring incidence rate adjusted according to the age of advanced cancers, that would be expected to decrease after the introduction of screening.

- mortality rates specific to cancer would be expected to decrease more rapidly in areas where screening is performed than in areas where level of screening is lower but patient management is similar.

However, the accumulation of epidemiological data shows that in populations where mammography screening has been widely practiced for a long time, there has been little or no decrease in the incidence of advanced cancers, and that reduction in breast cancer mortality  in areas with early introduction and high screening penetration is similar to whose in the areas with late introduction and low screening penetration. Reductions in cancer mortality rates in the different randomized trials were expected to be proportional to reductions in advanced cancer rates, which was not the case, and in addition we are faced with the unexpected problem of significant over-diagnosis and consequent over-treatment, as evidenced by the increase in the rate of radical mastectomies seen in many countries. (See 12 studies on the non-reduction of advanced cancers [1].

Parallelism with the history of neuroblastoma

Screening for neuroblastoma in children was abandoned in the early 2000s because these two criteria of effectiveness were not met. In addition, over-diagnosis - the detection of non-evolving occult neuroblastoma that would not have endangered the individual over the course of his or her life - is a major undesirable consequence of screening.

The epidemiological picture of mammography screening closely looks like whose of neuroblastoma screening.

Results of first trials, the studies

Breast cancer screening advocates have favored studies that were consistent with the results of the first trials (New York [2] and the two Swedish counties [3] trials), while minimizing over-diagnosis. These trials had never been contested, but a re-evaluation by the authors of this publication shows major problems (4 mainly) of methodology, which enabled advocates of screening to significantly overestimate the benefit in terms of mortality reduction and minimize over-diagnosis.

So-called IBM studies [4] [5] [6] [7] [8] [9] [10] [11] (based on the incidence of breast cancer mortality) compared breast cancer mortality in screened and unscreened populations over time. These studies based on breast cancer mortality incidence and case-control studies were then used to evaluate the effectiveness of screening. But they are flawed by contradictions.

They had a strong "ecological" bias and did not meet the methodological criteria defined in the CIRC Manual published in 2002 [12]. Adjustments are needed in these studies (on lead time and on the increase in breast cancer incidence), moreover two Norwegian IBM studies which were the most compliant ref 4, 8 with the CIRC criteria found no significant influence of screening mammography on the risk of death from breast cancer.

Two contradictions emerge: how can screening be effective when, in populations where most women have participated in screening mammography for decades, the incidence rates of advanced breast cancer have not decreased at all or only moderately? Screening proponents have nevertheless tried to show that such decreases in advanced cancer rates are indeed observed, but the methodologies are not appropriate. Another contradiction is the impossibility to link the astonishing mortality reductions of 25-31% in breast cancer mortality that have been put forward with the lack of real observed reduction in breast cancer mortality in countries that have rapidly adopted mass screening.

Another unsolved problem is the natural history of cancer. Every small asymptomatic breast cancer detected by screening does not automatically become symptomatic advanced breast cancer. And a serious high-grade cancer is not the result of a small one that was prevented from becoming serious by anticipation. Unfortunately, current evidence indicates that the ability of screening mammography is insufficient to detect these more advanced breast cancers. We do not know the growth patterns of breast cancer and the mechanisms involved in metastatic spread, contrary to a fundamental principle before screening can be implemented, namely that the natural history of the disease, including the development from latent to declared disease, must be well known".

To summarize

The early Swedish trials overestimated the reductions in breast cancer mortality related to screening, partly because of biases in the statistical analyses, some of which were deliberately due to under-reporting of breast cancer in the selected groups. On the other hand, many models assumed that if screening did not take place, the majority of asymptomatic cancers detected by screening would have progressed to symptomatic advanced cancers. The accumulation of demographic data in well-selected populations invalidates the relevance of these initially admitted models, based on a linear evolution of the disease.

The promoters of screening have voluntarily rejected studies that did not show a decline in mortality, such as the Euroscreen Group convened by the the CIRC for to asses the procedure in 2015 [13]. But since then, many major, independent and large-scale studies (from Harding in the USA 2015 [14] [15] or the impact studies by Bleyer and Miller [16] [17] ) have added further evidence of a questionable, minor effectiveness of mammography screening, which in any case does not reflect what was expected.

The example of the Netherlands

 study Autier, The Netherlands

analysis study Netherlands

From 1988 onwards, the national breast cancer screening program in the Netherlands invited women between 50 and 75 years of age to undergo a biennial mammography screening, with a consistently high participation rate of about 80% among women. Despite a high participation in screening for 23 years, the incidence of stage II-IV breast cancer has not improved over time for women aged 50 and over, ranging from 168 per 100,000 in 1989 to 166 per 100,000 in 2012.

The Dutch mammography screening program would have had only a marginal effect on breast cancer mortality.

Similar results are found in other studies (Europe, USA, Australia); furthermore, no association was found between the magnitude of the downward trends in breast cancer mortality and the timing of the implementation of mammography screening in the different states of the United States as evidenced in the Bleyer and Miller impact study ref [16] [17].

According to the authors, population studies using a variety of models would provide convincing evidence that it is rather therapeutic advances that have a favourable influence on breast cancer mortality.

The particularity of the cancers detected during mammography screening, is that less aggressive cancers with better prognosis are selected.

Mammography has a high sensitivity for ductal carcinoma in situ, for example, and a relatively low sensitivity for certain aggressive cancers such as triple negative breast cancer [18] [19]. Mammography practically does not detect in situ or invasive lobular carcinomas, which account for 8-14% of all breast cancers [20] . Lobular carcinomas infiltrate the tissue without forming masses, making them difficult to detect by mammography.

Compared to interval cancers, i.e., those that progress rapidly between mammograms and have aggressive characteristics, invasive cancers detected by mammography have the clinical and pathological characteristics of less aggressive tumors.

Moreover, after examining the features of these tumors and the extent of the disease at the time of diagnosis, the risk of dying from a detected breast cancer is lower than the risk of dying from an interval cancer [21][22][23]. Thus, the fact that a breast cancer has been detected by mammography is in itself a good prognostic factor. Therefore, the method of detection should be considered when estimating a patient's prognosis [24].

“Precursor” lesions

The chronological model of successive stages describes reasonably well the progression from normal tissue to the benign lesion and then to the malignant tumor observed in the majority of colorectal and cervical cancers. It was believed that the model would also be applicable to other cancers, including breast cancer.

However, epidemiological and clinical observations, such as the stable incidence of metastatic (i.e. stage IV) breast cancer in populations that are screened, call into question the validity of this model for describing the natural history of many breast cancers. Rather, these observations are consistent with the hypothesis of a "cancer spectrum", whereby breast cancers consist of a set of malignant lesions that are highly heterogeneous in their ability to invade surrounding tissues and metastasize to distant lymph nodes or organs [25].

Most of these lesions would constitute a reservoir of asymptomatic tumors that would remain indolent or would grow slowly, with rare progression to symptomatic disease. Because of their long detectable sub-clinical period (i.e., the period without clinical signs in the breast), these lesions would represent the main findings of screening mammography and other breast imaging techniques [26].

At the other end of the spectrum, tumors with high metastatic potential would grow rapidly and metastases would already be present in lymph nodes and distant organs when the tumor is detected [27].

Due to the short residence time of rapidly evolving and aggressive cancer, the window of opportunity for its detection before metastatic spread would be very narrow.

Grade, size and stage could reflect the intrinsic malignant potential of a cancer. Since high-grade cancers develop and metastasize rapidly, their size at diagnosis is generally greater than that of low-grade cancers that develop slowly and are not expected to metastasize.


Overdiagnosis of cancer corresponds to the excess of cancers in women invited for screening divided by the total number of cancers that would have been diagnosed in the absence of screening (on a population with the same profile, same age group, without screening).

Over-diagnosis occurs in women attending screening.

If over-diagnosis is calculated using the number of cancers detected at screening as the denominator, then for every 100 breast cancers detected by screening, 30-50 would be over-diagnosed [28] [29].

An interesting observation is that, compared to breast cancers diagnosed in the absence of screening, interval breast cancers are not more deadly and do not have clinicopathological features of more aggressive cancer [30] [31].

If interval cancers are similar to cancers diagnosed in the absence of any screening, and if the cancers detected have a better prognosis on average than interval cancers, it follows logically  that a proportion of cancers detected are non-lethal cancers that would never have been symptomatic during a woman's lifetime.  Under the microscope, these lesions have the morphological features of cancer, but  if screening had not taken place, they would have remained asymptomatic during the woman's lifetime. This phenomenon is called "over-diagnosis", and is first highlighted by the Oslo experience comparing a group of unscreened women to a group of screened women, in which 22% of cancers in excess were reported.

(Editor's note Zahl P-H, Gøtzsche PC, mæhlen J. Natural history of breast cancers detected in the Swedish mammography screening programme: a cohort study. The Lancet Oncology. 2011 Nov;12(12):1118–24.).

The first evaluations, the Canadian one of Miller [32] and the Oslo one [33] considered only invasive cancers alone, without any quantification of in situ cancers. This group, however, is also a major cause of over-diagnosis, especially since it is known that their treatment neither prevents the occurrence of invasive cancers nor reduces their recurrence (see next chapter on CIS).

The sudden drop in incidence in 2002 coincided with the mass discontinuation of hormone replacement therapy after the publication of the Women's Health Initiative trial [34]. The incidence of DCIS in women aged 50 years and older increased from 10 per 100,000 in the 1980s to about 90 per 100,000 in the 2000s. In 2008, the excess incidence was estimated to be 31% of all breast cancers diagnosed in the United States, consisting mainly of DCIS and stage I cancers. [35]

If a large proportion of breast cancers diagnosed in  excess after 1986 were due to advancements in diagnosis, therefore proportional decreases in incidence ratio should have been observed for older women.

Screening by mammography or other imaging modalities preferentially detects these lesions that are not or are mildly aggressive. This is the case for most low-grade DCIS, for instance, but also for atypical epithelial lesions as well. The natural evolution of these lesions is poorly known. However, assuming that cancer develops later, doctors are inclined to treat most of these abnormalities.

CIS (carcinomas in situ)

They represent an over-diagnosis because these lesions are treated despite our limited knowledge of their long-term evolution. Trials and studies show that the greater detection of CIS has not led to a reduction in breast cancer mortality. Before the period of screening, CIS accounted for less than 5% of all breast cancers. In areas where screening mammography is used widely, 15-20% of breast cancers are CIS [36].

Moreover, there is a lack of a real consensus among anatomist-pathologists on the classification of such lesions during biopsies, with a tendency to upgrade them to poorer prognosis categories, for fear of underestimating the disease [37].

Most CIS are non-mandatory precursors of invasive cancer, because the spectacular increase in their detection accompanied by surgical removal was not followed by a proportional decreases in the incidence of invasive cancers [38] [39] .

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 (10-100% over 20 years) increase in the incidence of major coronary events [40].

Over-diagnosis often has a considerable disruptive effect on the interpretation of clinical results expressed as percentages (instead of raw rates) or as overall survival (instead of mortality rates or staging specific survival), artificially improving these data which in no way reflect the effectiveness of the screening practice.

Overtreatment, mastectomies

In randomized trials, rates of radical mastectomy were highest among women invited for screening. In the Netherlands and the United Kingdom, rates of radical mastectomy increased in the years following the implementation of screening programs. In the United States, a rebound in the rate of radical mastectomies has been observed since 2005. [41] [42][43] [44][45] [46] 

The resurgence of radical mastectomies has been encouraged by the use of preoperative magnetic resonance imaging (MRI) to assess the extent of the cancer and look for malignant lesions in the contralateral breast or multifocal.

MRI has the potential to discover satellite cancer sites, but the clinical outcome of these sites is unknown. Multiple examinations attest to the unfavorable benefit-risk balance of preoperative MRI [47] [48].

Editor's note: see our study on mastectomies in France, published in the October 2017 issue of the Revue Médecine, to be found here :

Study in Médecine/oct 2017

With the analysis here: CR study analysis

The effect of the treatment

A recent study by Prof. Autier suggests that access to effective adjuvant and chemotherapeutic treatments modifies the reduction in the risk of death from breast cancer initially reported by randomized trials[49].

Consequently, with the advent of effective treatments, the necessary number of women needed to be screened to prevent a single death from breast cancer is increasing.

In other words, the ability of the treatments to reduce breast cancer mortality, makes the mortality reduction from screening all the more negligible and inexistent, and even more women will need to be screened to be able to achieve an avoided death that would be prevented by screening rather than by the available treatments.

The imbalance between the effectiveness of screening and over-diagnosis enhances the potential for powerful treatments to reduce breast cancer mortality on their own.

Criticism of the method

A critical review of the literature promptly demonstrates that most studies are based on short-term designs focusing on the ability of different technologies to detect asymptomatic lesions labeled "cancer" after microscopic examination. In addition, many studies have unfortunately considered percentages of lesions by size, or percentages of more favorable grades of the detected cancers, and have reported on statistics of overall survival. These findings are not adequate to evaluate the effectiveness of screening, since we have seen that over-detection is more favorably linked to the lower grades, and artificially increases survival rates by finding lesions that would not have been lethal anyway.

These higher detection rates unfortunately do not translate into fewer advanced breast cancers or fewer breast cancer deaths. The emphasis on detection rates, even in randomized trials, may simply reveal the greater ability of a particular detection method to draw from the reservoir of slowly progressing or progressing tumors, to select those same lesions, according to the adage of "the more you look, the more you find" [50].

The transition from analog to digital mammography has not been accompanied by a decrease in the rate of interval or advanced breast cancers, while false positives, over-diagnosis, and costs have increased, mainly due to increased CIS detection [51].

Priority should be given to studies that can demonstrate that the use of a new screening technique can actually reduce the incidence of interval cancer or advanced cancers, at the cost of acceptable rates for recall and biopsy and limited over-diagnosis.

What is happening elsewhere, public health policies

The Swiss Medical Board has concluded that most of the original evidence is outdated and therefore screening mammography programs should be terminated [52].

The opposite point of view is held by the American College of Radiology, which recommends annual mammography for asymptomatic women from the age of 40 and without limit [53].

Editor's note: we are updating this statement in 2019, as we are now observing across the Atlantic a de-escalation of screening recommendations through the new guidance notes of the American College of Physicians (ACP)[54].

The evaluation of the screening mammography program in Norway rightly concludes that "at the individual level ... each invited woman must weigh information about potential benefits and harms according to her own values, health and life situation to decide whether or not to participate in the program"[55].

The polarization of visions:

In this context of polarized opinions and the pursuit of incentive messages to women, and the weakness of available loyal information, it is difficult to imagine that we can achieve to a consensual information on the harms and benefits of breast cancer screening given to the women and whole society. An alternative might be to inform about divergences between different experts, letting the women make their choice in any case[56] [57].

Personalized screening, prioritized according to risk[58]

- Although attractive, a personalized breast cancer screening strategy encounters major difficulties. First, there is little evidence that more frequent screening (i.e., annually with other imaging of breast techniques) would be more effective than biennial or triennial screening.

- A second difficulty is related to the available data on the factors (or combination of factors) that would identify the asymptomatic women as having an intermediate risk of death from breast cancer. Since screening is supposed to prevent breast cancer death but increases the likelihood of being cancer carrier (over-diagnosis), the risk factors should be those for the risk of breast cancer death, not those for the risk of breast cancer occurrence. Studies to document risk-based screening strategies have generally focused on factors associated with the occurrence of breast cancer and rarely on factors associated with death from breast cancer. Risk factors for breast cancer and risk factors for dying from breast cancer do not necessarily coincide [59].

- Much remains to be learned about genetic determinants and the importance of breast density (related to cancer occurrence and mortality). We also need to learn more about the relationship between this breast density factor and other risk factors such as obesity or atypical breast hyperplasia [60].

Breast imaging technologies such as ultrasound, tomosynthesis and MRI find cancers missed by mammography in dense breasts, and the use of alternative breast imaging techniques is encouraged in women with dense breasts, particularly in the USA. However, there is still no evidence that such an approach would decrease the risk of death from breast cancer, while it certainly increases the risk of false positives, unnecessary biopsies and over-diagnosis. A good option to get around these many problems associated with breast density is to avoid screening mammography before the age of 50, especially since the risk/benefit balance in this age group is uncertain[61].

Currently there are risk calculators that attempt to assess which women would be at risk, including various factors.

 Particular case of women at high risk of breast cancer

Women at high risk of breast cancer, such as women carrying the BRCA1/2 mutation, have a 30 % or greater lifetime risk of developing breast cancer, with an early onset before the age of 50, a high risk of contralateral breast cancer and, depending on the hereditary trait (e.g. BRCA2 mutation), an increased risk for ovarian cancer. They require close monitoring of the breasts: MRI is recommended in these cases [62].

There is no evidence that more frequent mammography screening or screening with other modalities actually reduces the risk of death from breast cancer in women with an intermediate risk of breast cancer (10-29% risk), including women with extremely high breast density,  .

Since 1985, advances in the management of breast cancer patients have resulted in a marked reduction in breast cancer mortality, even in patients with disseminated (stage IV) disease. In contrast, epidemiological data indicate a marginal contribution of mammography screening on the decline in mortality from breast cancer. The more effective the treatments, the less favorable the benefit/risk balance of screening mammography.

Very high-risk women with a BRCA1 or BRCA2 mutation have a 50-85% lifetime risk of breast cancer; these high-risk women represent a context of psychological and medical problems that have nothing in common with women at medium or intermediate risk of breast cancer.

The critical question remains unanswered: to what extent can regular breast surveillance  with mammography and MRI prevent the onset of an advanced cancer and death from breast cancer in high-risk women?

Conducting a randomized trial to evaluate the effectiveness of screening in high-risk women (comparing a group of women who are followed and a group who are not) is obviously ethically indefensible, and many high-risk women will probably not consent to be in the control group of such a trial.

Recent studies indicate rather systematically  that annual surveillance of high-risk women by MRI, with or without other techniques (e.g. mammography and/or ultrasound), avoid to miss interval cancers within a year of screening [63][64][65][66]

These findings suggest that annual MRI screening by experienced radiology teams could be an alternative to more radical options for reducing risks

To conclude

- Evidence shows that in populations where mammography has had high penetration over the past 20-30 years, screening mammography has had little or no influence on the burden of advanced breast cancer and no influence on metastatic breast cancer in women. The authors claim that meta-analyses of randomized trials of breast cancer screening largely overestimated the reductions in breast cancer risk that could be expected from screening mammography.

- Over-diagnosis is a source of considerable harm to women who undergo screening mammography. The medical community's recognition  of the concept of over-diagnosis of cancer has been slow, but  it is no longer acceptable to minimize the burden of it nor its consequences in 2017. According to the authors, probably in 10 years, if we continue in this way, one woman in six will be diagnosed with breast cancer during her lifetime, without any certainty that the risk of death from breast cancer will actually be reduced, but with the certainty that the overall harm induced by screening will only increase.

- While screening mammography has some influence on breast cancer mortality, this influence diminishes, as patient care progresses. This means that the more effective patient care, the more women need to be screened to finally see a breast cancer death prevented by screening.

 - Current evidence indicates that the capacity of screening mammography (and probably other breast imaging technics ) is restricted for the "catch up" of breast cancers before they reach more advanced stages. This essentially reflects our ignorance of the growth patterns of breast cancer and of the mechanisms involved in metastatic spread. This ignorance runs counter to the seventh principle on screening formulated by Wilson and Jungner [67], according to which "the natural history of the disease, including the development from latent to declared disease, should be well understood".

 - According to the authors, since 2016 there has been strong evidence  that improvements in patient care have played a major role in the observed reductions in breast cancer mortality in Europe, Oceania and North America, while the impact of screening mammography has been marginal. Decision-makers struggling to implement affordable and equitable national cancer control plans must be aware of the serious data gaps that are often presented as incontestable to them.

- To justify the perpetuation of screening mammography, some may support the cynical view that all medical activity leads to harm due to over-diagnosis: for example, treating the high blood pressure of a person who would never have died of cardiovascular disease. But breast cancer screening is not about simply taking a pill, but about the threat of life-long damage due to mutilation, over-treatment and psychological distress, with no gain in terms of reduction of the risk of death from breast cancer.

- An increasingly high detection rate indicates little about the ability to prevent interval or advanced cancers; it simply reflects additional over-diagnosis. Prior to the implementation of new breast imaging technologies studies based on the incidence of advanced cancers and the assessment of over-diagnosis will need to be conducted

 -Screening according to the risk level should be researched so that screening recommendations can be tailored to the risk of death from  breast cancer in women.

- Regarding costs: most of the economic evaluations carried out in the 1990s ignored or underestimated the costs due to over-diagnosis, which are estimated to be approximately US$ 4 billion per year in the U.S. [68].


 1] 12 studies on the non-reduction of advanced cancers

  • Autier, M. Boniol, A. Koechlin, C. Pizot, M. Boniol

Effectiveness of and overdiagnosis from mammography screening in the Netherlands: population study BMJ, 359 (2017), p. j5224

  • Autier, M. Boniol, R. Middleton, JF Dore, C. Héry, T. Zheng et al.  Advanced breast cancer incidence following population-based mammographic screening.Ann Oncol, 22 (8) (2011), p. 1726-1735
  • Bleyer, HG Welch Effect of three decades of screening mammography on breast-cancer incidence N Engl J Med, 367 (21) (2012), pp. 1998-2005
  • NA de Glas, AJ de Craen, E. Bastiaannet, EG Op ‘t Land, M. Kiderlen, W. van de Water, et al. Effect of implementation of the mass breast cancer screening programme in older women in the Netherlands: population based study.
  • Autier, M. Boniol The incidence of advanced breast cancer in the West Midlands, United Kingdom, Eur J Cancer Prev, 21 (3) (2012), pp. 217-221
  • Nederend, LE Duijm, AC Voogd, JH Groenewoud, FH Jansen, MW Louwman  Characteristics and prognosis of interval cancers after biennial screen-film or full-field digital screening mammography, Breast Cancer Res, 14 (1) (2012), p. R10
  • ML Lousdal, IS Kristiansen, B. Moller, H. Stovring Trends in breast cancer stage distribution before, during and after introduction of a screening programme in Norway Eur J Public Health, 24 (6) (2014), pp. 1017-1022
  • RH Johnson, FL Chien, A. Bleyer  Incidence of breast cancer with distant involvement among wme in united states, 1976 to 2009  J Am Med Assoc, 309 (8) (2013), pp. 800-805
  • Esserman, Y. Shieh, I. Thompson  Rethinking screening for breast cancer and prostate cancer, Jama, 302 (15) (2009), pp. 1685-1692
  • Jorgensen, PC Gøtzsche, M. Kalager, P. Zahl Breast Cancer Screening in Denmark: A Cohort Study of Tumor Size and Overdiagnosis Ann Intern Med, 166 (5) (7 mars 2017), pp. 313-323
  • HG Welch, DH Gorski, PC Albertsen  Trends in Metastatic Breast and Prostate Cancer--Lessons in Cancer Dynamics N. Engl JMed, 373 (18) (2015), pp. 1685-1687
  • Di Meglio, RA Freedman, NU Lin, WT Barry, O. Metzger-Filho, NL Keating, et al. Time trends in incidence rates and survival of newly diagnosed stage IV breast cancer by tumor histology: a population-based analysis Breast Cancer Res Treat, 157 (3) (2016), p. 587-596

[2] S. Shapiro, W. Venet, P. Strax, L. Venet, R. Roeser

Selection, follow-up, and analysis in the Health Insurance Plan Study: a randomized trial with breast cancer screening Natl Cancer Inst Monogr, 67 (1985), pp. 65-74.

[3] L. Tabar, CJ Fagerberg, A. Gad, L. Baldetorp, LH Holmberg, O. Grontoft, et al. Reduction in mortality from breast cancer after mass screening with mammography. Randomised trial from the Breast Cancer Screening Working Group of the Swedish National Board of Health and Welfare Lancet, 1 (8433) (1985), pp. 829-832

[4] S. Njor, L. Nystrom, S. Moss, E. Paci, M. Broeders, N. Segnan, et al. The impact of mammographic screening on breast cancer mortality in Europe: a review of observational studies J Med Screen, 19 (Supplément 1) (2012), pp. 33-41

[5] AH Olsen, E. Lynge, SH Njor, M. Kumle, M. Waaseth, T. Braaten, et al.Breast cancer mortality in Norway after the introduction of mammography screening Int J Cancer, 132 (1) (2013), p. 208-214

[6] L. Tabar, M.-F. Yen, B. Vitak, H.-HT Chen, RA Smith, SW Duffy Mammography service screening and mortality in breast cancer patients: 20-year follow-up before and after introduction of screening Lancet, 361 (9367) (2003), pp. 1405-1410

[7] BN Hellquist, SW Duffy, S. Abdsaleh, L. Bjorneld, P. Bordas, L. Tabar, et al. Effectiveness of population-based service screening with mammography for women ages 40 to 49 years: evaluation of the Swedish Mammography Screening in Young Women (SCRY) cohort Cancer, 117 (4) (2011), pp. 714-722

[8] LE Johns, DA Coleman et AJ Swerdlow, SM Moss Effect of population breast screening on breast cancer mortality up to 2005 in England and Wales: an individual-level cohort study Br J Cancer, 116 (2) (2017), pp. 246-252

[9] M. Kalager, M. Zelen, F. Langmark, HO Adami Effect of Screening Mammography on Breast-Cancer Mortality in Norway N Engl J Med, 363 (13) (2010), pp. 1203-1210

[10] H. Jonsson, L. Nyström, S. Törnberg, P. Lenner The Impact of Mammographic Screening on Breast Cancer Mortality in Europe: A Review of Observational Studies J Med Screen, 8 (2001), pp. 152-160

[11] Preliminary evaluation of the Swedish organized service Reduction of breast cancer mortality due to organized screening of mammography services: Further confirmation with extended data  Cancer Biomarqueurs Epidemiol Prev, 15 (1) (2006), pp. 45-51

[12] International Agency for Research on Cancer. IARC Handbooks of Cancer Prevention. Vol. 7. Breast cancer screening. Lyon, France: IARC Press, 2002.

[13] Lauby-Secretan, D. Loomis, K. Straif  Breast Cancer Screening - IARC Working Group Perspective N Engl J Med, 373 (15) (2015), p. 1479





[18] M. Bare, N. Tora, D. Salas, M. Sentis, J. Ferrer, J. Ibanez, et al. Mammographic and clinical characteristics of different phenotypes of screen-detected and interval breast cancers in a nationwide screening program

Breast Cancer Res Treat, 154 (2) (2015), pp. 403-415

[19] A. Caldarella, D. Puliti, E. Crocetti, S. Bianchi, V. Vezzosi, P. Apicella, et al.

Volumetric breast density and risk of advanced cancers after a negative screening episode: a cohort study J Cancer Res Clin Oncol, 139 (2) (2013), pp. 181-185

[20] G. Arpino, VJ Bardou, directeur général Clark, RM Elledge Infiltrating lobular carcinoma of the breast: tumor characteristics and clinical outcome Breast Cancer Res, 6 (3) (2004), pp. 1-8

[21] S. Mook, LJ Van ‘t Veer, EJ Rutgers, PM Ravdin, AO van de Velde, FE van Leeuwen,et al. Independent prognostic value of screen detection in invasive breast cancer J Natl Cancer Inst, 103 (7) (2011), p. 585-597

[22] T. Lehtimaki, M. Lundin, N. Linder, H. Sihto, K. Holli, T. Turpeenniemi-Hujanen, et al. Long-term prognosis of breast cancer detected by mammography screening or other methods Breast Cancer Res, 13 (6) (2011), p. R134

[23] P. Klemi, H. Joensuu, S. Toikkanen, J. Tuominen, O. Rasanen, J. Tyrkko, et al. Aggressiveness of breast cancers found with and without screening

BMJ, 304 (1992), pp. 467-469

[24] B. Hayse, RJ Hooley, BK Killelea, NR Horowitz, AB Chagpar, DR Lannin  Breast cancer biology varies by method of detection and may contribute to overdiagnosis Chirurgie, 160 (2) (2016), pp. 454-462

[25] S. Hellman

 Karnofsky Memorial Lecture Natural History of Small Breast Cancers J Clin Oncol, 12 (10) (1994), pp. 2229-2234

[26] HG Welch, WC Noir

Overdiagnosis in cancer

J Natl Cancer Inst, 102 (9) (2010), pp. 605-613

[27] MJ van de Vijver, YD He, LJ van’t Veer, H. Dai, AA Hart, DW Voskuil, et al. A gene-expression signature as a predictor of survival in breast cancer N Engl J Med, 347 (25) (2002), pp. 1999-2009

[28]  Effectiveness of and overdiagnosis from mammography screening in the Netherlands: population based study BMJ 2017 ; 359 doi: (Publié 05 décembre 2017)Citer ceci: BMJ 2017; 359: j5224

[29] Gilbert Welch, MD, MPH, Philip C. Prorok, Ph.D., A. James O’Malley, Ph.D., et Barnett S. Kramer, MD, MPH Breast-Cancer Tumor Size, Overdiagnosis, and Mammography Screening Effectiveness

N Engl J Med 2016; 375: 1438-1447 13 octobre 2016 DOI: 10.1056 / NEJMoa1600249

[30] LH Holmberg, L. Tabar, HO Adami, R. Bergstrom Survival in breast cancer diagnosed between mammographic screening examinations Lancet, 2 (8497) (1986), pp. 27-30

[31] M. Kalager, RM Tamimi, M. Bretthauer, HO Adami Prognosis in women with interval breast cancer: population based observational cohort study BMJ, 345 (2012), article e7536



The natural history of invasive breast cancers detected by screening mammography, Archives of Internal Medicine, 24/11/08 Per-Henrik Zahl, Jan Mæhlen, H. Gilbert Welch

[34] JE Rossouw, GL Anderson, RL Prentice, AZ LaCroix, C. Kooperberg, ML Stefanick, et al.Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results From the Women's Health Initiative randomized controlled trial Jama, 288 (3) (2002), pp. 321-333

[35] A. Bleyer, HG Welch Effect of Three Decades of Screening Mammography on Breast-Cancer Incidence N Engl J Med, 367 (21) (2012), pp. 1998-2005

[36] RS Punglia, SJ Schnitt, JC Semaines Treatment of ductal carcinoma in situ after excision: would a prophylactic paradigm be more appropriate? J Natl Cancer Inst, 105 (20) (2013), pp. 1527-1533

[37] JG Elmore, HD Nelson, MS Pepe, GM Longton, AN Tosteson, B. Geller, et al. Variability in Pathologists’ Interpretations of Individual Breast Biopsy Slides: A Population Perspective Ann Intern Med, 164 (10) (2016), pp. 649-655



[40] SC Darby, M. Ewertz, P. McGale, AM Bennet, U. Blom-Goldman, D. Bronnum, et al.Risk of ischemic heart disease in women after radiotherapy for breast cancer N Engl J Med, 368 (11) (2013), p. 987-998

[41] PC Gotzsche, KJ Jorgensen Screening for breast cancer with mammography Cochrane Database Syst Rev, 6 (2013), Article Cd001877

[42] Nederend, LE Duijm, MW Louwman, RM Roumen, FH Jansen, AC Voogd  Trends in surgery for screen-detected and interval breast cancers in a national screening programme Br J Surg, 101 (8) (2014), pp. 949-958

[43] Tataru, D. Robinson, H. Moller, E. Davies  Trends in the treatment of breast cancer in Southeast England following the introduction of national guidelines Journal of Public Health (Oxf), 28 (3) (2006), pp. 215-217

[44] AE Dragun, B. Huang, TC Tucker, WJ Spanos  Increasing Mastectomy Rates Among all Age Groups for Early Stage Breast Cancer: A 10‐Year Study of Surgical Choice Breast J, 18 (4) (2012), pp. 318-325

[45] Mahmood, AL Hanlon, M. Koshy, R. Buras, S. Chumsri, KH Tkaczuk, et al. Spectrum of very early breast cancer in a setting without organised screening Ann Surg Oncol, 20 (5) (2013), pp. 1436-1443

[46] KL Kummerow, L. Du, DF Penson, Y. Shyr, MA Crochets ,Nationwide Trends in Mastectomy for Early-Stage Breast Cancer JAMA Surg, 150 (1) (2015), pp. 9-16[46]

[47] N. Houssami, R. Turner, M. MorrowPreoperative Magnetic Resonance Imaging in Breast Cancer: Meta-analysis of Surgical OutcomesAnn Surg, 257 (2) (2013), p. 249-255

[48] A. Parsyan, A. Alqahtani, B. Mesurolle et S. Meterissian The role of magnetic resonance imaging mammography in the surgical management of the index breast cancer

World J Surg, 37 (9) (2013), pp. 2134-2139

[49] P. Autier

Efficient treatments reduce the cost-efficiency of breast cancer screening Ann Intern Med, 164 (4) (2016), pp. 297-298

[50] ER Haut, PJ Pronovost

Surveillance bias in outcomes reporting JAMA, 305 (23) (2011), pp. 2462-2463

[51] B. Hayse, RJ Hooley, BK Killelea, NR Horowitz, AB Chagpar, DR Lannin Breast cancer biology varies by method of detection and may contribute to overdiagnosis Surgery, 160 (2) (2016), pp. 454-462

[52] N. Biller-Andorno, P. Juni

Abolishing mammography screening programs? A view from the Swiss Medical Board N Engl J Med, 370 (21) (2014), pp. 1965-1967

[53]  American College of Radiology 

Acr Practice Parameter for Screening and Diagnostic Mammography Performance



Data on reducing breast cancer mortality in women aged 70-74 are mitigated; although data from randomized controlled trials are limited in this age group, there is agreement on screening women at average risk from age 50 to 74. There is no reduction in all-cause mortality among women aged 50 to 69 years. Most guidelines recommend biennial mammography as an acceptable or preferred option for women undergoing screening.

Decreasing the interval exposes women to an increased risk of over-diagnosis. Below and above this age range, the disadvantages outweigh the expected benefit.

There is insufficient evidence of effectiveness for screening by clinical palpation alone or in combination with mammography.

[55] Norway Research Council

Research-based evaluation of the Norwegian Breast Cancer Screening Programme, Final Report: Evaluation Division for Society and Health


[56] N. Biller-Andorno, P. Juni

Abolishing mammography screening programs? A view from the Swiss Medical Board

N Engl J Med, 370 (21) (2014), pp. 1965-1967

[57] KW Lin, LO Gostin A public health framework for screening mammography: Evidence-based vs politically mandated care JAMA, 315 (10) (2016), pp. 977-978

[58] OW Brawley

Risk-Based Mammography Screening: An Effort to Maximize the Benefits and Minimize the Harms

Ann Intern Med, 156 (9) (2012), pp. 662-663

[59] HD Nelson, B. Zakher, A. Cantor, R. Fu, J. Griffin, ES O’Meara, et al. Risk factors for breast cancer for women aged 40 to 49 years: a systematic review and meta-analysis.

Ann Intern Med, 156 (9) (2012), pp. 635-648

[60] NF Boyd, LJ Martin, L. Sun, H. Guo, A. Chiarelli, G. Hislop, et al. Body size, mammographic density, and breast cancer risk.

Cancer Biomarqueurs Epidemiol Prev, 15 (11) (2006), pp. 2086-2092

[61] J. Melnikow, JJ Fenton, EP Whitlock, DL Miglioretti, MS Weyrich, JH Thompson, et al. Supplemental Screening for Breast Cancer in Women With Dense Breasts: A Systematic Review for the U.S. Preventive Services Task Force Ann Intern Med (2016)

[62] T. Schenberg, G. Mitchell, D. Taylor, C. Saunders

Le dépistage par IRM du cancer du sein chez les femmes à haut risque; est le programme australien d’accès au dépistage de l’IRM du sein répondant aux besoins des femmes à haut risque de cancer du sein?

J Med Radiat Sci, 62 (3) (2015), pp. 212-225

[63] T. Schenberg, G. Mitchell, D. Taylor, C. Saunders MRI screening for breast cancer in women at high risk; is the Australian breast MRI screening access program addressing the needs of women at high risk of breast cancer?

J Med Radiat Sci, 62 (3) (2015), pp. 212-225

[64] S.A. Narod

Screening of women at high risk for breast cancer

Prev Med, 53 (3) (2011), pp. 127–130

[65] A.M. Chiarelli, M.V. Prummel, D. Muradali, V. Majpruz, M. Horgan, J.C. Carroll, et al.

Effectiveness of screening with annual magnetic resonance imaging and mammography: results of the initial screen from the ontario high risk breast screening program

J Clin Oncol, 32 (21) (2014), pp. 2224–2230

[66] C.C. Riedl, N. Luft, C. Bernhart, M. Weber, M. Bernathova, M.K. Tea, et al.

Triple-modality screening trial for familial breast cancer underlines the importance of magnetic resonance imaging and questions the role of mammography and ultrasound regardless of patient mutation status, age, and breast density

J Clin Oncol, 33 (10) (2015), pp. 1128–1135


[68] MS Ong, KD Mandl

National expenditure for false-positive mammograms and breast cancer overdiagnoses estimated at $4 billion a year Health Aff (Millwood), 34 (4) (2015), p. 576-583


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.

Leave a Reply