Introduction
This position paper on screening for breast cancer (BC) has been proposed by the Executive Board and the Scientific Committee of the European Society of Breast Imaging (EUSOBI) and approved by 30 national breast radiology bodies/sections (Table 1). The aim is to give a clear message in favour of screening mammography to national/local governments, policy makers, referring physicians and the general population.
Table 1
List of 30 national breast radiology bodies who signed a Memorandum of Understanding with the European Society of Breast Imaging and co-authored this paper
Austria | WG on Breast Imaging, Austrian Roentgen Society, Österreichische Röntgengesellschaft (ÖRG) |
Belgium | Senology Section of the Belgian Society of Radiology |
Bosnia and Herzegovina | Association of Radiology of Bosnia and Herzegovina |
Bulgaria | Bulgarian Society of Breast Imaging |
Croatia | Croatian Society of Radiology Working Group of Breast |
Czech Republic | Association of Czech Breast Radiologists |
Denmark | Danish Society of Breast Imaging |
Estonia | Breast Imaging Subgroup of Estonian Society of Radiology |
Finland | Radiological Society of Finland/Breast Radiologists of Finland |
France | Société d'Imagerie de la Femme (SIFEM) |
Germany | AG Mammadiagnostik / Breast Imaging Working Group of the German Roentgen Society |
Greece | Hellenic Breast Imaging Society |
Hungary | Section of Breast Diagnostics, Hungarian Society of Radiologists |
Iceland | The Breast Imaging Group of The Radiological Society of Iceland |
Ireland | Irish Breast Radiology Group |
Israel | Israel Breast Imaging Society |
Italy | Italian College of Breast Radiologists by SIRM (Società Italiana di Radiologia Medica) |
Lithuania | Lithuanian Radiology Association |
Moldova | Department of Breast Imaging in the Society of Imagists of the Republic of Moldova |
The Netherlands | Dutch College of Breast Imaging (DCBI) |
Norway | Norwegian Society of Breast Imaging |
Poland | Sekcja Diagnostyki Obrazowej Chorób Piersi; Polskie Towarzystwo Radiologiczne |
Portugal | Breast Imaging Section of Portuguese Society of Radiology and Nuclear Medicine (SPRMN) |
Romania | Romanian Society of Breast Imaging |
Serbia | School of Breast Imaging |
Slovakia | The Section of Breast Imaging of Slovac Radiologic Society |
Spain | Spanish Society of Breast Imaging, Sociedad Espaňola de Diagnostico e Interventencionismo de la Mama (SEDIM) |
Sweden | Swedish Breast Imaging Society |
Switzerland | Breast screening representative of the Swiss Radiological Society |
Turkey | Turkish Society of Radiology Breast Imaging Working Group |
Breast cancer as a major health issue and the role of mammography in early diagnosis
All over the world, BC remains a major issue for public health. Increasing numbers of new cases and deaths are observed in both developed and less developed countries, only partially attributable to the increasing population age. In the 28 member states of the European Union, there were 361,608 new BC cases in 2012 and these are estimated to have increased to 373,733 in 2015 (+3.4 %); deaths were 91,585 and 95,357, respectively (+4.1 %) [1]. No major differences in this trend can be appreciated across European countries.
Anzeige
Notwithstanding its intrinsic limitations in terms of sensitivity and specificity, mammography remains the main tool for population-based mass screening with demonstrated effectiveness in reducing mortality and allowing for conservative treatment, as already stated by EUSOBI [2]. Tumour stage at diagnosis of BC still significantly impacts on overall survival even in the current era of effective systemic therapy. Thus, early diagnosis remains crucial. This principle has been recently confirmed by an interesting population-based study from the Netherlands Cancer Registry, which evaluated more than 170,000 BC patients. The proportion of patients receiving neoadjuvant/adjuvant systemic therapy increased from 53 % in 1995–2005 to 60 % in 2006–2012. However, in 2006–2012 the mortality for larger tumours remained greater than that for smaller tumours, significantly for the comparison of T1c and T1a stage, and was independent from nodal status [3].
The evidence in favour of screening mammography has been recently summarized by the International Agency for Research on Cancer (IARC) [4]. Upon randomized controlled trials, the reduction in BC mortality due to screening mammography is confirmed for women between 50 and 69 years of age. Considering 20 cohort studies and 20 case-control studies, the estimated reduction in BC mortality is 40 % for women aged 50–69 years who take up the invitation and 23 % when also including those who do not accept the invitation, as a societal effect of the screening policy. From cohort studies, a mortality reduction has also been estimated for women aged 40–49 years and 70–74 years, though the evidence from published studies was considered to be “limited”. Available data did not allow the IARC working group to define an optimal screening interval. However, we should consider that the majority of European countries opted for biennial screening in the 50- to 69-year-old cohort. When 40- to 49-year-old cohort is invited, the yearly interval is generally adopted in consideration of a potential higher speed of BC growth and of a lower sensitivity of mammography due to the higher breast density.
The average cumulative risk for a false-positive recall in organized screening programmes has been evaluated by the IARC working group to be about 20 % for women aged 50–69 years who have ten screens in 20 years, while the needle biopsy rate for a false-positive finding is lower than 1 % per round [4]. In addition, it should be noted that screening mammography allows for both downscaling clinico-pathological features of invasive BCs and reducing the impact of loco-regional and adjuvant treatments [5‐8].
With regard to overdiagnosis (i.e. the rate of screen-diagnosed BCs otherwise unnoticed during the patient’s lifetime), the IARC working group accepted the estimate provided by the Euroscreen working group [9], equal to 6.5 % (range 1–10 %), which was calculated on the basis of the difference in the cumulative probability of a BC diagnosis among women receiving or not receiving screening mammography, taking into account lead time and underlying increasing incidence. If these factors are carefully considered, a similar estimate of overdiagnosis (4–11 %) is also obtained from randomized controlled trials [4]. Notably, while overdetection (resulting from a specific action by the breast radiologist evaluating a finding as suspicious) should be distinguished from overdiagnosis (which also implies an essential role of the pathologist) [10], further efforts should be dedicated to the reduction of the most important negative consequences of overdiagnosis, i.e. overtreatment.
Anzeige
Risk of radiation-induced breast cancer
Radiation-induced BCs from mammography were estimated, based on models including different factors. For the 50- to 69-year age group, taking into account a latency time of 10 years and a dose of 2.5 mGy per screening round, the risk of radiation-induced BC death has been estimated to be 1 per 100,000 screened women. The risk of radiation-induced BC due to screening mammography is at least 100 times lower than the probability of avoiding a BC death [4]. Applying a mortality reduction rate of 43 %, biennial screening mammography performed in 100,000 women saves 350 lives [11]. For the 40- to 49-year age group, the problem of radiation effects must be more carefully considered and depends on the estimated magnitude of radiation- induced BCs. Importantly, most of radiation-induced BCs will be cured [12].
Screening models
On the basis of the available evidence, the EUSOBI and the above-listed national breast radiology bodies strongly support screening mammography of the female population at average BC risk, typically from 50 to 69 years of age; extension of this up to 73–75 years, biennially, is a second priority. Extension from 40 or 45–49, with annual screening, can be evaluated as a third priority, country-by-country. Age selection and screening interval should be adapted to national demographics and local priorities. Importantly, these societies strongly discourage the use of methods for screening such as thermography or other optical imaging tools as an alternative to mammography [13]. Moreover, these societies also discourage the use of ultrasound as a primary screening tool in asymptomatic European women at average risk of BC.
Preference should be given to population-based screening programmes on a regional/national basis with double reading rather than spontaneous mammographic screening with a single reading, given the advantages of the former in terms of higher specificity and positive predictive value [14, 15], lower cost, as well as structured quality controls and central data management. This concept has also been recently reinforced by the IARC working group in the above-mentioned paper [4].
In a wider framework, the EUSOBI and the above-listed national breast radiology bodies are aware of the open debate in other contexts such as that in the USA where the Society of Breast Imaging and the American College of Radiology support annual screening mammography from the age of 40 years by informing women on the advantages of early BC diagnosis [16]. The recent recommendations of the American Cancer Society [17] can be a reference for the US context: (1) regular screening mammography starting at age 45 years (strong recommendation); (2) annual screening mammography from 45–54 years of age (qualified recommendation); (3) from 55 years of age, transition to biennial or continuing annually (qualified recommendation); (4) opportunity to begin annual screening from 40–44 years (qualified recommendation); (6) continue screening mammography as long as women’s overall health is good and they have a life expectancy of ≥10 years (qualified recommendation); (7) no suggestion for screening clinical breast examination at any age (qualified recommendation).
Breast density
The EUSOBI and the above-listed national breast radiology bodies are aware of the masking effect of increased breast density, strongly impacting on the sensitivity of screening mammography, declining from 86–89 % for almost entirely fatty breasts to only 62–68 % for extremely dense breasts [18]. Studies aimed at reducing this negative effect by means of supplemental screening tools, such as manual or automated breast ultrasound, are welcome, especially when evaluating the cost-effectiveness of the additional tools on the large scale of population-based screening programmes. These societies also take into consideration the role of breast density as an independent BC risk factor, although this factor can be over-rated [19, 20], especially when reported as a communication to the women. In studies with a control group not limited to fatty breasts, the relative risk of women with dense breasts dropped to 2 or less [21, 22]. At any rate, these societies consider the general adoption of direct digital mammography to be the first priority to improve the sensitivity in women with increased breast density.
The potential of digital breast tomosynthesis
These societies also consider the increasing evidence in favour of digital breast tomosynthesis (DBT) as a screening tool. Three prospective studies showed that DBT used as an adjunct [23‐25] or alternative [26] to two-dimensional (2D) digital mammography allows for a superior diagnostic performance when compared to the latter alone. Overall, DBT increases the detection rate from 0.5 to 2.7 per 1,000 screened women and reduces the recall rate from 0.8 to 3.6 per 100 screened women [27]. Of note, DBT is now proposed along with synthetic 2D views, practically solving the problem of an increased radiation exposure when DBT is performed as an adjunct to 2D digital mammography [28‐30]. All these aspects will probably also confer to DBT the status of future “routine mammography” in the screening setting. However, before introducing DBT in BC screening outside trials approved by ethical committees, we need evidence for a statistically significant and clinically relevant reduction in the interval cancer rate. This cautiousness is due to the need to avoid an increase in overdiagnosis and costs, in the absence of the demonstration of cost-effectiveness of screening DBT (proof of which may require very long studies). First results on a reduction from 0.7 to 0.5 interval cancers per 100 screened women were very recently reported from a large study in the USA [31], but further evidence is needed. Moreover, the probable increase in reading time associated with the use of DBT in screening [32] and its effects on sustainability of screening programmes should be considered before routine implementation.
Preference for digital instead of film-screen mammography
Overall, looking at the course of technological evolution of mammography in the last decades and at the current trend in favour of DBT, these societies strongly support the adoption of direct digital mammography (not phosphor-plate computer radiography) instead of film-screen mammography in all countries. In fact, digital mammography implies many substantial advantages, including lower dose, higher image quality, possibility of post-processing, digital archiving, image transmission and no chemical pollution. We suggest that new mammographic units should be based on direct digital mammography technology and, when possible, equipped with DBT in readiness for the next evolution.
Need for certified and subspecialty-trained radiologists in the context of breast centres
Screening mammograms, with or without DBT, should be read by radiologists qualified as screening mammography readers. Proficiency tests at the regional/national/European level are encouraged in order to guarantee a standardized reading quality together with minimum screening numbers read per year.
Anzeige
It is essential that there is a continuity of care from screening mammography to diagnostic breast imaging, to needle sampling and treatment planning either in the context of a dedicated breast centre or in a screening centre that has a well-organized relationship with a diagnostic imaging facility. Whenever possible, radiologists should operate in the context of integrated breast units with the help of organized/structured cooperation among BC specialists.
Quality assurance programmes regarding breast radiology units/sections are also encouraged in the context of forthcoming new European guidelines for BC screening, diagnosis and treatment.
Preference for core or vacuum-assisted biopsy
Preference should be given to needle sampling of breast lesions using core biopsy or vacuum-assisted biopsy instead of fine needle aspiration [33], in consideration of the lower false-negative rate and/or inadequate sampling, unless strict cooperation with a cytologist allows for a demonstrable equally high diagnostic performance. This preference does not apply for sampling of lymph nodes suspected to be metastatic at ultrasound of axilla, where fine needle aspiration has been shown to be effective [34].
Women at increased risk for breast cancer
These societies are in favour of including, whenever possible, dedicated pathways for high-risk women (lifetime risk equal to or higher than 20 %), offering magnetic resonance imaging according to national or international guidelines and recommendations [35‐37]. In this regard, policies will be different, considering the heterogeneity of health systems across countries. Studies considering risk stratification for different screening strategies of women at increased BC risk are welcome.
Anzeige
Summary statement
EUSOBI and 30 national breast radiology bodies strongly support mammography as a population-based mass screening tool which results in a relevant reduction in BC mortality and leads to a favourable decrease in both loco-regional and adjuvant treatments in women attending these programmes. People and institutions questioning its validity despite a large body of evidence accumulated in more than three decades put women’s lives at risk.
Acknowledgements
The scientific guarantor of this publication is Francesco Sardanelli. The authors of this manuscript declare no relationships with any companies whose products or services may be related to the subject matter of the article. The authors state that this work has not received any funding. No complex statistical methods were necessary for this paper. Institutional Review Board approval was not required because the presented article is a special report. Methodology: Special report
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.