2.1 Mammography

The principle of mammography is to use the physical properties of X-rays and the different densities of breast tissue to project an image of the breast onto the image receptor.¹¹ Currently, there are three main types of mammography available for screening: film mammography (FM), digital mammography (DM), and digital breast tomosynthesis (DBT).¹² FM is an X-ray image of the breast and has been officially recommended as a screening technique for BC by the guidelines from the National Cancer Institute of the United States since 1977.¹³ X-rays produce a latent image on the film, which is visualized by the chemical processing of the film emulsion. DM uses an electronic detector to replace the conventional screen-film system and displays the images digitally on a high-resolution monitor.¹⁴ DM has been available since 2000, which improves screening accuracy by decoupling image acquisition from the image display and optimizing them independently.^{12, 14} In this study, many European countries have replaced FM with DM for screening.¹⁰ DBT is a novel technique that produces a three-dimensional (3D) image of the X-ray attenuation coefficient by obtaining a series of projection images at different angles around the breast.^{8, 9, 15} DBT can reduce parenchymal overlap that masks cancers or creates false lesions and can potentially improve the accuracy of BC screening.¹⁶ Compared with FM, the size of cancer detected by DBT is smaller, and the rate of negative lymph node invasion is higher.¹⁷ A recent meta-analysis including 17 studies (1,009,790 participants) demonstrated an incremental cancer detection rate of 1.6/1000 screenings (95% confidence interval [CI]: 1.1–2.0, p < 0.001) and an absolute reduction in recall rate of 2.2% (95% CI: 3.0–1.4, p < 0.001) for DBT compared to that of two-dimensional mammography.¹⁶ However, there is no difference between DBT and DM in interval cancer detection rate,^18-20 and research on the direct effects of DBT on BC mortality, radiation-induced cancer, and quality of life is not yet available.²¹

The latest randomized controlled trial (RCT) for the effectiveness of mammography screening was carried out in Sweden, and 133,065 women aged 40–74 years were recruited. After 29 years of follow-up, women in the mammographic screening arm had a 27% (95% CI: 11%–41%) reduction in BC mortality compared with women receiving usual care.²² The relative risk of BC mortality for women who had undergone screening compared with that of those who had not was 0.80 (95% CI: 0.73–0.89) for the UK Independent Panel.²³ Pooled estimates from meta-analyses of RCTs (including six meta-analyses) also demonstrated that mammographic screening reduced BC mortality by at least 20%.²⁴

Although mammography is an effective tool for detecting BC at an early stage, its effectiveness is affected by several host factors, including breast density, age, and lesion size.¹² Breast density makes the greatest impact, because of similar X-ray attenuation of dense fibroglandular tissue with tumors, in increasing the masking effect due to overlapping tissue or image artifacts and decreasing the contrast between the lesion and the surrounding tissue.²⁵ The sensitivity of mammography was 30%–64% for extremely dense breasts compared with 76%–98% for mainly fatty breasts.⁹ Given these limitations, it is essential to explore supplemental screening tools in addition to mammography.

2.2 Ultrasound

Breast ultrasound, which can help distinguish benign/malignant breast masses by providing a multifaceted view of the mass and detecting blood flow, is an easy-to-operate, radiation-free, and painless way to screen for BC.²⁶ It can be performed by using a handheld ultrasound (HHUS) unit or an automated breast ultrasound unit, which is also called 3D ultrasound.¹²

Ultrasound is typically used as a supplemental tool for further evaluation of suspicious areas or as an alternative when mammography cannot be performed.²⁷ This is particularly true in young women or those with dense breasts, where mammography is less effective.²⁸ The Japan Strategic Anticancer Randomized Trial²⁹ demonstrated that screening with mammography plus ultrasound in women aged 40–49 years significantly improved the cancer detection rate (5.0 per 1000 vs. 3.3 per 1000) and sensitivity (91.1%, 95% CI: 87.2%–95.0% vs. 77.0%, 95% CI: 70.3%–83.7%) than those of mammography alone. For women with dense breasts, the sensitivity of mammography alone was 74% (95% CI: 69%–79%), while the sensitivity of combined mammography and ultrasound could be 96% (95% CI: 93%–97%).³⁰

Lately, mammography instruments have not been available in many low-resource settings. Ultrasound as a stand-alone screening test has been put forward, and several trials have been conducted in many countries. The American College of Radiology Imaging (ACRIN) conducted a prospective RCT of ultrasound as the primary screening option and demonstrated that the cancer detection rate with ultrasound was comparable to that of mammography.³¹ However, ultrasound also has certain drawbacks, such as being less specific and having a higher biopsy rate than mammography.⁹ Further, screening with manually performed HHUS is time-consuming and operator-dependent, and the consistency of reporting results from health professionals with different levels of diagnostic skills is poor.¹²

2.3 MRI

MRI can also be used for BC screening. It has a high soft-tissue resolution and is not affected by breast density. Thus, MRI has higher sensitivity for lesions that are difficult to characterize or detect by mammography.^{32, 33} However, due to the high cost and long examination time, MRI is temporarily designated as a choice for women at increased risk of BC.³⁴ In a multicenter RCT of screening with MRI versus mammography in women with familial risk, Saadatmand et al.³⁵ reported that MRI was capable of detecting more cancer cases (59 per 1000 vs. 22 per 1000; p = 0.0017) and smaller tumors (median size, 9 mm vs. 17 mm; p = 0.010) than those by mammography.³⁵

Abbreviated MRI is a novel technique with shorter image acquisition and interpretation times, which may increase the availability and reduce the cost of breast MRI. For instance, Kuhl et al. conducted a prospective observational study among 443 women and found that abbreviated MRI was associated with a significant reduction in image acquisition time compared to full breast MRI (3 min vs. 17 min).³⁶ In addition, the diagnostic accuracy of abbreviated MRI was equivalent to that of conventional MRI.³⁷

2.4 Clinical breast examination

CBE refers to breast and axillary palpation performed by the clinician on the patient, which is used in low- or middle-income countries as an alternative screening technique to mammography.^{38, 39} Although 5%–10% of BC cases can be detected by CBE alone, many factors affect the sensitivity of CBE, such as the clinician's experience, patients' age, and body mass index.⁴⁰ In 1998, a cluster RCT conducted in Mumbai, India, indicated that after 20 years of follow-up, biennial CBE led to a significant reduction in mortality by nearly 30% among women 50 years or older, while no significant reduction was seen in women younger than 50 years.⁴¹ There is still controversy regarding the effectiveness of CBE as a screening tool for BC.

2.5 Nonimaging novel screening techniques

Growing tumors actively or passively excrete debris into the bloodstream, including cell-free RNA and tumor DNA, which can be used as biomarkers for molecular-level tumor screening in the early detection of BC. Liquid biopsies help to detect tumor components in circulating blood or plasma.⁴² Some preliminary studies have shown that circulating proteins,^{43, 44} microRNAs,^{45, 46} autoantibodies,^{47, 48} and nucleic acid methylation⁴⁹ may be promising biomarkers for early detection of BC. However, most studies are still in the exploratory phase, and clinical trial data are scarce.⁵⁰ One study showed that the sensitivity of using a combination of circulating tumor DNA and cancer-associated proteins for asymptomatic BC was only 33%, suggesting that screening with liquid biopsy is still challenging and test methods with higher sensitivity need to be developed.⁵¹

3.1 Europe

In total, we collected data on the implementation of screening programs in 29 European countries,¹⁰ most of which have implemented organized population-based BC screening programs, except for the Russian Federation, Bulgaria, Greece, the Czech Republic, and Slovakia (where population-based screening programs were being piloted). The target age range of women enrolled in the mammography screening program was 50–69 years in 11 of 29 countries, which was the widest target age for BC screening recommended by the European Council.¹⁰ DM has completely replaced FM as a screening technique in 72.4% (21/29) of the countries. Seven countries also used ultrasound as a screening tool, including Finland, the Czech Republic, Austria, Belgium, Monaco, Italy, and San Marino. In the French program, CBE was used in addition to mammography.

The screening interval was 2 years in most countries except Malta and the United Kingdom (UK) (3 years). Examination coverage varied between 19.1% and 83.6% among countries, with the highest examination coverage observed in the United Kingdom and the lowest in Slovenia. A total of eight countries concentrated in Northern and Western Europe had examination coverage of over 60%, including Finland, Sweden, Norway, Denmark, the United Kingdom, Ireland, the Netherlands, and Luxemburg. Additional details are provided in Table 2.

The second report on the implementation of the Council Recommendation on Cancer Screening consolidated the current implementation and performance indicators of cancer screening programs in the European Union (EU) Member States⁶³ (Table S1). Of almost 15 million tests in the BC screening programs in the EU member states, the positive rate in women aged 50–69 years was 5.2% (range: 1.6%–11.8%). The average detection rate for invasive cancer was 5.2 per 1000 breasts screened, ranging from 1.8 (Portugal) to 8.1 (UK) per 1000 breasts screened, and the average detection rate for CIS was 1.0 per 1000 women screened, ranging from 0.3 (Portugal) to 2.1 (UK), with the proportion of CIS among all cancers being 16.3%, ranging from 1.2% (Poland) to 33.1% (Slovenia). Open surgery with benign results following a positive screening test, which was a rare undesirable outcome of BC screening, was 0.8 (range: 0.2–4.7) per 1000 women screened (Table S1).

3.2 Americas

For North America, we focused on Canada⁵² and the United States (US).⁵³ BC screening was primarily conducted through an organized population-based program in Canada, whereas it was opportunistic in the United States. In both countries, mammography was adopted as the screening technique, and the screening interval was set as 1–2 years, but the target age ranges were different; 50–69 years in Canada and 40 to >75 years in the United States. Examination coverage rates in Canada and the United States were 47.3% and 51.3%, respectively. The positive rate, CIS detection rate, and invasive cancer in Canada were 13.5%, 1.2 per 1000 women, and 4.6 per 1000 women, respectively (Table S2). However, the official data on performance indicators for the Canadian BC screening program were not updated in 2006. In addition, performance indicators are not available in the United States due to the implementation of opportunistic BC screening.

Latin America includes Central America, South America, and the Spanish-speaking countries of the Caribbean, and none of these countries meets all the criteria of the organized programs.¹² Of the 11 countries with available data, seven had a target age range for screening of 40–69 years, four used mammography as the screening technique, and five used mammography or CBE. Examination coverage in Latin America was unsatisfactory, with only four countries exceeding 30%, including Argentina, Chile, Brazil, and Uruguay. Additional details are provided in Table 2.

3.3 Asia

Japan⁵⁴ was the first Asian country to implement a national screening program with CBE in 1987 and later introduced mammography as a screening technique. The target age for screening in Japan was 40 years or older, and the screening interval was 2 years. China designed and implemented two large pilot BC screening programs in 2008, including the Chinese National BC Screening Program (CNBCSP)⁵⁵ and the Chinese BC Multi-technique Independent Screening Trial (MIST),⁶⁴ using techniques including CBE, mammography, and ultrasound. In urban and rural areas, the target age ranges were 35–69 years and 35–59 years, respectively, with examination coverage rates of 54.4% and 63.1%.⁵⁵ Malaysia⁵⁶ initiated opportunistic BC screening in 2009. CBE was performed triennially for women aged 20–39 years and annually for women at average risk of 40 years or older, and women at high risk had to undergo mammography. The examination coverage rate in Malaysia was 51.8%. Korea⁵⁷ initiated population-based screening in 1999 with a target age of 40 years or older, and screening options included CBE and mammography at a screening interval of 2 years. Singapore⁵⁸ initiated population-based screening in 2002 with a target age of 50–69 years, using mammography biennially. Vietnam⁵⁹ began pilot CBE screening in 2008, and the examination coverage rate was 15%–20%.

3.4 Oceania

The information we obtained about the implementation of BC screening programs in Oceania is limited and presented in Table 2. Australia⁶⁰ and New Zealand⁶¹ launched a population-based organized BC screening program in 1991 and 1999, respectively, with target age ranges of 50–74 and 45–69 years. Palau⁶² initiated a population-based screening program in 1997, which included CBE screening for women aged 21–64 years and mammography screening for those aged 40–74 years.