Should screening using digital breast tomosynthesis vs. digital mammography be used in organised screening programmes for early detection of breast cancer in asymptomatic women?

The majority of the GDG agreed that the balance of desirable and undesirable effects probably favours DBT. The overall certainty in the evidence is very low because of the uncertainty in the effect for interval breast cancer, and the absence of data regarding the downstream impact on breast cancer mortality. 

The evidence shows that when using DBT there are greater numbers of breast cancers detected. In addition, there may be less overdiagnosis when using DBT since a higher proportion of invasive breast cancers are detected compared with non-invasive ductal carcinoma (DCIS) which are clinically less relevant cancers. Evidence also found that there may be fewer women returning for assessment because of a false positive result. However, adding DBT may have little to no effect on interval breast cancers, but this evidence is very uncertain. Overall, the desirable effects of adding DBT were judged as moderate and probably outweigh the harms that were judged as small.

There are moderate costs associated with DBT over DM, which could increase health inequities if implemented. However, implementation of DBT would be facilitated by increased availability of DBT machines, adequate human resources (radiologists and technical personnel), and financial resources.

The majority of the GDG agreed with the recommendation: 8 members voted for a conditional recommendation for the DBT, 1 member voted for a strong recommendation for DBT, 3 members voted for a conditional recommendation for either DBT or DM, and 1 member abstained.

The GDG developed a specific recommendation for women with high mammographic breast density on the use of DBT in the context of an organised population-based screening programme. See recommendation Screening in women with high breast density: tomosynthesis vs mammography. 

The availability of DBT machines, adequate human (radiologists and technical personnel) and financial resources, and the ability to reduce appropriately recall rate may favour the implementation of this recommendation.

Other implementation considerations include:

• There may be variability in the quality of DBT machines currently available and in their methods of capturing images. For example, a wide-angle form of DBT image capture and single view DBT format may result in different breast cancer detection rates.
• New quality assurance standards of technologies and screening programmes must be considered in choosing DBT over DM. In addition, specific standards for synthesised 2D imaging, and their use in comparison to previously captured DM screening images are necessary in order to implement this recommendation.
• There will be significantly increased data storage needs for screening programmes using DBT as compared to DM. The IT ability to share images is also important for adequate implementation.
• Health equity in access to screening should be considered for countries choosing DBT-based screening programmes because of different resource settings and the capacity for different countries to be able to pay for DBT over DM, which may lead to increased health inequalities.

• Quality control or other standardisation of the technology for better image storage and quality of tomosynthesis should be implemented. Useful information is available within the European quality assurance scheme for breast cancer services.
• Screening monitoring and evaluation programmes should be able to distinguish tests done with DBT and with DM, stratified standard indicators should be computed. 

• Gather evidence pertaining to the obstacles in implementing DBT-based screening programmes. In order to accomplish this, screening programmes should be able to produce stratified indicators (see monitoring and evaluation considerations).
• Research regarding distribution of tumour grade/biology/prognostic measures in the additionally detected cancers might help to clarify the amount of possible overdiagnosis.
• Further research is needed to build the evidence on benefits and harms of DBT vs DM through comparison of direct outcomes, including impact on interval cancer incidence, stage of breast cancer at detection and mortality reduction.
• There is a need for research to assess the cost-effectiveness of a breast cancer screening programme using DBT in order to inform decision-making on breast cancer screening.  Appropriate models will assist in evaluating cost-effectiveness.
• Research is needed to define the quality parameters that need to be fulfilled for DBT-based breast cancer screening programmes.