Tomosynthesis has the potential for more accurate breast cancer screenings

Although still a relatively new form of mammography, tomosynthesis holds the potential to radically increase the ease and accuracy of breast cancer screenings. Sophie Peacock speaks to Dr Marina Alvarez Benito about how early-stage detection could be revolutionised by making 3D digital mammography the standard of care.

Since obtaining FDA and CE certification in 2011, digital breast tomosynthesis (DBT) has become a widely used technique, with many medical centres investing in the equipment.

Unlike traditional 2D mammography, tomosynthesis produces a 3D digital image of breast tissue. DBT uses equipment similar to conventional mammography: the patient is positioned in the machine the same way, with her breast compressed. The device within the X-ray beam tube then performs a small scan at an angle, obtaining information on the whole volume of the breast and providing a set of images, each displaying a millimetre.

In a study performed in Norway by Per Skaane, DBT was found to give far superior results to digital mammography or conventional 2D mammography, demonstrating increased sensitivity and so increased cancer detection, in particular for very small lesions.

Turn up the volume
A DBT scan can obtain as many images as there are millimetres in the compressed breast - so if the compressed breast is 40mm, then 40 images of that projection can be obtained. This enables the viewing of the whole breast tissue without compression. In a 2D image from traditional mammography, the whole breast volume is viewed as a single plane. However, tomosynthesis can provide much sharper images and avoid overlapping of structures.

Another extremely important impact of tomosynthesis's high accuracy is that it drastically reduces the need for referring patients for additional tests, diminishing distress for the patient.

A report released earlier this year in the UK showed that, in 2015 and for the third consecutive year, the number of eligible women aged 50-70 attending breast screenings had fallen, partly due to fear of false positives and overtreatment.

In the guidelines of a current clinical trial, the charity Cancer Research UK points out that DBT can take longer than a traditional mammography and, therefore, might be more uncomfortable. But could wide distribution of equipment that promises clearer, more sensitive results in one session help boost these figures again?

"Here in Cordoba, we have implemented a tomosynthesis device at our breast cancer screening unit; there are now several other public centres that offer DBT," explains Dr Marina Alvarez Benito, director of the breast cancer unit at Reina Sofía Hospital in Cordoba, Spain.

Previously, there were four screening rooms in Cordoba, but these have now been merged into a single unit with a centre dealing solely with breast cancer screening. The Preliminary results for DBT have been positive and clearly demonstrate the advantages of tomosynthesis over traditional mammography.

"We know that DBT is a safe method with a controlled radiation dose within acceptable limits, and that we can use it to improve sensitivity and specificity," Alavarez says. "In Spain, the screening programme for breast cancer is offered to all women - in our area, specifically between 50 and 69. Almost 80% of the target population takes part." The results of the programme so far have shown a high detection rate of cancer and of tumours diagnosed in their initial stages.

Healthy control
As screening programmes are aimed at engaging the healthy population, it's important to yield clear and accurate results. "We wanted to start to work with tomosynthesis, but we needed to begin with a screening programme that would allow a high number of women to benefit," Alvarez explains. "The results obtained in a national programme can have a major impact for the entire population."

The objective of using mammography in the healthy population is to be able to detect small lesions before they are clinically evident. With tomosynthesis, more cases are detectable and, therefore, the referral rate is reduced, enabling patients to start treatment sooner.

"I think these are the two most important, interesting aspects in a screening programme. We must convince regulatory authorities that it is working well in every respect, because a large number of women are participating and the results are encouraging," Alvarez says. "If we want to change a technology, we should obtain similar or better results."

A very high volume of women participate in breast screening programmes, so a large number of mammographies or tomosynthesis screenings must be performed daily. All tests need to be read and reported, and any change in this operating procedure can have a major impact.

"Regulatory authorities are reluctant to give approval until you demonstrate your plans are feasible," says Alvarez. "We were able to convince them that a tomosynthesis screening programme would be a good opportunity for Cordoba because the results we obtained at this unit would enable us to continue to make progress in the remainder of the early-screening unit."

Photo synthesis
Synthesised mammography is the latest development in tomosynthesis, where software enables all the 1mm images of compressed breast volume to be combined into a 2D image. This means that with a single shot, one can obtain a 3D and a 2D image.

"We have several arms in our study," Alvarez says. "One arm is of digital mammography, with two readings of that mammography; another arm is for tomosynthesis with synthesised mammography; the third includes all the technologies - DBT with synthesised mammography and conventional 2D digital mammography."

She explains that 98% of the women participating in the screening agreed to tomosynthesis when offered it. "Technicians are performing the study without problems, and radiologists have become familiar with the new reading system," she says. "I think this is an achievement in itself."

Tomosynthesis has been able to increase the detection rate by about 25% and reduce referrals to the reference unit by 50%.
"We have been able to see malignant lesions that are only detectable with tomosynthesis and not conventional mammography. I believe these highly encouraging results will unquestionably have an impact on how screening programmes will be organised in the future," Alvarez says.

Supported by the Maimonides Biomedical Research Institute in Cordoba, she is now performing a research study where participants undergo conventional mammography as well as tomosynthesis.

I don't recall
Alvarez's next challenge is to implement the same standards of her screening unit elsewhere in order to enable tomosynthesis to reach other private units.

DBT is not without its limitations, however. As well as requiring greater digital storage and archive space for its images, the equipment and procedures for tomosynthesis come at a slightly higher cost than traditional 2D mammography.

"But I think there are solutions for this," she suggests, "because once it is proven that a technology is safe and provides good results, the generalisation will be gradual but should involve no problem for widespread use."

Alvarez has now distributed the preliminary results of the 3,500 women included in the study - 17 breast carcinomas were detected - to a small circle of specialists working in the same area.

"We would like to wait until we have a larger volume of patients to state that our results are more consistent, but the disclosure of these results already provides, we believe, enough support to convince authorities that this technology is worthwhile."

The specificity of tomosynthesis is 99.94% and, while it is true that in 2D mammography specificity is as high as 94.43%, the best sensitivity and specificity results have been obtained in the DBT and the synthesised image arm. Alvarez explains that, in comparing the results in terms of recall and detection rate, tomosynthesis shows a referral or recall rate of 2.84%, whereas digital/2D mammography shows a rate of 5.88%.

She intends that these preliminary findings will be published and disclosed in forums with radiologists already working with DBT, as well as radiologists working on breast imaging diagnosis in screening programmes, so they can know how this project has been implemented and its results. The intention is to not only demonstrate the difference between tomosynthesis and conventional mammography, but also determine whether the latter will need to still be performed in the future.

"We knew that the advent of digital mammography would bring many new novelties," she says. "Tomosynthesis is one of the advances of digital mammography, and we are fully aware that it is has come to stay.