Added Scans Caught Cancers That Mammograms Missed in Women With Dense Breasts

CATEGORY: WOMEN’S HEALTH

Among women whose screening mammogram came back normal but whose breast tissue was dense, a mammogram enhanced with contrast dye found about 19 additional cancers for every 1,000 women screened and a shortened MRI found about 17, while an automated ultrasound found about 4, in a randomized trial of more than 9,000 UK women published in The Lancet.¹ Every one of those cancers had been invisible on a mammogram read as clear.

White on White: Why Dense Tissue Hides Tumors

Breast tissue counts as dense when it holds more glandular and fibrous tissue than fat—a common finding, taking in the two highest of the four standard density grades, which together cover close to half of women who have mammograms. The trouble is physics: on an X-ray, fat looks dark but dense breast tissue looks white, and so does a tumor. Searching for one in the other is like looking for a snowball in a snowstorm. Dense tissue is also itself linked to a somewhat higher risk of developing breast cancer, so the women hardest to screen are among those with the most to gain from a better look.

Three Extra Scans, Assigned by Chance

The BRAID trial, run at ten UK hospitals, enrolled women aged 50 to 70 with dense breasts whose recent screening mammogram was normal. Each was assigned by chance—the design that makes the groups alike in everything but the scan—to one of three additional tests: a shortened breast MRI, which uses magnets rather than X-rays; contrast-enhanced mammography, a mammogram taken after an iodine dye is injected into an arm vein, so that the extra blood vessels a growing cancer recruits light up on the image; or an automated whole-breast ultrasound. This first planned report covers roughly 6,300 women who had completed their extra scan, about 2,000 in each group.

Nineteen, Seventeen, Four

Contrast-enhanced mammography detected 19.2 cancers per 1,000 women scanned and the shortened MRI 17.4, while the automated ultrasound found 4.2. Most of what the top two methods found was invasive cancer—15.7 and 15.0 per 1,000 respectively, against 4.2 for ultrasound—and the tumors they caught were around half the size of those the ultrasound picked up, which is the point of screening: finding cancers while they are small. In all, 85 cancers turned up in women whose mammograms had shown nothing.

Callbacks, Dye Reactions, and a Question Still Open

The sharper eyes came at a cost. Nearly one woman in ten scanned with MRI or contrast-enhanced mammography (9.7% of each group) was called back for more testing, against 4% with ultrasound, and biopsies followed for 4.9% and 4.4% versus 1.5%. Set the callback numbers beside the cancer numbers and the meaning of a callback is clear: far more women were recalled than were found to have cancer, so a callback after a screening scan is a prompt for a closer look, not a diagnosis. Among the roughly 2,000 women who received the contrast dye, 24 had a reaction to it, one of them serious. The researchers were direct about what the trial has not yet shown: longer follow-up is needed to learn how these scans affect interval cancers (those surfacing between screening rounds), overdiagnosis—the detection of cancers that would never have caused harm—and, ultimately, deaths from breast cancer.

The Dutch Trial That Pointed the Same Way

BRAID does not stand alone. In a Dutch randomized trial of women with extremely dense breasts, published in the New England Journal of Medicine, offering a supplemental MRI after a normal mammogram cut the rate of interval cancers in half—2.5 per 1,000 women in the MRI-invited group versus 5.0 per 1,000 with mammography alone.² Fewer cancers surfacing between screens is exactly what would be expected if the extra scan is catching real disease earlier, and the new trial extends that picture by lining up three supplemental methods against each other in the same population.

Key Takeaways

• In women with dense breasts and a normal mammogram, contrast-enhanced mammography found 19.2 cancers per 1,000 women and shortened MRI 17.4—roughly four times what automated ultrasound found (4.2 per 1,000).
• The extra detection carried costs: nearly one in ten women scanned with the top two methods was called back, and far more callbacks ended without a cancer diagnosis than with one.
• The trial measured detection, not survival; whether finding these cancers earlier translates into fewer breast cancer deaths awaits longer follow-up.
• An earlier Dutch randomized trial found supplemental MRI halved the cancers that surface between screening rounds in women with extremely dense breasts.

References

1. Gilbert FJ, Payne NR, Allajbeu I, et al. Comparison of supplemental breast cancer imaging techniques—interim results from the BRAID randomised controlled trial. The Lancet. 2025;405(10493):1935–1944. link
2. Bakker MF, de Lange SV, Pijnappel RM, et al. Supplemental MRI Screening for Women with Extremely Dense Breast Tissue. New England Journal of Medicine. 2019;381(22):2091–2102. link