TOPLINE:

Digital breast tomosynthesis (DBT) combined with digital mammography (DM) detected 50% more breast cancers at baseline compared with DM alone, showed no reduction in the incidence of interval cancers, and maintained a 16% excess in the incidence of cumulative cases.

METHODOLOGY:

• Researchers pooled data from two randomised controlled trials within the MAITA consortium and included 100,743 women aged 45-69 years (664 cases of breast cancer) screened at baseline, of whom 82,938 were followed up for the cumulative incidence at subsequent rounds.
• Participants were randomly assigned at baseline to undergo either DBT combined with DM (experimental group) or DM alone (control group); however, both groups underwent DM screening at the follow-up round, with women aged 45-49 years followed up for 21 months and those aged 50-69 years followed up for 33 months after recruitment or until the next screening test.
• The analysis compared histologic and molecular characteristics of cancers detected in the experimental and control groups separately at the first round and after a negative screening test, which included interval cancers and cancers detected at the following round when both groups underwent DM.
• Co-primary outcomes included the comparison of the efficacy of breast cancer screening with DBT vs DM in reducing interval cancers and the incidence of advanced breast cancers.

TAKEAWAY:

• At baseline, DBT combined with DM detected 50% more cancers than DM alone (incidence rate ratio [IRR], 1.50; 95% CI, 1.29-1.75), with the increase consistent across most tumour types except for large tumours (≥ 20 mm) and grade 3, human HER2-positive, and triple-negative cancers.
• The overall incidence of interval cancers was similar between the experimental and control groups (0.14% vs 0.15%; IRR, 0.95; 95% CI, 0.69-1.32; P = .763), indicating no significant reduction in cancers detected after a negative baseline screening test and before the follow-up round.
• At the second screening round using DM in both groups, the overall detection was lower in the experimental group (IRR, 0.82; 95% CI, 0.67-1.01), with the reduction appreciable for large tumours. A lower detection of cancers was observed in the experimental group with better prognosis, but both groups showed similar detection for node-positive, grade 3, hormone receptor-negative, HER2-positive, and Ki67-high cancers.
• The cumulative incidence over the study period showed a 16% excess of cancers in the experimental group (IRR, 1.16; 95% CI, 1.04-1.30; P = .008).

IN PRACTICE:

"The cancers detected by DBT and not by DM are early detection of clinically relevant cancers; only longer follow up will answer whether the remaining excess detection is overdiagnosis or early diagnosis of cancers that would occur more than two years later," the authors wrote.

"Screening with DBT likely detects tumors that would have presented as larger tumors in later screening rounds if screening was conducted with DM," they added.

SOURCE:

This study was led by Pamela Mancuso, Azienda Unità Sanitaria Locale-IRCCS di Reggio Emilia, Reggio Emilia, Italy. It was published online on March 19, 2026, in the European Journal of Cancer.

LIMITATIONS:

Subgroups of cancers counted few cases per group, resulting in imprecise estimation. Stratified analyses could not be conducted due to the lack of information on breast density and other risk factors. External radiologists blinded to the groups have not yet reviewed the diameter of cancers without pathologic assessment. The follow-up was incomplete. The study was conducted using a screening protocol inconsistent with recent European guidelines, and the experimental group included the outdated DM in addition to DBT.

DISCLOSURES:

This study was partially funded by the Italian Ministry of Health. Five authors reported receiving speaker fees or grants or having other ties with GE Healthcare or various other sources.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.