Table 2 Studies on breast MRI screening in women with dense breasts

Bakker 2019 [27], DENSE, The Netherlands

Women aged 50−75 with negative mammography and ACR BI-RADS density d

Multicentre RCT: MRI vs mammography

8061 in the MRI-invitation group (4783 accepted, 59%) 32,312 in the mammography-group

1^st

Interval cancer rate: 2.5‰ (MRI-group) vs 5.0‰ (mammography group) at intention-to-treat analysis; 0.8‰ (MRI group) vs 5.1‰ at complier average causal effect analysis

Supplemental MRI screening resulted in the diagnosis of significantly fewer interval cancers compared to mammography alone during a 2-year screening period.

Comstock 2020 [97], EA1141 ECOG-ACRIN, USA, Germany

Women aged 40−75 scheduled to screening with DBT with ACR BI-RADS density c or d

Multicentre, intraindividual, comparative, cross-sectional study with longitudinal follow-up: abbreviated MRI vs DBT

1444 women underwent both examinations

Invasive cancer detection rate: 11.8‰ (MRI) vs 4.8‰ (DBT)

Abbreviated breast MRI, compared with DBT, was associated with a significantly higher rate of invasive breast cancer detection.

Veenhuizen 2021 [96], DENSE, The Netherlands

ACR BI-RADS density d who already underwent first round of screening with MRI in the DENSE trial

Multicentre RCT: MRI vs. mammography

3436 women attending the second MRI screening, out of 4783 attending the first round

2^nd

Cancer detection rate: 5.8‰ (2^nd round) vs 16.5‰ (1^st round). False-positive rate: 26.3 ‰ (2^nd round) vs 79.8‰ (1^st round). All MRI-detected cancers in the 2^nd round were early stage (0–I) and node negative.

The results of 2^nd round play in favour of the MRI detection of prevalent cancers at 1^st round, not detected by mammography. The availability of prior MRI examination can partly explain the reduction in the false-positive rate.

den Dekker 2021 [98], DENSE, The Netherlands

Positive patients with MRI BI-RADS 3, 4, or 5 years at the first round in the DENSE trial

Multivariable logistic regression

454 women recalled for diagnostic work-up

1^st

79 true-positives and 375 false-positives. The prediction model based on all clinical characteristics and MRI findings could have prevented 46% of false-positive recalls and 21.3% of benign biopsies without missing any cancers.

Prediction models based on clinical characteristics and MRI findings may be useful to reduce the false-positives first-round screening MRI rate and benign biopsy rate in women with extremely dense breasts.