The effects of breast density on the benefits of mammograms with adjunctive ultrasonography in breast screening.

ABSTRACT

OBJECTIVE: Various efforts have been made to improve the accuracy of breast cancer screening. This study aimed to report differences in the contribution of ultrasonography to cancer screening assessments of dense and non-dense breasts. METHODS: The participants in this study were 29,640 Japanese women in their 40 s who underwent breast cancer screening at the Iwate Cancer Society between 2018 and 2021. This included women who chose mammography alone or mammography with adjunctive ultrasonography (overall assessment). They were classified into two groups according to the breast density in mammography dense breasts and non-dense breasts. Recall rate, breast cancer detection rate, and positive predictive value of the two screening-type groups were evaluated for each breast density group. RESULTS: Of the 29,640 women analyzed, 18,861 (63.6%) underwent mammography alone and 10,779 (36.3%) were by overall assessments. The number of women recalled was higher in the overall assessment group than in the mammography-alone group (2.9% vs. 1.9%, p < 0.01). The proportion of women in whom breast cancer was detected was higher in the overall assessment group than in the mammography-alone group (0.31% [n = 33] vs. 0.15% [n = 28], p < 0.01). For non-dense breasts, there were no significant differences in either the recall rate or the breast cancer detection rate between those who underwent mammography alone and those who underwent overall assessment. Conversely, for dense breasts, the recall rate after mammography alone was lower than that after overall assessment (1.8% vs. 3.8%, p < 0.01), and the breast cancer detection rate was higher after overall assessment than after mammography alone (0.40% vs. 0.18%, p < 0.01). CONCLUSION: We found the benefits of adjunctive ultrasonography with mammography to differ depending on breast density. This could be used to tailor the selection of screening modalities to individuals.