Addition of Screening Breast US to Digital Mammography and Digital Breast Tomosynthesis for Breast Cancer Screening in Women at Average Risk
Menée à partir de données portant sur 1 003 femmes ayant bénéficié entre 2016 et 2018 d'une tomosynthèse mammaire numérique en combinaison ou non avec une mammographie numérique (âge moyen : 56 ans), cette étude évalue l'intérêt, du point de vue du taux de détection de lésions cancéreuses, de la sensibilité, de la spécificité et du taux d'interprétation anormale, d'ajouter une échographie mammaire à ces examens de dépistage
Background : Digital breast tomosynthesis (DBT) with or without digital mammography (DM) is the primary method of breast cancer screening. However, the sufficiency of DBT screening for women at average risk and the need for supplemental whole-breast US needs further investigation.
Purpose : To evaluate the added value of supplemental US screening following combined DM/DBT.
Materials and Methods : A retrospective database search identified consecutive asymptomatic women who underwent DM/DBT and radiologist-performed screening breast US simultaneously between March 2016 and December 2018. The cancer detection rate (CDR) per 1000 screening examinations, sensitivity, specificity, and abnormal interpretation rate of DM/DBT and DM/DBT combined with US were compared.
Results : A total of 1003 women (mean age, 56 years ± 8.6 [standard deviation]) were included. Among them, 12 cancers (mean invasive tumor size, 14 mm; range, 6–33 mm) were diagnosed. With DM/DBT and DM/DBT combined with US, the CDRs were 9.0 per 1000 screening examinations (nine of 1003 women; 95% CI: 4.1, 17) and 12 per 1000 screening examinations (12 of 1003 women; 95% CI: 6.2, 21), respectively, and the abnormal interpretation rates were 7.8% (78 of 1003 women; 95% CI: 6.2, 9.6) and 24% (243 of 1003 women; 95% CI: 22, 27). In women with negative findings at DM/DBT, supplementary US yielded a CDR of 3.2 per 1000 examinations (three of 925 women; 95% CI: 0.7, 9.4), sensitivity of 100% (three of three women; 95% CI: 29, 100), specificity of 82% (760 of 922 women; 95% CI: 80, 85), and abnormal interpretation rate of 18% (165 of 925 women; 95% CI: 15, 21). The three additional US-detected cancers were identified in women with dense breasts; no benefit was observed in women with nondense breasts.
Conclusion : The addition of breast US to digital mammography and digital breast tomosynthesis yielded an additional 0.7–9.4 cancers per 1000 women at average risk, with a substantial increase in the abnormal interpretation rate.
Radiology , résumé, 2020