1085. Screening the Augmented Breast: Is Digital Breast Tomosynthesis Necessary for All Views With and Without Implant Displacement?
Authors * Denotes Presenting Author
  1. Derek Nguyen *; Duke University Medical Center
  2. Lars Grimm; Duke University Medical Center
  3. Jeffrey Nelson; Duke University Medical Center
  4. Karen Johnson; Duke University Medical Center
  5. Sujata Ghate; Duke University Medical Center
To determine the appropriate screening protocol for imaging the augmented breast to maximize cancer detection and minimize radiation exposure.

Materials and Methods:
This retrospective cohort study identified 218 women with implant augmentation who underwent digital breast tomosynthesis (DBT) examinations assigned BI-RADS assessments of 0, 3, 4, or 5 from July 28, 2014 to December 31, 2021. Our standard breast implant screening protocol includes digital breast tomosynthesis (DBT) with synthesized 2D (S2D) implant-displaced (ID) and non-ID craniocaudal (CC) and mediolateral oblique (MLO) views. Three fellowship-trained breast radiologists (experience: 8 - 20 years), who were blinded to pathology outcomes, independently reviewed all examinations to determine if the initial identified abnormality could be visualized on two different protocols: our standard protocol (S2D and DBT ID and non-ID views) and an alternative protocol (S2D and DBT ID and S2D non-ID only). Follow-up imaging and pathology outcomes were recorded for all abnormalities. Kappa index was used to determine agreement between readers. Cancer detection and false positive rates were calculated for both protocols to determine the added value of DBT non-ID views. Patient’s average glandular dose (AGD) and associations between cancer detection with patient and implant characteristics for both protocols were evaluated using two-sample t-test and fisher exact test.

A total of 195 unique patients (mean age 55 years ± 10) with 223 abnormal findings were included. There was moderate agreement between the readers, ?=0.582 (95% CI, 0.581-0.583). Subsequent biopsy was performed for 86 abnormalities: 58 (67%) benign, 8 (9%) high-risk, and 19 (22%) malignant. The false positive (69%) and high-risk lesion (9%) rates were similar between the two protocols. No significant difference in the malignancy rate between S2D non-ID views only (18/223, 8%) and DBT with S2D non-ID views (19/223, 9%), <em>p</em> = 0.92. Inclusion of the DBT non-ID views found one additional malignancy: a 0.8 cm architectural distortion seen only on the DBT non-ID MLO view in a patient with heterogeneously dense tissue and saline, retropectoral implants. Inclusion of the DBT non-ID view significantly increased the patient’s reported total examination AGD (with DBT: 21.9 mGy ± 5.0 vs S2D only: 12.6 mGy ± 5.0), <em>p</em> <0.001. This remained true when stratified by breast thickness (6.0-7.9 cm, 8.0-9.9 cm, >10.0 cm), <em>p</em> < 0.001.

The inclusion of DBT non-ID views is not significantly associated with increased cancer detection but does significantly increase overall examination AGD. For women with implant augmentation, switching the non-ID views from DBT with S2D to 2D-only in screening protocols will reduce radiation exposure and patient discomfort without affecting cancer detection.