Objective:

To explore the clinical

imaging

and pathological features of breast cancer with false negative digital breast tomosynthesis (DBT) and to avoid missed cancer.

Methods:

The database was retrospectively reviewed to collect pathologically confirmed breast cancer from July 2018 to October 2020 in Xinhua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine

which underwent preopera

tive diagnostic DBT with a Breast Imaging Reporting and Data System (BI-RADS) 1

2

or 3 diagnoses. Two experienced breast imaging specialists reviewed DBT images of these cases. Clinical

imaging and pathological features were collected.

Results:

A total of 31 cases with 32 cancer lesions of 378 cases were included in this study. Median size of these lesions was 0.8(1.0) cm

22 lesions no larger than 1 cm accounting for 68.8%. 2 lesions were not covered in the field of view

6 lesions were incorrectly classified

and 24 lesions were not visible on DBT images. The positive rate of preoperative ultrasound and magnetic resonance imaging (MRI) was 56.25% (18/32) and 100.00% (32/32)

respectively. Pathological types of these lesions include ductal carcinoma in situ of the breast (DCIS) 14 lesions (43.75%)

papillary carcinoma 8 lesions (25.00%)

and invasive carcinoma 10 lesions (31.25%).

Conclusion:

DBT false-negative breast cancer is not uncommon. Most of these lesions are small

and occurred in relatively dense breasts

which can be DCIS or invasive cancer. Supplemental ultrasonography and MRI may help reduce missed diagnosis in DBT false-negative cases.