The Diagnostic Performance of Wide-Angle Digital Breast Tomosynthesis in Comparison to Hand-Held 2D Ultrasound in the Evaluation of Probably Benign & Malignant Lesions in Dense Breast Parenchyma

Abstract

Background:

Imaging of dense breast represents a diagnostic challenge for interpreting radiologists. Tomosynthesis and ultrasound are used as additional imaging tools to compensate the lower sensitivity of 2D mammography when examining a dense breast.

This study compares the diagnostic performance of 50° wide-angle digital breast tomosynthesis (wide-angle DBT) to hand-held breast ultrasound (HHUS) in the evaluation of probably benign and malignant breast lesions in dense breast.

Patients and Methods:

A retrospective review of 131 women with dense breasts (ACR C and D) was conducted during a 12-months period (October 2018 -October 2019). Out of the 131 women; 40 cases (30.5%) were reported to have 64 probably benign and malignant lesions (BI-RADS 3, 4 and 5) detected by combined wide-angle DBT and HHUS. Patients with negative results (BI-RADS 1) or typically benign lesions (BI-RADS 2) were excluded from this study.

Median age for the included 40 cases=46.8 years (range =30-67). Women above the age of 40 (n=35, 87%) were examined with combined 2D mammogram + wide-angle DBT and ultrasound. Women below 40 (n=5, 13%) were initially examined by ultrasound and obtained complementary wide-angle DBT mammography based on clinical necessity. Separate BI-RADS were given for each modality and independently correlated with histopathological results for BI-RADS 4 and 5, or routine follow up images for BI-RADS 3.

Results:

Among the 64 lesions; 10 (15.6%) were pathologically proven malignant in 5 cases (two cases were presenting with multi-focal disease), and 54 (84.4%) benign lesions pathologically proven or followed up in 35 patients.

Sensitivity was 100% for wide-angle DBT (10/10) and 80% for ultrasound (8/10). Specificity and positive predictive value for ultrasound was 94.4% (51/54) and 72.7% (8/11), respectively. For wide-angle DBT, specificity was 90.7% (49/54) and positive predictive value was 66.7% (10/15). Despite the high sensitivity, in our study, DBT showed equal diagnostic accuracy to that of breast ultrasound (92.2%, 59/64).

Both modalities were not similar in depicting malignant lesions; two more lesions were initially identified at wide-angle DBT images presented as <1cm area of architectural distortion (p=0.07). Other two diagnostic cases presented with acute inflammatory symptoms; they showed architectural distortion in DBT, which were false-positives for the later, yet, ultrasound showed mastitis and abscess formation. Ultrasound significantly identified more benign lesions than DBT mammography (p=0.001).

Conclusion:

Wide-angle DBT and HHUS are indispensable in identifying probably benign and malignant lesions in dense breasts. Our study showed the role of Wide-angle DBT in detecting small (<1 cm) malignant lesions presenting as architectural distortion. However, the radiologists experience with DBT and the significant clinical information added by ultrasound may impact the diagnostic performance. Therefore, supplemental HHUS to wide-angle DBT will increase accuracy and true negative results for a lower benign biopsy rate.