Additional US or DBT after digital mammography: which one is the best combination?
Elizalde A (1), Pina L (2), Etxano J (2), Slon P (2), Zalazar R (2), Caballeros M (2).
Digital mammography (DM) is widespread used for the detection of breast cancer, but its sensitivity drops in dense breasts. It is well known that additional breast ultrasound (US) and digital breast tomosynthesis (DBT) increase the sensitivity of DM. However, to our knowledge, there are no articles comparing the role of both additional techniques.
To assess the diagnostic performance of DM and the different combinations of DM + additional DBT and DM + additional US in an enriched sample of patients.
MATERIAL AND METHODS
Retrospective study in an enriched sample of 1042 patients. Out of them, 84 patients had histologically proven malignant lesions and 258 patients had benign lesions. Finally 700 patients with normal explorations or benign lesions without biopsy confirmation (but stable for at least 12 months) were included.
All of them underwent DM, US, and DBT examinations that were retrospectively reviewed by one expert radiologist, blinded to the final diagnoses. The DBT examinations were performed using one single view with wide angle (50°).
The reader categorized the cases as benign (BI-RADS 1 or 2) or malignant (BI-RADS 3-5) for DM and the different combination of techniques. The sensitivity (SE) and specificity (SP) were calculated with the PEPI software and the ROC curves of the different techniques and combinations were calculated by using the SPSS 15.0 software.
The SE and SP of DM were 69.05% and 88.20%, respectively. Additional DBT significantly increased the AUC of DM as well as additional US or the combination DM + DBT + US (P < 0.05). However there were no significant differences between the AUC of DM + US and DM + DBT (P = 0.7).
Additional US, DBT, or both, in combination with DM, significantly increased the AUC of DM. However, there were no significant differences between DM + DBT and DM + US.
CITATION Acta Radiol. 2016 Jan;57(1):13-8. doi: 10.1177/0284185114563641. Epub 2014 Dec 18.