The predictive value of digital mammography, breast ultrasound and their combination in the diagnosis of early breast cancer

Aim. To assess the predictive value of digital mammography, breast ultrasound and a combination of these methods in the diagnosis of early breast cancer - BC (carcinomas in situ and invasive tumors up to 1.0 cm in size). Outcomes and methods. We retrospectively reviewed clinical records of 110 patients of the FSBI “N.N. Blokhin National Medical Research Center of Oncology» of the Ministry of Health of Russia Federation who were examined and treated for early BC (ductal carcinoma in situ, lobular carcinoma in situ, and invasive breast cancer no larger than 1.0 cm in size without regional or systemic tumor spreading). All patients were examined using digital mammography across two projections, with targeted magnification if it was necessary, as well as breast ultrasound with B-mode, Doppler ultrasound and elastography. We analyzed a frequency of establishing various BI-RADS categories according to data of mammography and breast ultrasound. We also assessed breast tissue density, a presence of nodules and calcifications and their radiological characteristics, as well as a frequency of false-negative results of mammography and breast ultrasound, their radiological characteristics, and causes of obtaining the false-negative results. According to the data of mammography and breast ultrasound, the study group was divided into 2 subgroups: patients who were diagnosed with BC (BI-RADS 5) and patients with changes in the breast which were not clearly interpreted as malignant (BI-RADS 0-4). Statistical analysis was carried out using the SPSS 20.0 program; differences were considered statistically significant with p<0,05. Results. According to the data of mammography, BI-RADS category 5 and 4 were determined in 80 (75,5%) and 9 (8.5%) patients, respectively; in other cases, the findings were interpreted as BI-RADS category 0-3. BC diagnosis (BI-RADS 5) was established significantly more often compared to BI-RADS 4 at a low breast tissue density (82.6% vs 50%, p=0.004), the presence of breast nodules (70% vs 7.7%, p<0.0001) and calcifications (68.8% vs 30.7%, p=0.002) and malignant changes in the microcalcifications (52.7% vs 25%, p=0.002). According to the data of breast ultrasound, BI-RADS category 5 and 0-4 were determined in 78 (73.6%) and 28 (26.4%) patients, respectively. The most pathognomonic ultrasound signs of early BC in the study group included: the presence of irregular hypoechoic masses with uneven not circumscribed margin, of non-parallel orientation, with an echogenic halo, acoustic shadowing and diffuse vascular changes. In 8 out of 26 patients whose BI-RADS category, based on mammography data only, was interpreted as 0-4, additional breast ultrasound increased the BI-RADS category to 5. The findings that allowed to establish a correct radiological diagnosis were: irregular shape of masses, uneven not circumscribed margin, hypoechoic structure, non-parallel orientation, presence of an echogenic halo and acoustic shadowing. Conclusions. The combination of mammography and breast ultrasound increased the likelihood of the identification of malignant breast neoplasms and increased the frequency of interpretation of the revealed findings as BI-RADS category 5 from 75,5 to 83%. Along with this, the frequency of determining BI-RADS category 4-5 increased from 85.6 to 93.5%. The rationale of using this combination of diagnostic methods for screening BC in Russia can be assessed by conducting a larger study and after an analysis of its economic efficiency.

ductal carcinoma in situ, lobular carcinoma in situ, breast microcarcinomas, early diagnosis, mammography, breast ultrasound, BI-RADS, errors in radiological diagnosis of early breast cancer