Comparison of mammography, sonography and MRI in patients receiving neoadjuvant chemotherapy for breast cancer: Final results of a prospective study

10538 Background: Imaging techniques used to evaluate response during neoadjuvant chemotherapy for breast cancer include mammography (Mx), sonography (US) and MRI. This study prospectively evaluated the ability of each technique to determine pathologic response in breast cancer patients undergoing neoadjuvant chemotherapy. Methods: Forty women with operable breast cancer (T ≥ 2 cm, N0–1, M0) were treated with four cycles of anthracycline-based and taxane-based neoadjuvant chemotherapy as part of a phase II clinical trial. The longest diameter of each tumor (n = 48 neoplastic foci) was measured by Mx, US and MRI at baseline, after two cycles and after 4 cycles of chemotherapy, before surgery. Tumor size at pathology was determined and considered as gold standard of response. Differences among techniques in measuring tumor diameters were evaluated by means of t-test. Results: At baseline, US provided statistically higher measures than Mx (mean difference: 6.3 mm, p < 0.0001) or MRI (mean difference: 5.6 mm, p < 0.0001). No difference was observed between Mx and MRI (mean difference: 0.59 mm, p = 0.5). After two cycles of chemotherapy, tumor diameter measured by US was significantly longer than that measured by Mx (mean difference: 8.8 mm, p < 0.0001) and significantly shorter than that measured by MRI (mean difference: −5.1 mm, p = 0.0009). Mx provided longer measures than MRI (mean difference: 4.4 mm, p = 0.0034). After four cycles of chemotherapy, US provided significantly shorter measures than Mx (mean difference: 11.8 mm, p < 0.0001) and MRI (mean difference: −3.3 mm, p = 0.007), whereas tumor size measured by Mx was significantly longer than that measured by MRI (mean difference: 6.75 mm, p = 0.0027). In addition, the tumor diameter measured by the pathologist was longer than that measured by US (mean difference: 2.6 mm, p = 0.09) and significantly shorter than that measured by Mx (mean difference: −7.8 mm, p < 0.0001). No statistically significant difference was observed between MRI and pathological measures (mean difference: −1.0, p = 0.5). Conclusions: This study provides further evidence that, among imaging techniques, MRI is the best method to evaluate the tumor size after neoadjuvant chemotherapy for breast cancer. No significant financial relationships to disclose.