Abstract P3-07-50: Early and accurate prediction of pathological response by magnetic resonance imaging and ultrasonography in patients undergoing neoadjuvant chemotherapy for operable breast cancer

Background: Neoadjuvant chemotherapy (NAC) reduces tumor size, and increases the frequency of breast-conserving surgery in operable breast cancers. Response predictions to NAC are made based on diagnostic imaging. Although various studies have reported the optimal timing for diagnostic imaging, this still remains unclear. Purpose: To identify the optimal timing of diagnostic imaging for the response prediction to NAC, and to evaluate the accuracy of response prediction. Methods: We evaluated 146 cases enrolled in the JONIE-1 study (a randomized controlled trial comparing zoledronic acid plus chemotherapy with chemotherapy alone as a NAC in patients with HER2-negative primary breast cancer). The chemotherapy regimen was FEC100×4 courses followed by weekly paclitaxel 80×12 courses (± zoledronic acid). Statistical analysis of the association between the tumor reduction ratio and the histopathological response and the prediction of pathological complete response (pCR) was performed using JMP software. The maximum tumor diameter was evaluated using magnetic resonance imaging and ultrasound on each patient 3 times (before NAC, after FEC treatment, after NAC) and tumor reduction ratios were calculated. Results: The average age of the patients was 49.8 years old. The menopause status was pre-menopause in 84 patients, and post-menopause in 58 patients. Regarding the subtype classification, 116 patients were of the luminal type (Lum) and 26 patients were triple negative (TN), and the Ki-67 labeling index had a median of 25% (1%-93%). Pathological examination demonstrated that 16 patients had pCR(11.3%, Lum, 9;TN: 7), and 126 patients had non-pCR (88.7%, Lum:107; TN:19). Seven patients had clinical-CR (4.8%, Lum: 4; TN: 3) at post-FEC, and 26 patients (17.8%, Lum: 20; TN: 6) at post-NAC. The prediction of pCR at post-FEC and post-NAC was evaluated by single variable analysis, resulting in an AUC (0.75645) p=0.0017 at post-FEC, and AUC (0.76563) p=0.0001 at post-NAC. The sensitivity / specificity / positive predictive value / negative predictive value were 0.625 / 0.873 / 0.385 / 0.948 at post-FEC, 0.250 / 0.976 / 0.571 / 0.911 at post-NAC, respectively. In TN cases, the values were 0.714 / 0.947 / 0.833 / 0.900 in post-FEC, and 0.429 / 1.000 / 1.000 / 0.826 in post-NAC. Conclusions: Diagnostic imaging evaluation performed after FEC treatment was useful for the prediction of pCR. Furthermore, the reliability was high in Triple Negative Sub type, but is affected by the existence of residual tumors in Luminal type. Citation Format: Kaise H, Ishikawa T, Miura D, Hasegawa Y, Horiguchi J, Hayashi M, Takao S, Kim SJ, Tanino H, Miyashita M, Konishi M, Shigeoka Y, Yamagami K, Suzuki M, Taguchi T, Akazawa K, Kohno N. Early and accurate prediction of pathological response by magnetic resonance imaging and ultrasonography in patients undergoing neoadjuvant chemotherapy for operable breast cancer. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P3-07-50.