Original Article

Evaluating the response of neoadjuvant chemotherapy for treatment of breast cancer: are tumor biomarkers and dynamic contrast enhanced MR images useful predictive tools?

Zijing Zhang, Wei Zhang, Yiting Jin, Hongying Wang, Fei Gu, Jian Zhou, Zhengyin Lao, Zude Xu, Feng Tang, Liping Zou, Weijun Tang, Rong Lu, Qiang Zou


Objective: In order to evaluate the therapeutic response to neoadjuvant chemotherapy (NAC) for breast cancer, this research focused on the changes in expression of tumor biomarkers and the correlations associated with changes of magnetic resonance imaging (MRI) pre- and post-NAC. We also compared the accuracy of MRI and pathology in terms of residual tumor extent after NAC.
Methods: MRI was performed before and after four courses of cyclophosphamide, epirubicin and paclitaxel (CET) NAC on 114 patients treated in Huashan Hospital (Fudan University) from December 2009 to January 2013. All patients were pathologically diagnosed with invasive breast cancer via core needle biopsy. A series of tumor biomarkers, including P-glycoprotein (P-gp) and Ki-67, was tested by immunohistochemistry in both core needle biopsy and surgical specimens. The changes in tumor biomarker expression and the shrinkage of tumor on MRI were observed. The residual tumor extent after NAC was compared in terms of MRI and histopathology, and the accuracy of MRI was evaluated by both residual tumor extent and by NAC therapeutic effect. Together, these methods enabled a prognostic estimate of NAC.
Results: The P-gp expression before NAC was used to evaluate the therapeutic effect of NAC. The upregulation of P-gp expression after NAC was associated with poor therapeutic effect (P=0.0011). The expression of Ki-67 was significantly down-regulated (P<0.0001) but it had no association with NAC response (P=0.9645). The mean extent of residual tumor after NAC as seen on MRI was 20.83 mm (±4.14 mm, 95% CI) and that of surgically removed specimens, 18.89 mm (±3.71 mm, 95% CI). The sensitivity of MRI was 95.1%, the specificity was 28.6%, the positive predictive value was 79.6%, and the negative predictive value was 66.7%.
Conclusions: P-gp status was an important factor affecting the pathological complete response (pCR) rate. The change in P-gp expression, from negative to positive following NAC treatment, indicated the emergence of drug resistance resulting from chemotherapy. The down-regulation of Ki-67 was associated with the decline of tumor proliferation. However, compared to the pre-NAC P-gp status, the pre-NAC Ki- 67 status had little prognostic value. Additionally, the evaluation of the efficacy of NAC by either MRI or histopathology was inconclusive.

Download Citation