Mammographic assessment of breast density as a tool for predicting the response to neoadjuvant therapy in breast cancer patients
Abstract
Background and aims. Breast cancer (BC) is the most frequently diagnosed cancer and the leading cause of cancer-related death among women worldwide. For locally advanced diseases and high-risk tumors, neoadjuvant therapy (NAT) is the treatment of choice. Some studies show that mammographic density (MD) tumor margins and the presence of microcalcifications play a prognostic role in BC patients. Hence, the objective of this retrospective study was to assess if MD could predict the response to NAT among different molecular subtypes of BC patients undergoing NAT at The “Prof. Dr I. Chiricuta” Oncology Institute of Cluj-Napoca, Romania (IOCN). Furthermore, the association between MD, tumor margins and the presence of microcalcifications with clinico-pathological data was analyzed.
Methods. Eighty-four breast cancer patients diagnosed and treated at IOCN were included in this study. The morphological characteristics of the tumors were framed according to the BIRADS lexicon. The presence or absence of microcalcifications was also assessed. First, the significance of associations between breast density, margins and microcalcifications and clinico-pathological parameters of the patients were tested with Fisher or Fisher-Freeman-Halton Exact Test. Next, using multinomial logistic regression, we modelled the associations between the pathological response measured by Miller Payne and Residual cancer burden (RCB) systems and the BI-RADS. Variables having significant univariate tests were selected as candidates for the multivariable analysis (adjusted model).
Results. Breast densities were significantly associated with the age of the patients (p=0.01), number of positive lymph nodes (p=0.037), margins (p=0.002) and combined categories of Miller-Payne (p=0.034) and RCB pathological response (p=0.021). Margins was significantly associated with ki67 proliferation index (p=0.029), estrogen receptor (ER) (p=0.007), progesterone receptor (PR) (p=0.019), molecular subtype (p<0.001) and the number of clinically observed positive lymph nodes at diagnosis (p=0.019).
Conclusions. In our cohort, BC patients with lower MD had higher odds of achieving pCR following NAT, suggesting the role of MD as a clinical prognostic marker. Larger multicenter studies are warranted to validate the prognostic value of MD, which could aid in patients stratification based on their likelihood to respond to NAT