Variables relacionadas con la respuesta a terapia sistémica primaria en el cáncer de mama

Abstract

OBJECTIVE: The purpose of the study is to identify features related to breast cancer as predictive factors of response to the Primary Systemic Therapy (PST), studying the correlation between clinical, radiological and pathological response and trying to identify groups of "responsive patients" and "non- responsive patients" to the PST. MATERIAL AND METHOD: A retrospective study was performed from May 2010 to February 2015 and 64 patients were included. Clinical, epidemiological, diagnostic, pathological, immunological and therapeutic features were registered. The PST response was assessed in clinical response (by physical exploration); radiological breast response and axillary response (MRI); and pathologic breast response and axillary response (Miller and Payne system). Data were analyzed using both and Student tests. The results with p<0.05 were considered statisticaly significant. Univariate and multivariate analysis were performed to determine the predictive value of features. RESULTS: 65.6% of the 64 patients showed clinical response; 68.8% radiological breast response; 70.3% radiological axillary response; 73.4% obtained pathological breast response and 53.1% pathological axillary response; 90.5% of patients who achieved clinical response also obtained pathological breast response vs. 40.9% in patients without clinical response (p = 0.001) and 88.1% of patients who achieved clinical response also obtained radiological response compared to 31.8% of cases without clinical response (p = 0.001). HER2 and triple negative tumors were associated with higher rates of clinical response (p = 0.018); Ki67 positive tumors achived a 76,2% of clinical response vs. 45.5% in the negative cases (p = 0.014); 86.4% of RP- tumors achived clinical response while RP+ tumors obtained a 54.8% (p = 0.011). The "well-defined" MRI phenotypes (types 1, 2, and 3) showed higher percentage of clinical response (p = 0.001), pathologic breast response (p = 0.022), radiological breast response (p = 0.001) and radiological axillary response (p = 0.008) when we compare them with the "no-defined" MRI phenotypes. In multivariate analysis, nuclear grade 3 was significantly associated with pathologic breast response (p = 0.045; OR: 2.82, CI: 1.022 to 7.816) and MRI phenotype "well-defined" showed association with clinical response (p = 0.001; OR: 0.41, CI 0.24 to 0.70); radiological breast response (p = 0.016, OR: 0.564, CI 0.354 to 0.899); radiological axillary response (p = 0.048; OR: 0.631, CI 0.401 to 0.996) and pathologic axillary response (p = 0.011; OR: 0.520, CI 0.315 to 0.860). CONCLUSIONS: The clinical, radiological and pathologic response in breast cancer after PST achive a good correlation among them. The RNM phenotype independently predicts clinical, radiological breast and axillary response as well as pathologic axillary response. Nuclear grade independently predicts pathologic breast response. Patients with MRI phenotype " well-defined " and nuclear grade 3 are more likely to be defined as "responsive patients".