Comparison of MRI With PET / CT in the Evaluation of Response to Neoadjuvant Therapy Based on the Molecular Subtypes of Breast Cancer

Eighty-eight patients with breast cancer who had undergone surgery after NAC were retrospectively analyzed between January 2018 and May 2020, at the general surgery clinic of Istanbul Medeniyet University Göztepe Training and Research Hospital. Demographic data, clinical findings, tumor markers, radiologic findings and pathology data were documented. With the use of MRI and / or PET-CT obtained before and after NAC, tumor staging was performed, and tumor size and the presence of metastatic lymph nodes and radiological response to NAC were evaluated. The response to NAC was evaluated according to the Response Evaluation Criteria in Solid Tumors (RECIST) and contrast enhancement patterns described by Tozaki et al. on MRI, and the classification PET Response Criteria in Solid Tumors (PERCIST) on PET-CT. The postoperative pathology study reported the followings: the size and histopathologic type of the tumor, its hormone receptor status, staining intensities (Allred score), HER-2 positivity, Ki-67 proliferation percentages, and the molecular subtypes; the presence and number of metastatic lymph nodes, the response of the tumor and, the metastatic lymph node, if any, to NAC according to the Sataloff classification. Patients with a Sataloff score of "A" were considered to have a pCR, those with the scores of "B" or "C" were considered to have a partial response. When examining the Sataloff scores, the presence of pCR in both the primary tumor and axilla was evaluated and compared with the presence of radiological complete response on MRI and / or PET-CT. In addition, the sensitivities of MRI and PET-CT in predicting pCR in patients undergoing NAC were statistically compared based on the molecular subtypes. Patients whose MRI and PET-CT findings were consistent with histopathologic data were included in the "Radiology-Pathology agreement " group. The sensitivity of imaging methods was determined by examining the rates of agreement in the molecular subgroups. The Statistical Package for Social Sciences software for Windows version 20 (SPSS Inc.; Chicago, IL, USA) was used for all analyses. Normality was assessed by the Shapiro-Wilk test. Categorical variables were compared using chi-square or, where appropriate, Fisher's exact test. The Mc-Nemar test was used for paired categorical data. Trends in tables larger than 2x2 were examined with the linear-by-linear association test. Kruskal-Wallis test was used to compare non-normal distributed variables in three groups. Cohen's kappa (κ) was calculated to test the agreement between the classifications of radiology and pathology.