The aim of this study was to define the concordance between tissue microarrays (TMAs) of different sizes and whole slide for 15 different antibodies in endometrial cancer and study the use of TMAs in preoperative endometrial samples. Cores of preoperative and hysterectomy specimens of 14 endometrial cancer and three atypical hyperplasia cases were collected in TMA blocks. Two 0.6-mm and two 2.0-mm cores were used from each sample. Different antibodies were tested in TMAs and compared with results of whole slides of hysterectomy. Tested antibodies were as follows: ER, PR, p53, Ki-67, MLH1, PMS2, MSH2, MSH6, ARID1A, stathmin, IMP3, L1CAM, PTEN, β-catenin, and p16. Seventeen cases with four cores per paraffin block (both 0.6 and 2.0 mm in duplicate) and 15 different antibodies resulted in a total of 1020 cores for both preoperative and hysterectomy specimen. Overall, 2.0-mm cores were more assessable for evaluation than 0.6-mm cores (96.0 versus 79.5%, p < 0.01). For most antibodies, a substantial to good agreement between hysterectomy TMA and whole slide was present, with lowest agreement for p16 and stathmin and perfect agreement for mismatch repair proteins. Preoperative TMAs showed for most antibodies moderate to perfect agreement with hysterectomy TMAs. In conclusion, 2.0-mm cores are the preferred size for immunohistochemical studies in endometrial cancer. For all tested antibodies, TMAs are a good alternative for whole slide analysis in scientific studies with large patient cohorts, even in preoperative endometrial samples. However, caution is required for interpretation of TMA results of p16 and stathmin.
Keywords: Biomarkers; Endometrial cancer; Endometrial sampling; Immunohistochemistry; Tissue microarray.