Validation of tissue microarray biomarker expression of breast carcinomas in Saudi women.

BACKGROUND: Analysis of immunohistochemical expression of a large number of tumor tissue samples with conventional techniques is costly, tedious and slow. Tissue microarray (TMA) technology can facilitate the sampling of over 500 tumors on a one-glass slide, which then can be analyzed by fluorescence in situ hybridization (FISH), RNA in situ hybridization, or immunohistochemistry. We attempted to validate this technique in breast cancer specimens by comparing the staining result obtained by TMA with the conventional whole-section method. PATIENTS AND METHODS: Eighty cases of breast cancer of diverse subtypes were used to build a breast cancer biomarker evaluation model. Serial sections of the recipient block were immunostained with a panel of 4 anti- bodies (ER, PR, HER2 and p53). Concordance of immunohistochemical expression between TMA sections and whole-sections was expressed as a kappa statistic.
RESULTS: Target tumors were accurately sampled by two cores in 19 of 26 donor blocks, and by only one core in 5 blocks. Failure to sample tumor was seen in 2 blocks. Concordance between the staining results of TMA and whole sections was good for PR (kappa=0.67) and ER (kappa=0.67) and very good for p53 (kappa=0.91) and HER2 (kappa=0.91), when all the 26 recipient blocks were included. The rate improved to excellent for p53 (kappa=1.0) and did not change for the other markers when concordance analysis was limited to recipient blocks that had been sampled by two cores.
CONCLUSION: TMA is a reliable technique for examining a large set of tumors. It shows immunostaining scores comparable to those obtained by conventional whole-section evaluation in breast cancer. However, some alterations are not detected due to heterogeneity of biomarker expression inherent in these tumors, but this shortfall can be improved.