Abstract
Breast cancer research frequently requires frozen cancer tissue samples for potential biomarker detection. Ninety-one breast cancer frozen tumor tissue samples were examined for factors contributing to the proportion of the tumor cells per sample. The strongest predictor of high proportion tumor cells was tumor grade as determined by the Scarff–Bloom–Richardson scoring system, followed by tumor size. Because the low-grade tumors had fewer tumor cells per specimen, the type of surgery and experience level of the sample collector was not significant. This variable should be considered by investigators when evaluating potential biomarkers.