| pubmed-article:2421876 | pubmed:abstractText | Human breast carcinomas have been one of the most difficult tumor types to culture in agar-based clonogenic assays. This fact has limited their clinical applicability. We have used statistically motivated experimental designs to systematically improve the clonal culture of enzymatically monodispersed primary human carcinoma cells in an anchorage-independent agar system. Based upon an initial comparison of two basal media, we selected one which gave the best colony growth and then sought to optimize the individual additives in the medium. Hydrocortisone, fetal bovine serum, and red blood cells all improved both plating efficiency and median size of colonies derived from breast carcinoma cells. Next, the concentrations of these three components were simultaneously idealized using response surface methodology. By these methods, it was found that the optimal concentration of hydrocortisone was 0.35 microgram/ml, fetal bovine serum was 6.5%, and red blood cells was 2.1 X 10(7) cells/ml. Using these culture conditions, we have achieved plating efficiencies of 0.39% and 0.19% for colonies with diameters greater than 50 (50 cells) or 70 (130 cells) micron, respectively. | lld:pubmed |
