Background and purpose: The usefulness of colony forming assays (CFA) has been established for almost 40 years (Puck and Marcus, J.Exp.Med. 103: 653-666, 1956). Although time-consuming and not successful for all cell lines, it is generally considered to be the gold standard of assays for testing the sensitivity of cell lines to ionizing radiation or other cytotoxic agents in vitro. We recently found for several cell lines that the plating efficiencies of both control and irradiated cells is dependent upon the density of cells seeded for colony formation; that is, increasing cell inoculum levels resulted in a non-linear relationship with colony formation, even at relatively low colony numbers.
Material and methods: All data from a human melanoma cell line, transfected with c-myc or N-ras, as well as from normal human diploid fibroblasts, were taken to see how this phenomenon influenced outcome and interpretation of clonogenic assays. Survival was recalculated using all data, or only data with a linear relationship between inoculum level and colony formation.
Results: It is found that when data with a non-linear relationship between inoculum level and colony formation are included, survival can be underestimated due to inhibition of colony formation in treated cultures.
Conclusion: For validity, colony forming assays must be standardized to assure a constant relationship between the cell density and colony forming efficiency. This usually requires a much lower density of colonies than has been typically published for many cell survival studies.