Can information on the selectivity of a potential anticancer drug be obtained from in vitro cytotoxicity assays? Answer: No

Vast efforts are being made to identify new drugs, both synthetic and derived from natural sources, to be added to the arsenal of anticancer drugs presently in use in the clinic. None of these well-known drugs, such as cisplatin, doxorubicin, taxol, etc. are truly specific for cancer cells. They are cytostatic or cytotoxic for both cancer and normal cells, because their mechanisms of action involve targets vulnerable in all dividing cells. However, it is common to find reports of plant compounds with a high selectivity index, typically using short-term MTT assays comparing their IC₅₀'s in tumour and “normal” cell lines. We propose that the “selectivity” observed may be explained by general differences in the metabolic and proliferation rates of tumour and non-tumour cells. To investigate this, we compared the results of in vitro cytotoxicity assays using a) doxorubicin and paclitaxel, b) two assays, sulphorrhodamine and MTS, c) both tumour (HT-29, HeLa, MDA-MB231) and normal foreskin-derived fibroblasts and d) different culture medium conditions to limit cell proliferation. Both drugs were apparently “selective” (lower IC₅₀'s) for the tumour cells. However, when cell proliferation was reduced, either by reducing the concentration of bovine serum or nutrients in the culture medium, “selectivity” could also be shown, for example, against faster growing HT-29 cells compared to slower growing HT-29 cells. Selectivity indexes of over 20 could be achieved in the same cell line merely by altering its growth rate. This type of assay does not allow us to distinguish truly specific anticancer drugs that target processes essential for tumor growth or progression, from potentially useful cytotoxic drugs but with the same limitations of toxicity as the present arsenal.

Keywords: Cytotoxicity assay, selectivity, tumour cell, MTS, sulphorrhodamine