PurposeThe purpose of this study is to suggest a probable problem in chemosensitivity tests performed in practice and to speculate on practicable measures for more accurate chemosensitivity evaluation.MethodsThree colorectal cancer cells (RSC, RRC1, and RRC2) were treated with 5-fluorouracil (5-FU). Inhibition percentage (%inhibition) of cancer cells and relative quantitation of thymidylate synthase (TS) mRNA were measured on day 2, day 5 after replacement of 70% media on day 2, day 7, and day 3 after replacement of all media on day 7. Doses that produced 50% inhibition (Dm) were calculated to evaluate drug effect. Relative quantitation of TS mRNA and correlations between TS mRNA levels and 5-FU concentrations were analyzed.ResultsRRC1 was more resistant than RRC2 on day 7, but Dm value of RRC2 increased three days after replacement of media from 12.3 to 18.1. Mean TS mRNA levels of RSC on D2 and D7 were significantly lower than those of RRC1 and RRC2, respectively (P = 0.004, P = 0.004 on D2; P = 0.010, P = 0.006 on D7). TS mRNA levels in RRC1 were significantly reversely correlated with 5-FU concentrations on day 2 (correlation coefficient = -0.867, P = 0.015). On the other hand, correlations were not significant in RRC2 (r = 0.067).ConclusionEvaluating %inhibition of cancer cells at one point in chemosensitivity tests seems to be inadequate in determining chemotherapeutic regimens. Multilateral approaches, such as trials evaluating cancer cell survival before and after media replacement and correlations between TS mRNA levels and 5-FU concentrations, needs to be implemented for the practical application of chemosensitivity tests.