Summary Low pH and hypoxia are a common feature of many solid tumours. This study examined the effect of these two conditions on the cytotoxic properties of the bifunctional agent RB 6145, the prodrug of RSU 1069. The effect of acidic pH on RB 6145 toxicity was examined in six human tumour (Coleman, 1988), much effort has been directed at finding ways to specifically target such cells in tumours. The use of hypoxic cell radiosensitisers such as the 2-nitroimidazole, misonidazole, has provided some therapeutic gain, though neurotoxicity limited its dose and thus its effectiveness. (Dische, 1985;Overgaard et al., 1989). RSU 1069 [x-(1-aziridinomethyl)-2-nitro-lH-imidazole-1-ethanol] was one of the newer generation nitroimidazoles developed by Adams et al. (1984a,b), and one which exhibited two functions: radiosensitisation due to the electron affinic properties of the molecule and alkylation by the aziridine moiety. It was very efficient both as a radiosensitiser and as a chemosensitiser, but also produced severe gastrointestinal toxicity (Horwich et al., 1986). RB 6145 [a-([(2-bromoethyl)-aminoJmethyl)-2-nitro-JH-imidazole-1-ethanol hydrobromide] was developed as an analogue (see Figure 1) and prodrug of RSU 1069 (Jenkins et al., 1990). It retained most of the radiosensitisation and bioreductive cytotoxicity of RSU 1069 but had much lower toxicity (Cole et al., 1990(Cole et al., , 1991(Cole et al., , 1992Bremner, 1993), and is currently awaiting clinical evaluation.The pharmacokinetics of RB 6145 and its metabolites have been carefully examined in mice (Binger and Workman, 1991). The major metabolites were shown to be RSU 1069 (the pharmacologically active aziridine ring metabolite) and an oxazolidinone metabolite, with much lower levels of at least two other analogues, RSU 1137 and RSU 1111 (see Figure 1)

Human tumour cell lines can exhibit qualitatively similar age-responses despite having markedly different G1 checkpoint responses. This suggests that modulation of the G1 arrest response may not prove to be a useful clinical strategy because it may not lead to significant cell age specific changes in radiosensitivity. The mathematical model of the radiation response of mitotically selected synchronized cells was a useful way to quantitatively describe cell age heterogeneity in these populations, and demonstrated the important impact of this heterogeneity on measured age-responses.