Correlation of acute lethal potency with in vitro cytotoxicity

The in vitro cytotoxicity (measured as ID(50) value) of 27 compounds believed to act by interference with cell basal functions/structures has been determined and compared with their acute lethal toxicity (measured as oral and ip LD(50) values) using a previously described approach to the choice of the most appropriate LD(50) value from the Registry of Toxic Effects of Chemical Substances database. A weak, but significant, positive correlation of log oral LD(50) value with log ID(50) value was obtained (r = 0.49; 0.02 < P < 0.05) for 21 compounds for which oral and ip LD(50) values were obtainable. A better correlation between LD(50) and ID(50) was obtained when the log ip LD(50) value was used (r = 0.68; P < 0.001). Three compounds, for which metabolism is a major determinant of toxicity in vivo, yielded anomalous results in the ip LD(50)/ID(50) comparison. A further improvement in the correlation of log ip LD(50) with log ID(50) was obtained when data from these compounds were excluded (r = 0.82; P < 0.001). Close correlations of log ip LD(50)/log ID(50) were obtained with groups of six anti-metabolites and six alkylating agents (r = 0.94 and 0.96, respectively; P < 0.001 in each case). The locations of the regression lines for these two groups were significantly different (P < 0.01). It is concluded that the in vitro cytotoxicity of compounds that exert their toxicity by interference with cell basal functions/structures is correlated with their intrinsic lethal potency, but that information on absorption, metabolism and disposition are required before the cytotoxicity data can be used to assess apparent potency. The data provide evidence that the precise relation of LD(50)/ID(50) values is determined by the mode of toxicity.