Interspecies extrapolation: a reexamination of acute toxicity data

Area of Science:

• Pharmacology and Toxicology
• Quantitative Biology
• Biostatistics

Background:

• Accurate prediction of toxic doses across species is crucial for drug development and risk assessment.
• Previous allometric scaling analyses have limitations in accounting for measurement error and incorporating all available data.
• Cancer chemotherapeutic agents provide a valuable dataset for investigating interspecies toxicity scaling.

Purpose of the Study:

• To reanalyze acute toxicity data for cancer chemotherapeutic agents to derive robust coefficients for allometric dose scaling.
• To extend previous analyses by incorporating hamster data and explicitly addressing uncertainties, including measurement errors.
• To test hypotheses regarding the variability of allometric scaling power (b) across chemicals versus a universal scaling law.

Main Methods:

• Reanalysis of acute toxicity data from Freireich et al. and Schein et al.
• Application of Monte Carlo sampling to account for measurement errors in confidence interval derivation and hypothesis testing.
• Statistical modeling to test two hypotheses: variable allometric power (b) vs. a single scaling law.

Main Results:

• Under the hypothesis of variable scaling power, 95% of cases showed a body weight exponent (b) between 0.42-0.97, with a mean of 0.74.
• Assuming a single scaling law, the maximum likelihood estimate for the scaling power was 0.74.
• Both surface area scaling (b=0.67) and a 0.75 power scaling law are statistically consistent with the data when accounting for variance heterogeneity.

Conclusions:

• Allometric scaling of toxic doses across species is supported by the analyzed data.
• Both surface area-based scaling (b≈0.67) and a 0.75 power law provide a good fit, indicating flexibility in scaling approaches.
• The findings highlight the importance of considering measurement error and data completeness in interspecies toxicity scaling for cancer chemotherapeutics.