[Is the prediction of LD50 using cell cultures generally valid?]

Area of Science:

• Pharmacology and Toxicology
• Computational Chemistry

Context:

• In vitro assays like IC50 are crucial for drug discovery and toxicity assessment.
• Estimating in vivo LD50 values in animal models is essential but ethically challenging.
• Linear regression models offer a quantitative approach to correlate different biological measurements.

Purpose:

• To investigate the linear correlation between in vitro IC50 and in vivo LD50 values across diverse substance classes.
• To assess the predictability of LD50 from IC50 using literature and laboratory data.
• To explore the potential of reducing animal toxicity testing through in vitro data analysis.

Summary:

• A linear regression model was employed to compare literature and laboratory-derived IC50 and LD50 values (p.o. in mice and rats).
• A similar relationship between IC50 and LD50 was observed across different substance classes, cell types, and toxicity testing methods.
• The study found that LD50 values deviated minimally from the regression line, indicating a general validity of the IC50-LD50 correlation for predicting LD50 within a specific dosage range.

Impact:

• Establishes a general validity for the positive linear correlation between IC50 and LD50.
• Demonstrates the potential for predicting LD50 values from in vitro IC50 data, reducing reliance on animal testing.
• Opens new avenues for ethical and efficient toxicity assessment in drug development.