An accurate method for estimating an approximate lethal dose with few animals, tested with a Monte Carlo procedure

Area of Science:

• Toxicology
• Computational toxicology
• Statistical modeling

Background:

• Accurate chemical toxicity classification requires reliable, yet not overly precise, toxicity indicators.
• Current LD50 estimations often use excessive animal testing and precision.
• Existing approximate lethal dose (ALD) methods lack sufficient validation regarding accuracy and precision.

Purpose of the Study:

• To develop and validate a novel method for estimating ALD, addressing the systematic error present in the Deichmann and LeBlanc (1943) method.
• To compare the performance of the new ALD estimation method against Lorke's (1983) method using simulation.

Main Methods:

• A new ALD estimation method was developed, correcting for systematic error related to log tolerance variance.
• Monte Carlo simulations were employed, modeling log-normal distribution of individual tolerance.
• The new method and Lorke's method were tested using simulated toxicological data with varying tolerance variances.

Main Results:

• The new method demonstrated no systematic error in simulations.
• For log tolerance standard deviations below 0.85, the new method was more precise than Lorke's.
• The new method required fewer animals on average (under 10) compared to Lorke's method (13 animals).

Conclusions:

• The newly developed ALD estimation method is a viable alternative to existing approaches, offering improved precision and reduced animal usage.
• This method is particularly advantageous for substances with lower log tolerance variance.
• Further validation is recommended to confirm its efficacy in real-world toxicological assessments.