ViVo: A Temporal Modeling Framework That Boosts Statistical Power and Minimizes Animal Usage

Area of Science:

• Oncology
• Pharmacology
• Biostatistics

Background:

• Preclinical tumor studies face challenges with high variability and small sample sizes, limiting statistical power and obscuring treatment effects.
• Existing methods often struggle to detect subtle therapeutic impacts due to these limitations.

Purpose of the Study:

• To develop a robust framework for analyzing preclinical tumor growth kinetics and treatment efficacy.
• To improve statistical power and reveal treatment effects masked by variability in small-sample studies.

Main Methods:

• An exponential framework was developed to estimate tumor growth rates (r) independent of initial tumor burden.
• Tumor Growth Rate (TGR) matrices were introduced to map treatment effects across defined time windows.
• The framework was validated using public xenograft data and four additional mouse models.

Main Results:

• Exponential fits demonstrated high agreement with raw tumor growth data (median R² = 0.937).
• The proposed approach consistently outperformed conventional endpoint and daily nonparametric analyses in detecting treatment effects, especially with small sample sizes (n=3-7).
• The framework successfully predicted tumor weights for early-euthanized animals, enhancing statistical consistency and supporting the 3Rs principles.

Conclusions:

• The novel exponential framework provides a powerful and statistically robust method for analyzing preclinical tumor growth and therapeutic efficacy.
• The open-source ViVo web platform facilitates accessible and standardized analysis of in vivo tumor kinetics, promoting wider adoption in cancer research.