On the analysis of count data of birth-and-death process type: with application to molecularly targeted cancer...

Area of Science:

• Biostatistics
• Cancer Therapy
• Immunology

Background:

• Monoclonal antibodies are crucial for targeted cancer therapy, with in vitro studies assessing their potency on tumor cell populations.
• Current analysis of cell count data often uses log-linear models, which are biologically inappropriate for tumor growth dynamics.

Purpose of the Study:

• To propose a novel statistical approach for analyzing in vitro cancer cell count data using a birth-and-death process model.
• To provide a more biologically relevant framework for evaluating monoclonal antibody efficacy in cancer research.

Main Methods:

• Utilizing a birth-and-death process model to describe tumor cell population dynamics.
• Applying a naive regression model for birth and death rates, with estimation via quasi-likelihood methods.
• Approximating variance using quadratic or linear functions of the mean for statistical package compatibility.

Main Results:

• The proposed birth-and-death model demonstrates that a log-linear relationship persists for average counts and covariates.
• The method yields consistent and robust estimations, as confirmed by simulation studies.
• The approach effectively analyzes experimental data for anti-lymphoma monoclonal antibodies.

Conclusions:

• A birth-and-death process model offers a superior alternative to log-linear models for analyzing in vitro cancer cell growth data.
• This novel statistical framework enhances the accurate assessment of monoclonal antibody potency in cancer therapy research.
• The proposed method is validated and applicable to real-world experimental datasets in oncology.