Comparing time activity curves using the Akaike information criterion.

Area of Science:

• Biostatistics
• Pharmacokinetics
• Radiopharmaceutical Science

Background:

• Curve comparison is crucial in science but traditional methods like sum of squares or R(2) are insufficient.
• Existing statistical tests for curve comparison have notable disadvantages.
• Accurate curve comparison is vital for understanding complex biological processes and drug behavior.

Purpose of the Study:

• To introduce a novel method for comparing scientific curves based on model selection.
• To apply this method to analyze time activity data of (111)In- and (90)Y-labelled anti-CD66 antibodies.
• To demonstrate the superiority of this approach over less advanced methods.

Main Methods:

• The core method involves transforming curve comparison into a model selection problem.
• The corrected Akaike Information Criterion (AICc) is utilized for model selection.
• The method is exemplified using serum time activity data from (111)In- and (90)Y-labelled anti-CD66 antibodies.

Main Results:

• The study demonstrates that the biokinetics of (111)In- and (90)Y-labelled anti-CD66 antibodies differ significantly.
• A distinct difference is observed in the contribution of the longer half-life component between dosimetry and therapy.
• The proposed AICc-based model selection method provides a more nuanced comparison of the antibody data.

Conclusions:

• The corrected Akaike Information Criterion offers a robust framework for comparing scientific curves.
• Significant differences in biokinetics exist for (111)In- and (90)Y-labelled anti-CD66 antibodies.
• This advanced method is recommended over traditional, less sensitive techniques for curve comparison in scientific research.