# The Use of Survival Dose-Rate Dependencies as Theoretical Discrimination Criteria for In-Silico Dynamic Radiobiological Models

## TL;DR

This paper introduces a method using survival dose-rate effects to improve the calibration of radiobiological models for cancer therapy.

## Contribution

A novel theoretical framework using dose-rate dependencies to filter unrealistic model dynamics during radiobiological model calibration.

## Key findings

- The proposed discriminators filtered 99% of parameter sets, improving model calibration for Abrams cells.
- SiHa data validated the discriminators and suggested universal aspects of cellular repair.
- The method can be applied to diverse biological data using dynamic in-silico models.

## Abstract

Cell repair dynamics are crucial in optimizing anti-cancer therapies. Various assays (eg, comet assay and γ-H2AX) assess post-radiation repair kinetics, but interpreting such data is challenging and model-based data analyses are required. However, ambiguities in parameter calibration remain an unsolved challenge. To address this, we propose combining survival dose-rate effects with computer simulations to gain knowledge about repair kinetics.

After a literature review, theoretical discriminators based on common fractionation/dose-rate-related effects were defined to discard unrealistic model dynamics. The Multi-Hit Repair (MHR) model was calibrated with canine osteosarcoma Abrams cell line data to study the discriminators’ efficacy in scenarios with limited survival data. Additionally, survival dose-rate-dependent data from the human SiHa cervical cancer cell line were used to illustrate the survival behavior at diverse dose-rates and the capability of the MHR to model these data.

SiHa data confirmed the validity of the proposed discriminators. The discriminators filtered 99% of parameter sets, improving the calibration of Abrams cells data. Furthermore, results from both cell lines may hint universal aspects of cellular repair.

Dose-rate theoretical discrimination criteria are an effective method to understand repair kinetics and improve radiobiological model calibration. Moreover, this methodology may be used to analyze diverse biological data using dynamic models in-silico.