Addition of Repopulation on Tumor Control Probability (TCP) Model in Prostate Cancer
Salma Aqilah Zahroh (a*), Rena Widita (a)

a) Department of Physics, Faculty of Mathematics and Natural Sciences, Institut Teknologi Bandung, Jalan Ganesha 10, Bandung, 40132, Indonesia
salmaqilahz[at]gmail.com

Abstract

When designing fractionation schedules for radiotherapy, one critical factor to account for is overall treatment time. Accelerated repopulation of tumor cells can significantly impact biological effectiveness of treatment. To address this, incorporating a repopulation factor into tumor control probability (TCP) model becomes essential. In this study, we analyzed data from 9,690 patients treated with external beam radiotherapy. Patients were categorized into three risk groups: 28% low risk, 53% intermediate risk, and 19% high risk. TCP model was fitted to clinical outcomes, specifically 5-year biochemical relapse-free survival (5y-bRFS). We employed maximum likelihood estimation (MLE) with Nelder-Mead simplex algorithm to maximize likelihood function produced by TCP model. Our results show that dose required to counteract daily repopulation is 0.54 Gy/day, with a kick-off time of 23 days. Our analysis reveals that there was no significant difference in kick-off time and repopulation rate across different risk groups. Incorporating repopulation factor into the TCP model yielded a good fit to data, as indicated by Akaike Information Criterion (AIC). Strategies to mitigate tumor cell repopulation during radiotherapy include employing accelerated fractionation, which reduces overall treatment time and minimizes opportunities for accelerated tumor repopulation.

Keywords: Maximum likelihood estimation, Nelder-Mead simplex algorithm, Prostate cancer, Repopulation, Tumor control probability model

Topic: Biophysics and Medical Nuclear Physics