Program Information

Towards a More Realistic Model of Tumor Control Probability (TCP) for High-Dose-Rate Prostate Monotherapy

C Tien

CJ Tien*, Z Chen, R Nath, DJ Carlson, Yale Univ. School of Medicine, New Haven, CT

Presentations

SU-I-GPD-T-3 (Sunday, July 30, 2017) 3:00 PM - 6:00 PM Room: Exhibit Hall

Purpose: High-dose-rate brachytherapy as a monotherapy for prostate cancer has increased in recent years. We created a model of tumor control that considers intrafraction DNA repair adjusted for decaying source strength and a simplified representation of tumor hypoxia.

Methods: Theoretical tumor control probability (TCP) curves were initially generated using TG-137 consensus parameter values: N₀=4.1x10⁶, α=0.15 Gy⁻¹, α/β=3.1 Gy, and a double-strand-break repair half-time of 17 minutes. The Lea-Catcheside dose protraction factor was calculated with dose-rate of 20 Gy per 15 minutes (10 Ci). Source activity and reductions in the dose protraction factor were then recalculated for 0-90 days of decay. A single-compartment binary hypoxia model was implemented with clinically-effective hypoxia reduction factors (HRF) ranging from 1.0-1.2.

Results: A simple TCP model that neglects the effects of DNA repair (G=1) and tumor hypoxia predict TCP>99% for a single HDR fraction 19 Gy. However, clinical outcomes data at 6 years Kaplan-Meier yielded only 66% (low-risk) and 63% (intermediate-risk) biochemical control. A TCP model which includes dose protraction scaled for source activity predicts 62%-69% control with 18.8-19 Gy delivered at 6.75 Ci. For single-fractions, prescriptions of 17.5 Gy (G=1) or 19.1-21.5 Gy (G<1) were required for TCP>90%. For six-fraction, 6.4 Gy/Fx (G=1) or 6.55-6.8 Gy/Fx (G<1) was required for TCP>90%. Clinical data fit by 6.75 Ci TCP curve was used as baseline hypoxia response with HRF=1.1. For single-fraction, well-oxygenated targets (HRF=1) required 17.9 Gy while more hypoxic targets (HRF=1.2) required 22.0 Gy for TCP>90%. For six-fractions, prescriptions varied between 6.0-7.25 Gy/Fx to achieve TCP>90%.

Conclusion: TCP modeling which includes intrafraction DNA-repair is more consistent with reported clinical outcomes than simplified models with no dose protraction considerations. The effects of hypoxia were investigated for a range of HRF values. Due to their impact, TCP models should aim to incorporate both hypoxia and intrafraction repair.


Contact Email: