The "Goldilocks" Pulse: Mechanistic Constraints on Dose-per-Pulse and Repetition Rate for Flash Sparing
Poster
AAPM 2026 Poster #2-final.pdf Hosted by AAPM · 1.9 MBAbstract
Purpose
To mechanistically define the beam parameter space required for FLASH radioprotection, moving beyond Mean Dose Rate (MDR) to identify critical constraints in pulse structure (dose-per-pulse and frequency).
Methods
We utilized a Mechanistic Lethal-Potentially Lethal (M-LPL) model extended with oxygen-dependent fixation kinetics to simulate pulsed delivery. We analyzed 10 Gy irradiation across three microenvironments: Perinecrotic Tumor (0.2 mmHg), Normal Stem-Cell Niche (3.0 mmHg), and Perivascular Tissue (30 mmHg). A "Split-Step" solver performed sensitivity analysis, sweeping pulse counts from n = 1 to 320 across frequencies of 5–320 Hz to quantify differential sparing (Λ = ln [SFFLASH/SFCONV]).
Results
Simulations identified three response regimes. (1) Invariance of Extremes: Perinecrotic (0.2 mmHg) and Perivascular (30 mmHg) targets exhibited negligible sparing (Λ ≈ 0) regardless of beam structure, constrained physically by the OER "floor" and "plateau," respectively. (2) Conditional MDR Validity: For Normal Niches (3.0 mmHg), MDR predicts sparing only when repetition frequency exceeds tissue reoxygenation rates (>> 80 Hz). (3) Dose-per-Pulse Dominance: At low frequencies (< 20 Hz), MDR fails to predict survival. Instead, maximal sparing is achieved with a small number of pulses (low n) due to the extremely high dose-per-pulse, which drives deep transient hypoxia. Increasing n at low frequencies allows inter-pulse reoxygenation, rapidly abolishing the effect.
Conclusion
The M-LPL model demonstrates that MDR is an insufficient metric for FLASH efficacy. Clinical translation must prioritize high dose-per-pulse delivery, particularly for low-frequency sources (e.g., standard linacs), to ensure protection of the stem-cell niche.