Towards CT-Based Proton Range Uncertainties Specifically for Pediatric Cancer Patients
Abstract
Purpose
Stopping-power ratio (SPR) uncertainty in proton therapy is typically estimated from computed tomography (CT) scans of adult-sized phantoms and applied to adult and pediatric patients. Adult uncertainties may not represent pediatric patients due to differences in size and tissue composition. This study quantified pediatric size-specific proton range uncertainties and compared them with adult-based estimates.
Methods
A modular ring phantom (Sun Nuclear) with diameters of 10, 20, 30, and 40 cm was scanned with both single-energy CT (SECT) and dual-energy CT (DECT). From the DECT scans, virtual monoenergetic images (VMIs) were created. Tissue-equivalent inserts were scanned at the central and peripheral positions to generate size-specific CT-to-SPR conversion curves. Proton range uncertainties were estimated by combining uncertainties from: (i) beam hardening and CT stability, (ii) modelling CT energy spectrum, (iii) variability in tissue composition, and (iv) pediatric tissue uncertainty (evaluating adult-based calibrations on pediatric tissues). Composite uncertainties were determined using tissue-specific weighting for brain (10–20 cm) and pelvic (20–40 cm) patients. Adult uncertainties used the conventional head–body method, while pediatric uncertainties were size-specific in 10 cm intervals. To illustrate clinical impact, a clinical proton plan (originally optimized with 3.5% range uncertainty) of an 8-year-old brain cancer patient was re-optimized using the pediatric-specific uncertainty while maintaining plan objectives.
Results
Adult range uncertainties were 2.8% (brain) and 3.3% (pelvic) with SECT, and 2.0% and 2.4% with VMI. Pediatric-specific range uncertainties were 1.3% (SECT) and 1.1% (VMI) for 10-20 cm brain, 1.8% and 1.6% for 20-30 cm pelvic, and 2.7% and 1.9% for 30-40 cm pelvic. For the patient case, re-optimization using the pediatric-specific range uncertainty of 1.1% reduced the brainstem D0.03cc from 50.5 Gy to 49.5 Gy.
Conclusion
Adult-based range uncertainties overestimated the uncertainty for pediatric patients, especially for smaller patient sizes. Applying pediatric-specific range uncertainties has potential to reduce dose to organs-at-risk.