SU-E-T-292: Sensitivity of Fractionated Lung IMPT Treatments to Setup Uncertainties and Motion Effects




Evaluate the sensitivity of intensity-modulated proton therapy (IMPT) lung treatments to systematic and random setup uncertainties combined with motion effects.


Treatment plans with single-field homogeneity restricted to ±20% (IMPT-20%) were compared to plans with no restriction (IMPT-full). 4D Monte Carlo simulations were performed for 10 lung patients using the patient CT geometry with either ±5mm systematic or random setup uncertainties applied over a 35 × 2.5Gy(RBE) fractionated treatment course. Intra-fraction, inter-field and inter-fraction motions were investigated. 50 fractionated treatments with systematic or random setup uncertainties applied to each fraction were generated for both IMPT delivery methods and three energy-dependent spot sizes (big spots – BS σ=18-9mm, intermediate spots – IS σ=11-5mm, small spots – SS σ=4-2mm). These results were compared to a Monte Carlo recalculation of the original treatment plan, with results presented as the difference in EUD (ΔEUD), V95 (ΔV95) and target homogeneity (ΔD1–D99) between the 4D simulations and the Monte Carlo calculation on the planning CT.


The standard deviations in the ΔEUD were 1.95±0.47(BS), 1.85±0.66(IS) and 1.31±0.35(SS) times higher in IMPT-full compared to IMPT-20% when ±5mm systematic setup uncertainties were applied. The ΔV95 variations were also 1.53±0.26(BS), 1.60±0.50(IS) and 1.38±0.38(SS) times higher for IMPT-full. For random setup uncertainties, the standard deviations of the ΔEUD from 50 simulated fractionated treatments were 1.94±0.90(BS), 2.13±1.08(IS) and 1.45±0.57(SS) times higher in IMPTfull compared to IMPT-20%. For all spot sizes considered, the ΔD1-D99 coincided within the uncertainty limits for the two IMPT delivery methods, with the mean value always higher for IMPT-full. Statistical analysis showed significant differences between the IMPT-full and IMPT-20% dose distributions for the majority of scenarios studied.


Lung IMPT-full treatments are more sensitive to both systematic and random setup uncertainties compared to IMPT-20%.

This work was supported by the NIH R01 CA111590.