Purpose: Intensity modulated proton therapy (IMPT) utilizes complex patterns of spot placement to generate highly conformal dose distributions and rapid fall-offs. Therefore, a high confidence level in the positional accuracy of each spot is demanded. However, the dosimetric impact of uncertainties associated with delivered spot positions, particularly in regions that involve several complex anatomical structures such as head-and-neck has not been investigated.
Methods: Fifteen patients with head and neck cancers were randomly selected for this study. Patients were treated with ≥95% of the largest CTV receiving the prescription dose (Rx: 59.4 Gy) according to a single integrated boost regimen. For each patient plan, random uncertainties in spot position were simulated by perturbing the spot position according to uniform (or Gaussian) distribution with interval width equal to the vendor-set tolerances (1.2 mm/1.45 mm in X/Y direction at monitoring chamber for Varian ProBeam). Impact of spot position uncertainties on target coverage was studied using the minimum, maximum (dose to 0.01cm3), and dose to 95% of CTV volume (D95%).
Results: Distributions for target coverage metrics were normal with standard deviations ranging between 0.5% to 5% of Rx dose (180cGy/fx). D95% exhibited the lowest sensitivity, and was within 2% of the clinical plan for all patients. Larger deviations in Dmin (<10%) and Dmax (<6%) were observed suggesting potential for target under-dosage, or unwanted hot spots. No significant correlation between coverage variation and target volume was found.
Conclusion: While in comparison to MLC based intensity modulated radiation therapy (IMRT), IMPT is more sensitive to positional uncertainties, spot positional uncertainties within the vendor supplied 1.2 mm tolerance do not pose significant clinical impact.