Purpose: The vagina has not been widely recognized as organ at risk in brachytherapy for cervical cancer. No widely accepted dose parameters are available. This study analyzes the uncertainties in dose reporting for the vaginal wall using tandem-ring applicators.
Methods and materials: Organ wall contours were delineated on axial magnetic resonance (MR) slices to perform dose-volume histogram (DVH) analysis. Different DVH parameters were used in a feasibility study based on 40 magnetic resonance imaging (MRI)-based treatment plans of different cervical cancer patients. Dose to the most irradiated, 0.1 cm(3), 1 cm(3), 2 cm(3), and at defined points on the ring surface and at 5-mm tissue depth were reported. Treatment-planning systems allow different methods of dose point definition. Film dosimetry was used to verify the maximum dose at the surface of the ring applicator in an experimental setup.
Results: Dose reporting for the vagina is extremely sensitive to geometrical uncertainties with variations of 25% for 1 mm shifts. Accurate delineation of the vaginal wall is limited by the finite pixel size of MRI and available treatment-planning systems. No significant correlation was found between dose-point and dose-volume parameters. The DVH parameters were often related to noncontiguous volumes and were not able to detect very different situations of spatial dose distributions inside the vaginal wall. Deviations between measured and calculated doses were up to 21%.
Conclusions: Reporting either point dose values or DVH parameters for the vaginal wall is based on high inaccuracies because of contouring and geometric positioning. Therefore, the use of prospective dose constraints for individual treatment plans is not to be recommended at present. However, for large patient groups treated within one protocol correlation with vaginal morbidity can be evaluated.