On the validity of the superposition principle in dose calculations for intracavitary implants with shielded vaginal colpostats

Intracavitary vaginal applicators typically incorporate internal shielding to reduce dose to the bladder and rectum. While dose distributions about a single colpostat have been extensively measured and calculated, these studies neglect dosimetric perturbations arising from the contralateral colpostat or the intrauterine tandem. Dosimetric effects of inhomogeneities in brachytherapy is essential for both dose-based implant optimization as well as for a comparison with alternate modalities, such as intensity modulated radiation therapy. We have used Monte Carlo calculations to model dose distributions about both a Fletcher-Suit-Delclos (FSD) low dose-rate system and the microSelectron high dose-rate remote afterloading system. We have evaluated errors, relative to a Monte Carlo simulation based upon a complete applicator system, in superposition calculations based upon both precalculated single shielded applicator dose distributions as well as single unshielded source dose distributions. Errors were largely dominated by the primary photon attenuation, and were largest behind the shields and tandem. For the FSD applicators, applicator superposition showed differences ranging from a mean of 2.6% at high doses (>Manchester Point A dose) to 4.3% at low doses (<Manchester Point A dose) compared to the full geometry simulation. Source-only superposition yielded errors higher than 10% throughout the dose range. For the HDR applicator system, applicator superposition-induced errors ranging from 3.6%-6.3% at high and low doses, respectively. Source superposition caused errors of 5%-11%. These results indicate that precalculated applicator-based dose distributions can provide an excellent approximation of a full geometry Monte Carlo dose calculation for gynecological implants.