The effect of seed anisotropy on brachytherapy dose distributions using I-125 and Pd-103

We have evaluated the effect of the anisotropy of individual seeds on dose distributions for permanent prostate implants using I-125 and Pd-103. The d ose distributions were calculated for various implants using both the line source and point source calculational formalisms, for two different models of I-125 and Pd-103 seeds. The dose distributions were compared using cumul ative dose volume histograms (DVH) and cumulative difference dose volume hi stograms (Delta DVH) for the prostate target volume and for the rectum surf ace. The DVHs could not distinguish between the dose distributions from iso tropic and non-isotropic seeds. However, the Delta DVHs were useful in dete rmining the fraction of the target volume for which the difference between the dose distribution for line sources and for point sources exceeded a thr eshold value. The dose distributions were calculated (1) for all the seeds oriented co-linearly, along either the x-, y-, or z-axis, and (2) for the s eeds at randomized orientations, more closely resembling the clinical situa tion. For all cases, there was a significant difference in the effect of se ed anisotropy from the different seed types. For the geometrically simpler test cases with a small number of seeds, the effect of anisotropy on the do se distribution was too large to ignore for any of the seed types investiga ted. For the idealized pre-plan case, the effect was much smaller. For clin ical prostate implants, the calculations done with seeds oriented co-linear ly along the z-axis (needle implant axis) were a reasonable approximation f or those from simulations of seeds with randomized orientations. Again, the effect of anisotropy varied drastically between different seed models, and also between different clinical cases. How ever, the effect of anisotropy must be considered in the context of all the other uncertainties in clinica l brachytherapy treatments. (C) 2001 American Association of Physicists in Medicine. [DOI: 10.1118/1.1350674].