Purpose: The dosimetric merit of a permanent prostate implant relies on two factors: the quality of the plan itself, and the fidelity of its implementation. The former factor depends on source type and on source strength, while the latter is a combination of skill and experience. The purpose of this study is to offer criteria by which to select a source type (I-125 or Pd-103) and activity. Methods and Materials: Given a prescription dose and potential seed positions along needles, treatment plans were designed for a number of seed types and activities, specifically for I-125 with activities ranging from 0.3 to 0.7 mCi, and for Pd-103 with activities in the range of 0.8 to 1.6 mCi. To avoid human planner bias, an automated computerized planning system based on integer programming was used to obtain optimal seed configurations for each seed type and activity. To simulate the effect of seed-placement inaccuracies, random seed-displacement "errors" were generated for all plans. The displacement errors were assumed to be uniformly distributed within a cube with side equal to 2 sigma. The resulting treatment plans were assessed using two volumetric and two dosimetric indices. Results: For I-125 implants a coverage index (CI) of 98.5% or higher can be achieved for all activities (CI is the fraction of the target volume receiving the prescribed or larger dose). The external volume index (EI) (i.e., the amount of healthy tissue, as percentage of the target volume, receiving the prescribed or larger dose) increases from 13.9% to 20% as the activity increases from 0.3 to 0.7 mCi. For implants using Pd-103, the external volume index increases from 10.2% to 13.9% whenever CT exceeds 98.5%. Volumetric and dosimetric indices (coverage index, external volume index, D90, and D80) are all sensitive to seed displacement, although the activity dependence of these indices is more pronounced for I-125 than for Pd-103 implants. Conclusions: For both isotopes, the lower activities studied systematically result in lower EIs. If seeds can be placed within approximately 0.5 cm of their intended position Pd-103 Should be preferred because its EI is lower than that of I-125. For all activities the coverage indices and D90 are within the required range. If seed placement uncertainties are larger than 0.5 cm, I-125 provides slightly better target coverage; however, in terms of external volume (healthy tissue) covered, Pd-103 is superior to I-125. (C) 2000 Elsevier Science Inc.
