Patient-specific dose correction for prostate postimplant evaluation with flexible timing of postimplant imaging

Purpose The dosimetric effect of edema on prostate implants have long been realized, but large uncertainties still exist in the estimation of dose actually received by the prostate. This study attempted to develop a new method to accurately estimate dose delivered to the prostate accounting for the variation of prostate volume and seed distribution, edema half-lives, and times for postimplant evaluation. Methods and materials A series of prostate seed implants for Cs-131, Pd-103, and I-125 with various prostate volumes were simulated in a water phantom using the TG-43 algorithm on the Varian Eclipse treatment planning system. Dose analysis was performed to derive a quantitative relationship between the prostate peripheral dose and the prostate radius with the variation of prostate volume and seed distribution. Using this relationship to calculate dynamically, the total dose accumulated in the prostate (D-T) accounting for the variation of prostate volume and seed distribution and edema half-lives. Moreover, the total dose can be estimated statically based on the prostate volume that can be determined in a computerized tomography (CT) image taken at a time after implantation. The statically estimated total dose (D-CT) was compared with D-T to determine optimal imaging times as well as dose correction factors for other imaging times. Results An inverse power law was established between the prostate peripheral dose and prostate radius. The value of the power was 1.3 for Cs-131 and I-125, and 1.5 for Pd-103, respectively. D-T was derived dynamically using the inverse power law. Given the edema half-lives, T-E, of 4, 9.3, and 25 days and the volume expansion of 1.1 and 2.0 times of the prostate without edema, the optimal times for postimplant imaging were: 7, 9, and 16 days for T-E = 4 days; 10, 14, and 28 days for T-E = 9.3 days; and 12, 19, 45 days T-E = 25 days, for Cs-131, Pd-103, and I-125, respectively. D-CT calculated using the prostate volume determined on the optimal days agreed with D-T to 0.0%-1.8% and within 0.3% for most cases. For various prostate volumes, edema half-lives, and nonoptimal times, D-CT was able to achieve a 1% accuracy. Conclusion The postimplant dose calculation based on the proposed inverse power law for prostate seed implants with edema has improved the accuracy of postimplant dosimetry with accurate and patient-specific dose corrections accounting for prostate size, edema half-life, and postimplant imaging times. Optimal times for postimplant imaging have been accurately determined, and the high accuracy of postimplant dose calculation can be achieved for both optimal imaging times and nonoptimal imaging times.