Optimum timing for image-based dose evaluation of I-125 and (PD)-P-103 prostate seed implants

Purpose/Objective: Image-based dose evaluation of permanent brachytherapy i mplants for prostate cancer is important for optimal patient management aft er implantation. Because of edema caused by the surgical procedure in the i mplantation, if the dose evaluation is based on the images obtained too ear ly after implantation, dose coverage will usually be underestimated. Conver sely, if the images are obtained too late, the dose coverage will be overes timated. This study uses a biomathematical model to simulate edema and its resolution on 29 patients, so that the optimum time to obtain image scans a nd perform dose evaluation can be investigated and estimated. Methods and Materials: Edema of a prostate and its resolution has been show n to follow an exponential function V(t) = V(0)(1 + Delta V[e(-0.693t/Te)-1 ]) where Delta V is the initial relative increase in the prostate volume du e to edema land is related to edema magnitude), and T-e (edema half-life) i s the time for the edema to decrease by half in volume. In this study, edem a was simulated by increasing the volume of preimplant prostate (obtained f rom ultrasound volume study) to a given magnitude of edema. Similarly, the locations of planned seeds were changed to their corresponding locations in the edematous prostate proportionally. The edema was then allowed to resol ve according to the exponential function. The correct dose distribution was calculated by taking into account the dynamic variations of the prostate v olume, seed locations, and source strengths with respect to time. Dose volu me histograms (DVHs) were then generated from this dose distribution. The c onventional postimplant DVHs, which assume the prostate volume and seed loc ations are as in the image scans and constant in time, were also calculated based on the simulated image scans for various days postimplantation. The conventional DVHs of prostate on various days after implantation were compa red to the DVH calculated assuming dynamic conditions. The optimum timing f or conventional postimplant dose evaluation was identified as the time at w hich a minimum difference between the conventional DVH and the dynamic mode l DVH was achieved. The analysis was done on 29 prostate seed implant patie nts for both I-125 and Pd-103. Th, edema magnitude was assumed to be 30%, 4 0%, 50%, 75%, and 100% of original prostate volume, and the half-life of ed ema was assumed to be 4, 7, 10, 15, 20, and 25 days. In this study, the ori ginal volume of prostate varied from 17 cm(3) to 91 cm(3), and number of se eds in the implants varied from 57 to 119. Results: The optimum timing was mainly dependent on the half-lives of edema and radionuclides, and varied slightly with edema magnitude, prostate volu me, and number of seeds. It can be expressed as a function of edema half-li fe in the form of C-0 + C(1)exp(-C2Te). However, if the dose evaluation was performed based on the image scans taken too early or too late, the error became larger, as the edema magnitude was larger. By averaging all 29 patie nts and various edemas, it was found that for 125I Seed implants, if the po stimplant dose evaluation is performed based on image scans taken between 5 and 9 weeks, the average error will be less than 5%, with a maximum possib le error less than 10% in 80% coverage dose; for Pd-103 Seed implants, if t he postimplant dose evaluation is performed based on image scans taken betw een 2 and 4 weeks, the average error will be less than 5%, with a maximum e rror less than 15% in 80% coverage dose. Because of edema, a conventional p reimplant plan also overestimates dose coverage of prostate. On the average , a standard preimplant planning overestimates dose coverage by about 6% fo r I-125 implants and 14% for Pd-103 implants in our study. Conclusion: Based on the dynamic model, the optimum timing of image scans f or postimplant dose evaluation of prostate seed implantation is 7 weeks pos timplantation for I-125 implants and about 3 weeks for Pd-103 implants. The time-window for reasonable accuracy (+/- 5%) is +/- 2 weeks for I-125 and +/- 1 week for Pd-103 around the optimum timing. During preimplant procedur e, the minimum prescribed coverage dose should be increased by an amount of about 6% for I-125 implants and about 14% for Pd-103 implants to compensat e for the effect of edema. (C) 1999 Elsevier Science Inc.