The impact of edema on planning 125I and 103Pd prostate implants.

Abstract

Permanent transperineal interstitial 125I and 103Pd prostate implants are generally planned to deliver a specific dose to a clinically defined target volume; however, the post-implant evaluation usually reveals that the implant delivered a lower or higher dose than planned. This difference is generally attributed to such factors as source placement errors, overestimation of the prostate volume on CT, and post-implant edema. In the present work we investigate the impact of edema alone. In routine prostate implant planning, it is customary to assume that both the prostate and seeds are static throughout the entire treatment time, and post-implant edema is not taken into consideration in the dosimetry calculation. However, prostate becomes edematous after seed implantation, typically by 50% in volume [Int. J. Radiat. Oncol., Biol., Phys. 41, 1069-1077 (1998)]. The edema resolves itself exponentially with a typical half-life of 10 days. In this work, the impact of the edema-induced dynamic change in prostate volume and seed location on the dose coverage of the prostate is investigated. The total dose delivered to the prostate was calculated by use of a dynamic model, which takes edema into account. In the model, the edema resolves exponentially with time, as reported in a separate study based on serial CT scans [Int. J. Radiat. Oncol., Biol., Phys. 41, 1069-1077 (1998)]. The model assumes that the seeds were implanted exactly as planned, thus eliminating the effect of source placement errors. Implants based on the same transrectal ultrasound (TRUS) images were planned using both 125I and 103Pd sources separately. The preimplant volume and planned seed locations were expanded to different degrees of edema to simulate the postimplant edematous prostate on day 0. The model calculated the dose in increments of 24 h, appropriately adjusting the prostate volume, seed locations, and source strength prior to each time interval and compiled dose-volume histograms (DVH) of the total dose delivered. A total of 30 such DVHs were generated for each implant using different combinations of edema half-life and magnitude. In addition, a DVH of the plan was compiled in the conventional manner, assuming that the prostate volume and seeds were static during treatment. A comparison of the DVH of the static model to the 30 edema corrected DVHs revealed that the plan overestimated the total dose by an amount that increased with the magnitude of the edema and the edema half-life. The maximum overestimation was 15% for 125I and 32% for 103Pd. For more typical edema parameters (a 50% increase in volume and a 10 day half-life) the static plan for 125I overestimated the total dose by about 5%, whereas that for 103Pd overestimated it by about 12%.