Abstract
In high dose rate interstitial brachytherapy for prostate cancer, we compared a Geometrical optimization method and a Dose point optimization method in terms of coverage index (CI) and dose nonuniformity ratio (DNR) computed from dose volume histograms (DVHs) of respective dose distributions, under several different conditions of dose prescription points. In the Geometrical optimization method, we set the prescribed isodose curves as those crossing the prescription points placed at 3-, 5-and 10-mm distant from the applicators, respectively. When the distance was as close as 3mm, the CI was less than 90%. With 5 and 10mm, the CIs were large enough (97% and 100%); while in the case of 10 mm, the DNR was as high as 0.84, that is, yielding a too high dose volume. In the Dose point optimization method, when the prescription points were placed on the surface of CTV, the CI was as low as 72%, indicating that the Dose point optimization method failed to prescribe the given dose on CTV surface. When the dose points were placed at the distances of 5-and 10-mm outside of CTV, the CIs were 91%, and 100%; while the DNRs were as high as 0.64 and 0.85, respectively, indicating dose volumes their were too high. Whichever “optimization” program is used, it does not always ensure a good dose distribution, and so radiation oncologists should examine the dosedistribution carefully without overestimating the“computer-optimized” one.
