Does inverse planning applied to Iridium192 high dose rate prostate brachytherapy improve the optimization of the dose afforded by the Paris system?

Background and purpose: The purpose of the work is to analyse for Ir-192 prostate brachytherapy (BT) some of the different steps in optimizing the dose delivered to the CTV, urethra and rectum. Materials and methods: Between 07/1998 and 12/2001, 166 patients were treated with Ir-192 wires providing a low dose rate, according to the Paris system philosophy and with the 2D version of the treatment planning Isis(R). 40-45 Gy were delivered after an external beam radiotherapy of 40 Gy. The maximum tolerable doses for BT were 25 Gy to the anterior third of the rectum on the whole length of the implant (R dose) and 52 Gy to the urethra on a 1 cm length (U-max). A U-max/CTV dose ratio > 1.3 represented a pejorative value as the planned dose of 40-45 Gy could not be achieved. On the other side a ratio < 1.25 was considered optimal and the intermediate values satisfactory. A R/CTV dose ratio < 0.55 which is easily obtained was also stated as an optimal situation. From the CT Scan images performed for these implants, a theoretical study investigated the possibilities of complementary optimization afforded by a 3D treatment planning. This work was based on an inverse planning philosophy and a stepping source technology (SST) for high dose rate Ir-192 sources. Results: At the end of a learning curve reaching a plateau after the first 71 patients, 90% of the implants with Ir-192 wires were stated at least satisfactory for a total rate of 82% for the whole population. When the 3D dosimetry for SST was used, the initial values > 1.25 decreased significantly with optimization required on CTV contours and additional constraints on urethra while the R/CTV ratio was maintained under 0.55. For initial U-max/CTV > 1.3 or > 1.25 but <= 1.3 indeed, the mean respective values of 1.41 +/- 0.16 and 1.28 +/- 0.01 decreased to 1.28 +/- 0.24 and 1.17 +/- 0.09 (P < 0.001), allowing to increase the total dose to the CTV by 4 Gy. Conclusions: The Paris system which assumes a homogeneous distribution of a minimum number of catheters inside the CTV allowed to anticipate a satisfactory dosimetry in 82% of cases. However, this precision rate could be improved until 95% with an optimization approach based on an inverse planning philosophy. These new 3D optimization methods, ideally based on good quality implants at first allow to deliver the highest doses with minimal probability of creating cold spots inside the CTV or unacceptable hot spots outside. (c) 2004 Elsevier Ireland Ltd. All rights reserved.