Is intraoperative real-time dosimetry in prostate seed brachytherapy predictive of biochemical outcome?

To analyze intraoperative (IO) dosimetry using transrectal ultrasound (TRUS), performed before and after prostate low-dose-rate brachytherapy (LDR-BT), and compare it to dosimetry performed 30 days following the LDR-BT implant (Day 30).

A total of 236 patients underwent prostate LDR-BT using ¹²⁵I that was performed with a three-dimensional TRUS-guided interactive inverse preplanning system (preimplant dosimetry). After the implant procedure, the TRUS was repeated in the operating room, and the dosimetry was recalculated (postimplant dosimetry) and compared to dosimetry on Day 30 computed tomography (CT) scans. Area under curve (AUC) statistics was used for models predictive of dosimetric parameters at Day 30.

The median follow-up for patients without BF was 96 months, the 5-year and 8-year biochemical recurrence (BR)-free rate was 96% and 90%, respectively. The postimplant median D₉₀ was 3.8 Gy lower (interquartile range [IQR], 12.4-0.9), and the V₁₀₀ only 1% less (IQR, 2.9-0.2%) than the preimplant dosimetry. When comparing the postimplant and the Day 30 dosimetries, the postimplant median D₉₀ was 9.6 Gy higher (IQR [-] 9.5-30.3 Gy), and the V₁₀₀ was 3.2% greater (0.2-8.9%) than Day 30 postimplant dosimetry. The variables that best predicted the D₉₀ of Day 30 was the postimplant D₉₀ (AUC = 0.62, p = 0.038). None of the analyzed values for IO or Day 30 dosimetry showed any predictive value for BR.

Although improving the IO preimplant and postimplant dosimetry improved dosimetry on Day 30, the BR-free rate was not dependent on any dosimetric parameter. Unpredictable factors such as intraprostatic seed migration and IO factors, prevented the accurate prediction of Day 30 dosimetry.