A detailed radiobiological and dosimetric analysis of biochemical outcomes in a case-control study of permanent prostate brachytherapy patients.

The purpose of this study is to determine dosimetric and radiobiological predictors of biochemical control after recalculation of prostate implant dosimetry using updated AAPM Task Group 43 (TG-43) parameters and the radiobiological parameters recommended by TG-137. All biochemical failures among patients implanted with 125I Or 103Pd sources between 1994 and March 2006 were matched 2:1 with nonfailure controls. The individual matching was by risk group, radionuclide, prescribed dose, and time of implant (one match before and one after the failed patient) resulting in a median follow-up of 10.9 years. Complete dose volume histogram (DVH) data were recalculated for all 55 cases and 110 controls after updating the original source strength by the retrospectively determined ratios of TG-43. Differential DVH data were acquired in 179 increments of prostate volume versus percentage prescribed dose. At each incremental dose level i, the biologically equivalent dose BEDi, equivalent uniform dose EUDi, and tumor control probability TCPi were calculated from the implant dose plus any external beam delivered to the patient. Total BED, EUD, and TCP were then derived from the incremental values for comparison with single point dosimetric quality parameters and DVH-based averages. There was no significant difference between failures and controls in terms of total BED (143 vs 142 Gy), EUD (95 vs 94 Gy), or TCP (0.87 vs 0.89). Conditional logistic regression analysis factored out the matching variables and stratified the cohort into each case and its controls, but no radiobiological parameter was predictive of biochemical failure. However, there was a significant difference between radiobiological parameters of 125I and 103Pd due to less complete coverage of the target volume by the former isotope. The implant BED and TCP were highly correlated with the D90 and natural prescription doses and a series of mean DVH-based doses such as the harmonic mean and expressions of the generalized EUD. In this case-control study of prostate brachytherapy biochemical failures and nonfailures, there were no radiobiological parameters derived from detailed DVH-based analysis that predicted for biochemical control. This may indicate that in our approach, implant dosimetry is at or near the limits of clinically effective dose escalation.