Adjuvant and salvage radiotherapy after radical prostatectomy for prostate cancer

Purpose: The optimal role of radiotherapy (RT) to the prostate bed after radical prostatectomy (RP) is the subject of much debate. In this study, the results of adjuvant RT (ART) and salvage RT (SRT) were compared. Methods and Materials: A total of 146 lymph node-negative patients were treated postoperatively after RP with RT to the prostate bed between 1987 and 1998. Of these, 75 patients had an undetectable prostate-specific antigen (PSA) level and were treated with ART for adverse pathologic features only to a median dose of 60 Gy (range 51-70). A positive margin was identified in 96%, and two of the three with negative margins had seminal vesicle involvement (SVI). SRT was administered for either a persistently detectable PSA level after RP (n = 27) or for a delayed rise in PSA (n = 44) to a median dose of 70 Gy (range 60-78). Adjuvant androgen ablation was given to 37 patients; 2 who had received ART and 35 had who received SRT. The median duration of androgen ablation was 24 months. The primary end point was freedom from biochemical failure (bNED), which was considered to be an undetectable PSA level. The median follow-up was 53 months for all patients: 68 months for the ART patients and 35 months for the SRT patients. Results: For the ART group, 8 patients subsequently developed a rising PSA level. The 5-year bNED rate was 88%. SVI was the strongest predictor of outcome, with a 5-year bNED rate of 94% for those without SVI and 65% for those with SVI (p = 0.0002). SVI was the only significant factor in Cox proportional hazards regression analysis in the ART cohort. For the SRT group, 20 patients developed a rising PSA level after RT. The 5-year bNED rate was 66% for all SRT patients, and 43% and 78% in those with a persistently detectable PSA and those with a delayed rise in PSA, respectively. In the Cox proportional hazards regression analysis, this subdivision of SRT was statistically significant. Moreover, when the Cox model included all patients and variables, the timing of RT (ART vs. SRT) was an independent correlate of bNED, as was androgen ablation. Conclusion: For RP patients with high-risk pathologic features, the timing of postoperative RT and the PSA status after RP were strong determinants of outcome. Because of the potential confounding factors, direct comparisons of ART and SRT are problematic; however, ART is extremely effective and offers the surest approach for maintaining biochemical control.