The indications for adjuvant external beam radiotherapy (EBRT) after radical prostatectomy (RP) are poorly defined. We performed a retrospective comparison of our institution's experience treating prostate cancer with RP vs. RP followed by adjuvant EBRT.
Between 1987 and 1998, 617 patients with clinical Stage T1-T2N0M0 prostate cancer underwent RP. Patients who underwent preoperative androgen deprivation and those with positive lymph nodes were excluded. Of the 617 patients, 34 (5.5%) with an undetectable postoperative prostate-specific antigen (PSA) level underwent adjuvant prostatic fossa RT at a median of 0.25 year (range, 0.1-0.6) postoperatively because of poor pathologic features. The median total dose was 59.4 Gy (range, 50.4-66.6 Gy) in 1.8-2.0-Gy fractions. These 34 RP+RT patients were compared with the remaining 583 RP patients. Biochemical failure was defined as any postoperative PSA level > or =0.1 ng/mL and any postoperative PSA level > or =0.3 ng/mL (at least 30 days after surgery). Administration of androgen deprivation was also scored as biochemical failure when applying either definition. The median clinical follow-up was 8.2 years (range, 0.1-11.2 years) for RP and 8.4 years (range, 0.3-13.8 years) for RP+RT.
Radical prostatectomy + radiation therapy patients had a greater pathologic Gleason score (mean, 7.3 vs. 6.5; p < 0.01) and pathologic T stage (median, T3a vs. T2c; p < 0.01). Age (median, 65.7 years) and pretreatment PSA level (median, 7.9 ng/mL) were similar between the treatment groups. Extracapsular extension was present in 72% of RP+RT patients vs. 27% of RP patients (p < 0.01). The RP+RT patients were more likely to have seminal vesicle invasion (29% vs. 9%, p < 0.01) and positive margins (73% vs. 36%, p < 0.01). Despite these poor pathologic features, the 5-year biochemical control (BC) rate (PSA <0.1 ng/mL) was 57% for RP+RT and 47% for RP (p = 0.28). For patients with extracapsular extension, the 5-year BC rate was 52% for RP+RT vs. 30% for RP (p < 0.01). The 5-year BC rate for patients with seminal vesicle invasion was 60% for RP+RT vs. 18% for RP (p < 0.01). For those with positive margins, the 5-year BC rate was 64% for RP+RT vs. 27% for RP (p < 0.01). The use of adjuvant RT remained statistically significant on multivariate analysis when applying either biochemical failure definition. Adjuvant RT also remained statistically significant when including the postoperative PSA level (>30 days after surgery) in the multivariate analyses. In addition, 99 (17%) of the 583 RP patients required salvage prostatic fossa RT (median dose, 59.4 Gy) at a median interval of 1.3 years after surgery (range, 0.1-8.4) for a palpable recurrence (n = 10) or a detectable/rising postoperative PSA level (n = 89). The median PSA level before salvage RT was 0.8 ng/mL (mean, 3.2 ng/mL). The 5-year and 8-year BC rate, using the PSA <0.1 ng/mL definition, from the date of salvage RT was 41% and 35%, respectively. The 5-year and 8-year BC rate, using the PSA <0.3 ng/mL definition, was 46% and 36%, respectively. The 8-year local recurrence rate after salvage RT was 4%.
Adjuvant RT demonstrated improved efficacy against prostate cancer. For patients with poor pathologic features (extracapsular extension, seminal vesicle invasion, positive margins), adjuvant RT improved the biochemical outcome independent of other prognostic factors.