Clinical and pathologic factors predictive of biochemical control following post-prostatectomy irradiation

Purpose/Objective: Indications for post-prostatectomy radiation therapy are not well defined. We reviewed our experience treating post-prostatectomy patients with external beam irradiation to assess clinical and pathologic factors predictive of biochemical control. Materials and Methods: Between 1/87 and 3/93, 61 patients received post-operative tumor bed irradiation with a median dose of 59.4 Gy (50.4 - 68 Gy). Median follow-up was 4.1 years (7.6 months - 8.3 years) from irradiation. Patients were treated for the following reasons: 1) adjuvantly, within 6 months of surgery for extracapsular extension, seminal vesicle involvement, or positive surgical margins (n=38); 2) persistently elevated PSA post-operatively (n=2); 3) rising PSA >6 months after surgery (n=9); and 4) biopsy proven local recurrence (n=12). No patients had known nodal or metastatic disease. All patients had post-radiation PSA data available. Biochemical control was the endpoint studied using Kaplan-Meier life table analysis. Biochemical control was defined as the ability to maintain an undetectable PSA (< 0.4 ng/ml) after irradiation. If the PSA failed to become undetectable within 1 year following irradiation or if hormonal therapy was initiated, the patient was censored for failure. Presurgical and pre-radiation PSA values, Gleason score, pathologic T stage, and reason for irradiation were analyzed for their impact on biochemical control. Results: The 5 year actuarial survival and disease-free survival rates for the entire group were 87.7% and 69.0%, respectively. The 5 year actuarial rate of biochemical control was 47.2%. Of 33 patients who were scored a biochemical failure, 21.3% did not achieve an undetectable PSA within 1 year. Patients were divided into 4 presurgical PSA groups: ≤4, >4 and ≤1 0, >10 and ≤20, and > 20 ng/ml. The 3 year actuarial rates of biochemical control were 100% for group 1, 66.7% for group 2, 61.5% for group 3, and 28.6% for group 4. Pre-RT PSA values were also evaluated. Univariate Cox models indicated lower presurgical and pre-RT PSA values were predictive of biochemical control (p=0.017, p<0.001). Gleason scores were divided into 3 groups: 2-4, 5-7, and 8-10. The 3 year actuarial rates of biochemical control were 50.0%, 80.0%, and 35.3%, respectively. By pairwise log rank test, biochemical control was significantly lower in group 3 than in group 2 (p=0.031). Group 1 (n=2) was too small for analysis. When stratified by T stage, the 3 year actuarial rates of biochemical control were as follows: T2: 20.0%, T3a: 72.4%, T3b: 85.7%, T3c: 56.2%, and T4: 0% (p<0.001). Patients treated adjuvantly (group 1) had a 3 year actuarial rate of biochemical control of 83.8%. For patients whose PSA failed to normalize within 6 months of surgery (group 2), the 3 year actuarial rate of biochemical control was 0% (n=2). For patients treated for a rising PSA >6 months after surgery (group 3), the 3 year actuarial rate of biochemical control was 55.6%. The 3 year actuarial rate of biochemical control for patients treated for a biopsy proven recurrence (group 4) was 8.3%. By pair-wise log rank test, the rates of biochemical control were significantly different between groups 1 and 3 (p=0.036), groups 1 and 4 (p<0.001), and groups 3 and 4 (p=0.009). Conclusion: Biochemical control was achieved in approximately half of the patients treated with post-operative prostatic fossa irradiation. Elevated presurgical and pre-RT PSA values, Gleason score ≥8, and biopsy proven recurrence were significantly predictive of biochemical failure. Patients treated adjuvantly were more likely to achieve biochemical control when compared to those with persistent or rising PSA values. Further study is needed to identify the subset of post-prostatectomy patients most likely to benefit from the addition of irradiation