Is it possible to provide a prognosis after radical prostatectomy for prostate cancer by means of a PSA regression model?

Background: For over 15 years, studies have been done to evaluate the elimination kinetics of the prostate-specific antigen (PSA) after radical prostatectomy. Even though evaluation of PSA regression in the two-compartment model has become established, no clear data are currently available as to whether a statement can be made with regard to tumor prognosis from a computation of the PSA half-life (PSA-HL). This study focuses on the determination of the PSA-HL in the two-compartment model and on its correlation with the biochemical recurrence-free survival. In addition, a computer program is being developed to simplify the determination of PSA-HL.

Material And Methods: Seventy-seven prospective patients were examined who subsequently had a radical prostatectomy at our facility without neoadjuvant or adjuvant hormone deprivation. In addition to preoperative measurement of the PSA value (dO), PSA determinations were carried out postoperatively on days 5, 10 and 60, and at four-monthly intervals thereafter (mean follow-up: 16 months). By means of the computer program developed for this purpose, CTK.TumW, the PSA half-lives for the first (d0-d5, PSA-HL1) and second (d5-d10, PSA-HL2) compartments were subsequently determined and their effect on biochemical recurrence-free survival was assessed.

Results: PSA-HL1 and PSA-HL2 were 1.89 (+/- 0.03) and 3.39 (+/- 0.14) days, respectively. Whilst PSA-HL1 did not permit any prognostic statement, the median PSA-HL in the second compartment between patients with and without disease progression differed significantly (4.44 versus 3.12 days; p < 0.001). Discrimination analysis produced a cutoff of 3.8 days for the second compartment; patients with a PSA-HL2 > or = 3.8 days had a significantly worse biochemical recurrence-free survival after 18 months than the other patients (27% versus 93%; p < 0.001).

Conclusion: The PSA regression kinetics after radical prostatectomy follows a two-compartment model in which the prognostic value of the PSA-HL1 is limited. When a cutoff of 3.8 days is used, evaluation of the PSA-HL in compartment 2 (d5-10) appears to permit a prognostic statement. Due to the limited postsurgical follow-up, the disease process was only assessed as biochemical recurrence-free survival, and a longer follow-up will be necessary to generate data on progression-free survival.