Neoadjuvant androgen deprivation therapy before radical prostatectomy: who is unlikely to benefit?

To determine which patients undergoing radical prostatectomy for localized cancer are unlikely to benefit from neoadjuvant androgen withdrawal therapy.

Over a 5-year period, 173 patients underwent radical retropubic prostatectomy: 87 patients received no pre-operative androgen withdrawal therapy, for whom the clinical stages were T1b (17 patients), T1c (3), T2a (31) and T2b/c (36). Pre-operative prostate biopsies contained well-differentiated cancer in 43 and moderately differentiated cancer in 44 patients. Serum prostate-specific antigen (PSA) ranged from 0.4 ng/mL to 110.1 ng/mL (median 5.0). The presence of extracapsular disease, positive surgical margins and seminal vesicle involvement were correlated with pre-operative PSA, PSA density (PSAD), the natural logarithm of PSA [In(PSA)], biopsy grade, clinical stage, and sparing of neurovascular bundles.

Patients with a PSA < or = 4.0 ng/mL had a rate of extracapsular disease and positive margins of 8.3% and none had seminal vesicle invasion; corresponding rates for patients with a PSA > 4.0 ng/mL were 57%, 51% and 12%, respectively. Multivariate analysis revealed In(PSA) to be the best predictor of extracapsular disease (P < < 0.001), margin positivity (P < < 0.001) and seminal vesicle invasion (P = 0.0019).

These results suggest that in patients who have clinically localized, well or moderately differentiated prostate cancer, a PSA < or = 4.0 ng/mL is predictive for organ-confined disease. This group is unlikely to benefit from neoadjuvant androgen deprivation. Ln(PSA) may be able to improve the accuracy of mathematical multivariate models predicting the presence of extracapsular disease or margin positivity, but this will require verification in larger population-based studies.