Pattern of proliferative index (Ki-67) after anti-androgen manipulation reflects the ability of irradiation to control prostate cancer

Anti-androgen (AA) therapy will cause hormone-sensitive prostate cancer cells to undergo apoptosis and/or enter the resting phase of the cell cycle. Although the decrease of tumor burden would be an advantage for tumor control when irradiation is subsequently added, the cells in resting phase would seemingly be less vulnerable to the usual type of radiation-induced cell killing via DNA strand breakage. In this study of patients with prostate cancer, we examined the proliferative index via Ki-67 staining of biopsy material before, during, and after withdrawal of leuprolide. We studied 15 previously untreated patients with locally advanced prostate cancer. Prostate biopsies were taken at three times: 1) initial diagnosis; 2) after 3 consecutive months of intramuscular 7.5 mg depot; and c) 6 weeks after the last dose. External beam radiation (EBRT) then delivered 66 Gy in 33 sessions to local fields. We used the ASTRO definition of prostate-specific antigen (PSA) failure. We measured serum luteinizing hormone and total testosterone coinciding with each biopsy date. Immunohistochemical staining was performed using Ki-67 antibody clone MIB-1. The follow-up ranged from 36 to 73 months (median 52 months). We discerned two perturbation patterns of Ki-67 with hormonal manipulation. Pattern 1 demonstrated a drop of Ki-67 labeling after leuprolide was in effect and then after leuprolide withdrawal, the Ki-67 rebounded to less than 120% of baseline. Pattern 2 also showed an initial drop with leuprolide but rebounded to more than 120%. Among eight patients demonstrating pattern 1, only one patient had a PSA failure. In contrast among patients with pattern 2, six of seven failed biochemically (Fisher's exact, p = 0.018). All patients had a LH less than 1.0 during leuprolide effect that rose with its withdrawal. There was no correlation of PSA failure with whether total testosterone did or did not rise to more than 100 ng/dl by the time of the withdrawal phase biopsy. Neither the percent of PSA decline during leuprolide nor the minor PSA rebound 6 to 8 weeks after leuprolide withdrawal correlated with the Ki-67 pattern. The pattern of perturbation of immunohistochemical staining for Ki-67 predicts biochemical failure after moderate-dose EBRT in patients with prostate cancer. Several recent analyses of combined EBRT and AA suggest that some patients may benefit from more prolonged use of AA. Because AA can have substantial side effects and is expensive, a method to select patients likely to benefit from long-term AA would be useful. After neoadjuvant AA manipulation, the Ki-67 perturbation pattern, but not the early PSA changes, may help select patients for long-term AA. The Ki-67 pattern might also be used to select patients needing escalated radiation dosage. Further validation of these concepts beyond this pilot study is suggested.