Castration therapy rapidly induces apoptosis in a minority and decreases cell proliferation in a majority of human prostatic tumors.

Major differences in the long-term clinical response to castration therapy of prostatic carcinoma suggests intertumoral differences in cellular response and defines a need for identification of patients with an eventually positive outcome as well as those in need of additional treatment. Using morphometry, monoclonal antibodies against Bcl-2, c-myc, Ki-67, and p53 proteins, and an in situ method to visualize apoptotic cells, we examined the short-term response of prostatic tumors to castration in core biopsies from 18 prostatic cancer patients taken the day before and 7 days after castration. At the histological level, 3 tumors seemed practically unaffected by castration. In 15 tumors, castration induced vacuolization of tumor cell cytoplasm and decreases in nuclear area and Ki-67 index. In these 15 tumors, apoptotic index was significantly increased in 6, principally unaffected in 6, and decreased in 3. The 6 tumors responding with an increase in apoptotic index were WHO grade 1 or 2 and negative for p53, c-myc, and Bcl-2 or contained only few Bcl-2- or c-myc-positive tumor cells before therapy. The 12 tumors in which apoptotic index was unaffected or decreased were WHO grade 2 or 3 and immunopositive for one or more of p53, Bcl-2, and c-myc proteins before therapy. The Bcl-2 index was significantly increased in 10 patients. Prostatic tumors may respond in a variety of possibly predictable ways to castration therapy including a decrease in apoptotic index. The magnitude of these responses are not correlated in individual tumors, suggesting that the common classification of prostatic tumors as either androgen dependent (dying after castration) or independent (not responding at all to castration) may be an oversimplification.