Occult dissemination of tumor cells mainly determines the prognosis of patients with primary prostate cancer. The effect of androgen deprivation on micrometastatic tumor cells in these patients is currently unknown. We therefore used an immunocytochemical assay with monoclonal antibodies (MAbs) directed against epithelial cytoskeleton proteins (i.e., cytokeratins) to monitor the concentration of isolated tumor cells in the bone marrow of 36 prostate cancer patients (stage C), who underwent hormonal androgen deprivation with Flutamide and Leuprorelin acetate. Tumor cells in cytologic bone marrow preparations were detected using an assay that employed the MAb CK2 directed against cytokeratin (CK) 18 and the alkaline anti-alkaline phosphatase staining method. Prior to therapy, we detected between 1 and 38 CK-positive cells per sample of 2 x 10(6) nucleated cells in 21 patients, while the remaining 15 patients displayed tumor-free marrow samples. There was no significant correlation between the concentration of CK-positive cells and the volume of hypo-echogenic lesions as an indicator of the primary tumor volume or the serum level of prostate-specific antigen (PSA). After androgen deprivation, 20 of the 21 initially positive patients either became negative (n = 16) or showed at least a reduction in the concentration of CK-positive cells (n = 4). Moreover, only 2 of the 15 patients with negative pre-treatment findings became positive. All of the 7 patients with remaining tumor cells in the bone marrow after therapy showed no detectable amounts of PSA in their serum. Our findings suggest that serum PSA concentration is no indicator of micrometastatic disease in bone marrow. Neoadjuvant androgen deprivation appears to eliminate disseminated CK-positive tumor cells present in bone marrow, a preferred site of overt metastasis in prostate cancer patients.
