This is a response to RFA CA-91-10 and is a collaboration among cytogenetics, medicine, molecular biology, pathology, and urology. Prostate cancer (PCA) is the most commonly diagnosed invasive cancer and the second most common cause of cancer deaths in US males. It is increased in incidence among blacks and among increasingly aged subjects. Survival is shorter among blacks. The clinical course ranges from rapidly fatal to survival often 10-20 years after diagnosis. Most PCA is discovered only at autopsy. Most PCA diagnosed during life is or becomes symptomatic and may result in severe pain from metastases to bone in up to 70-80% of PCA. Survival may be long, with or without severe symptoms, even many years after the identification of metastases. There is a great need to be able to predict which PCAs will progress slowly and which will progress rapidly, i.e., may be better followed without aggressive therapy. Our goal is the precise identification of biologically different subpopulations. The most commonly recognized predictor of clinical behavior is histopathological differentiation. A more powerful predictor is needed; this may require an approach that is qualitatively different from (hopefully complementary to) the currently employed predictors. We want to identify cytogenetic and molecular genetic aberrations that will be useful as more precise prognostic indicators. An important obstacle to this goal in the past has been the exceptional difficulty encountered in the propagation of PCA cells both in vivo and in vitro. Recently, we reported a method for the propagation of a large proportion of primary PCAs in athymic animals. We have since transplanted xenografts serially. This permits expansion of the available tumor, a useful approach to the procurement of malignant PCA cells free of benign epithelial cells and human stromal cells, and the study of viable PCA cells with many techniques over the course of time. We also have methods that have permitted the short-term culture of PCA adequate for cytogenetic analyses. We shall attempt to correlate several clinical parameters with experimental parameters. Clinical parameters routine in our center include race, age, Gleason grade, ultrasound image, stage of disease, metastatic survey, serum prostate specific antigen before and at intervals after surgery, rate of growth of measurable tumors in patients, time to recurrence of disease, and survival (both total and disease specific). Experimental parameters to be correlated and evaluated as possibly important predictors both as single parameters and as multiple complementary parameters include Gleason patterns (quantified morphometrically), size of tumor (quantified morphometrically), ploidy, cytogenetic aberrations (both conventional and by fluorescence in situ hybridization). rate of growth in immunodeficient animals, the loss and mutation of selected tumor suppressor genes, and the protein expression and gene amplification of selected growth factors and their receptors.
Showing the most recent 10 out of 40 publications
