Combined study of prostatic carcinoma by classical cytogenetic analysis and comparative genomic hybridization.

Conventional cytogenetic analysis of prostatic carcinoma (PC) is characterized by inefficient growth of tumor cells during in vitro culture, leading to a lack of aberrant karyotypes in many of investigated tumors. In this study we have combined a modified short-term tissue culture method for conventional banding analysis and comparative genomic hybridization (CGH) to examine genetic changes in PC, and to evaluate the effect of the in vitro culture on chromosomal changes by comparing results of the two methods. Cytogenetic analysis was performed on 34 PCs using both, conventional and molecular methods. Tumor tissues were obtained predominantly from untreated primary tumors from 48 patients. For karyotyping all tumor samples were short-term cultured using a feeder layer technique. Additionally DNA from uncultured tumor material from 17 of those patients was isolated and screened for copy number changes using CGH. Conventional banding analysis: clonal aberrations were detected in 65% of the tumor samples. Most of the chromosomal findings were numerical changes, including loss of chromosomes Y (32%), 18, 19 and 21 (each 12%). Less frequent, trisomy of chromosome 7 and monosomy of chromosomes 9, 12 and 22 (each 9%) was found. Additionally an inversion of chromosome 9p and a deletion at chromosome 7q was found in two cases. In 35% no clonal aberrations could be detected. CGH: DNA copy number changes were detected in 65% of the analyzed tumors. Predominantly losses of DNA sequences were found. The most common losses were found at chromosome regions 13q21q33 (29%), 6q11q23 (24%), 16q, and 18 (each 18%), and the most common gains at 19 (18%). In six tumors no copy number changes were found. Both methods showed a similar aneuploidy rate, suggesting that the feeder layer technique is quite a suitable method for in vitro culture of PC cells. However, the two techniques produced substantially differing results for most of the tumor samples, and in some cases the discrepancies are quite striking. Therefore eventual culture effects need to be taken into account when comparing results from conventional cytogenetics and CGH. Some contrary findings from the two methods are discussed.