8q gain is an independent predictor of poor survival in diagnostic needle biopsies from prostate cancer suspects.

PURPOSE The main procedure to confirm a suspected diagnosis of prostate cancer is histologic analysis of ultrasound-guided sextant prostate biopsies. As it is difficult to reliably assess tumor stage and grade in such minute samples, the clinical significance of some tumor foci remains unclear. Genetic markers that could augment pretreatment prognostic information would improve the clinical management of the disease. EXPERIMENTAL DESIGN We have analyzed by comparative genomic hybridization a consecutive series of prostate needle biopsies obtained prospectively from 100 prostate cancer suspects. For 25 of these patients, a second independent biopsy core was analyzed to assess possible tumor heterogeneity. Additionally, a three-color fluorescent in situ hybridization assay was done in paraffin-embedded biopsy cores to validate the comparative genomic hybridization findings and to confirm their prognostic value. RESULTS Sixty-one of 100 biopsy samples had morphologic evidence of prostate cancer and 41 (67%) of these displayed genomic copy number changes as opposed to none of the morphologically normal biopsies. The presence of losses, amplifications, and the total number of genomic imbalances were significantly associated with poorly differentiated tumors. Kaplan-Meier curves with log-rank test showed that patients whose tumors displayed 8q gains had a significantly worse survival even when tumor grade was taken into account (P = 0.008). Restricting the analysis to cases with Gleason score 7, the most troublesome category in terms of prognostic information, gains at 8q were still significantly associated with poor survival (P = 0.011), something that was confirmed by fluorescent in situ hybridization in an independent series of biopsies with much longer follow-up time (P = 0.023). CONCLUSIONS We show that whole genomic information can be obtained from minute needle biopsies of prostate cancer suspects and that genetic data can provide additional prognostic information before a therapeutic decision is taken.