ABSTRACT
Prostate cancer is characterized by recurrent deletions that can considerably vary in size. We hypothesized that large deletions develop from small deletions and that this "deletion lengthening" might have a "per se" carcinogenic role through a combinatorial effect of multiple down regulated genes. In vitro knockdown of 37 genes located inside the 6q12-q22 deletion region identified 4 genes with additive tumor suppressive effects, further supporting a role of the deletion size for cancer aggressiveness. Employing fluorescence in-situ hybridization analysis on prostate cancer tissue microarrays, we determined the deletion size at 6q and 16q in more than 3,000 tumors. 16q and 6q deletion length was strongly linked to poor clinical outcome and this effect was even stronger if the length of both deletions was combined. To study deletion lengthening in cancer progression we eventually analyzed the entire cancers from 317 patients for 6q and 16q deletion length heterogeneity and found that the deletion expanded within 50-60% of 6q and 16q deleted cancers. Taken together, these data suggest continuous "deletion lengthening" as a key mechanism for prostate cancer progression leading to parallel down regulation of genes with tumor suppressive properties, some of which act cooperatively.
ABSTRACT
Deletion of chromosome 8p is the second most frequent genomic alteration in prostate cancer. To better understand its clinical significance, 8p deletion was analyzed by fluorescence in-situ hybridization on a prostate cancer tissue microarray. 8p deletion was found in 2,581 of 7,017 cancers (36.8%), and was linked to unfavorable tumor phenotype. 8p deletion increased from 29.5% in 4,456 pT2 and 47.8% in 1,598 pT3a to 53.0% in 931 pT3b-pT4 cancers (P < 0,0001). Deletions of 8p were detected in 25.5% of 1,653 Gleason ≤ 3 + 3, 36.6% of 3,880 Gleason 3 + 4, 50.2% of 1,090 Gleason 4 + 3, and 51.1% of 354 Gleason ≥ 4 + 4 tumors (P < 0,0001). 8p deletions were strongly linked to biochemical recurrence (P < 0.0001) independently from established pre- and postoperative prognostic factors (P = 0.0100). However, analysis of morphologically defined subgroups revealed, that 8p deletion lacked prognostic significance in subgroups with very good (Gleason ≤ 3 + 3, 3 + 4 with ≤ 5% Gleason 4) or very poor prognosis (pT3b, Gleason ≥ 8, pN1). 8p deletions were markedly more frequent in cancers with (53.5%) than without PTEN deletions (36.4%; P < 0,0001) and were slightly more frequent in ERG-positive (40.9%) than in ERG-negative cancers (34.7%, P < 0.0001) due to the association with the ERG-associated PTEN deletion. Cancers with 8p/PTEN co-deletions had a strikingly worse prognosis than cancers with deletion of PTEN or 8p alone (P ≤ 0.0003). In summary, 8p deletion is an independent prognostic parameter in prostate cancer that may act synergistically with PTEN deletions. Even statistically independent prognostic biomarkers like 8p may have limited clinical impact in morphologically well defined high or low risk cancers.
Subject(s)
Chromosome Deletion , Kallikreins/blood , PTEN Phosphohydrolase/genetics , Prostate-Specific Antigen/blood , Prostatic Neoplasms/genetics , Prostatic Neoplasms/mortality , Aged , Biomarkers, Tumor/blood , Chromosomes, Human, Pair 8 , Disease Progression , Follow-Up Studies , Gene Deletion , Humans , Immunohistochemistry , In Situ Hybridization, Fluorescence/methods , Male , Middle Aged , Neoplasm Grading , Neoplasm Staging , Prognosis , Prostatic Neoplasms/blood , Prostatic Neoplasms/pathology , Tissue Array Analysis/methods , Transcriptional Regulator ERG/metabolismABSTRACT
Early-onset prostate cancer (EO-PCA) represents the earliest clinical manifestation of prostate cancer. To compare the genomic alteration landscapes of EO-PCA with "classical" (elderly-onset) PCA, we performed deep sequencing-based genomics analyses in 11 tumors diagnosed at young age, and pursued comparative assessments with seven elderly-onset PCA genomes. Remarkable age-related differences in structural rearrangement (SR) formation became evident, suggesting distinct disease pathomechanisms. Whereas EO-PCAs harbored a prevalence of balanced SRs, with a specific abundance of androgen-regulated ETS gene fusions including TMPRSS2:ERG, elderly-onset PCAs displayed primarily non-androgen-associated SRs. Data from a validation cohort of > 10,000 patients showed age-dependent androgen receptor levels and a prevalence of SRs affecting androgen-regulated genes, further substantiating the activity of a characteristic "androgen-type" pathomechanism in EO-PCA.
