TMPRSS2:ERG fusions, in combination with deletion of the phosphatase and tensin homolog (PTEN) tumor suppressor, have been suggested to cooperatively drive tumor progression in prostate cancer. We utilized a novel heterogeneity tissue microarray containing samples from 10 different tumor blocks of 189 prostatectomy specimens to study heterogeneity of genomic PTEN alterations in individual foci. PTEN alterations were found in 48/123 (39%) analyzable individual tumor foci, including 40 foci with deletions, 7 with deletion and rearrangement, and 1 focus with rearrangement only. PTEN was homogeneously aberrant in only 4 (8%) and heterogeneously in 44 (92%) of the foci. We found a specific sequence of molecular events from PTEN breakage followed by deletion of DNA sequences flanking the breakpoint, resulting in homozygous deletion. The observation that 16 of 19 foci with homogeneous ERG positivity had focal PTEN alterations but none of 10 foci with PTEN alterations had focal ERG positivity (Po0.0001) suggests that PTEN alterations typically develop subsequent to ERG fusions. We demonstrate a high level of intratumoral heterogeneity of PTEN alterations with deletions and rearrangements that challenges potential PTEN routine diagnosis testing in biopsies. The observation that PTEN alterations develop subsequent to ERG fusion strongly suggests that ERG expression may directly drive development of PTEN aberrations.

Approximately 50% of prostate cancers are characterized by TMPRSS2 (transmembrane protease serine 2)-ERG (avian v-ets erythroblastosis virus E26 oncogene homolog) gene fusions resulting in an androgen-regulated overexpression of the transcription factor ERG. Some studies have suggested prognostic or predictive relevance of ERG status in prostate cancer. Such concepts could be impaired by extensive ERG heterogeneity in analyzed tumors. The aim of this study was to analyze the extent of heterogeneity for TMPRSS2-ERG fusion in prostate cancer. To enable large-scale studies on the extent of heterogeneity of biomarkers in prostate cancer, a heterogeneity tissue microarray containing samples from 10 different tumor blocks of 190 large prostate cancers selected from a consecutive series of 480 radical prostatectomies was developed. ERG expression was analyzed by immunohistochemistry. Positive ERG immunostaining was found in arrayed cancer-containing samples from 103 of the 178 analyzable patients (58%). ERG immunostaining was homogeneously positive in 29 prostate cancers (16%), whereas heterogeneous ERG positivity was seen in 74 cancers (42%). ERG heterogeneity was within one tumor focus (intrafocal heterogeneity) in 69 cases (93% of heterogeneous cases) and between different tumor foci (interfocal heterogeneity) in 5 cases (7%). Marked intrafocal heterogeneity challenges the concept of TMPRSS2-ERG fusion always representing an early step in prostate cancer development. Marked heterogeneity also compromises the concept of analyzing ERG status for treatment decisions in diagnostic needle core biopsies.

Prostate cancer is notorious for its heterogeneity, which poses a problem for the applicability of diagnostic molecular markers. However, heterogeneity analysis can provide valuable information on the chronology in which molecular alterations arise. Here, we constructed a heterogeneity tissue microarray (TMA) comprising samples from 10 different tumor areas of 189 prostate cancers each in order to study the sequence of two frequent molecular alterations, i.e. 6q15 deletion and TMPRSS2:ERG fusion. Previous work shows a marked inverse relationship between these alterations, suggesting that presence of one of these alterations might impact development of the other. 6q15 deletion was analyzed by fluorescence in situ hybridization and ERG-expression by immunohistochemistry. Only 6.6% of 334 ERG-positive but 28.4% of 440 ERG-negative TMA spots showed 6q15 deletions (p < 0.0001). A breakdown of these data to the level of tumor foci revealed 6q deletions in 138 tumor foci that were large enough to have at least 3 analyzable TMA spots. These included 42 tumor foci with homogeneous ERG positivity and 16 with homogeneous 6q15 deletions. Remarkably, six of the 42 homogeneously ERG-positive tumor foci (14.3%) harbored small 6q15-deleted areas, but none of the 34 6q15-deleted foci showed areas of ERG positivity (p = 0.022). In conclusion, our data suggest that ERG-fusion can precede 6q15 deletion, but not vice versa. The complete absence of ERG-positive tumor areas in 6q15-deleted tumor foci further suggest that the functional consequences of 6q15 deletions may prevent the development of TMPRSS2:ERG fusions.