Clinical impact of PTEN genomic deletions and TMPRSS2-ERG gene fusions in prostate cancer

AACR Annual Meeting-- Apr 14-18, 2007; Los Angeles, CA 5674 The inactivation of tumor suppressor genes, and the disruption of transcription factor function play a central role in the initiation and progression of many human cancers. Supportive evidence for this model in prostate cancer (CaP) has been provided by the high frequency PTEN tumor suppressor gene losses in tumors, and the recent demonstration of activating genomic fusion rearrangements of ETS transcription factor genes, such as ERG with the TMPRSS2 locus in ~60% of CaP. The strong association between PTEN haploinsufficiency and CaP in both the knockout mouse and in sporadic human cancers, together with the high specificity and frequency of TMPRSS2-ETS fusions in human CaP, implies that both pathways are directly involved in the pathogenesis of CaP. To date there has been no large-scale systematic analysis of the influence of genomic PTEN deletion and TMPRSS2-ETS gene fusions on the clinical parameters of CaP progression, including incidence-free survival, biochemical response and other clinical-pathologic determinants of outcome. Moreover, no studies have been performed to determine the frequency with which PTEN and TMPRSS2-ETS rearrangements occur together in CaP, or to examine clinical phenotypes associated with genomic mutations affecting both pathways. This study was designed to retrospectively assess the overall frequency with which PTEN deletions and TMPRSS2-ERG fusion rearrangements occur simultaneously, and their clinical impact through FISH analysis using well-annotated high-throughput tissue microarrays (TMAs) derived from 107 CaP patients. In addition, four-color FISH analysis of the TMPRSS2-ERG fusion was used to precisely delineate the interstitial deletion of genomic content from this region of chromosome 21. Analysis of the TMAs by interphase FISH indicated that PTEN deletion was present in 47/107 (44%) of CaP, with a homozygous PTEN deletion present in 5 of the 47 tumors. Kaplan-Meier survival analysis of PTEN deleted (hemizygously or homozygously) vs. not deleted identified three subgroups with different prognosis with respect to time to biochemical relapse after surgery. Overall the analysis demonstrated a significant association (P=0.0003) between PTEN deletion and a shorter time interval to PSA recurrence. Significantly, homozygous PTEN deletion in five tumors was associated with a much earlier onset of biochemical recurrence based on PSA values. It was noteworthy that 35% of tumors had neither a PTEN deletion nor a TMPRSS2-ERG fusion event. While the ERG and PTEN pathways are distinct, these data suggest that PTEN loss and ETS activation by TMPRSS2-ERG fusions do not occur independently. Given the uncertainty regarding the biological impact of ETS fusion transcripts in CaP, alone or in combination with PTEN genomic losses, these new genotype-phenotype correlative studies will provide valuable prognostic information for future CaP treatment and long-term management.