We read with interest this article by Jhavar et al.[1]; we feel that this study essentially corroborates our study on 455 archival prostate cancer lesions from the Swedish Watchful Waiting and the USA-based Physicians’ Health Study cohort [2]. Our study established an 87-gene expression signature that distinguishes TMPRSS2-ERG fusion prostate cancer as a discrete molecular entity. To elucidate that nature of the TMPRSS2-ERG signature, we used computational analysis and identified a relationship between the gene signatures of TMPRSS2-ERG cancers and oestrogen receptor (ER) signalling. We functionally validated the role of ER signalling in TMPRSS2-ERG fusion prostate cancer.

However, although both studies used approaches to carefully select tumour areas, and interrogated for TMPRSS2-ERG signatures, direct comparisons between our study and that of Jhavar et al. are not valid, because of the different methods used. The studies differed in cohort size, array platforms, methods for computational analyses, selection of tissue areas for analysis, and sample types (FFPE versus fresh-frozen tissues). Despite these essential differences, Jhavar et al. came to similar conclusions as reported by us, i.e. that a proportion of differentially regulated genes in ERG-rearranged cancers are known to be modulated by oestradiol (and appeared in the context of an oestradiol-containing pathway). In addition, there were significant overlaps between the differentially expressed gene signatures. Hence, we think that these findings strongly support our results. Furthermore, the authors raised the issue of further analysing ERG-rearrangement-positive cancers to test whether they show sensitivity to oestrogen exposure. We have addressed this topic by a series of functional studies in vitro. Indeed, oestrogen treatment has an effect on prostate cancer cells harbouring the TMPRSS2-ERG fusion, depending on the type of ER (ERα and β) that is stimulated. Thus, we draw the attention of your readers to our study [2], which has been corroborated by the results of Jhavar et al.[1].

  • 1
    Jhavar S, Brewer D, Edwards S et al. Integration of ERG gene mapping and gene-expression profiling identifies distinct categories of human prostate cancer. BJU Int 2008 [Epub ahead of print]
  • 2
    Setlur SR, Mertz KD, Hoshida Y et al. Estrogen-dependent signaling in a molecularly distinct subclass of aggressive prostate cancer. J Natl Cancer Inst 2008; 100: 81525