Prostate cancer is the most commonly diagnosed cancer in men in many developed countries. It is often indolent, particularly when diagnosed in the oldest age groups: some patients live with their disease for many years or even decades, eventually dying of a different cause. However, aggressive, metastatic prostate cancers are also found, and these have a poor prognosis. Outcome prediction is therefore very important when planning the treatment for an individual patient, and the current methods, which use a combination of prostate specific antigen (PSA) levels, Gleeson score, and tumour stage, are imprecise. Genetic analysis should lead to the development of more accurate predictive tools. One particularly interesting gene is SMAD4; its loss or down-regulation has already associated with an aggressive phenotype in human prostate cancer.

Mice in which the Pten gene has been deleted in the prostate only (Pten^pc-/-) develop slow-growing, non-metastatic prostate adenocarcinomas. A group of researchers led from the Dana-Farber Cancer Institute and Harvard Medical School, Boston, MA, has now investigated the relationship between SMAD4 and prostate cancer progression using these mice. They generated a mouse model with prostate-specific deletion of both Pten and SMAD4. In contrast with the Pten^pc-/- mice, these Pten^pc-/- SMAD4 ^pc-/- mice developed highly aggressive, metastatic prostate adenocarcinomas and died by 32 weeks of age.

The researchers then compared the profile of RNA transcripts in primary prostate tumours from the Pten^pc-/- SMAD4 ^pc-/- mice with tumours from the original mice and from mice with prostate-specific deletion of the tumour suppressor gene p53 as well as Pten. Genes in the broad functional categories of “cell movement” and “cell growth and proliferation” were found to be up-regulated in tumours from the Pten^pc-/- SMAD4 ^pc-/- mice compared to those from the other two mouse models. More detailed analysis suggested that two genes in particular were involved in the proliferative response in the absence of SMAD4 expression: cyclin D1 and Spp1, and this was confirmed in further experiments. Chromatin immunoprecipitation confirmed that SMAD4 could bind to the promoter region of cyclin D1, and a similar binding element was found in the Spp1 promoter. Increasing Spp1 expression was found to increase the metastatic potential of human prostate cancer cell lines.

Taken together, these findings suggested that it might be possible to derive a four-gene signature for the metastatic potential of prostate cancer, with under-expression of Pten and SMAD4 and over-expression of cyclin D1 (or Ccnd1) and Spp1 correlated with an aggressive phenotype. The researchers used published expression profile data from a large cohort of prostate tumours [2] to test this hypothesis, finding that a risk score obtained from combining the expression levels of the four genes could stratify the cohort into two groups with significantly different risk of cancer progression (defined as biochemical recurrence or BCR). More importantly, although the four-gene score alone was only as accurate as the Gleason score in predicting progression, a combination of the two scores gave improved accuracy.

Realising that their results may be complicated by the inherent heterogeneity of individual prostate tumours, the researchers then stained sections of 450 randomly selected tumours for expression of the protein products of these four genes, and found that these tissue-specific expression profiles could enhance the prognostic potential of the Gleeson test further. Validation in further patient cohorts showed that outcome prediction was most accurate when the four-gene profile was added to the standard clinical combination of Gleeson score, age of diagnosis and tumour stage. The researchers expect this work to feed into the development of a molecular assay for predicting the prognosis of prostate cancer, which would be an important new clinical tool.

References

[1] Ding, Z., Wu, C-J, Chum G.C. and 25 others (2011). SMAD4-dependent barrier constrains prostate cancer growth and metastatic progression. Nature, published online ahead of print 2 February 2011. doi:10.1038/nature09677

[2] Glinsky, G.V., Glinskii, A.B., Stephenson, A.J., Hoffman, R.M. and Gerald, W.L. (2004). Gene expression profiling predicts clinical outcome of prostate cancer. J. Clin. Invest. 113(6): 913-23.