**3. DNA biomarkers**

#### **3.1 Epigenetic markers**

Epigenetic alterations, i.e., alterations in gene expression without changes in the DNA sequence, include global genomic hypomethylation, promoter hypermethylation of CpG islands and loss of imprinting. The most common somatic genome alteration during prostate cancer development appears to be the hypermethylation in the regulatory region of certain genes, most commonly in the promoter of the π-class glutathione-*S*-transferase (*GSTP1*) gene (Lee et al., 1994). GSTP1 is a member of a large family of glutathione transferases that protect DNA from free radicals (Hayes & Strange, 1995). Men with a positive preoperative serum analysis for *GSTP1* CpG island hypermethylation were at significant risk to experience PSA recurrence within the first several years following radical prostatectomy (Bastian et al., 2005). A high frequency of *GSTP1* methylation in the urine of men with high-stage cancer was also recently reported (Woodson et al., 2008). Other candidate genes have been examined for hypermethylation along with *GSTP1*. A recent study suggested that men with advanced prostate cancer and biochemical recurrence experience a significant increase in promoter hypermethylation between initial diagnosis (first blood analysis) and time to progression (second blood analysis) in the four genes with the highest methylation frequencies (*GSTP1*, *APC*, *RAR*β*2* and *RASSF1*α) in prostate cancer patients compared to age-matched controls (Roupret et al., 2008). This study suggests that multiple gene methylation analysis in circulating cell DNA could be a good biomarker for early detection of prostate cancer recurrence.

#### **3.2 Gene fusion proteins**

Based on a bioinformatics strategy Tomlins et al. described for the first time in prostate cancer a series of genetic rearrangements between the 5'-untranslated region of TMPRSS2 (21q22) and some members of the ETS family of transcription factors, such as ERG (21q22), ETV1 (7p21) and ETV4 (17q21), which have important roles in several oncogenic pathways (Tomlins et al., 2005). Approximatively 50% of prostate cancers from serum PSA-screened cohorts harbor recurrent gene fusions (Kumar-Sinha et al., 2008), which can be detected by fluorescent *in situ* hybridisation (FISH). Conflicting results have been reported regarding the prognosis value of prostate cancer harboring TMPRESS2:ERG gene fusions. Earlier studies reported associations with high stage, metastasis, and prostate cancer-specific death (Attard et al., 2008; Demichelis et al., 2007; Nam et al., 2007), but more recent reports found no association with outcome (Gopalan et al., 2009; Leinonen et al., 2010; Mehra et al., 2007), an association with favorable outcome (Saramaki et al., 2008), or a similar percentage of TMPRSS2:ERG gene fusion in minute and nonminute adenocarcinomas (Albadine et al., 2009), all suggesting its lack of value as a marker of aggressive prostate cancer. The analysis of the relationship between TMPRSS2:ERG fusion and morphological features of prostate cancer has produced diverging results. Most studies have found no statistically significant association between TMPRSS2:ERG rearrangement and Gleason score, while some have demonstrated an association with either higher (Attard et al., 2008; Demichelis et al., 2007) or lower Gleason scores (Fine et al., 2010; Gopalan et al., 2009). Taken together, it seems like the TMPRSS2:ERG fusion gene is an early event related to development of prostate cancer rather than a marker for progressive disease. Of note, the TMPRSSE:ERG fusion has potential for noninvasive prognosis of prostate cancer. Although RNA-based urinary tests demonstrate in general a high specificity and sensitivity to detect prostate cancer, no significant relationship was found between the presence of fusion transcripts and Gleason score or clinical stage (Hessels et al., 2007; Rice et al., 2010).

#### **3.3 Loss of heterozygosity**

6 Prostate Cancer – Diagnostic and Therapeutic Advances

radiation therapy, and /or development of distant metastases (Shariat et al., 2007). Larger studies are needed to validate the promising role of uPA and uPAR as biomarkers of

Epigenetic alterations, i.e., alterations in gene expression without changes in the DNA sequence, include global genomic hypomethylation, promoter hypermethylation of CpG islands and loss of imprinting. The most common somatic genome alteration during prostate cancer development appears to be the hypermethylation in the regulatory region of certain genes, most commonly in the promoter of the π-class glutathione-*S*-transferase (*GSTP1*) gene (Lee et al., 1994). GSTP1 is a member of a large family of glutathione transferases that protect DNA from free radicals (Hayes & Strange, 1995). Men with a positive preoperative serum analysis for *GSTP1* CpG island hypermethylation were at significant risk to experience PSA recurrence within the first several years following radical prostatectomy (Bastian et al., 2005). A high frequency of *GSTP1* methylation in the urine of men with high-stage cancer was also recently reported (Woodson et al., 2008). Other candidate genes have been examined for hypermethylation along with *GSTP1*. A recent study suggested that men with advanced prostate cancer and biochemical recurrence experience a significant increase in promoter hypermethylation between initial diagnosis (first blood analysis) and time to progression (second blood analysis) in the four genes with the highest methylation frequencies (*GSTP1*, *APC*, *RAR*β*2* and *RASSF1*α) in prostate cancer patients compared to age-matched controls (Roupret et al., 2008). This study suggests that multiple gene methylation analysis in circulating cell DNA could be a good biomarker for

Based on a bioinformatics strategy Tomlins et al. described for the first time in prostate cancer a series of genetic rearrangements between the 5'-untranslated region of TMPRSS2 (21q22) and some members of the ETS family of transcription factors, such as ERG (21q22), ETV1 (7p21) and ETV4 (17q21), which have important roles in several oncogenic pathways (Tomlins et al., 2005). Approximatively 50% of prostate cancers from serum PSA-screened cohorts harbor recurrent gene fusions (Kumar-Sinha et al., 2008), which can be detected by fluorescent *in situ* hybridisation (FISH). Conflicting results have been reported regarding the prognosis value of prostate cancer harboring TMPRESS2:ERG gene fusions. Earlier studies reported associations with high stage, metastasis, and prostate cancer-specific death (Attard et al., 2008; Demichelis et al., 2007; Nam et al., 2007), but more recent reports found no association with outcome (Gopalan et al., 2009; Leinonen et al., 2010; Mehra et al., 2007), an association with favorable outcome (Saramaki et al., 2008), or a similar percentage of TMPRSS2:ERG gene fusion in minute and nonminute adenocarcinomas (Albadine et al., 2009), all suggesting its lack of value as a marker of aggressive prostate cancer. The analysis of the relationship between TMPRSS2:ERG fusion and morphological features of prostate cancer has produced diverging results. Most studies have found no statistically significant association between TMPRSS2:ERG rearrangement and Gleason score, while some have demonstrated an association with either higher (Attard et al., 2008; Demichelis et al., 2007)

aggressive prostate cancer.

early detection of prostate cancer recurrence.

**3.2 Gene fusion proteins** 

**3. DNA biomarkers 3.1 Epigenetic markers** 

> The loss of heterozygosity (LOH) is a frequent genetic alteration in prostate cancer, in particular on chromosome arms 7q, 8p, 10q, 12p, 13q, 16q, 17q and 18q (Dong, 2006). Studies on chromosomal deletions of 8p22 by fluorescence in situ hybridization technique revealed 8p22 deletion to be the strongest parameter to predict disease progression in patients undergoing surgery (Matsuyama et al., 2001). If some LOH have been shown to be associated with early stages of prostate cancer (Lu & Hano, 2008), others seem to indicate the presence of tumor suppressor genes whose inactivation is correlated with aggressive and metastatic tumors (Dong et al., 2000; Kibel et al., 2000; Matsuyama et al., 2007). A recent study reported the development of a noninvasive method to detect early stages of prostate cancer using LOH analysis of 7q31, 8p22, 12p13, 13q14, 16q23.2 and 18q21. Indeed LOH could be found in cells from urine obtained by prostatic massage (Thuret et al., 2005). In patients who underwent radical prostatectomy, LOH was confirmed from the prostatic tissue with a concordance of 86%. This noninvasive approach warrants further investigation to bring pronostic information on prostate cancer aggressiveness.
