TMPRSS 2 : ETS Fusions and Clinicopathologic Characteristics of Prostate Cancer Patients from Eastern China

Since first observed gene fusions in prostate cancer of androgen-regulated trans-membrane-serine protease gene (TMPRSS2) and erythroblast transformation-specific (ETS) family members (Tomlins et al., 2005), much progress has been made not only in the understanding of fusion mechanism, but also in the transferring to clinical practice. Hessels detected TMPRSS2:ERG in postDRE urine and reported the test a sensitivity of 37%, a specificity of 93%, and positive predictive value of 94% (Hessels et al., 2007). An observation was published that 83% of castration-independent prostate cancer patients with TMPRSS2:ERG fusion had a decrease in PSA following treatment of Abiraterone (Attard et al., 2008). Up to date, more than 20 fusion types have been found. Among these, TMPRSS2: ERG fusion has a dominant prevalence of approximately 50%, compared with other fusion type less than 15% incidence. However, some studies focusing only on Asian people revealed a prevalence of TMPRSS2:ERG differed from these reports mainly concerned patients of western countries (Lee et


Introduction
Since first observed gene fusions in prostate cancer of androgen-regulated trans-membrane-serine protease gene (TMPRSS2) and erythroblast transformation-specific (ETS) family members (Tomlins et al., 2005), much progress has been made not only in the understanding of fusion mechanism, but also in the transferring to clinical practice.Hessels detected TMPRSS2:ERG in post-DRE urine and reported the test a sensitivity of 37%, a specificity of 93%, and positive predictive value of 94% (Hessels et al., 2007).An observation was published that 83% of castration-independent prostate cancer patients with TMPRSS2:ERG fusion had a decrease in PSA following treatment of Abiraterone (Attard et al., 2008).
Up to date, more than 20 fusion types have been found.Among these, TMPRSS2: ERG fusion has a dominant prevalence of approximately 50%, compared with other fusion type less than 15% incidence.However, some studies focusing only on Asian people revealed a prevalence of TMPRSS2:ERG differed from these reports mainly concerned patients of western countries (Lee et

TMPRSS2:ETS Fusions and Clinicopathologic Characteristics of Prostate Cancer Patients from Eastern China
Jun Dong 1 , Li Xiao 2,3 , Lu Sheng 1 , Jun Xu 1 , Zhong-Quan Sun 1,3 ﹡   al., 2010; Miyagi et al., 2010; Rawal et al., 2013).Six investigations studied Chinese cases, but the outcomes of which were quite dissimilar.The lowest incidence of TMPRSS2:ERG is 7.5%, while the highest is 90% (Dai et al., 2008;Mao et al., 2010;Sun et al., 2010;Xiang et al., 2011;Ren et al., 2012;Wang et al., 2012).Considering the prostate cancer prevalence and aggressiveness, as well as genomic alterations, vary in different ethnic origin and geographic locations (Grönberg et al., 2003;Sim et al., 2005), it is necessary to figure out the prevalence of fusion between TMPRSS2 and ETS family members in prostate cancers patients from eastern China and to explore differences in fusion rates in Different areas of China, other parts of Asia, Europe, and USA.
In this study, we evaluated TMPRSS2:ERG, TMPRSS2:ETV1, and TMPRSS2:ETV4 fusions in more than 100 eastern Chinese prostate cancer patients, using multi-probe fluorescence in situ hybridization (FISH) assay.The specimens were obtained by 12 core needle biopsies and radical prostatectomy, which were analyzed separately.The fusion result was studied with patients' clinicopathologic factors (age, PSA, Gleason score, TNM stage).And the fusion pattern was also investigated.All patients provided written, informed consent in accordance with the institutional guidelines.

Study population and Specimens selection for FISH
Paraffin blocks needle biopsies were collected from 91 cases of eastern Chinese prostate cancer patients who were diagnosed in Huadong Hospital, Shanghai China from October 2010 to February 2012.Their age ranged from 55-90, with the median age of 75.The number of early stage (≤pT2b, without lymph node and bone metastasis) patients is 31 (34.1%), the local advanced (>pT2b, with or without regional lymph node but without bone metastasis) patients is 36 (39.6%) and metastatic patients is 24 (26.4%).The number of Gleason score<7, =7, >7 patients is 9, 43, 39, respectively.14 cases PSA<10ng/ml, 16 cases PSA 10-20 ng/ml, and 61 cases PSA >20ng/ml.Additional, we observed 18 cases of Paraffin blocks gained from radical prostatectomy during the same period.Every case was diagnosed by two independent pathologists from our pathology department.
Tissue from paraffin blocks were obtained for the following procedure.For every fusion types, we selected at least three 3um tissue sections from each needle biopsy case.The sections came from different cores which contained cancer.We tried to choose scattered cores in order to reduce the influence of heterogeneity of prostate cancer.For radical prostatectomy paraffin blocks, sections were collected from different cancer foci.

Criteria for fusion positivity
For TMPRSS2:ERG fusion, the normal signal pattern is two yellow (red/green fusion) signals in a cell (Figure 1A).And the positive signal patterns are one yellow/one green in a cell, which revealed a deletion pattern of fusion (Figure 1B), or one yellow/one green/one red, which illustrated an insertion pattern (Figure 1C).The criteria to determine a prostate cancer case TMPRSS2:ERG fusion positive was that five or more cells with the positive signal pattern were found in a random count of 400 cells.
The normal signal pattern for TMPRSS2:ETV1 or TMPRSS2:ETV4 fusion is two red/two green in a cell (Figure 2).And the positive pattern is an appearance of at least one yellow signal.After a random count of 400 cells, if there were more than two cells with positive signal pattern, we decided the case was TMPRSS2:ETV1 fusion positive.And the criteria for TMPRSS2:ETV4 fusion positive was at least one.

Statistical analysis
The chi-square test or Fisher's exact test were used to compare the results of TMPRSS2:ERG fusion in accordance with the patients' clinicopathologic factors (age, PSA, Gleason score, TNM stage).A P value of< 0.05 was considered to be statistically significant.All statistical analyses were conducted using SPSS v. 20.0.

Discussion
8 years after the first discovery of gene fusions in prostate cancer, relevant studies had already covered a large amount of samples, although mostly the western populations.Methods of PCR, FISH and immunohistochemistry have been introduced to investigate the types, prevalence and mechanism of gene fusions.TMPRSS2:ERG fusion has a prevalence ranging from 40%-70% depending on the clinical cohorts observed.Next to TMPRSS2:ERG, SPINK1-positive fusion and fusion including ETV1 have incidence of 10%-15%, respectively (Rubin et al., 2011).Other fusion types have been reported less than 5%.But studies focused on Asian patients observed different frequencies of gene fusions, especially TMPRSS2:ERG.Miyagi tested transcripts of TMPRSS2:ERG, TMPRSS2:ETV1, SLC45A3:ETV1, HERV-K:ETV1, C15ORF21:ETV1, HNRPA2B1:ETV1 by RT-PCR and found only 54 cases (28%) of TMPRSS2: ERG fusion positive and 2 cases (1%) of HNRPA2B1:ETV1 fusion positive
Whether Chinese patients have a similar frequency like Japanese and Korean is still unknown.In present study, we found eastern Chinese prostate cancer patients had a lower incidence of TMPRSS2:ERG, TMPRSS2:ETV1 and TMPRSS2:ETV4 fusion similar to Japanese and Korean, TMPRSS2:ERG incidence is 14.3% (13/91) (95% confidence interval, 7.1%-21.5%)by biopsy specimens or 11.1% (2/18) (95% confidence interval, 0%-25.6)by radical prostatectomy samples and TMPRSS2:ETV1 and TMPRSS2:ETV4 were found in none of the cases.However, the frequency of gene fusion varies among different studies of China, and the reasons maybe complex, probably owing to the sensitivity of the technique used, the number of samples included in the study, the criteria be used to determined a positive signal and the patients from different areas (northern, southern, eastern and western) of China (Dai et al., 2008;Mao et al., 2010;Sun et al., 2010;Xiang et al., 2011;Ren et al., 2012;Wang et al., 2012).
It has already been reported that smaller studies used PCR revealed higher incidence of ERG fusion than large samples (Cerveira et al., 2006;Soller et al., 2006;Wang et al., 2006).Two investigators from China reported TMPRSS2:ERG fusion rate were 50.3% (16/30) and 53.10% (17/32) (Dai et al., 2008;Xiang et al., 2011).But a relatively small amount of samples decreasing each one's conviction.The criteria determining a positive signal for TMPRSS2:ERG fusion influenced greatly on the outcomes.There were situations that the green and red signal could be distinguished but were very close in position (Figure 3).Actually, when the green and red signal was not separated for more than two signal space, it could not be considered as a fusion positive signal pattern.What's more, only when the signal between TMPRSS2 and ERG was missing, it can be considered as fusion of deletion pattern.The lost of the other signal only means the lost of other sequence on the chromosome rather than the key one.So in Sun study, the criteria 'one yellow/one green (or one red) represented abnormal signal patterns indicative of partial deletion' may include cases which were not fusion positive (Sun et al., 2010).Using FISH technique, Wang and Sun identified TMPRSS2:ERG fusion rate of prostate cancer patients from northern and southern China was 46% and 90%, respectively (Sun et al., 2010;Wang et al., 2012).Ren used RT-PCR to detect TMPRSS2:ERG 10 out of 54 (18.5%) cases of prostate adenocarcinoma from eastern China (Ren et al., 2012).The frequency was similar to our finding (14.3%) .
Collecting biopsy specimens as study object may underestimate the incidence of fusion positivity when prostate cancer posses multifocality with heterogeneity.Report had shown about 75% prostatectomy specimens had multiple cancer foci (Meiers et al., 2007).And 41% to 67% cases may differed among individual foci in the presence of gene fusion or the mechanism of fusion (Barry et al., 2007;Mehra et al., 2007;Clark et al., 2008;Furusato et al., 2008).Miyagi found 11 cases with multi foci, 6 of which is TMPRSS2:ERG fusion positive.And 5 of the positive cases showed heterogeneity among foci (Miyagi et al., 2010).The two fusion positive radical prostatectomy cases in our study also hold multi-foci with different fusion mechanisms.Heterogeneity did exist in Asian prostate cancer patients.However, compared with an investigation of 12 core needle biopsies in USA, Whose TMPRSS2:ERG fusion prevalence was 46% (46/100), we still hold a lower incidence (Mosquera et al., 2009).
Our study also concerned different characteristics related to TMPRSS2:ERG fusion pattern.We found deletion pattern count for 86.7% (13/15) of positive cases.While insertion pattern incidence was 46.7% (7/15).Correspondingly, Mao report 5 cases of deletion pattern and 2 cases insertion pattern (Mao et al., 2010).What Lee found was 64.2% (34/53) of deletion pattern.We also observed that 5 of 13 deletion pattern cases had metastasis.Except one metastatic case harbored both deletion and insertion pattern, there was no insertion pattern accompanied with metastasis.Tomlins had suggested that only one focus in a prostate cancer case is seeding metastatic deposits, based on the observation that all metastatic foci in an individual patient are uniformly positive or negative for ETS fusion (Tomlins et al., 2009).Because our small sample and lack of study on the metastatic foci, we could not think deletion pattern related to metastasis in Chinese cases.But it is worth further study.
There were no differences between fusion positive and negative cases in the distribution of age, PSA, Gleason score and TNM stage in our study.A majority of studies suggested that age did not correlate with TMPRSS2:ERG fusion (Magi-Galluzzi et al., 2011).In some studies, TMPRSS2:ERG fusion correlated with a more aggressive clinical outcomes, but others reported the opposite results (Cerveira et al., 2006).Large population-based investigation is needed and fusion types and fusiontranscript isoforms should be included in the analysis.
In conclusion, our study reveal eastern Chinese prostate cancer patients have a significant lower incidence of TMPRSS2:ETS fusion rate consistent with other results from eastern Chinese patients and other Asian countries patients.Our study also find TMPRSS2: ERG fusion positive cases harbor more deletion pattern than insertion pattern.There is a possibility that deletion pattern may correlate with distant metastasis in eastern Chinese patients.As low incidence, it seems that the application of TMPRSS2:ERG fusion in diagnosis and treatment will be limited in eastern Chinese.But, we still need larger