Citation
Telomerase Pathway Rna Expression as a Prognostic Marker for Lethal and Metastatic Prostate Cancer

Material Information

Title:
Telomerase Pathway Rna Expression as a Prognostic Marker for Lethal and Metastatic Prostate Cancer
Creator:
Su, Fangyu
Publisher:
University of Florida
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Epidemiology
Committee Chair:
GERKE,TRAVIS A
Committee Co-Chair:
STRILEY,CATHERINE L
Committee Members:
YAGHJYAN,LUSINE

Subjects

Subjects / Keywords:
cancer
gleason
prostate
telomerarse
telomere

Notes

General Note:
Background: Current tools to predict the likelihood of prostate cancer metastasis and recurrence are highly sensitive but not specific, and may contribute to high rates of overtreatment for non-aggressive cancers. Telomerase activation has been implicated in tumor formation across several cancer types. We examined whether telomerase-related gene expression from tumor tissue at diagnosis could improve prediction accuracy for prostate cancer metastasis and mortality. We also sought to replicate a previously reported interaction between telomerase activity and TP53 expression on aggressive disease status. Methods: This study is a secondary analysis of data from 545 patients who were participants in the Mayo Clinic Radical Retropubic Prostatectomy (RRP) Registry. Cases were defined as patients who experienced clinical metastasis after prostatectomy (n=212) over 16.7 median follow-up years and controls were patients who had no clinical signs of disease progression for at least 7 years. A subset of 5 genes, TERT, TERC, TEP1, DKC1 and PINX1, were selected for this investigation on the basis of their known relationship with telomerase activation. A gene expression score was derived from these genes to assess the ability of telomerase to predict metastatic or lethal disease. The prognostic performance of each gene and a composite gene score was compared with Gleason score, and a previously identified interaction with TP53 was investigated. Results: In both crude and Gleason-adjusted analyses, DKC1 appeared as significant predictor of metastasis. In the adjusted model, each standard deviation increase in DKC1 expression increased the odds of metastasis by a factor of 2.35 (95% CI1.31, 4.32). Genes TERC, TEP, and PINX1 did not reach significance in univariable or multivariable logistic regression models. Interactions with TP53 on metastatic disease were observed for TERT and DKC1. The odds of metastatic disease increased by a factor of 3.31 (95% CI 1.40, 8.13) for each standard deviation increase in DKC1 in individual with low TP53 expression. For intermediate TP53 expression, the odds ratio was 0.42 (95% CI: 0.20, 0.90) for each standard deviation increase in TERT. Discussion: Telomerase gene TERT and DKC1 may be able to improve prognostic scoring in prostate cancer patients, particularly for those with TP53 deficient tumors.

Record Information

Source Institution:
UFRGP
Rights Management:
All applicable rights reserved by the source institution and holding location.
Embargo Date:
8/31/2018

Downloads

This item has the following downloads:


Full Text

PAGE 1

TELOMERASE PATHWAY RNA EXPRESSION AS A PROGNOSTIC MARKER FOR LETHAL AND METASTATIC PROSTATE CANCER By FANGYU SU A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2016

PAGE 2

2016 Fangyu Su

PAGE 3

To my f amily and f riends

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank my committee members Dr. Travis Gerke, Dr. Lusine Yaghjyan, and Dr. Catherine W. Striley who provided constructive advice throughout the thesis process. I extend my sincere appreciation to my adviser Dr. Gerke for his program. During the thesis preparation, Dr. Gerke has provided me with inspiring advice regarding statistical methods and thesis writing. I would l ike to thank the Department of Epidemiology. Each member of the faculty and staff have offered invaluable knowledge and experience in Public Health. I would like to thank my family who support and encourage me to pursue my goals. Their love makes me belie ve I can accomplish it all.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 6 LIST OF FIGURES ................................ ................................ ................................ .......... 7 ABSTRACT ................................ ................................ ................................ ..................... 8 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 10 Gleason Grading System ................................ ................................ ........................ 10 Telomeres and Telomerase ................................ ................................ .................... 11 Telo merase Related Genes ................................ ................................ .................... 12 Purpose of the Study ................................ ................................ .............................. 13 2 METHODS ................................ ................................ ................................ .............. 15 Gene Selection and Study Design ................................ ................................ .......... 15 Patient Population and Analysis ................................ ................................ ............. 15 Statistical Analysis ................................ ................................ ................................ .. 16 3 RESULTS ................................ ................................ ................................ ............... 18 Characteristics of Study Population ................................ ................................ ........ 18 Trends of Telomerase Genes and TP53 Expression ................................ .............. 18 Telomerase Gene Performances ................................ ................................ ............ 19 The Interaction Between Telomerase and TP53 ................................ ..................... 19 Telomerase and Prediction of Metastasis ................................ ............................... 20 4 DISCUSSION ................................ ................................ ................................ ......... 27 Limitations ................................ ................................ ................................ ............... 29 Implications ................................ ................................ ................................ ............. 29 REFERENCES ................................ ................................ ................................ .............. 31 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 36

PAGE 6

6 LIST OF TABLES Table page 1 1 Summary description of telomerase genes. ................................ ....................... 14 3 1 Characteristics of metastasis and non metastasis group ................................ ... 21 3 2 Randomly selected training and test groups. ................................ ...................... 22 3 3 Unadjusted and adjusted odds ratio for telomerase RNA by Gleason Score. .... 22 3 4 Odds ratio of interaction with TP53 of telomerase RNA in unadjusted and adjusted model. ................................ ................................ ................................ .. 23 3 5 AUC of Gleason score and telomerase related gene. ................................ ........ 24

PAGE 7

7 L IST OF FIGURES Figure page 2 1 Selected diagram. Study sorted by Cases and Controls. Patients: 1987 2001 .. 17 3 1 RNA expression distribution of telomerase genes and TP53 between case and control groups. ................................ ................................ ............................. 25 3 2 ROC curve of Gleason Score and Telomerase genes for prostate cancer metastasis.. ................................ ................................ ................................ ........ 26

PAGE 8

8 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science TELOMERASE PATHWAY RNA EXPRESSION AS A PROGNOSTIC MARKER FOR LETHAL AND METASTA TIC PROSTATE CANCER By Fangyu Su Au gust 2016 Chair: Travis Gerke Major: Epidemiology Background: Current tools to predict the likelihood of prostate cancer metastasis and recurrence are highly sensitive but not specific and may contribute to high rates of overtreatment for non aggressive cancers Telomerase activation has been implicated in tumor formation across several cancer types. We examined whether telomerase related gene expression from tumor tissue at diagnosis could improve prediction accuracy for prostate cancer metastasis and mortality We also sought to replicate a previously reported interaction between telomerase activity and TP53 expression on aggressive disease status. Methods: This study is a secondary analysis of data from 545 patients who were participants in the Mayo Clinic Radical Retropubic Prostatectomy (RRP) Registry. Cases were defined as patients who experienced clinical metastasis after prostatectomy (n=212) over 16.7 median follow up years and controls were patients who ha d no clinical signs of disease progression for at least 7 years. A subset of 5 genes TERT, TERC, TEP1, DKC1 and PINX1 were selected for this investigation on the basis of their known relationship with telomerase activation. A gene expression score was derived

PAGE 9

9 from these genes to assess the ability of telomerase to predict meta static or lethal disease The prognostic performance of each gene and a composite gene score w as compared with Gleason score, and a previously identified interaction with TP53 was investi gated Results: In both crude and Gleason adjusted analyses DKC1 appeared as significant predictor of metastasis In the adjusted model, each standard deviation increase in DKC1 expression increased the odds of metastasis by a factor of 2.35 ( 95% CI 1.31 4.32) Genes TERC, TEP, and PINX1 did not reach significance in univariable or multivariable logistic regression models I nteraction s with TP53 on metastatic disease w ere observed for TERT and DKC1 T he odds of metastatic disease increased by a factor of 3.31 (95% CI 1.4 0 8.13 ) for each standard deviation increase in DKC1 in individual with low TP53 expression For intermediate TP53 expression, the odds ratio wa s 0.42 (95% CI: 0.20 0.90) for each standard deviation increase in TERT Discussion: Telomera se gene TERT and DKC1 may be able to improve prognostic scoring in prostate cancer patients, particularly for those with TP53 deficient tumors

PAGE 10

10 CHAPTER 1 INTRODUCTION Prostate cancer is the most common non skin cancer and second leading cause of cancer related death among American men [1] The American Cancer Society estimated there will be 180,890 new cases and 26,120 deaths from prostate cancer in 2016 [2] Changes in incidence rate will depend heavily on screening guidelines and intensity; a recent study estimated that there are approximately 45 million undiagnosed prostate cancers in the US [3] Prognosis for prostate cancer is generally favorable: the five year relative overall survival is 98.9% [1] However, the five year survival of metastatic prostate cancer is 67.3% [4] The American Cancer Society also estimated that the US had 2.9 million prostate cancer survivors in 2014 and that this number will increase to 4.1 million by 2024 [5] G leason Grading System Currently, Gleason score is considered the most accurate marker for prostate cancer prognosis [6, 7] Gleason score is a grading system which is used by a pathologist to evaluate prostate tumor tissue from biopsy or surgery. Gleason score has been regarded as a reliable approach to help evaluate cancer aggressiveness and to offer a reference guide in therapy and survival prediction [8] While Gleason score has become one of the most essential tools for predicting metastasis after radical prostatectomy, it lacks specificity and may not provide clear prognostication for men diagnosed with low risk disease [9] A study has estimated that 45.8% of new diagnoses involve low risk tumor s [10] Nonetheless most prostate cancers are treated. R adical prostatectomy and external beam radiation therapy (EBRT) were commonly used [11, 12] A ctive surveillance may be a

PAGE 11

11 more desirable the rapeutic strategy for low risk prostate cancer s due Quality of Life concerns [13] While Gleason scores range from 2 to 10, a score of 5 is the lowest grade commonly used assigned after reading the biopsy samples [7] Due to poor correlations with prostatectomy grade and poor reproducibility, tumors with Gleason score below 4 are rarely considerable diagnosed [14] A Gleason score of 8 o r higher is very sensitive for aggressive disease, however, large amount of patients, nearly 68 %, are diagnosed with Gleason score <8 [7, 10] Thus, Gleason score alone may be insufficient to predict likely cancer outcomes. Th erefore, researchers are seeking novel biomarkers to better understand cancer development and to help predict tumor metastasis. By dealing with this problem, overtreatment in low risk cancer can be relieved [15] Telomere s and Telomerase Telomeres are found at the ends of chrom osomes. Studies have shown that telomere shortening is associated with age related diseases, such as late stage cancer, type 2 diabetes [16] and heart disease [17] Evidence suggests that tumor tissue has less extensive telomere shortening. By comparison, tissue specimen from normal prostate or benign prostati c hyperplasia exhibit s substantial telomere shortening [18, 19] For both normal and tumor cells, s hortened telomeres can result in chromosomal damage and instability which, in turn, can lead to cell apoptosis [20, 21] However, reactivation of telomerase makes cancer cells long lived allowing them to continuously multiply without telomere erosion [19] Telomerase is a reverse transcriptase that acts to repair and maintain telomeres. In most somatic cells, telomerase is inactivated or at low levels, but high levels of telomerase activities are detected in cancer tissue and early stages of oncogenic processes [22]

PAGE 12

12 Telomerase Related Genes There are several genes that have been implicated in telomerase activity in the literature. And the specific genes involved in telomerase activity may vary by tissue type. No gene expression signature of telomerase activity in the prostate exists. We selected a set of 5 genes TERT, TERC, TEP1, DKC1 and PINX1 [23 27] to represent the telomerase pathway. TERT (telomeras e reverse transcriptase) and TERC (telomerase RNA component) are major telomerase encoding genes [28, 29] though s tudies in most somatic cells and adult stem cells have not observed adequate expression levels of telomerase (TERT and TERC) to protect telomere length and tumor growth [30] TEP1 (Telomerase associated protein 1) and TERT were previously associated with prostate cancer risk [31] One of DKC1 (Dyskeratosis Congenital 1, dyskerin ) stabilize and maintain telomerase activity. Studies show that DKC1 is highly expressed in p rostate cancer cases, particularly in high stage and recurring cases [24] and similar trends have been observed in other cancers [32] This suggests that overexpression of DKC1 in tumor tissue might be reflective of tumor aggressiveness. PINX1 (PIN2/TERF1 Interacting, Telomerase Inhibitor 1) is a protein coding gene which is regarded as a potential telomerase inhibitor [33, 34] Even though the mechanism by which PINX1 inhibits telomerase remains unclear, PINX1 was suggested to be a tumor suppressor due to its location on chromosome 8 near microsatellite marker D8S277 [26, 33] Previous studies have shown that tumor suppressor TP53 is involved in telomer e and telomerase monitoring [35 ] and that TP53 deletion in mouse models is associated with chrom osome instability [36] and telomere dysfunction [37] When the tumor suppressor TP53 is deficient in tumors, short telomeres may cause chromosomal instability [29, 37,

PAGE 13

13 38] Telomere dysfunction that leads to telomerase reactivation has been regarded as a contributing factor to cancer development [39] Several studies have indicated tha t telomerase plays an important role in cancer development [22, 40] For example, TERT mutations in glioblastoma and thyro id cancers are associated with increased age at diagnosis [41] With the message that telomerase changes in the cancer cell occur at specific points in the oncogenic process, it is possible that telomerase could serve as a prognostic marker at diagnosis for the development of a ggressive prostate cancer. Few studies have investigated this possibility in a large cohort of prostate cancer patients with long term follow up [22] Table 1 1 describes the biological characteristics of the selected telomerase genes. Purpose of the Study There is an urgent need to improve the identification of aggressive prostate cancer beyond Gleason score. The purpose of this study was to: 1) examine the a ssociation between the expression of telomerase related genes and the risk of cancer metastasis, 2) evaluate the prognostic power of Gleason score combined with telomerase pathway gene expression; 3) examine the interacting role of tumor suppressor TP53 an d telomerase expression on metastatic progression.

PAGE 14

14 Table 1 1 Summary description of telomerase genes. Gene Type of Gene Cytogenetic band Biological process/Molecular function References TERT Protein coding gene 5p15.33 telomere maintenance, transcription, RNA templated Chang et al, 2002 [42] ; Maida et al, 2009 [43] TERC RNA gene 3q26.2 telomerase RNA component, involved in telomere maintenance Marrone et al, 2007 [44] TEP1 Protien coding gene 14q11.2 telomere maintenance via recombination, RNA dependent DNA replication Koyanagi, Y., et al, 2000 [45] ; Chang et al, 2002 [42] DKC1 Protien coding gene Xq28 telomere maintenance, rRNA processing Heiss, N. S., et al, 1999 [46] PINX1 Protein Coding gene 8p23.1 regulation of telomerase activity, negative regulation of telomerase activity Banik, S. S. and C. M. Counter, 2004 [47] ;

PAGE 15

15 CHAPTER 2 METHODS Gene S election and Study D esign RNA microarrays have been widely used to measure g ene expression profil es from a variety of cellular sources including tumor tissue samples. This study leverages data from the Affymetrix Human Exon 1.0 ST array, which provides 1.4 million probes to measure the whole gene sections for RNA. Besides TERT and TERC, the major transcriptase of t elomerase additional related genes TEP1, DKC1 and PINX1 were selected to present the telom erase activation pathway based on the lit erature [23 27] We analyzed existing data from a nested case control study that recruited prostate cancer patients from the Mayo Clinic Radical Retropubic Prostatectomy (RRP) Registry [48] Patients from the registry provided tumor tissue specimen s for molecular profiling from radical prostatectomy Raw and normalized data from this study were accessed from the publicly available Natio Expression Omnibus database (GSE46691) [9] Due to t he limited annotation regarding telomerase genes in the available normalized data, we re normalized the raw CEL file s from Gene Expression Omnibus (GEO) using the RMA algorithm implemented in the [49] Patient P opulation and A nalysis During the study, 639 patients were enrolled in the Mayo Clinic RRP Tumor Registry (GSE46691) [9] P atients received radical prostatecto my and clinical data were recorded. The patients follow up was completed through 2008. After ruling out low quality samples, 5 45 prostate cancer patients with available RNA data were included in our study. The selecting procedure [9] is shown in Figure 2 1.

PAGE 16

16 C ancer pat ients who experienced metastasis were assigned into the case group and patients who had no detectable clinical metastasis in 5 years of follow up were classified into the cont rol group. To assess external validity of the prediction models, two thirds of 545 patients (n=363) were randomly chosen for model training, and the remaining cases and controls (n=182) were assigned as a test set. To evaluate the interaction factor, ter t iles of gene expression w ere selected as a cutoff point s to divide samples into three groups St atistical A nalysis First, we computed descriptive statistics to summariz e the data and report the frequencies of metastatic events among different levels of Gleason score. L ogistic regression and Receiver Operating Characteristic (ROC) methods were used to measure the prognostic accuracy of the models To avoid the multiple comparison mistake, the P value of the tests were Bonferroni corrected. S tatistical software R 3.2.2 to normalize and analyze RNA data. Changes in the area under the curve (A UC) were measured in the test data to evaluate the added prognostic value of telomerase expression beyond Gleason score. An interaction between the t umor suppressor gene P53 and telomerase expression and cancer metastasis was also evaluated in the full dat a set [39]

PAGE 17

17 Figure 2 1 S elected diagram. Study sorted by Cases and Controls. Patients: 1987 2001 Case group ( n=212 ) Tissue, RNA Available Microarray Quality Control Mayo Clinic Radical Prostatectomy Tumor Registry N=639 n=545 Control group ( n= 333 ) Randomize d Case group ( n=135 ) Control group ( n=228 ) Training group Case group ( n=77 ) Control group ( n=105 ) Test group

PAGE 18

18 CHAPTER 3 RESULTS Characteristics of Study Population T he re were 545 participants who underwent radical prostatectomy for the treatment of primary prostate cancer and who ha d tumor tissue RNA expression quantified through microarrays. Cases were defined as those patients with distant metastasis or death from prostate cancer (n=212) and con trols survived at least 7 years with no evidence of biochemical recurrence ( BCR ) (n=333). Demographic and clinical information for this study has been described in detail elsewhere [9] ; of note, the median age of the participants was 66 years, and median follow up was16.7 years. Table 3 1 shows the c linical diagnostic patterns of the prostate cancer pati ents. The results reveal that 61.6% (n=130) of cases were described as Gleason score higher than 8, 6 cases were diagnosed with grade lower than 6, and nearly 90% of the patients in Gleason score 6 group were controls. 72% (n=195) of patients in Gleason grade 7 were controls. As the previous study shows, 52 (23.7%) and 95 (37.5%) of cases had T2 and T3/4 disease respectively. For the Prostate specific antigen (PSA) test, 106 (37.5%) cases were <10 ng/mL and 28 (21.4%) were >20 ng/mL. Table 3 2 illustrates the distributional characteristics of training and test group after the randomized reclassification. Trend s of Telomerase G ene s and TP53 E xpression Figure 3 1 illustrates the overall distribution of telomerase related gene expression level s between cases and controls. Few apparent differences appear comparing cases to controls. However, several trends appear within Gleason s trata For example, TERT and TERC are inversely related to Gleason score (Figure 3 1A and 3 1B ). One way

PAGE 19

19 analysis of va riance has shown the e xpression levels of TEP1 (Figure 3 1C) and PINX1 (Figure 3 1E) increas e with Gleason score ( P <0.01) F or the tumor suppressor TP53 (Figure 3 1F), higher expression was observed in Gleason score 6 relative to Gleason groups 7 and 8. Telomerase Gene Performance s Table 3 3 provides the odds ratio s estimated from logistic regression s f or each telomerase related gene in crude and Gleason adjusted model s In unadjusted analysis, after the Bonferroni correction we found gene DKC1 ( P <0.01) is statistically significant in predicting metastasis, with an odds ratio of 2.59 (95%CI: 1.49 4.57) for each s tandard deviation increase in expression. When DKC1 was categorized based on median score into high and low expression, the odds ratio was 1.74 (95%CI: 1.23 2.47). Results for PINX1 were suggestive in the crude model, with an odds ratio of 2.11 for each standard deviation increase in PINX1 expression Upon adjustment for Gleason score, only the DKC1 model was statistically significant ( P =0.004) with an estimated OR of 2.35 (95% CI: 1.31 4.32) for each unit increase in expression. The significance of DKC1 may suggest that a telomerase gene could complement Gleason grading for prognosis. Of note, however, primary telomerase genes TERT and TERC wer e not significant predictors in either of the models. The Interaction Between Telomerase a nd TP53 To evaluate the interaction of telomerase with TP53 on metastatic disease, we applied logistic regression to analyze the models from full dataset with and wit hout the adjustment of Gleason score. S ignificant interaction s with TP53 w ere observed for TERT and DKC1. Table 3 4 shows that, within high level s of TP53, intermediate and

PAGE 20

20 high expression of TERT has protective odds ratio s of 0.41 (95% CI: 0.19 0 .89 ) and 0.42 (95% CI: 0.20 0.90 ) respectively. In the adjusted model, intermediate level s of TERT give an odds ratio of 0.42 ( 95% CI: 0.18 0.96 ) when TP53 is in intermediate. For low levels of TP53, high expression of DKC1 produces an odds ratio of 2 .89 (9 5% CI: 1.30 6.66 ) in the crude model, and an OR of 3.31 (95% CI: 1 40 8. 1 3 ) in the Gleason adjusted model compared to low DKC1 expression For intermediate levels of TP53, gene DKC1 remains significant with an OR of 2.475 (95% CI: 1.14 5.37 ) for high expression in the crude model Full results for the remaining genes are provided in Table 3 4 Telomerase and Prediction of M etastasis To assess the area under the ROC curve (AUC), we split the prostate cancer patients into training and test sets And F igure 3 2 shows the ROC curves of each telomerase RNA genes compared with the ROC of Gleason score in the test data. The AUC value of Gleason is 0.73 (95% CI: 0.67 0.80). AUC values for TERT ( F igure 3 2A), TERC ( F igure 3 2B) and TEP1 ( F igure 3 2C) a re 0.73, 0.71 and 0.72 when respectively added to Gleason DKC1 (0.75) and PINX1 (0.75) are slightly better than Gleason score alone. The ROC of the integrated model which was a linear combination of all five telomerase genes was slightly lower than Glea son grade. Table 3 5 illustrates the AUC of each Gleason adjusted telomerase gene model and its comparison with the AUC of Gleason alone. In each model, each telomerase was combined with Gleason score and for the last model, all telomerase genes were linearly combined with Gl eason. There are no significant differences between Gleason only model and those with telomerase added.

PAGE 21

21 Table 3 1 Characteristics of metastasis and non metastasis group Cases Controls N (row %) N ( row %) Pathological Stage* pT2N0M0 52 (23.7) 167 (76.3) pT3/4N0M0 95 (37.5) 158 (62.5) pTanyN+M0 45 (61.6) 28 (38.4) P athologic Gleason Score 6 6 (9.5) 57 (90.5) 7 76 (28) 195 (72) 130 (61.6) 81 (38.4) Pre operative Prostate specific Antigen* <10 ng/mL 106 (37.5) 177 (62.5) 10 20 ng/mL 31 (26.7) 85 (73.3) >20 ng/mL 28 (21.4) 103 (78.6) Not available 6 (40) 9 (60) the distribution data is from the paper Erho, N., et al. [9]

PAGE 22

22 Table 3 2 Randomly selected training and test groups Training group Test group Cases Controls Cases Controls Gleason Category N N (row %) N (row %) N N (row %) N (row %) 6 47 5 (10.6) 42 (89.4) 16 1 (6.3) 15 (93.7) 7 134 79 (59.0) 55 (41.0) 77 51 (66.2) 26 (33.8) 182 51 (28.0) 131 (72.0) 89 25 (28.1) 64 (71.9) Table 3 3 Unadjusted and adjusted odds ratio for telomerase RNA by Gleason Score. Unadjusted Adjusted OR(95%CI) P value OR(95%CI) P value TERT 0.34 (0.11 1.06) 0.322 0.66 (0.196 2.20) 1.000 TERC 0.68 (0.42 1.10) 0.605 0.89 (0.53 1.50) 1.000 TEP1 3.24 (0.94 11.35) 0.322 1.54 (0.41 5.89) 1.000 DKC1 2.59 (1.49 4.57) 0.004 2.35 (1.31 4.32) 0.024 PINX1 2.11 (1.20 3.73) 0.05 1.35 (0.73 2.51) 1.000 *The P values were Bonferroni corrected.

PAGE 23

23 Table 3 4 Odds ratio of interaction with TP5 3 of telomerase RNA in unadjusted and adjusted model. TP53 low TP53 intermediate TP53 high Unadjusted Adjusted Unadjusted Adjusted Unadjusted Adjusted OR(95%CI) OR(95%CI) OR(95%CI) OR(95%CI) OR(95%CI) OR(95%CI) TERT Low 1 [ Reference ] 1 [ Reference ] 0.72 ( 0.35,1.46 ) 0.74 ( 0.35,1.59 ) 1.00 ( 0.49,2.03 ) 1.28 ( 0.60,2.75 ) Intermediate 1.00 (0 .49,2.04 ) 0.19 (0.55,2.57) 1.08 ( 0.52,2.22 ) 1.34 ( 0.61,2.94 ) 0.41 ( 0.19,0.89 ) 0.42 ( 0.18,0.96 ) High 0.51 (0.24,1.06) 0.56 ( 0.25,1.22 ) 0.92 ( 0.43,1.95 ) 1.31 ( 0.58,3.01 ) 0.42 ( 0.20,0.90 ) 0.47 ( 0.21,1.07 ) TERC Low 1 [ Reference ] 1 [ Reference ] 0.84 ( 0.40,1.78 ) 0.83 ( 0.37,1.84 ) 0.69 ( 0.35,1.37 ) 0.79 ( 0.38,1.63 ) Intermediate 0.68 ( 0.34,1.38 ) 0.79 ( 0.37,1.68 ) 0.91 ( 0.42,1.93 ) 1.30 ( 0.57,2.97 ) 0.61 ( 0.28,1.31 ) 0.84 ( 0.37,1.93 ) High 0.70 ( 0.34,1.45 ) 0.84 ( 0.38,1.82 ) 0.67 ( 0.31,1.45 ) 0.94 ( 0.41,2.19 ) 0.71 ( 0.34,1.46 ) 0.93 ( 0.42,2.03 ) DKC1 Low 1 [ Reference ] 1 [ Reference ] 1.03 ( 0.42,2.50 ) 1.24 ( 0.48,3.23 ) 1.13 ( 0.52,2.53 ) 1.42 ( 0.61,3.37 ) Intermediate 1.86 ( 0.85,4.21 ) 1.76 ( 0.76,4.20 ) 1.44 ( 0.66,3.16 ) 1.84 ( 0.78,4.33 ) 1.09 ( 0.51,2.36 ) 1.19 ( 0.52,2.74 ) High 2.89 ( 1.30,6.66 ) 3.31 ( 1.40,8.13 ) 2.4 8 ( 1.14,5.37 ) 1.90 ( 0.83,4.37 ) 1.29 ( 0.63,2.63 ) 1.28 ( 0.59,2.78 ) TEP1 Low 1 [ Reference ] 1 [ Reference ] 0.75 ( 0.33,1.72 ) 0.95 ( 0.39,2.34 ) 0.49 ( 0.22,1.06 ) 0.65 ( 0.28,1.50 ) Intermediate 0.81 ( 0.37,1.79 ) 0.85 ( 0.36,2.01 ) 1.46 ( 0.70,3.04 ) 1.52 ( 0.69,3.39 ) 1.94 ( 0.92,4.10 ) 1.62 ( 0.72,3.62 ) High 1.14 ( 0.53,2.49 ) 1.10 ( 0.48,2.55 ) 0.94 ( 0.44,2.02 ) 0.70 ( 0.31,1.62 ) 1.43 ( 0.68,3.01 ) 1.38 ( 0.62,3.08 ) PINX1 Low 1 [ Reference ] 1 [ Reference ] 1.57 ( 0.62,4.16 ) 1.39 ( 0.50,3.97 ) 0.99 ( 0.34,2.92 ) 1.17 ( 0.37,3.79 ) Intermediate 2.01 ( 0.71,5.95 ) 1.83 ( 0.59,5.89 ) 1.57 ( 0.25,9.74 ) 6.81 ( 1.60,28.91 ) 3.01 ( 0.86,10.53 ) 2.98 ( 0.76,11.60 ) High 1.28( 0.49,3.46 ) 1.14 ( 0.40,3.31 ) 0.75 ( 0.22,2.55 ) 0.74 ( 0.20,2.83 ) 0.90 ( 0.27,2.96 ) 0.56 ( 0.15,2.04 ) p < .05; ** p < 0.01; *** p < .001 based on multinomial logistic regression results.

PAGE 24

24 Table 3 5 AUC of Gleason score and telomerase related gene. AUC (95% CI) P value Gleason score 0.7346 (0.6667 0.8025) TERT + Gleason score 0.7349 (0.6605 0.8094) 0.981 TERC + Gleason score 0.7132 (0.636 0.7903) 0.159 TEP1 + Gleason score 0.7223 (0.6461 0.7985) 0.449 DKC1 + Gleason score 0.7491 (0.6767 0.8216) 0.317 PINX1 + Gleason score 0.7478 (0.6748 0.8208) 0.399 All telomerase Adjusted 0.7284 (0.6532 0.8036) 0.701

PAGE 25

25 Figure 3 1 RNA expression distribution of telomerase genes and TP53 between case and control group s

PAGE 26

26 Figure 3 2 ROC curve of Gleason Score and Telomerase genes for prostate cancer metastasis. (a) ROC curve of TERT ROC of TERC adjusted (AUR curve=0.7132) vs. Gleason; (c) ROC of TEP1 adjusted (AUR curve=0.7223) vs. Gleason; (d) ROC of DKC1 adjusted (AUR curve=0.7491 ) vs. Gleason ; (e) ROC of PINX1 adjusted (AUR curve=0.7478 ) vs. Gl eason; and(f) ROC of all telomerase adjusted (AUR curve=0.7284 ) vs. Gleason

PAGE 27

27 CHAPTER 4 DISCUSSION The purpose of t his study was to assess the hypothesis that telomerase gene expression in primary prostate cancer has the capability to predict the development of clinical cancer metastasis. The secondary purpose wa s to examine the previously reported interaction between the tumor suppressor TP53 and telomerase related gene expression. In addition to test s of association thro ugh logistic regressions we evaluated whether telomerase markers could improve the discrimination of Gleason score by evaluating changes in AUC. Our findings are 1) most selected telomerase genes did not significantly improve the prediction of primary pro state cancer metastasis; 2) high DKC1 expression was a significant ly predicted prostate cancer progression ; and effect on metastasis was enhanced under low TP53 expression. The TP53 telomerase interaction we observed is consistent with previous a nimal studies [27, 37, 39] When TP53 was knocked out in prostate tumors, it revealed an increase in telomere structural instability and telomerase reactivation [39] The present study showed that, in lower levels of TP53 expression, high expression of telomerase related genes like DKC1 were correlated with adverse outcomes And under highly expressed TP53, the effects of other telomerase gene s on outcome were insignificant. Our finding indicates that TERT may play as a protective role in lethal and metastatic prostate cancer. From the expression trends, TERT is dec reased with cancer progression We hypothesize that at prostate cancer initiation TERT is highly expressed in tumor cell. However, expression is reduced by tumor multiplication. interaction ma y have inhibit ory effect s on TERT expression f ollowing tumorigenesis This hypothesis is aligned with previous studies in vivo which proved

PAGE 28

28 that the reactivation of TERT occurs in most human tumors [50] and that TERT Catenin was one of the hallmarks of oncogenesis [22] The with other genes or factors in the human environment are complex. In the present study, TERT expression did not d iffer by Gleason score Beyond the explanation that this may result from the complex molecular environmental factors that monitor telomere expression, it may also imply that TERT serves as a mechanism of mediation in cancer progression. DKC1 was observed to deliver significant prognostic value in this study. This prostate cancer [24] particularly in high grade and recurrent cancer. Thus, DKC1 overexpression has poten tial capability to assist predicting metastasis as a biomarker. Moreover, as a gene to maintain telomere stability, the mechanism of DKC1 in cancer cell implies potential as a therap eutic marker [51] We found that even though Gleason scor e is the most accurate prognostic marker in use, the AUC of 0.735 shows that as a predictive model, the Gleason score leave s room for improvement Several previous studies have aimed to discover gene expression biomarkers in order to fulfil l the need to pr edict aggressive prostate cancer development [9, 52 54] Since Gleason score is graded based on the pathologic patterns of cancer tissue, gene expression and protein activities are likely to affect the medical prognosis of prostate cancer. Therefore, gene expression signature s may measure much of the same information as Gleason score grading.

PAGE 29

29 One considerable strength of the present study is the large sample size and long follow up period. Since most previous studies which focused on the di scovery of telomerase were animal based a complete human analysis may be a robust and external ly reliable source for testing the association of telomerase expression. Limitation s T h is study has several limitations. First, the original study does not offer sufficient demographic and clinical information regarding the recruited patients for further control and stratification in our analysis. Such stratification could have examined the interaction s of genes as well as the predicted power of cancer diagn osis. Second, the up time period after they had been received radical prostatectomy. The clinical and medication therapies other than prostatectomy which have not be en recorded in the aggressive cancer recurrence. Nevertheless, the sensitivity of therapy reaction of prostate cancer is not as high as other type of cancers such as breast ca ncer [55] so the effect of adjuvant therapies may be minimal Moreover, the patients were all treated with radical prostatectomy although the grade of biomedical indicators may not telomerase expression may be impacted by other unknown RNA networks. Implications A primary finding of t he present study concerns the significant prognostic value of DKC1 for prostate cancer progression Further research may characterize the activity of D KC1 in the prostate in order to better improve the accuracy of prostate cancer prognosis and clarify its potential clinical role Moreover, telomerase gene expression

PAGE 30

30 might not enough to represent the telomerase activities in cancer development. A protein measurement to discover the roles of telomerase in cancer metastasis would be a valuable complementary effort to examine the hypothesis. In summary the present study discovered an association between telomerase gene expression and cancer metastasis. We provided evidence that gene DKC1 may be associated with early metastasis and demonstrated its interaction with the tumor suppressor TP53. It provided a progressive step to discover the development of predicting prostate cancer recurrence, yet more studies are needed to focus on the sensitivity and specificity improvement in prediction of prostate cancer metastasis, and influences of risk and protective factors.

PAGE 31

31 REFERENCE S 1. SEER Cancer Statistics Factsheets: Prostate Cancer. National Cancer Institute. Bethesda, MD, http://seer.cancer.gov/stat facts/html/prost.html 2. Siegel, R.L., K.D. Miller, and A. Jemal, Cancer statistics, 2016. CA: A Cancer Journal for Clinicians, 2016. 66 (1): p. 7 30. 3. Jahn, J.L., E.L. Giovannucci, and M.J. Stampfer, The high prevalence of undiagnosed prostate cancer at autopsy: implications for epidemiology and treatment of prostate cancer in the Prostate specific Antigen era. Int J Cancer, 2015. 137 (12): p. 2795 802. 4. Antonarakis, E.S., et al., The natural history of metastatic progression in men with prostate specif ic antigen recurrence after radical prostatectomy: long term follow up. BJU International, 2012. 109 (1): p. 32 39. 5. Society, A.C., Cancer Treatment and Survivorship Facts & Figures 2014 2015. Atlanta: American Cancer Society, 2014. 6. Epstein, J.I., et a l., The 2005 International Society of Urological Pathology (ISUP) Consensus Conference on Gleason Grading of Prostatic Carcinoma. Am J Surg Pathol, 2005. 29 (9): p. 1228 42. 7. Pierorazio, P.M., et al., Prognostic Gleason grade grouping: data based on the m odified Gleason scoring system. BJU Int, 2013. 111 (5): p. 753 60. 8. King, C.R., et al., Extended prostate biopsy scheme improves reliability of Gleason grading: implications for radiotherapy patients. Int J Radiat Oncol Biol Phys, 2004. 59 (2): p. 386 91. 9. Erho, N., et al., Discovery and Validation of a Prostate Cancer Genomic Classifier that Predicts Early Metastasis Following Radical Prostatectomy. PLoS ONE, 2013. 8 (6): p. e66855. 10. Cooperberg, M.R., et al., Contemporary trends in low risk prostate ca ncer: risk assessment and treatment. J Urol, 2007. 178 (3 Pt 2): p. S14 9. 11. Carter, H.B., et al., Gleason score 6 adenocarcinoma: should it be labeled as cancer? Journal of Clinical Oncology, 2012. 30 (35): p. 4294 4296. 12. Cooperberg, M.R., et al., The changing face of low risk prostate cancer: trends in clinical presentation and primary management. Journal of Clinical Oncology, 2004. 22 (11): p. 2141 2149. 13. Hayes, J.H., et al., Active surveillance compared with initial treatment for men with low risk prostate cancer: A decision analysis. JAMA, 2010. 304 (21): p. 2373 2380.

PAGE 32

32 14. Epstein, J.I., et al., The 2005 International Society of Urological Pathology (ISUP) consensus conference on Gleason grading of prostatic carcinoma. The American journal of surgic al pathology, 2005. 29 (9): p. 1228 1242. 15. Cooperberg, M.R., J.M. Broering, and P.R. Carroll, Time trends and local variation in primary treatment of localized prostate cancer. Journal of Clinical Oncology, 2010. 28 (7): p. 1117 1123. 16. Salpea, K.D., et al., Association of telomere length with type 2 diabetes, oxidative stress and UCP2 gene variation. Atherosclerosis, 2010. 209 (1): p. 42 50. 17. Cawthon, R.M., et al., Association between telomere length in blood and mortality in people aged 60 years or o lder. The Lancet, 2003. 361 (9355): p. 393 395. 18. Shay, J.W., Are Short Telomeres Predictive of Advanced Cancer? Cancer discovery, 2013. 3 (10): p. 1096 1098. 19. Fras, C., et al., Telomere shortening is associated with poor prognosis and telomerase activity correlates with DNA repair impairment in non small cell lung cancer. Lung Cancer. 60 (3): p. 416 425. 20. Shay, J.W. and W.E. Wright, Role of telomeres and telomerase in cancer. Semin Cancer Biol, 2011. 21 (6): p. 349 53. 21. Zhang, X., et al., Telo mere shortening and apoptosis in telomerase inhibited human tumor cells. Genes Dev, 1999. 13 (18): p. 2388 99. 22. Hoffmeyer, K., et al., Wnt/beta catenin signaling regulates telomerase in stem cells and cancer cells. Science, 2012. 336 (6088): p. 1549 54. 2 3. Cao, Y., T.M. Bryan, and R.R. Reddel, Increased copy number of the TERT and TERC telomerase subunit genes in cancer cells. Cancer Science, 2008. 99 (6): p. 1092 1099. 24. Sieron, P., et al., DKC1 overexpression associated with prostate cancer progression Br J Cancer, 2009. 101 (8): p. 1410 1416. 25. Li, D. M. and H. Sun, TEP1, encoded by a candidate tumor suppressor locus, is Cancer research, 1997. 57 (11): p. 2124 2129. 26. Ha wkins, G.A., et al., Mutational analysis of PINX1 in hereditary prostate cancer. The Prostate, 2004. 60 (4): p. 298 302. 27. Bolton, E.M., et al., Noncoding RNAs in prostate cancer: the long and the short of it. Clin Cancer Res, 2014. 20 (1): p. 35 43.

PAGE 33

33 28. F lores, I., R. Benetti, and M.A. Blasco, Telomerase regulation and stem cell behaviour. Current Opinion in Cell Biology, 2006. 18 (3): p. 254 260. 29. Blasco, M.A., Telomeres and human disease: ageing, cancer and beyond. Nat Rev Genet, 2005. 6 (8): p. 611 22. 30. Harrington, L., et al., Human telomerase contains evolutionarily conserved catalytic and structural subunits. Genes Dev, 1997. 11 (23): p. 3109 15. 31. Gu, C., et al., Genetic variants in the TEP1 gene are associated with prostate cancer risk and recur rence. Prostate Cancer Prostatic Dis, 2015. 18 (4): p. 310 6. 32. Montanaro, L., et al., Dyskerin expression influences the level of ribosomal RNA pseudo uridylation and telomerase RNA component in human breast cancer. J Pathol, 2006. 210 (1): p. 10 8. 33. Z hou, X.Z. and K.P. Lu, The Pin2/TRF1 Interacting Protein PinX1 Is a Potent Telomerase Inhibitor. Cell, 2001. 107 (3): p. 347 359. 34. Diotti, R. and D. Loayza, Shelterin complex and associated factors at human telomeres. Nucleus, 2011. 2 (2): p. 119 135. 35. Wu, S., et al., Telomerase reverse transcriptase gene promoter mutations help discern the origin of urogenital tumors: a genomic and molecular study. Eur Urol, 2014. 65 (2): p. 274 7. 36. Veronese, L., et al., Telomeres and chromosomal instability in chron ic lymphocytic leukemia. Leukemia, 2013. 27 (2): p. 490 493. 37. Begus Nahrmann, Y., et al., p53 deletion impairs clearance of chromosomal instable stem cells in aging telomere dysfunctional mice. Nat Genet, 2009. 41 (10): p. 1138 43. 38. Londoo Vallejo, J. A., Telomere instability and cancer. Biochimie, 2008. 90 (1): p. 73 82. 39. Ding, Z., et al., Telomerase reactivation following telomere dysfunction yields murine prostate tumors with bone metastases. Cell, 2012. 148 (5): p. 896 907. 40. Rane, J.K., et al., Telomerase Activity and Telomere Length in Human Benign Prostatic Hyperplasia Stem like Cells and Their Progeny Implies the Existence of Distinct Basal and Luminal Cell Lineages. Eur Urol, 2015. 41. Killela, P.J., et al., TERT promoter mutations occur freq uently in gliomas and a subset of tumors derived from cells with low rates of self renewal. Proc Natl Acad Sci U S A, 2013. 110 (15): p. 6021 6.

PAGE 34

34 42. Chang, J.T., et al., Differential regulation of telomerase activity by six telomerase subunits. Eur J Bioche m, 2002. 269 (14): p. 3442 50. 43. Maida, Y., et al., An RNA dependent RNA polymerase formed by TERT and the RMRP RNA. Nature, 2009. 461 (7261): p. 230 5. 44. Marrone, A., et al., Functional characterization of novel telomerase RNA (TERC) mutations in patien ts with diverse clinical and pathological presentations. Haematologica, 2007. 92 (8): p. 1013 20. 45. Koyanagi, Y., et al., Telomerase activity is down regulated via decreases in hTERT mRNA but not TEP1 mRNA or hTERC during the differentiation of leukemic c ells. Anticancer Res, 2000. 20 (2a): p. 773 8. 46. Heiss, N.S., et al., Dyskerin localizes to the nucleolus and its mislocalization is unlikely to play a role in the pathogenesis of dyskeratosis congenita. Hum Mol Genet, 1999. 8 (13): p. 2515 24. 47. Banik, S.S. and C.M. Counter, Characterization of interactions between PinX1 and human telomerase subunits hTERT and hTR. J Biol Chem, 2004. 279 (50): p. 51745 8. 48. Nakagawa, T., et al., A Tissue Biomarker Panel Predicting Systemic Progression after PSA R ecurrence Post Definitive Prostate Cancer Therapy. PLoS ONE, 2008. 3 (5): p. e2318. 49. Lockstone, H.E., Exon array data analysis using Affymetrix power tools and R statistical software. Brief Bioinform, 2011. 12 (6): p. 634 44. 50. Su, Z., et al., Telomeras e mRNA transfected dendritic cells stimulate antigen specific CD8+ and CD4+ T cell responses in patients with metastatic prostate cancer. The Journal of Immunology, 2005. 174 (6): p. 3798 3807. 51. Ng, S.S.M., et al., A novel glioblastoma cancer gene therap y using AAV mediated long term expression of human TERT C terminal polypeptide. Cancer Gene Ther, 2007. 14 (6): p. 561 572. 52. Sboner, A., et al., Molecular sampling of prostate cancer: a dilemma for predicting disease progression. BMC Medical Genomics, 20 10. 3 (1): p. 1 12. 53. Andren, O., et al., How Well Does the Gleason Score Predict Prostate Cancer Death? A 20 Year Followup of a Population Based Cohort in Sweden. The Journal of Urology, 2006. 175 54. Bibikova, M., et al., Expression signatures that cor related with Gleason score and relapse in prostate cancer. Genomics, 2007. 89 (6): p. 666 72.

PAGE 35

35 55. Million Women Study, C., Breast cancer and hormone replacement therapy in the Million Women Study. The Lancet, 2003. 362 (9382): p. 419 427.

PAGE 36

36 BIOGRAPHICAL SKETCH the Department of Public Administration at Beijing University of Chinese Medicine, China. He came to the University of Florida to pursue his Master of Science degree in the Department of Epidemiology. His r esearch focuses on the genetic e pidemiology and prostate cancer.