Expression of biomarkers modulating prostate cancer progression ...

5 downloads 0 Views 186KB Size Report
Prostate Cancer and Prostatic Diseases (2005) 8, 54–59. doi:10.1038/sj.pcan.4500768. Published online 18 January 2005. Keywords: Bcl-2; COX-2; MVD; ...
Prostate Cancer and Prostatic Diseases (2005) 8, 54–59 & 2005 Nature Publishing Group All rights reserved 1365-7852/05 $30.00

www.nature.com/pcan

Expression of biomarkers modulating prostate cancer progression: implications in the treatment of the disease G Di Lorenzo1*, S De Placido1, R Autorino2, M De Laurentiis1, C Mignogna3, M D’Armiento3, G Tortora1, G De Rosa3, M D’Armiento2, M De Sio1, AR Bianco1 & FP D’Armiento3 1

Dipartimento di Endocrinologia e Oncologia Molecolare e Clinica, Universita` degli Studi di Napoli Federico II, Naples, Italy; 2Cattedra di Urologia, Seconda Universita` degli Studi di Napoli, Naples, Italy; and 3Dipartimento di Scienze Biomorfologiche e Funzionali, Universita` degli Studi di Napoli Federico II, Naples, Italy

Objectives: To determine whether COX-2, bcl-2 and neoangiogenesis are related to human prostate cancer relapse after definitive surgical treatment and progression toward androgen independence and to evaluate the association between the patterns of these tumoral biomarkers and other standard clinico-pathological parameters (such as Gleason score, PSA, TNM stage). Materials and Methods: We retrospectively analyzed the records on 126 prostate cancer samples from patients treated at our University Hospital from 1995 to 2002. The 72 patients with clinically localized disease (group 1) had undergone radical prostatectomy. Another 54 patients (group 2) had metastatic androgen-independent disease. Archived material relating to the subjects was then immunostained for bcl-2, COX-2 and CD-31, using an anti-bcl-2 monoclonal primary antibody, an antiCOX-2 polyclonal rabbit antibody and an anti-CD-31 monoclonal mouse antibody to evaluate neoangiogenesis (MVD, microvessel density). Results: We found that bcl-2, COX-2 and MVD expression increased from group 1 to group 2. The intergroup difference was significant only for high MVD (Po0.05). On the other hand, high MVD, high bcl-2 and high COX-2 expression was correlated with a higher PSA level (Po0.01), whereas only a high MVD was also related with Gleason score (Po0.05). We used univariate analysis to evaluate the prognostic impact of biologic and clinico-pathologic parameters on the diseasefree-survival of 72 patients treated by radical prostatectomy. A total of 30 patients (41.6%) experienced biochemical relapse; bcl-2, COX-2 and MVD significantly correlated with disease relapse in these patients. In fact, we observed disease relapse in 24/45 (53%) with high bcl-2 expression, in 15/21 (71%) with a high MVD count and finally, in 30/58 (52%) with high COX-2 expression. Finally, PSA value and Gleason score were the only two biologic markers significantly associated to disease relapse in a multivariate analysis. Conclusions: Our results strongly support a role for bcl-2, COX-2 and angiogenesis in the development and progression of prostate cancer. Of course, we are aware of the small sample size considered in our study. Further investigations would better clarify the prognostic and therapeutic implications of these findings. Prostate Cancer and Prostatic Diseases (2005) 8, 54–59. doi:10.1038/sj.pcan.4500768 Published online 18 January 2005

Keywords: Bcl-2; COX-2; MVD; prognostic factors

*Correspondence: G Di Lorenzo, Dipartimento di Endocrinologia e Oncologia Molecolare e Clinica, Universita` degli Studi di Napoli Federico II, Naples, Italy. E-mail: [email protected] Received 13 July 2004; revised 18 August 2004; accepted 15 September 2004; published online 18 January 2005

Introduction Prostate cancer is characterized by an initial period during which the tumor growth is androgen dependent and localized disease responds to a variety of therapies (radical prostatectomy, radiotherapy and brachytherapy).

Prognostic factors in prostate cancer G Di Lorenzo et al

Up to 56% of clinically localized prostate cancers are associated with disease recurrence after radical prostatectomy1 and an accurate prediction of the patients most likely to have recurrence is mandatory when considering adjuvant therapy. Several immunohistochemical markers have been evaluated for prognostic significance in surgically treated, clinically localized prostate cancer. Advanced prostate cancer initially responds to hormone therapy, but the response usually lasts only about 18 months in the metastatic setting. After the failure of first-line hormone therapy, standard treatment options are limited and survival is between 6 and 12 months for patients with androgen-independent hormone-refractory prostate cancer (HRPC).2 Currently available chemotherapies do not confer a significant survival benefit in these cases.3 As a result, there is a demand for new therapies that more specifically target the cellular events involved in the development of malignancy and in normal host processes required for tumor progression. The development of prostate cancer, the progression from a normal prostate epithelium to androgen-dependent cancer and, eventually, to HRPC is a multistep process involving a series of changes in the function of various growth regulatory signals. Several peptide growth factors and their receptors have been implicated in the molecular mechanisms underlying the development of androgen independence.4,5 In this context, transduction of signals from the cell surface to the nucleus that regulate cell growth, differentiation and survival and become subverted during carcinogenesis and progression provides a particularly attractive target. These signal transduction pathways depend on a vast array of proteins, many of which possess tyrosine kinase activity.6 We have recently demonstrated that EGFR expression is significantly associated with disease progression in patients treated with radical prostatectomy. In a Cox multivariate analysis in 58 prostate cancer patients, we found that EGFR expression had an independent prognostic effect on disease-free survival (DFS), considering the PSA value, the stage of disease and the Gleason score as other variables.7 Other molecular mechanisms, for example, activation of bcl-2 and COX-2 and increased angiogenesis, have also been implicated in the progression to androgen independence in human prostate cancer. Bcl-2 is an antiapoptotic mediator that has been found to be involved in the molecular biology of a wide range of human cancers since its discovery as an oncoprotein implicated in human follicular lymphoma, almost 15 years ago.8 It belongs to a large family of apoptosisregulatory proteins that act as both antagonists and agonists of programmed cell death. Selective dimerization between pairs of these proteins determines how a cell responds to an apoptotic signal.9 Bcl-2 is generally overexpressed in androgen-independent prostate cancer.10 Experimental and clinical studies have suggested that increased bcl-2 expression confers androgen independence and hormonal treatment resistance.11,12 In preclinical prostate cancer models, treatment with bcl-2 antisense oligonucleotides inhibited bcl-2 expression, delayed the development of androgen independence and enhanced chemotherapy-induced apoptosis.13–15 Furthermore, it has been suggested that bcl-2 affects

tumorigenesis also by stimulating tumor-induced angiogenesis.16 The growth and dissemination of solid tumors depend on angiogenesis.17,18 In prostate cancer, microvessel density (MVD), as a marker of angiogenesis, correlates with metastatic disease and with overall survival.19 Various potential angiogenic factors are expressed in prostate tumors, including epidermal growth factor (EGF), basic fibroblast growth factor (bFGF), plateletderived growth factor (PDGF), transforming growth factors (TGF-alfa and TGF-beta) and vascular endothelial growth factor (VEGF).20–22 Cyclooxygenase-2 (COX-2) is the inducible form of the enzyme that converts arachidonic acid to prostaglandins.23 A variety of mechanisms have been proposed to explain how COX-2 upregulation affects carcinogenesis, that is, by promoting angiogenesis,24 by enhancing cellular motility25 and by increasing cancer cell resistance to apoptosis.26 Usually, COX-2 expression is either weak or absent in normal prostate tissue and overexpressed in prostate cancer cells.27–29 However, in a recent study, COX-2 was upregulated in proliferative inflammatory atrophy of the prostate but not in prostate cancer.30 The aims of the present study were to determine whether COX-2, bcl-2 and neo-angiogenesis are related to human prostate cancer relapse after definitive surgical treatment and progression toward androgen independence and to evaluate the association between the patterns of these tumoral biomarkers and other standard clinico-pathological parameters (such as Gleason score, PSA, TNM stage).

55

Materials and methods Patients We retrospectively analyzed the records on 126 prostate cancer samples from patients treated at our University Hospital from 1995 to 2002. The 72 patients with clinically localized disease (group 1) had undergone radical prostatectomy. Another 54 patients (group 2) had metastatic androgen-independent disease. The median follow-up for group 1 was 73 months. A cutoff value of 7 distinguished between low and high Gleason scores. DFS in group 1 was defined as the time from radical surgery to the detection of an increase in serum PSA above 0.4 ng/ml. Three consecutive increases of PSA were required to confirm disease relapse, but the time-torelapse is the time when elevated PSA was first detected.31

Immunohistochemistry and evaluation of immunoperoxidase staining Specimens were fixed in 10% neutral-buffered formalin, paraffin-embedded and processed routinely; 5-mm-thick serial sections were placed on poly-l-lysine-coated slides. The slides were then treated for 30 min at 201C with methanol containing 0.3% hydrogen peroxide to block endogenous peroxidase activity. We used an anti-bcl-2 monoclonal primary antibody (clone 124; Dako, Milan, Italy) to evaluate bcl-2 expression. We used an anti-COX-2 Prostate Cancer and Prostatic Diseases

Prognostic factors in prostate cancer G Di Lorenzo et al

56

polyclonal rabbit antibody (H-62, Santa Cruz Biotechnology, California, USA) at a dilution of 1 : 200 in PBS to evaluate COX-2 expression. Four pathologists (FD’A, CM, MD’A, GDR) who had no knowledge of clinical data independently evaluated immunostaining. A total of 100 consecutive malignant cells were counted in the area where staining was most intense to identify a ‘hot spot’. The percentage of positive tumor cells was graded as: 0, none; þ , 5–30%; þ þ , 31–60%; þ þ þ , above 60%. Immunostaining intensity was rated as: 0, none; þ , low expression; þ þ / þ þ þ , high expression. We used an anti-CD-31, monoclonal mouse antibody (clone JC/70A, Biogenex, California, USA) to evaluate neoangiogenesis. MVD within the designated hot spot was scored on a  200 light microscopic high-power field. Microvessels stained with CD-31 were counted in five adjacent highpower fields, according to a previously described method,20 and the average score of the five areas was recorded as: 0, none; þ , 1–45 neovessels (low); þ þ , 46– 90 neovessels (moderate); þ þ þ , 91–141 neovessels (high) (Figure 1).

Statistical analysis The association between Gleason score, PSA, T status, N status, bcl-2 and COX-2 expression and neoangiogenesis was evaluated with contingency tables. Statistical significance was evaluated with the w2 test. Univariate analysis of the role of each prognostic variable on DFS was estimated with the Kaplan–Meier product limit method.32 The statistical significance of the differences in DFS distribution among prognostic groups was evaluated with the log-rank test.33 The Cox proportional hazards regression model34 was used to determine which parameters (Gleason score, T status, N status, bcl-2, COX-2, angiogenesis) exerted an independent effect on DFS. The relative hazards (RH) with 95% confidence intervals (CI) were estimated. All P-values represent two-sided tests of statistical significance.

P-values p0.05 were considered significant. All analyses were performed with the BMDP New System statistical package version 1.0 for Microsoft Windows (BMDP Statistical Software, Los Angeles, CA, USA).

Results Table 1 summarizes the clinico-pathological characteristics of the 126 patients with prostate cancer. As shown in Table 2, in group 1 patients, bcl-2 staining was intense in 45/72 (62.5%) samples, COX–2 staining was intense in 58/72 (78%) samples and MVD was high in 21/72 (29%) specimens while in group 2 patients, the results for bcl-2, COX-2 and MVD intense staining were 38/54 (70%), 45/54 (83%) and 34/54 (63%) samples, respectively. We next evaluated the correlation among the various biologic parameters and clinico-pathologic characteristics in the 126 prostate cancer patients. As shown in Table 3, high bcl-2 (Figure 1), COX-2 (Figure 2) and MVD Table 1 patients

Clinicopathologic characteristics of prostate cancer Group 1

Group 2

72

54

38 34

25 29

PSA-values (ng/ml): p10 410

24 48

16 38

TNM T2N0 T3/T4N+ M+

51 21 0

0 0 54

No. pts. Gleason score o7 X7

As described in Materials and methods group 1 patients were treated with radical prostatectomy, whereas group 2 patients had hormone-refractory metastatic disease. PSA values were the serum PSA values at the diagnosis.

Table 2 patients

Bcl-2, COX-2 and MVD expression in prostate cancer Group 1 (total 72 pts.)

Figure 1 High expression of bcl-2. Prostate Cancer and Prostatic Diseases

Group 2 (total 54 pts.)

Bcl-2 Negative Low expression (+) High expression (++/+++)

4 (5.5%) 23 (32%) 45 (62.5%)

0 (0%) 16 (30%) 38 (70%) P ¼ ns

COX-2 Negative Low expression (+) High expression (++/+++)

1 (1%) 15 (21%) 56 (78%)

0 (0%) 9 (17%) 45 (83%) P ¼ ns

MVD 0 1–45 (low) 46–90 (moderate) 91–131 (high)

0 21 30 21

0 4 16 34

(0%) (29%) (42%) (29%)

(0%) (7%) (30%) (63%) Po0.05

Bcl-2, COX-2 and MVD immunostaining were defined as described in Materials and methods. P ¼ statistical differences between the three group. P-values p0.05 were considered significant. ns ¼ not significant.

Prognostic factors in prostate cancer G Di Lorenzo et al

57

Table 3 Correlation between Bcl-2, COX-2, angiogenesis and other clinicopathological parameters in 126 prostate cancer samples (72 from group 1 and 54 from group 2) Gleason score o7

Gleason score X7

PSA p10

PSA 410

bcl-2 Low (39 pts)

25 (64%)

14 (36%)

15 (38%)

14 (36%)

bcl-2 High (83 pts)

35 (42%)

48 (58%)

18 (22%)

65 (78%) Po0.01

COX-2 Low (24 pts)

16 (67%)

8 (33%)

12 (50%)

12 (50%)

COX-2 High (101 pts)

47 (47%)

54 (53%)

28 (28%)

73 (72%) Po0.01

Figure 3

MVD Low (25 pts)

20 (80%)

5 (20%)

13 (52%)

12 (48%)

MVD Moderate (46 pts)

23 (50%)

23 (50%)

12 (26%)

34 (74%)

MVD High (55 pts)

17 (31%)

38 (70%) Po0.05

14 (25%)

41 (75%) Po0.01

As stated, all variables were evaluated by the Cox proportional hazard model using forward logistic regression analysis. PSA value and Gleason score were the only characteristics significantly associated with DFS in group 1 patients (Table 5).

P-values p0.05 were considered significant.

Figure 2 High expression of COX-2.

(Figure 3) were significantly correlated with high PSA values (Po0.01). A high MVD was also correlated with Gleason score (Po0.05). We used univariate analysis to evaluate the prognostic impact of biologic and clinico-pathologic parameters on the DFS of 72 patients treated by radical prostatectomy. In all, 30 patients (41.6%) experienced biochemical relapse. The median interval from operation to PSA failure was 34 months. As shown in Table 4 and Figure 4, PSA value, Gleason score, bcl-2, COX-2 and MVD significantly correlated with disease relapse in these patients. In fact, we observed disease relapse in 24/45 (53%) with high bcl-2 expression, in 15/21 (71%) with a high MVD count and finally, in 30/58 (52%) with high COX-2 expression. In contrast, disease relapse occurred only in 5/23 (22%) with low bcl-2 expression, in 2/18 (11%) with low MVD count and in 13/33 (40%) with moderate MVD.

High expression of CD-31.

Discussion Prognostic indicators are needed to help correctly treat men with clinically localized prostate cancer. There is a need for additional biomarkers that might better predict the biologic nature of a tumor, predicting which tumor will act in a more aggressive fashion. Such biomarkers would play an important role in future treatment planning. The mechanisms responsible for androgen-independent prostate cancer are not clear. Most androgenindependent prostate cancer cells express androgen receptors, which suggests that in these patients, the androgen receptor signaling pathways may be activated even in the absence of androgens.35 Peptide growth factors and their receptors have been implicated in prostate cancer by virtue of the fact that they modulate androgen receptor signaling and determine progression to a hormone-refractory state. Experimental studies have demonstrated functional crosstalk between growth factors and growth factor receptors of the EGFR family and the androgen receptor-activated pathway.36–38 Our previous findings suggest that EGFR is an independent predictor of DFS in operable prostate cancer patients and is implicated in the progression to a hormone-refractory state.8 On the other hand, several studies have implicated bcl-2, angiogenesis and COX-2 in prostate cancer development and progression to an androgen-independent state.16,19,27 Bcl-2 was first associated with androgen-independent prostate cancer by McDonnell et al,11 who found that all the androgen-independent tumors evaluated immunohistochemically expressed bcl-2, whereas 70% of androgen-dependent samples were negative for bcl-2 staining and the other 30% showed weak staining. Similar results were reported by Colombel et al,10 who detected higher bcl-2 expression in prostatic intraepithelial neoplasia (PIN) and in androgen-independent tumors Prostate Cancer and Prostatic Diseases

Prognostic factors in prostate cancer G Di Lorenzo et al

58

Table 4 Relationship between bcl-2, MVD, COX-2 expression, clinicopathological parameters and biochemical relapse in group 1 prostate cancer patients (total 72 pts) Disease relapse

bcl-2+

bcl-2

MVD+

COX-2+

Low 5/23 (22%) High 24/45 (53%)

1/4 (25%)

30/72 (41.6%)

1–45 (low) 2/18 (11%) 46–90 (moderate) 13/33 (40%) 91–141 (high) 15/21 (71%)

Low 0/13 (0%) High 30/58 (52%)

Biochemical relapse was defined as described in Materials and methods.

a

Table 5 Multivariate analysis of disease-free survival in patients treated by radical prostatectomy

b

1.0

1.0 PSA < 10

0.8 0.6

0.6

PSA ≥ 10

0.4

Gleason < 7

0.8

0.2

0.2

p = 0.00061

p = 0.0032

0.0

0.0 0

20

40 MONTHS

60

0

c

20

40 MONTHS

60

d

1.0

COX2+

1.0

CD31+

0.8

0.8 0.6

0.6

COX2++/+++

p = 0.023

p = 0.00042

0.0

0.0 0

20

40 MONTHS

0

60

20

40 MONTHS

60

e 1.0 0.8

Bcl-2 low

0.6 Bcl-2 high

0.4 0.2

p = 0.015

0.0 0

20

40 MONTHS

60

Figure 4 Kaplan–Meier DFS estimates of group 1 prostate cancer patients in relationship to: PSA value (a), Gleason score (b), COX-2 (c), CD-31 expression (d) and bcl-2 (e). The P-values for each estimate were as follows: (a) P ¼ 0.00061; (b) P ¼ 0.0032; (c) P ¼ 0.021; (d) P ¼ 0.00042 and (e) P ¼ 0.015. P-values p0.05 were considered as statistically significant.

after androgen ablation treatment than in benign hyperplasia and untreated tumors, respectively. Based on immunocytochemistry, increased bcl-2 expression and the progression from pathologically localized to disseminated disease were associated with an androgenindependent phenotype of prostate cancer.39 Although upregulation of bcl-2 is a predictor of aggressive clinical behavior, it is not the only mechanism driving disease progression. Angiogenesis is believed to play an important role in tumor progression and metastasis.16–18 Although these processes are very complex, several authors have suggested that measuring the MVD of tumors might be of prognostic value. In particular, studies of prostate carcinomas found MVD to independently predict outcome.19 Cyclooxygenases are key enzymes in the conversion of arachidonic acid to various proinflammatory prostaglandins. COX-2 expression has been reported in prostate cancer tissue.24–27 However, there are some exceptions Prostate Cancer and Prostatic Diseases

4.15 (1.2–14.5)

0.026

4.35 4.68 1.00 1.26

0.039 0.14 0.99 0.54

(1.6–10.1) (0.6–37.1) (0.4–2.4) (0.6–2.6)

CD31+++

0.2

0.2

Pretreatment PSA value (p10 vs 410 ng/ml) Gleason score (o7 vs X7) COX-2 (low vs high expression) Bcl-2 (low vs high expression) MVD 1–45 vs 46–90 vs 91–141 CD-31 positive vessels

CD31++

0.4

0.4

Relative hazard (95% confidence interval)

Gleason ≥ 7

0.4

P-value

Variable

to this observation. Zha et al30 reported that COX-2 is not consistently overexpressed in prostate cancer and found that proinflammatory atrophic lesions in prostate, which are said to be precursors of prostate cancer, express COX-2. The aim of our study was to evaluate bcl-2 and COX-2 expression and angiogenesis in two groups of prostate cancer patients representative of two different stages of prostate cancer progression. We found that bcl-2, COX-2 and MVD expression increased from group 1 to group 2. We used univariate analysis to evaluate the prognostic impact of biologic and clinico-pathologic parameters on the DFS of 72 patients treated by radical prostatectomy. In all, 30 patients (41.6%) experienced biochemical relapse; bcl-2, COX-2 and MVD significantly correlated with disease relapse in these patients. In fact, we observed disease relapse in 24/45 (53%) with high bcl-2 expression, in 15/21 (71%) with a high MVD count and finally, in 30/58 (52%) with high COX-2 expression. Collectively, these results strongly support a role for bcl-2, COX-2 and angiogenesis in the development of prostate cancer and more specifically in the progression to androgen-independent hormone refractory clinical behavior. Of course, we are aware of the small sample size considered in our study and this fact could limit our predictive model, making it probably unstable when applied to an alternative data set. Another important issue that needs further investigations is whether bcl-2, COX-2, angiogenesis overexpression correlate with the activation of growth factor receptors as EGFR and c-erbB2, with induced intracellular downstream signaling as MAPK and PI3K-AKT pathways. Measurement of bcl-2, COX-2 and angiogenesis in prostate cancer could allow identification of subsets of patients with early stage disease at high risk of relapse following radical prostatectomy that could potentially result in benefit of treatment with bcl-2, COX-2, and angiogenesis-targeted drugs in combination with

Prognostic factors in prostate cancer G Di Lorenzo et al

hormonotherapy in an attempt to prevent the progression of the disease and the development of androgenindependent hormone-refractory disease.

Acknowledgements We thank the excellent technical assistance of Mr Gaetano Borriello, Antonella Verbeni, Ersilia Cerbone. We are grateful to Jean Gilder (Scientific Communication) for editing the text.

References 1 Walsh PC, Partin AW, Epstein JI. Cancer control and quality of life following anatomical radical retropubic prostatectomy: results at 10 years. J Urol 1994; 152: 1831. 2 Crawford ED, Rosenblum M, Ziada AM, Lange PH. Overview: hormone refractory prostate cancer. Urology 2000; 54: 1–7. 3 Autorino R et al. Role of chemotherapy in hormone refractory prostate cancer. Old issues, recent advances and new perspectives. Urol Int 2003; 70: 1–14. 4 Barton J, Blackledge G, Wakeling A. Growth factors and their receptors: new targets fro prostate cancer therapy. Urology 2001; (Suppl 58): 114. 5 Steiner MS. Role of peptide growth factors in the prostate: a review. Urology 1993; 42: 99–110. 6 Porter AC, Vaillancourt RR. Tyrosine kinase receptor-activated signal trasduction pathways which lead to oncogenesis. Oncogene 1998; 17: 1343. 7 Di Lorenzo G et al. Expression of epidermal growth factor (EGFR) correlates with disease relapse and progression to androgen-independence in human prostate cancer. Clin Cancer Res 2002; 8: 3438–3444. 8 Tsujimoto Y, Croce CM. Analysis and structure, transcripts and protein products of bcl-2, the gene involved in human follicular lymphoma. Proc Natl Acad Sci USA 1986; 83: 5214–5218. 9 Reed JC. Bcl-2 and the regulation of programmed cell death. J Cell Biol 1994; 124: 1–6. 10 Colombel M et al. Detection of the apoptosis-suppressing oncoprotein bcl-2 in hormone refractory human prostate cancer. Am J Pathol 1993; 143: 380–400. 11 McDonnell TJ et al. Expression of the protooncogene bcl-2 in the prostate and its association with emergence of androgenindependent prostate cancer. Cancer Res 1992; 52: 6940–6944. 12 Raffo AJ et al. Overexpression of bcl-2 protects prostate cancer cells from apoptosis in vitro and confers resistance to androgen depletion in vivo. Cancer Res 1995; 55: 4438–4445. 13 Gleave ME et al. Progression to androgen independence is delayed by adjuvant treatment with antisense bcl-2 oligodeoxynucleotides after castration in the LNCaP prostate model. Clin Cancer Res 1999; 5: 2891–2898. 14 Leung S et al. Synergistic chemosensitization and inhibition of progression to androgen-independence by antisense bcl-2 oligodeoxynucleotide and paclitaxel in the LNCaP prostate tumor model. Int J Cancer 2001; 91: 846–850. 15 Miyake H, Tolcher A, Gleave ME. Chemosensitization and delayed androgen-independent recurrence of prostate cancer with the use of antisense bcl-2 oligonucleotides. JNCI 2000; 92: 34–41.

16 Fernandez A et al. Angiogenic potential of prostate carcinoma cells overexpressing bcl-2. JNCI 2001; 93: 208–213. 17 Liotta LA, Steeg PS, Stevenson WGS. Cancer metastasis and angiogenesis: an imbalance of positive and negative regulation. Cell 1991; 64: 327–336. 18 Folkman J, Shing Y. Angiogenesis. J Biol Chem 1992; 267: 10931–10934. 19 Weidner N et al. Tumor angiogenesis correlates with metastasis in invasive prostate carcinoma. Am. J Pathol 1993; 143: 401–409. 20 Shih SC et al. Molecular profiling of angiogenesis markers. Am J Pathol 2002; 161: 35–41. 21 Jackson MW, Bentel JM, Tilley WD. Vascular endothelial growth factor (VEGF) expression in prostate cancer and benign prostatic hyperplasia. J Urol 1997; 157: 2323–2328. 22 Giri D, Ropiquet F, Ittmann M. FGF is an autocrine and paracrine prostatic growth factor expressed by prostatic stromal cells. J Cell Physiol 1999; 180: 53–60. 23 Smith WL, DeWitt DL, Garavito RM. Cyclooxygenases: structural, cellular and molecular biology. Annu Rev Biochem 2000; 69: 145–182. 24 Hussain T, Gupta S, Mukhtar H. Cyclooxygenase-2 and prostate carcinogenesis. Cancer Lett 2003; 191: 125–135. 25 Liu XH et al. Inhibition of cyclooxygenase-2 suppresses angiogenesis and the growth of prostate cancer in vivo. J Urol 2000; 164: 820–825. 26 Tsujii M, DuBois RN. Alterations in cellular adhesion and apoptosis in epithelial cells overexpressing prostaglandin endoperoxide synthase 2. Cell 1995; 83: 493–501. 27 Gupta S et al. Overexpression of cyclooxygenase-2 in human prostate adenocarcinoma. Prostate 2000; 42: 73–78. 28 Kirschenbaum A et al. Expression of cyclooxygenase-1 and cyclooxygenase-2 in the human prostate. Urology 2000; 56: 671–676. 29 Yoshimura R et al. Expression of cyclooxygenase-2 in prostate carcinoma. Cancer 2000; 89: 589–596. 30 Zha S et al. Cyclooxygenase-2 is up regulated in proliferative inflammatory atrophy of the prostate, but not in prostate carcinoma. Cancer Res 2001; 61: 8617–8623. 31 Amling CL et al. Defining prostate specific antigen progression after radical prostatectomy. What is the most appropriate cut point? J Urol 2001; 165: 1146–1151. 32 Kaplan EL, Meier P. Nonparametric estimation from incomplete observation. J Am Stat Assoc 1958; 53: 457–481. 33 Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1996; 50: 163–170. 34 Cox DR. Regression models and life tables. J Roy Statist Soc 1972; 34: 187–220. 35 Culing Z et al. Expression, structure and function of androgen receptor in advanced prostate cancer. Prostate 1998; 35: 63–70. 36 Nazareth LV, Weigel NL. Activation of the human androgen receptor through a protein kinase A signaling pathway. J Biol Chem 1996; 271: 19900–19907. 37 Ye D, Mendelsoh J, Fan Z. Androgen and epidermal growth factor down-regulate cyclin-dependent kinase inhibitor p27 and costimulate proliferation of MDA Pca 2a and MDA PCA 2b prostate cancer cells. Clin Cancer Res 1999; 5: 2171–2177. 38 Jenster G. Ligand-independent activation of the androgen receptor in prostate cancer by growth factors and cytokines. J Pathol 2000; 191: 227–228. 39 Catz SD, Johnson JL. Bcl-2 in prostate cancer: a minireview. Apoptosis 2003; 8: 29–37.

59

Prostate Cancer and Prostatic Diseases