Expression and Localization of Hypoxia Proteins in Prostate Cancer: Prognostic Implications After Radical Prostatectomy

Cancer Expression and Localization of Hypoxia Proteins in Prostate Cancer: Prognostic Implications After Radical Prostatectomy Judith Jans, Jessica H. van Dijk, Susanne van Schelven, Petra van der Groep, Sofie H. Willems, Trudy N. Jonges, Paul J. van Diest, and J. L. H. Ruud Bosch OBJECTIVES

METHODS

RESULTS

CONCLUSIONS

To uncover novel biomarkers that will help predict which patients are at risk and may guide future treatment decisions. A significant proportion of prostate cancer patients treated with radical prostatectomy will experience a recurrence of the disease. Given the substantial role of hypoxia in prostate cancer development and treatment, this investigation focuses on the Hypoxia Inducible Factor (HIF-1␣) pathway. A tissue microarray was constructed of prostate cancer tissue collected from 71 patients undergoing radical prostatectomy. The expression of proteins involved in the HIF-1␣ pathway was investigated by an immunohistochemical approach and correlated to clinical features including the time to biochemical recurrence. Expression of GLUT1 correlated significantly (P ⬍.05) with a shorter time to biochemical recurrence after radical prostatectomy and was independent from the Gleason grade and stage of cancer. Furthermore, our studies revealed for the first time that accumulation of prolyl-4-hydroxylases 1 especially in the nucleus, is a significant indicator for a worse prognosis (P ⬍.001). This study confirms the upregulation of proteins involved in the HIF-1␣ hypoxia pathway in prostate cancer cells, indicative of a hypoxic tumor state. Importantly, we report the identification of 2 novel markers, GLUT1 and prolyl-4-hydroxylases 1, with prognostic significance for patients undergoing radical prostatectomy. UROLOGY 75: 786 –792, 2010. © 2010 Elsevier Inc.

P

rostate cancer is the most frequently diagnosed cancer in males and is responsible for approximately 6% of all cancer deaths.1 The most common treatment strategies for patients with localized disease include radiotherapy and radical prostatectomy. This disease can take a very heterogeneous course. Although the cancer is often indolent, it may become aggressively metastatic and ultimately result in death. To date, no satisfying criteria or molecular markers have been identified that can accurately predict progression of the disease. Currently, the most commonly used diagnostic marker is prostate-specific antigen (PSA).2 PSA is, however, not only elevated in prostate cancer but also in hyperplasia and inflammatory lesions. Low tissue oxygen pressure (hypoxia) is a common characteristic of cancers from different sites, including prostate cancer, and is often associated with disease pro-

From the Department of Urology, University Medical Center Utrecht, The Netherlands; Laboratory of Experimental Oncology, University Medical Center Utrecht, The Netherlands; and Department of Pathology, University Medical Center Utrecht Cancer Center, Utrecht, The Netherlands Reprint requests: Judith Jans, Ph.D., Department of Urology, University Medical Center, PO Box 85500, 3508 GA, Utrecht, The Netherlands. E-mail: j.j.m.jans@ umcutrecht.nl Submitted: February 11, 2009, accepted (with revisions): August 6, 2009

786

© 2010 Elsevier Inc. All Rights Reserved

gression.3-5 Hypoxia in cancer occurs when tumors outgrow the existing vasculature.6 Tumors respond to hypoxic conditions by switching on multiple genes involved in angiogenesis (eg, vascular endothelial growth factor [VEGF]), glycolysis (eg, ENO1), glucose transportation (eg, GLUT1), and many other processes that will enable them to survive in such conditions by either increasing oxygen supply or by adapting to the low oxygen pressure. A key regulator in the adaptation to hypoxic situations is the transcriptional activator hypoxia-inducible factor I (HIF-1), which is formed by heterodimerization of HIF-1␣ and HIF-1␤ and/or aryl hydrocarbon nuclear translocator.7,8 Although HIF-1␤ and/or aryl hydrocarbon nuclear translocator levels are not affected by oxygen pressure, HIF-1␣ levels increase when oxygen pressure drops. In line with the hypoxic nature of cancer; most tumors (including prostate tumors) show increased HIF-1␣ protein levels.3,4 The accumulation of HIF-1␣ is reported to be an early event in prostate cancer because levels are higher in high-grade prostatic intraepithelial neoplasia (PIN) compared with healthy tissue.9 In normoxia, HIF-1␣ is constantly degraded by the 26S proteasome. The proteins responsible for targeting HIF-1␣ for proteasomal degradation are von HippelLindau (VHL) and the prolyl-4-hydroxylases 1 (PHD1), 0090-4295/10/$34.00 doi:10.1016/j.urology.2009.08.024

PHD2, and PHD3. PHD proteins hydroxylate HIF-1␣ at proline residues 402 and 577, and require oxygen as a cofactor. VHL is an E3 ubiquitin ligase that subsequently specifically targets hydroxylated HIF-1␣ for degradation. During hypoxia, this post-translational modification of HIF-1␣ does not occur and, as a consequence, VHL cannot target HIF-1␣ for degradation, resulting in an accumulation of nuclear HIF-1␣.10 Given the essential role of hypoxia and the HIF-1␣ pathway in cancer progression, we investigated the expression patterns of HIF-1␣, 3 important proteins regulated by this transcription factor (VEGF, GLUT1, carbonic anhydrase IX [CA-IX]), and the key regulators of HIF-1␣: VHL and PHD1, PHD2, and PHD3. In particular, we correlate the levels of these factors in tissue samples derived from 71 radical prostatectomy patients with their disease progression and provide novel evidence for their prognostic significance.

MATERIAL AND METHODS Patients and Prostate Specimens A total of 71 formalin-fixed and paraffin-embedded prostate carcinoma specimens of patients who underwent radical prostatectomy were obtained from the archives of the Pathology Department of the University Medical Center Utrecht (UMCU). Most patients were treated at the UMCU, a few were treated in regional hospitals relying on the UMCU for pathologic assessment. A tissue microarray (TMA) was constructed by taking 4 cores (1-mm diameter) from each of the 71 specimens and arranging them in a new composite paraffin block using an arrayer (Beecher instruments, Sun Prairie, WI). In addition to the TMA, 3 specimens, each of benign prostate hyperplasia, PIN lesions, and normal prostate, were selected for analysis. Immunohistochemistry Immunohistochemistry was performed on 4-␮m thick TMA sections. All slides were deparaffinized and rehydrated. Antigen retrieval for detection of VHL, PHD1, PHD3, VEGF, GLUT1, CA-IX, Ki-67, and the androgen receptor (AR) was performed using citrate buffer (pH ⫽ 6), whereas retrieval in ethylenediaminetetraacetic acid (EDTA) buffer (pH ⫽ 9) was performed for PHD2, HIF-1␣, and p21 staining. The source of the antibodies and their respective dilutions were as follows: VHL (BD Biosciences, Erembodegem, Belgium; 1/100), PHD1 (Abcam, Cambridge, UK; 1/100), PHD2 (Abcam; 1/100), PHD3 (Novus Biologicals, Littleton, CO; 1/400), HIF-1␣ (BD Biosciences; 1/50), VEGF (R&D Systems, Minneapolis, MN; 1/50), GLUT1 (Dako, Glostrup, Denmark; 1/200), CA-IX (Abcam;1/1000), p21 (Dako; 1/20), Ki-67 (Immunotech, Woerden, The Netherlands; 1/100), and AR (Abcam; 1/25). Diaminobenzidine was used as a chromogen. Finally, the array sections were counterstained with hematoxylin. UROLOGY 75 (4), 2010

The cytoplasmic staining (PHD1, PHD2, PHD3, VHL, GLUT1, CA-IX, VEGF), divided into 4 arbitrary intensity categories and the percentage of positively stained nuclei (PHD1, PHD2, PHD3, VHL, HIF-1␣, AR, p21, Ki-67) were assessed visually by 2 observers. Cell Culture and Immune Fluorescence Prostate cancer cell lines PC3 and DU145 were cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum. Cells were transiently transfected with PHD1-EGFP-N1 using Fugene 6 (Roche Applied Science, Meylan, France) according to the manufacturers’ protocol. The plasmid containing PHD1EGFP-N1 was kindly provided by Metzen et al.11 Subsequently, cells were cultured under 1% oxygen in a hypoxia workstation or, as a control, cells were cultured under normoxic conditions. Two days after transfection, cells were fixed for 30 minutes in 4% formaldehyde and permeabilized with 0.3% Triton X-100 (Sigma, St. Louis, MO), followed by a 1-hour incubation in 2% BSA. Subsequently, cells were stained with mouse anti-GFP (1:200, 2-hour room, Roche Applied Science, Meylan, France), washed in PBS, and incubated with goat antimouse coupled to alexa488 (Molecular Probes, 1:200, Plano, TX) for 30 minutes. Nuclei were visualized with DAPI (0.2 ␮g/mL) and mounted in Vectashield (Vector Laboratories, Burlingame, CA). Statistical Analysis Pearson coefficients were calculated for correlations between data of a categorical nature (eg, staining intensity). Kruskal–Wallis tests were performed for continuous data. For recurrence-free survival analysis, all values except T-stage were dichotomized. Kaplan–Meier curves were generated using time to biochemical recurrence (defined as ⬎4 ng/mL), and results analyzed by log-rank statistics. Multivariate analysis was performed with Cox regression. All statistical analyses were performed using SPSS software (SPSS, Chicago, IL). Two-sided P values below .05 were regarded as significant.

RESULTS Clinico-Pathologic Characterization Clinical and pathologic information on 71 prostate cancer patients undergoing radical prostatectomy is summarized in Table 1. All patients had unrecordable PSA levels at the first follow-up after surgery. The median follow-up time was 62 months. Almost two-thirds (61%) of the patients remained recurrence-free within this follow-up period, whereas 32% of the patients developed a biochemical recurrence. Gleason score (P ⬍.01), clinical T stage (P ⬍.01), and positive resection margins (P ⫽ .037) correlate significantly with the time to biochemical recurrence in a univariate analysis (log-rank statistics) (Fig. 1). Conflicting reports have appeared on the role of perineural 787

Table 1. Patient characteristics Features Age (y) Follow-up (mon) No. patients Gleason score 5 6 7 8 9 T-stage (clinical) T1c T2 T3 T-stage (pathology) T2a T2c T3a T3b PSA level ⱕ4.0 4.1-10 10.1-20 ⬎20 Surgical margins Negative Positive V.S. invasion No Yes PSA recurrence No Yes Perineural growth No Yes

Median (Inter-Quartile Range)

N (%)

63 (56-67) 62 (48-91) 71 1 (1.4%) 28 (39%) 28 (39%) 7 (9.8%) 6 (8.4%) 19 (26.8%) 36 (50.7%) 11 (15.5%) 8 (11%) 39 (55%) 15 (21%) 8 (11%) 12 (16.9%) 29 (40.8%) 14 (19.7%) 9 (12.7%) 33 (46%) 37 (52%) 61 (86%) 8 (11%) 43 (60.6%) 23 (32.4%) 16 (23%) 54 (76%)

growth as a predictor of disease progression.12,13 In the current study, patients without perineural growth show an improved recurrence-free survival, although the difference was not statistically significant. The prognostic role of perineural growth may therefore exist but will be limited, and statistical significance of its role in our study was not reached because of a limited number of patients. A Network of Molecular Markers Four independent cores of each tumor were included in the TMA to circumvent sampling errors because of the heterogeneous nature of the tissue. The expression of the markers described in this study showed no substantial difference between the cores of a tumor. In this study, activation of the HIF-1 pathway in prostate cancer cells was confirmed by expression of HIF-1␣ as well as downstream targets including GLUT1, VEGF, and CA-IX. Univariate correlations were observed between proteins upstream as well as downstream of HIF-1␣, consistent with activation of the entire pathway. In particular, HIF-1␣ levels showed a positive correlation with nuclear 788

VHL (P ⬍.01) as well as the HIF-1␣ downstream target CA-IX (P ⬍.01). In addition, a significant association was found between the AR and several hypoxia markers including HIF-1␣, PHD2, PHD3, and CA-IX, yet no predictive value for biochemical recurrence could be ascribed to AR levels. Similarly, we were unable to predict a recurrence based on the proliferative markers p21 and Ki-67. HIF-1␣, Upstream Regulators and Downstream Targets as Prognostic Factors Expression patterns of proteins were correlated with the time to PSA recurrence to determine whether they can serve as a prognostic factor. No association was observed between the expression levels of (nuclear or cytoplasmic) HIF-1␣, VEGF, and CA-IX and time to recurrence (Fig. 1). The presence of cytoplasmic GLUT1, however, was a negative prognosticator in univariate (P ⬍.05) (Fig. 2) and multivariate survival analysis (HR 6.6; 95% confidence interval 1.74-24.7; P ⫽ .005) in a model that included Gleason score, T stage, and tumor margins. Intense staining (Fig. 2) was found in a subset of patients with a shorter time to biochemical recurrence and not in control sections of healthy prostate, PIN lesions, or benign prostate hyperplasia. A strong univariate correlation (P ⬍.001) was also observed between the time to recurrence and nuclear PHD1 of prostate cancer cells. PHD1 did not localize to the nucleus in normal prostate, PIN lesions, or benign prostate hyperplasia. In contrast, levels of VHL and cytoplasmic PHD1, PHD2, and PHD3 did not correlate with biochemical recurrence (Fig. 3). Localization of PHD1 in Prostate Cancer Cell Lines Given the differences in localization of PHD1 (nuclear vs cytoplasmic) between patients, we set out to determine whether the nuclear localization was a direct consequence of a hypoxic environment. To this end, prostate cancer cell lines PC3 and DU145 were transfected with PHD1 tagged with a green fluorescent protein and cultured under normoxic or hypoxic (1% oxygen) conditions. PHD1 localized to the nucleus regardless of the concentration of oxygen (Fig. 3), indicating that the nuclear localization is independent of oxygen concentration and is a characteristic present in these cell lines derived from metastasized prostate cancer.

CONCLUSIONS Although radical prostatectomy is a common treatment for patients with localized prostate cancer, approximately 1 in 3 patients will eventually experience a biochemical recurrence.14-16 Despite the vast amount of research performed on potential diagnostic and prognostic factors, it is currently not possible to predict the outcome for individual patients. In future, novel biomarkers may be used to provide a patient with an appropriate prognosis, determine who is at risk for a recurrence and, where required, offer additional treatment. UROLOGY 75 (4), 2010

Figure 1. Recurrence-free survival based on clinico-pathologic characteristics and expression of hypoxia-related proteins Kaplan–Meier curves showing the fraction of biochemical recurrence-free survival against time after radical prostatectomy depending on (A) Gleason score, (B) clinical T-stage, (C) presence of positive surgical margins, (D) hypoxia-inducible factor 1␣, (E) vascular endothelial growth factor, or (F) CA-IX.

Currently, the commonly used tools for detection and prognosis include clinical T stage, pathologic grade, and levels of PSA in serum. The ideal biomarker for diagnosis and prognosis of any disease has properties that include a high specificity, high sensitivity, and can be obtained in a noninvasive manner. Therefore, tissue-markers may not be the best choice for screening purposes. For radical prostatectomy patients, however, tumor tissue is readily available and might be analyzed for presence of prognostic markers. In addition to the tissue obtained after radical prostatectomy, immunohistochemical analysis may also be performed on biopsy material from patients before radiotherapy to assess prognosis or to target therapy. An important concern when using biopsy material, however, is the heterogeneity of the tumor tissue and, as a consequence, errors in sampling. Given the great importance of hypoxia in prostate cancer biology and treatment, the study presented here focuses on proteins involved in the HIF-1␣ hypoxia pathway as potential biomarkers after radical prostatectomy. Overexpression of HIF-1␣ was previously reported to correlate with poor prognosis in various types of cancer including breast, cervical, esophageal, and lung cancer.17-20 Recently, Vergis et al21 reported a significant association of cytoplasmic HIF-1␣ levels and time to biochemical recurrence in a cohort of prostate cancer UROLOGY 75 (4), 2010

patients treated with radiotherapy or radical prostatectomy. Here, we confirm the accumulation of nuclear HIF-1␣ in prostate cancer cells but were unable to show an association between HIF-1␣ levels and PSA recurrence. One of the differences in the study by Vergis et al was the anti-HIF-1␣ antibodies used; this may affect the outcome. Similar to our findings, Boddy et al22 found no significant association between HIF-1␣ levels and time to biochemical recurrence. In addition to nuclear accumulation of HIF-1␣, we also observed upregulation of all proteins of the investigated pathway. The expression of many of these was very strongly correlated, in line with the general hypoxic nature of prostate cancer cells. Expression of GLUT1, a transporter important for glucose uptake, has been investigated in many types of cancer, and in some cases was reported to be an indicator of poor prognosis.23-25 Expression of GLUT1 at the mRNA level has been reported in prostate cancer cell line.26 Whether GLUT1 overexpression indicates a poor prognosis in prostate cancer patients was until now unknown. We now provide evidence for GLUT1 as a prognostic factor for prostate cancer patients, as patients with elevated levels of GLUT1 have a significantly shorter time to biochemical recurrence after radical prostatectomy. Whether elevated GLUT1 expression is simply an accurate 789

Figure 2. GLUT1 expression in prostate cancer. Representative examples of (A) weak and (B) strong staining of GLUT1 in prostate cancer tissue, (C) healthy prostate, (D) prostatic intraepithelial neoplasia lesions, and (E) benign prostate hyperplasia. Kaplan–Meier curves showing the fraction of biochemical recurrence-free survival against time after radical prostatectomy, depending on dichotomized levels (F) GLUT1.

readout of the hypoxic state of the tumor and it is in fact this hypoxic state that determines the progression of the disease, or whether the elevated GLUT1 levels reflect the altered metabolic state of tumor cells, remains unknown. Further studies on expression of other glucose transporters within the GLUT1 protein family in prostate cancer may provide additional insight into the prognostic significance of the metabolic state of prostate cancer cells. In a previous publication, Boddy et al22 investigated expression of the prolyl-4-hydroxylase (PHD) family of proteins in addition to HIF in prostate cancer patients. In their study, no correlation between expression of these proteins and prognosis of the patient could be established. Given the relevance of the intracellular localization on proper functioning of proteins, we decided to take into account the accumulation of these proteins in the nucleus as well as in the cytoplasm and found a striking correlation between nuclear localization of PHD1 and the time to PSA recurrence after radical prostatectomy. Patients with a high fraction of PHD1positive nuclei had a significantly worse prognosis. In contrast, cytoplasmic levels of PHD1 were similar in patients with or without a recurrence. Although little is 790

known about PHD1 function, these findings resemble the observations made in head and neck squamous cell carcinoma for the protein family member PHD2; overexpression and nuclear accumulation of PHD2 correlates with a more aggressive phenotype of tumor cells.27 Although we see a similar trend for PHD2, this was not significant in our set of radical prostatectomy patients. These results could point toward a more general phenomenon in which relocalization of PHD proteins reflect a more aggressive and less differentiated state of the tumor, and our findings warrant further investigation of these proteins in other types of cancer. Our observation that PHD1 localizes to the nucleus independently of oxygen concentrations in 2 cell lines derived from metastasized androgen-independent cancers suggests that this may be an important, separate step in the progression of prostate cancer. In addition to our immunohistochemical analysis, we reviewed the pathology reports and confirmed that the T stage and Gleason score are significant prognostic indicators. In contrast, the prognostic relevance of perineural invasion did not reach statistical significance in our dataset. Because perineural growth may be an early indicator UROLOGY 75 (4), 2010

Figure 3. Expression and localization of prolyl-4-hydroxylases 1 (PHD1) in prostate cancer cells. Immunohistochemical analysis of localization of PHD1 in prostate cancer patients with (A) low levels of nuclear PHD1, (B) high levels of nuclear PHD1, (C) normal prostate, (D) prostatic intraepithelial neoplasia lesions, and (E) benign prostate hyperplasia. Bar graph showing the average intensity of (F) cytoplasmic and (G) nuclear PHD1 staining in prostate cancer cells of patients who experienced a biochemical recurrence and patients who did not experience a biochemical recurrence during the follow-up time. Error bars indicate standard error of the mean. (H) Kaplan–Meier curves showing the fraction of biochemical recurrence-free survival against time after radical prostatectomy depending on nuclear PHD1 levels. (I) Kaplan–Meier curves showing the fraction of biochemical recurrence-free survival against time after radical prostatectomy, depending on nuclear PHD1 levels in addition to a dichotomized Gleason score. Immunofluorescence images of PC3 cells under (J) normoxic (K) and hypoxic conditions.

or precursor of capsular invasion, a longer patient follow-up time may be required to reveal the prognostic potential. Alternatively, quantification of the extent of perineural growth may provide additional information. Taken together, our immunohistochemical approach identified 2 novel factors, cytoplasmic GLUT1 and nuclear PHD1, that—in addition to general pathology and surgical UROLOGY 75 (4), 2010

factors—predict biochemical recurrences after radical prostatectomy. Although a prospective study is required to confirm the relevance of these proteins as novel biomarkers for prostate cancer patients, our observations do illustrate the strength of an immunohistochemical approach for biomarker research because specific subcellular localization of proteins will not easily be detected otherwise. 791

Acknowledgment. The authors acknowledge E. Metzen (University Duisburg–Essen) for providing PHD1 antibodies. References 1. American Cancer Society. Global cancer facts and figures 2007; 2007. 2. Thompson IM, Ankerst DP, Chi C, et al. Assessing prostate cancer risk: results from the Prostate Cancer Prevention Trial. J Natl Cancer Inst. 2006;98(8):529-534. 3. Brizel DM, Scully SP, Harrelson JM, et al. Tumor oxygenation predicts for the likelihood of distant metastases in human soft tissue sarcoma. Cancer Res. 1996;56(5):941-943. 4. Hockel M, Schlenger K, Aral B, et al. Association between tumor hypoxia and malignant progression in advanced cancer of the uterine cervix. Cancer Res. 1996;56(19):4509-4515. 5. Brahimi-Horn MC, Chiche J, Pouyssegur J. Hypoxia and cancer. J Mol Med. 2007;85(12):1301-1307. 6. Hanahan D, Folkman J. Patterns and emerging mechanisms of the angiogenic switch during tumorigenesis. Cell. 1996;86(3):353-364. 7. Wang GL, Jiang BH, Rue EA, et al. Hypoxia-inducible factor 1 is a basic-helix-loop-helix-PAS heterodimer regulated by cellular O2 tension. Proc Natl Acad Sci U S A. 1995;92(12):5510-5514. 8. Kallio PJ, Pongratz I, Gradin K, et al. Activation of hypoxiainducible factor 1alpha: posttranscriptional regulation and conformational change by recruitment of the ARNT transcription factor. Proc Natl Acad Sci U S A. 1997;94(11):5667-5672. 9. Zhong H, Semenza GL, Simons JW, et al. Upregulation of hypoxiainducible factor 1alpha is an early event in prostate carcinogenesis. Cancer Detect Prev. 2004;28(2):88-93. 10. Lee JW, Bae SH, Jeong JW, et al. Hypoxia-inducible factor (HIF-1) alpha: its protein stability and biological functions. Exp Mol Med. 2004;36(1):1-12. 11. Metzen E, Berchner-Pfannschmidt U, Stengel P, et al. Intracellular localisation of human HIF-1alpha hydroxylases: implications for oxygen sensing. J Cell Sci. 2003;116(7):1319-1326. 12. van den Ouden D, Hop WC, Kranse R, et al. Tumour control according to pathological variables in patients treated by radical prostatectomy for clinically localized carcinoma of the prostate. Br J Urol. 1997;79(2):203-211. 13. Endrizzi J, Seay T. The relationship between early biochemical failure and perineural invasion in pathological T2 prostate cancer. BJU Int. 2000;85(6):696-698. 14. Pound CR, Partin AW, Epstein JI, et al. Prostate-specific antigen after anatomic radical retropubic prostatectomy. Patterns of recurrence and cancer control. Urol Clin North Am. 1997;24(2): 395-406.

792

15. Zincke H, Oesterling JE, Blute ML, et al. Long-term (15 years) results after radical prostatectomy for clinically localized (stage T2c or lower) prostate cancer. J Urol. 1994;152(5 Pt 2):1850-1857. 16. Catalona WJ, Smith DS. 5-year tumor recurrence rates after anatomical radical retropubic prostatectomy for prostate cancer. J Urol. 1994;152(5 Pt 2):1837-1842. 17. Trastour C, Benizri E, Ettore F, et al. HIF-1alpha and CA IX staining in invasive breast carcinomas: prognosis and treatment outcome. Int J Cancer. 2007;120(7):1451-1458. 18. Dellas K, Bache M, Pigorsch SU, et al. Prognostic impact of HIF-1alpha expression in patients with definitive radiotherapy for cervical cancer. Strahlenther Onkol. 2008;184(3):169-174. 19. Enatsu S, Iwasaki A, Shirakusa T, et al. Expression of hypoxiainducible factor-1 alpha and its prognostic significance in smallsized adenocarcinomas of the lung. Eur J Cardio Thorac Surg. 2006;29(6):891-895. 20. Lidgren A, Hedberg Y, Grankvist K, et al. Hypoxia-inducible factor 1alpha expression in renal cell carcinoma analyzed by tissue microarray. Eur Urol. 2006;50(6):1272-1277. 21. Vergis R, Corbishley CM, Norman AR, et al. Intrinsic markers of tumour hypoxia and angiogenesis in localised prostate cancer and outcome of radical treatment: a retrospective analysis of two randomised radiotherapy trials and one surgical cohort study. Lancet Oncol. 2008;9(4):342-351. 22. Boddy JL, Fox SB, Han C, et al. The androgen receptor is significantly associated with vascular endothelial growth factor and hypoxia sensing via hypoxia-inducible factors HIF-1a, HIF-2a, and the prolyl hydroxylases in human prostate cancer. Clin Cancer Res. 2005;11(21):7658-7663. 23. Younes M, Brown RW, Stephenson M, et al. Overexpression of Glut1 and Glut3 in stage I nonsmall cell lung carcinoma is associated with poor survival. Cancer. 1997;80(6):1046-1051. 24. Haber RS, Rathan A, Weiser KR, et al. GLUT1 glucose transporter expression in colorectal carcinoma: a marker for poor prognosis. Cancer. 1998;83(1):34-40. 25. Cantuaria G, Fagotti A, Ferrandina G, et al. GLUT-1 expression in ovarian carcinoma: association with survival and response to chemotherapy. Cancer. 2001;92(5):1144-1150. 26. Chandler JD, Williams ED, Slavin JL, et al. Expression and localization of GLUT1 and GLUT12 in prostate carcinoma. Cancer. 2003;97(8):2035-2042. 27. Jokilehto T, Rantanen K, Luukkaa M, et al. Overexpression and nuclear translocation of hypoxia-inducible factor prolyl hydroxylase PHD2 in head and neck squamous cell carcinoma is associated with tumor aggressiveness. Clin Cancer Res. 2006;12(4):1080-1087.

UROLOGY 75 (4), 2010