Hypoxia-Inducible Factor 1a Expression in Renal Cell Carcinoma Analyzed by Tissue Microarray

Transcription

1 european urology 50 (2006) available at journal homepage: Kidney Cancer Hypoxia-Inducible Factor 1a Expression in Renal Cell Carcinoma Analyzed by Tissue Microarray Anders Lidgren a, Ylva Hedberg b, Kjell Grankvist c, Torgny Rasmuson d, Anders Bergh b,börje Ljungberg a, * a Department of Surgical and Perioperative Sciences, Urology and Andrology, Umeå University, Umeå, Sweden b Department of Medical Biosciences, Pathology, Umeå University, Umeå, Sweden c Department of Medical Biosciences, Clinical Chemistry, Umeå University, Umeå, Sweden d Department of Radiation Sciences, Oncology, Umeå University, Umeå, Sweden Article info Article history: Accepted May 30, 2006 Published online ahead of print on June 13, 2006 Keywords: Hypoxia-inducible factor HIF-1a RCC Prognosis Tissue microarray Abstract Objectives: Angiogenesis is important for tumour progression and metastatic spread. Hypoxia-inducible factor 1a (HIF-1a) is a major factor regulating a number of other angiogenic factors. Renal cell carcinoma (RCC) is a malignancy with a variable clinical course, partly attributable to specific genetic alterations of the different RCC types. We therefore analysed HIF-1a expression using immunohistochemistry and related the results to RCC type and clinicopathologic variables. Material and methods: We semiquantitatively analysed HIF-1a expression using immunohistological staining of a prepared tissue microarray. There were 216 patients including 176 conventional, 26 papillary, and 14 chromophobe RCCs. Results: The HIF-1a staining was found mainly in the cytoplasm. The tumours were subdivided into HIF-1a LOW and HIF-1a HIGH on the basis of staining intensity. HIF-1a expression between the RCC types did not differ. Patients with conventional RCC showed a trend ( p = 0.055) towards a prolonged survival for those with HIF-1a HIGH -staining versus HIF-1a LOW -staining tumors. In conventional RCC there were significant differences in HIF-1a expression in relation to TNM stage, nuclear grade, and vein invasion. In patients with papillary RCC, difference in HIF-1a expression was observed only for nuclear grade. Conclusions: We studied HIF-1a expression in RCC using tissue microarray. In patients with conventional RCC, HIF-1a levels were significantly lower in locally aggressive tumors versus localized tumors, and patients with high HIF-1a levels tended to have a better prognosis. There seems to be a diverging regulation of angiogenesis between the different RCC types. Further studies of HIF and angiogenesis in RCC are encouraged. # 2006 European Association of Urology. Published by Elsevier B.V. All rights reserved. * Corresponding author. Department of Surgical and Perioperative Sciences, Urology and Andrology, Umeå University, S Umeå, Sweden. Tel ; Fax: address: borje.ljungberg@urologi.umu.se (B. Ljungberg) /$ see back matter # 2006 European Association of Urology. Published by Elsevier B.V. All rights reserved. doi: /j.eururo

2 european urology 50 (2006) Introduction Renal cell carcinoma (RCC) accounts for approximately 2% of all human cancers, with a worldwide incidence of 189,000 in the year 2000 [1]. One distinguishing feature of RCC is vascularisation, and a relationship between vascularity and clinical outcome has been implied [2]. A number of studies show that initiation of the angiogenic switch through angiogenic factors such as vascular endothelial growth factor (VEGF), and hypoxiainducible factor 1a (HIF-1a) is correlated to poor prognosis in malignancies [2 4]. However, our previous findings suggest that upregulation of HIF- 1a is beneficial for survival in patients with conventional RCC [5]. The reason for a better survival by upregulated HIF-1a in conventional RCC is unclear but might be due to the biology of that RCC type. Among the three dominating RCC types, conventional RCC generally is more vascularised than both papillary and chromophobe RCC [6,7]. This vascularisation is most likely due to the alterations of the von Hippel-Lindau (VHL) tumoursuppressor gene in conventional RCC, where its product (pvhl) fails to exert its role as an oxygendependent degrader of HIF-1a. In the absence of a functional pvhl, HIF-1a erroneously commences the angiogenic response to hypoxia by transcribing a number of oxygen-regulated factors such as VEGF, Flt-1, Glut-1, and erythropoietin (EPO) [8,9]. The aim of this study was to investigate the immunohistochemical expression pattern of HIF-1a in a large tumor material with different RCC types and to evaluate whether HIF-1a expression correlates to clinicopathologic parameters. 2. Material and methods 2.1. Patients The study was approved by the ethical committee of Umeå University, and each patient participated after providing informed consent. There were 216 patients included with histopathologically verified RCC (176 conventional, 26 papillary, and 14 chromophobe), 127 men and 89 women with mean age 65.3 yr (range: 25 87). All patients had nephrectomies performed at the department of Urology, Umeå University Hospital between 1982 and Staging procedures included physical examination, chest radiography, ultrasound, and computerized tomography of the abdomen. Tumour stage was reclassified according to the TNM stage classification system 2002 [10]. Nuclear grading was performed according to Skinner et al. [11]. RCC type was defined according to the Heidelberg consensus conference [6]. Tumour size was measured on the surgical specimens and/or on computerized tomographies. The tumours varied in size from 0.6 cm to 18 cm (median: 7.5). Vein invasion was defined as tumor invasion in major renal veins verified microscopically in renal hilum. All patients were followed up with clinical and radiologic examinations. Among the 216 patients, 110 had died of the disease, 60 died of noncancer-related causes, and 46 were alive at the end of the last follow-up with a median survival of 152 mo (range: ) Tissue microarray, antibody, and staining procedure Tumor sections were screened by a pathologist and arranged pair wise in tissue microarray (TMA) blocks, as described by Hedberg et al. [12]. Each tumor was represented by two validated tissue cores on a TMA. Of 237 tumors in the TMA, 21 were excluded because of necrosis, defect cores, or >25% difference between the cores when quantified. The formalinfixed tumour blocks used had been stored for 3 19 yr before TMA construction. Paraffin sections (4 mm thick) from the TMA blocks were deparaffinised and microwaved according to standard procedures before being processed in an automatic immunohistochemical staining machine (Ventana ; Ventana Inc, Tucson, AZ, USA). Expression of HIF-1a was detected by a monoclonal antibody diluted 1:2000 (NB ; Novus Biologicals, Littleton, CO, USA). To verify that the results were not dependent on tissue storage time, the percentage of positively stained tumour cells were plotted for each year (data not shown). The percentage was uniform throughout the study period, and the staining was therefore regarded as stable. HIF-1a expression was evaluated in four groups according to staining intensity (0 3). Tumours that lacked staining were defined as group 0. Subsequent groups (1 3) were all HIF-1a positive but were separated depending on intensity (1 = weak, 2 = medium, 3 = strong) Statistical analysis The Spearman test of correlation, the Fisher exact test, and the Mann-Whitney U test were used for statistical analyses. The log-rank test and the Kaplan-Meier method were used for survival analyses. Statistical analyses were performed with SPSS statistical software, version 11.0 (Statistical Package for Social Sciences; SPSS Inc, Chicago, IL, USA). 3. Results 3.1. HIF-1a expression on TMA The HIF-1a staining was present mainly in the cytoplasm. Nuclear staining was observed only sparsely and was not further evaluated. The tumours were subdivided according to the HIF-1a expression into two groups in which the negative (0, n = 43) and weak stained (1, n = 50) were gathered and defined as HIF-1a LOW. Tumors with medium (2, n = 87) and strong HIF-1a staining (3, n = 36) were defined as HIF-1a HIGH. Fig. 1 illustrates the two groups (HIF- 1a LOW and HIF-1a HIGH ) that constituted the base for

3 1274 european urology 50 (2006) Fig. 1 Immunohistochemical HIF-1a LOW and HIF-1a HIGH staining in conventional and papillary RCCs, respectively. further clinicopathologic analyses in both conventional and papillary RCC. The distribution of HIF- 1a LOW expression versus HIF-1a HIGH expression did not differ between the three RCC types (Table 1). There was no statistical difference between the RCC types comparing negative versus stained tumors or between those with strong HIF-1a staining compared with weaker HIF-1a staining (data not shown) Conventional RCC and clinicopathological parameters There was no correlation between HIF-1a expression and age or gender in conventional RCC. For patients with HIF-1a HIGH staining, a trend ( p = 0.055) towards better survival was observed compared to those with HIF-1a LOW (Fig. 2A). Table 1 illustrates the differences between HIF-1a LOW versus HIF-1a HIGH when compared with TNM stage ( p = 0.024), grade ( p =0.009), and vein invasion ( p = 0.041). Tumour size correlated inversely ( p = 0.012) to HIF-1a expression Papillary RCC and clinicopathological parameters In papillary RCC, HIF-1a expression did not correlate to age, gender, or tumour size. There was no difference in survival between HIF-1a LOW and HIF- 1a HIGH papillary RCCs (Fig. 2B). As shown in Table 1, papillary RCC differed significantly when HIF-1a Table 1 Distribution of low and high HIF-1a expression in 216 patients with RCC, subdivided into RCC type and related to TNM stage, nuclear grade, and vein invasion Tumour type Conventional (n = 176) Papillary (n = 26) Chromophobe (n = 14) HIF-1a LOW HIF-1a HIGH HIF-1a LOW HIF-1a HIGH HIF-1a LOW HIF-1a HIGH TNM stage TNM I+II TNM III TNM IV Nuclear grade Vein invasion No Yes

4 european urology 50 (2006) Discussion Fig. 2 (A) Kaplan-Meier diagram illustrating survival curves of 176 patients with conventional RCCs having HIF- 1a HIGH - (102) versus low HIF-1a LOW -stained tumors (74). (B) Kaplan-Meier diagram illustrating survival curves of 26 patients having HIF-1a HIGH - (15) versus low HIF-1a LOW - stained papillary RCCs (11). expression was compared with tumor grade ( p = 0.05) but not to TNM stage or vein invasion Chromophobe RCC and clinicopathologic parameters Since there were only 14 chromophobe RCCs, no further statistical analyses was performed. Over the past decade a distinct transition has emerged in the field of oncology, leaving traditional clinical factors for prognostic purposes and for exploration of molecular and genetic markers. The use of novel markers increases the possibility of discriminating between different submalignancies, further enhancing the prognostic value of the markers. Angiogenic factors and their mediators have gained increased interest because of an association with tumor progression and metastatic spread [13 15]. The importance of these factors has been described in a large number of solid tumors such as lung, breast, urinary bladder, and prostate cancer [13 16]. HIF-1a is regarded as the single most important transcription factor initiating angiogenesis by regulating transcription of several factors such as VEGF, platelet-derived growth factor, and EPO; HIF-1a has also been shown to modulate blood flow in newly formed blood vessels [17]. Furthermore, HIF-1a regulates a number of other target genes relevant to cancer development and progression, such as cell-cycle regulators, growth factors, and metabolic and apoptotic factors [5,6,15,17]. One could expect that HIF-1a should be more expressed in conventional RCC than in the other RCC types, since mutations in the VHL suppressor gene are most common in that RCC type because of its genetically altered chromosome 3p [6]. The function of the VHL gene product (pvhl) is to bind to HIF-1a in normoxic cells. VHL s primary function is to ubiquinate HIF-1a for proteasomal degradation under normoxic conditions, something that thus is disabled in the majority of conventional RCCs. Lack of the VHL protein leads to reduced degradation of HIF-1a, a state that is normally seen only in hypoxic cells. High levels of HIF-1a have been noted in conventional RCC [5,18]. Thus, the HIF-1a levels are mainly caused by genetic alterations of the VHL gene in addition to, or despite, stimulation through hypoxia. Papillary and chromophobe RCCs do not have this genetic alteration, and HIF-1a might be regulated mainly by hypoxia in these RCCs. In this study the HIF-1a staining was present mainly in the cytoplasm, whereas nuclear staining was observed only sparsely. One might have suspected that nuclear staining should have been more prominent also in RCC, but nuclear staining was not achieved with different staining techniques. The reason for that staining distribution is not known, but one suggestion might be a relatively constant relative hypoxic condition in kidney tissue [19]. In the present study, using immunohistochemical detection of HIF-1a, we found no difference in

5 1276 european urology 50 (2006) HIF-1a expression between the major RCC types. This result contradicts previous studies showing higher HIF-1a levels in conventional RCC compared with other RCC types as analyzed by Western blot [5,18]. The lack of conformity between the studies might be due to different methods used to detect HIF-1a protein expression. A similar conformity between the RCC types was also noticed by Jacobsen et al. [20,21] who analyzed VEGF protein using immunohistochemistry, whereas higher VEGF levels was shown in conventional RCC with the use of Western blot analysis. The TMA technique only displays small pieces of information from the original tumor, and the expression of the protein of interest is subjectively quantified, which could also explain the differences. However, Jacobsen et al. found conformity analyzing VEGF on large tissue sections compared with TMA [13]. Our study also shows that TMA is a feasible method for analyzing HIF-1a expression in a large number of tumors. The use of two tissue cores from the same tumor allowed the immunohistochemical information to be validated. However, when interpreting the data, one needs to recognize that limited tumor areas are analysed. In most cancers, high HIF-1a levels have been associated with an unfavourable prognosis [3,4]. In contrast to those findings, our present results indicate a strong trend towards a better prognosis for high HIF-1a expression in conventional RCC, supporting our previous findings from Western blot analysis, for which high HIF-1a expression was found to be an independent inverse prognostic factor in conventional RCC in contrast to findings in papillary RCC [5]. Similar results, with an inverse correlation between survival and other angiogenic variables such as vessel density [22,23] and endoglin staining [24], have been observed in conventional RCC. Inverse correlation between elevated HIF-1a expression and favourable clinical course has also been reported in lung cancer [25]. Thus an upregulation of HIF-1a is not necessarily all negative, at least not in conventional RCC. For papillary RCC no prognostic information was observed in the comparison of high and low levels of HIF-1a, supporting previous findings [5]. Another observation in this study is that locally advanced conventional RCCs have significantly lower HIF-1a expression compared with tumors confined to the kidney and also those with metastatic spread. Similar findings have been observed in studies on vascular density [22 24]. However, opposite results were reported in a previous study in which higher VEGF messenger RNA (mrna) levels were found in localized conventional RCC [26]. The generally low HIF-1a expression and elevated VEGF mrna levels in locally advanced conventional RCCs might reflect that a hypoxia-independent pathway, other than through HIF-1a, might be present in this RCC type. Supporting these observations was the finding that low HIF-1a expression was associated with tumor vein invasion, similar to the low VEGF expression, as analyzed with immunohistochemistry, associated with vein involvement [20]. Our study demonstrates that TMA is a useful method to analyze HIF-1a expression in RCC. HIF-1a levels were significantly lower in locally aggressive conventional RCC, and patients with high HIF-1a levels tended to have a better prognosis. The regulation of angiogenesis seems to be diverging in the different RCC types. Further studies are needed to clarify the various pathways and regulation of angiogenesis in RCC. Acknowledgements This study was supported by the Swedish Cancer Society, the Cancer Research Foundation in Northern Sweden, and the Medical Faculty, Umeå University, Umeå, Sweden. References [1] Parkin DM, Bray F, Ferlay J, Pisani P. Estimating the world cancer burden: Globocan Int J Cancer 2001;94: [2] Jacobsen J, Rasmuson T, Grankvist K, Ljungberg B. Vascular endothelial growth factor as prognostic factor in renal cell carcinoma. J Urol 2000;163: [3] Birner P, Schindl M, Obermair A, Plank C, Breitenecker G, Oberhuber G. Overexpression of hypoxia-inducible factor 1a is a marker for an unfavourable prognosis in earlystage invasive cervical cancer. Cancer Res 2000;60: [4] Schindl M, Schoppmann SF, Samonigg H, et al. Overexpression of hypoxia-inducible factor 1a is associated with an unfavourable prognosis in lymph node-positive breast cancer. Clin Cancer Res 2002;8: [5] Lidgren A, Hedberg Y, Grankvist K, Rasmuson T, Vasko J, Ljungberg B. The expression of hypoxia-inducible factor 1alpha is a favorable independent prognostic factor in renal cell carcinoma. Clin Cancer Res 2005;11: [6] Kovacs G, Akhtar M, Beckwith BJ, et al. The Heidelberg classification of renal cell tumours. J Pathol 1997;183: [7] Brieger J, Weidt EJ, Schirmacher P, Storkel S, Huber C, Decker HJ. Inverse regulation of vascular endothelial growth factor and VHL tumor suppressor gene in sporadic renal cell carcinomas is correlated with vascular growth: an in vivo study on 29 tumors. J Mol Med 1999;77:

6 european urology 50 (2006) [8] Maxwell PH, Wiesener MS, Chang GW, et al. The tumour suppressor protein VHL targets hypoxia-inducible factors for oxygen-dependent proteolysis. Nature 1999;399: [9] Maxwell PH, Pugh CW, Ratcliffe PJ. Activation of the HIF pathway in cancer. Curr Opin Genet Dev 2001;11: [10] International Union Against Cancer. In: Sobin LH, Wittekind CH, editors. TNM classification of malignant tumours (UICC). 6th ed. New York: Wiley-Liss; p [11] Skinner DG, Colvin RB, Vermillion CD, Pfister RC, Leadbetter WF. Diagnosis and management of renal cell carcinoma. A clinical and pathologic study of 309 cases. Cancer 1971;28: [12] Hedberg Y, Ljungberg B, Roos G, Landberg G. Expression of cyclin D1, D3, E and p27 in human renal cell carcinoma analysed by tissue micro array. Br J Cancer 2003;88: [13] Berber DP, Herbstritt L, Dengler WA, Marme D, Mertelsmann R, Fiebig HH. Vascular endothelial growth factor (VEGF) mrna expression in human tumor models of different histologies. Ann Oncol 1995;6: [14] Folkman J, D Amore PA. Blood vessel formation: what is its molecular basis? Cell 1996;87: [15] Charlesworth PJS, Harris AL. Mechanisms of disease: angiogenesis in urologic malignancies. Nature Clin Pract Urol 2006;3: [16] Fontanini G, Lucchi M, Vignati S, et al. Angiogenesis as a prognostic factor of survival in non-small-cell lung carcinoma: a prospective study. J Natl Cancer Inst 1997;89: [17] Wenger RH. Cellular adaptation to hypoxia: O2-sensing protein hydroxylases, hypoxia-inducible transcription factors, and O2-regulated gene expression. FASEB J 2002;10: [18] Wiesener MS, Münchenhagen PM, Berger I, et al. Constitutive activation of hypoxia-inducible genes related to overexpression of hypoxia-inducible factor-1a in clear cell renal carcinoma. Cancer 2001;61: [19] Bernhardt WM, Schmitt R, Rosenberger C, et al. Expression of hypoxia-inducible transcription factors in developing human and rat kidneys. Kidney Int 2006; 69: [20] Jacobsen J, Grankvist K, Rasmuson T, Bergh A, Landberg G, Ljungberg B. Expression of vascular endothelial growth factor protein in human renal cell carcinoma. BJU Int 2004;93: [21] Jacobsen J, Grankvist K, Rasmuson T, Ljungberg B. Different isoform patterns for vascular endothelial growth factor between clear cell and papillary renal cell carcinoma. BJU Int 2006;97: [22] Kohler HH, Barth PJ, Siebel A, Gerharz EW, Bittinger A. Quantitative assessment of vascular surface density in renal cell carcinomas. Br J Urol 1996;77: [23] Delahunt B, Bethwaite PB, Thornton A. Prognostic significance of microscopic vascularity for clear cell renal carcinoma. Br J Urol 1997;80: [24] Sandlund J, Hedberg Y, Bergh A, Grankvist K, Ljungberg B, Rasmuson T. Endoglin (CD105) expression in human renal cell carcinoma. BJU Int 2006;97: [25] Volm M, Koomagi R. Hypoxia-inducible factor (HIF-1) and its relationship to apoptosis and proliferation in lung cancer. Anticancer Res 2000;20: [26] Ljungberg B, Jacobsen J, Häggström-Rudolfsson S, Rasmuson T, Lindh Grankvist K. Tumour vascular endothelial growth factor (VEGF) mrna in relation to serum VEGF protein levels and tumour progression in human renal cell carcinoma. Urol Res 2003;31: