Preface. Histology. Ronald M. Bukowski, Robert J. Motzer, and Robert A. Figlin

Transcription

1 Preface Ronald M. Bukowski, Robert J. Motzer, and Robert A. Figlin Renal cancer comprises 3% of all malignant tumors, with an estimated incidence of 39,000 new cases with 13,000 deaths in 2006 (1). A study comparing 43,685 cases of renal cancer from with those diagnosed in (SEER database) demonstrated a marginal increase in the proportion of localized cancers and a decrease in advanced cases in the latter group. During the next 10-year period, however, the increase in localized and smaller tumors appears real, but overall survival (OS) differences are not yet apparent (2). While increased imaging and laboratory testing may generally explain the increased incidence, other environmental factors may also play a role (2). Historically, patients presented with the classic triad of symptoms including flank pain, hematuria, and a palpable abdominal mass; but recently, increasing numbers of individuals are being diagnosed when asymptomatic with an incidentally discovered renal mass. Advances in imaging and techniques have increased the percent of patients who are eligible for surgical intervention, but a significant percent of patients still present with surgically unresectable disease (3) or will subsequently develop metastatic disease. Histology The importance of histology in predicting the biologic characteristics and clinical behavior of renal cancers was recognized in the last decade. Renal cell carcinoma (RCC) represents a group of histologic subtypes with unique morphologic and genetic characteristics (4). Clear-cell renal carcinoma is the most common type of renal cancer, accounting for 70 85% of renal epithelial malignancies, and arises from the proximal convoluted tubule. Papillary renal cancer is the second most common type comprising 10 15% of renal tumors. Understanding histologic subtypes and associated gene alterations has provided the opportunity to develop targeted therapy, and has ultimately lead to the development of a new treatment paradigm. v

2 vi Preface von Hippel Lindau (VHL) Syndrome The von Hippel Lindau (VHL) syndrome provided a unique opportunity to study the development of clear-cell tumors and delineate the genetic characteristics of this tumor. In sporadic renal cancer, both the maternal and paternal VHL alleles are inactivated by acquired mutations, whereas in the VHL syndrome the first mutation is inherited. Loss of VHL function may occur in 60 80% cases of sporadic clearcell renal carcinomas (5). The VHL protein is the product of the VHL gene, functions as a tumor-suppressor gene, and is responsible for ubiquination of hypoxia-inducible factor-α (HIF-α) and its subsequent degradation by the proteosome (5). Under hypoxic conditions or in the presence of abnormal VHL function, HIF-α accumulates and activates the transcription of a variety of hypoxia-inducible genes. These include vascular endothelial growth factor (VEGF), platelet-derived growth factor-β (PDGF-β), transforming growth factor-α (TGF-α), and erythryopoietin (EPO). The VHL gene may control this process by suppressing angiogenesis, but loss of the VHL gene or its function allow increased secretion of factors such as VEGF and produces the vascular phenotype characteristic of clear-cell carcinoma. Blocking components of the VEGF pathway and/or the function of HIF-α is currently the major therapeutic strategy for treatment or this malignancy, replacing immunotherapy with cytokines. Systemic Therapy: Metastatic Disease Immunotherapy consisting of interleukin-2 (IL-2) and/or interferon alpha (IFNα) had been the standard approaches for treatment of metastatic RCC, in addition to clinical trials investigating new agents. Responses were best with high-dose intravenous IL-2 (21%) compared to low-dose intravenous IL-2 (11%) and subcutaneous IL-2 (10%), although no survival advantage was observed (6). Similar response rates were reported comparing high-dose IL-2 (23.2%) versus subcutaneous IL-2 plus IFNα (9.9%) and again, no improvement in time to progression (TTP) or survival (7) were seen. IFNα has been established as the standard comparative treatment arm for Phase III clinical trials of new agents for the treatment of metastatic renal cancer. Several randomized trials have demonstrated improvement in medial survival for treated patients (8), and in a retrospective review a median OS of 13.1 months and a median TTP of 4.7 months for IFNα patients were reported (9). A major advance in the field during the past 10 years has been the recognition that a variety of clinical characteristics can be used to categorize patients into groups with differences in prognosis. For previously untreated patients a prognostic model was developed by investigators at Memorial Sloan Kettering Cancer Center (9) and then validated and expanded. Five clinical characteristics were identified (9)

3 Preface vii and later validated at the Cleveland Clinic (10). These prognostic criteria have been utilized in Phase III clinical trials of the targeted agents, such as sorafenib, sunitinib, temsirolimus (CCI-779), and bevacizumab. The cloning of the VHL tumor-suppressor gene and the elucidation of its role in up-regulating growth factors associated with angiogenesis have provided insights into RCC biology, as well as defining a series of potential targets for novel therapeutic approaches. The highly vascularized nature of this neoplasm has ultimately been utilized to control its growth and survival. VEGF and its receptors (VEGFR) are overexpressed in RCC compared to normal renal tissue, and VEGFR-2 is believed to be the major receptor mediating the angiogenic effects of VEGF (11). The binding of VEGF to the extracellular domain of the VEGFR induces tyrosine autophosphorylation and subsequent increases in tumor-associated angiogenesis, endothelial cell proliferation, migration, and enhanced survival. During the past 5 years a number of agents inhibiting the VEGF pathway have been investigated in advanced RCC patients, and a series of these have produced significant clinical benefit including increases in progression-free and OS. This group of novel agents has formed the central part of the new treatment paradigm for this tumor. The purpose of the current textbook is to provide an overview of these developments, as well as provide insights into the other targeted approaches that may ultimately play a role in the treatment of patients with this tumor. Chapters include a discussion of the biologic rationale for each target, as well as potential clinical approaches to provide inhibition of the pathway. The clinical data supporting the current approaches utilizing agents, such as sunitinib, sorafenib, temsirolimus, and bevacizumab, are outlined. In addition, novel targets including tumor necrosis factor, EGFR, Smac/DIABLO, and EpH2A are discussed in detail. The approval of three new agents for treatment of advanced RCC in 2007, and the likelihood that two additional drugs will receive regulatory approval in , make RCC a disease where not only significant clinical progress has occurred, but also an area that will be exploited to increase our understanding of how angiogenesis inhibitors function biologically and clinically. The treatment paradigm for patients with localized and advanced RCC has changed dramatically in the last 5 10 years. Surgical advances are now mirrored by the dramatic changes in therapy available for metastatic disease. The collection of chapters in this text provides an update for urologists, medical oncologists, and researchers interested in the biology and therapy of this tumor. References 1. American Cancer Society. Cancer Facts & Figures Atlanta, GA: Author; Hock LM, Lynch J, Balaji KC. Increasing incidence of all stages of kidney cancer in the last 2 decades in the United States: an analysis of surveillance, epidemiology and end results program data. J Urol. 2002; 167: Russo P. Renal cell carcinoma: presentation, staging, and surgical treatment. Sem in Oncol. 2000; 27:

4 viii Preface 4. Kovacs G, Akhtar M, Beckwith BJ, et al. The Heidelberg classification of renal cell tumours. J Pathology. 1997; 183: Kim, W.Y., Kaelin, W.G. The role of VHL gene mutation in human cancer. J Clin Oncol. 2004; 22: Yang JC, Sherry RM, Steinberg SM, et al. Randomized study of high-dose and low-dose interleukin-2 in patients with metastatic renal cancer. J Clin Oncol. 2003; 21: McDermott DF, Regan MM, Clark JI, et al. Randomized phase III trial of high-dose interleukin-2 versus subcutaneous interleukin-2 and interferon in patients with metastatic renal cell carcinoma. J Clin Oncol. 2005; 23: Medical Research Council and Collaborators. Interferon alfa and survival in metastatic renal carcinoma: early results of a randomized controlled trial. Lancet. 1999; 353: Motzer RJ, Bacik J, Murphy BA, et al. Interferon-alfa as a comparative treatment for clinical trials of new therapies against advanced renal cell carcinoma. J Clin Oncol. 2002; 20: Mekhail TM, Abou-Jawde RM, BouMerhi G, et al. Validation and extension of the Memorial Sloan-Kettering prognostic factors model for survival in patients with previously untreated metastatic renal cell carcinoma. J Clin Oncol. 2005; 23: Ferrara N, Gerber HP, LeCouter J. The biology of VEGF and its receptors. Nat Med. 2003; 9(6):

5 Contents Targeted Therapy for Metastatic Renal Cell Carcinoma: Overview... 1 Ronald M. Bukowski, Robert A. Figlin, and Robert J. Motzer Molecular Genetics in Inherited Renal Cell Carcinoma: Identification of Targets in the Hereditary Syndromes Nadeem Dhanani, Cathy Vocke, Gennady Bratslavsky, and W. Marston Linehan Molecular Targets in Renal Tumors: Pathologic Assessment Ming Zhou Interferons and Interleukin-2: Molecular Basis of Activity and Therapeutic Results Thomas E. Hutson, Snehal Thakkar, Peter Cohen, and Ernest C. Borden The Molecular Biology of Kidney Cancer and Its Clinical Translation into Treatment Strategies William G. Kaelin Jr. and Daniel J. George VEGF: Biologic Aspects and Clinical Approaches W. Kimryn Rathmell and Brian I. Rini VEGF and PDGF Receptors: Biologic Relevance and Clinical Approaches to Inhibition John S. Lam, Robert Figlin, and Arie Belldegrun Sunitinib and Axitinib in Renal Cell Carcinoma Robert J. Motzer Sorafenib in Renal Cell Carcinoma Saby George and Ronald M. Bukowski ix

6 x Contents Additional Tyrosine Kinase Inhibitors in Renal Cell Carcinoma Brian I. Rini Integrin a5b1 as a Novel Therapeutic Target in Renal Cancer Vanitha Ramakrishnan, Vinay Bhaskar, Melvin Fox, Keith Wilson, John C. Cheville, and Barbara A. Finck Carbonic Anhydrase IX: Biology and Clinical Approaches Brian Shuch, Arie S. Belldegrun, and Robert A. Figlin Monoclonal Antibody G250 Recognizing Carbonic Anhydrase IX in Renal Cell Carcinoma: Biological and Clinical Studies J. C. Oosterwijk-Wakka, Otto C. Boerman, Peter F. A. Mulders, and Egbert Oosterwijk Chemokines in Renal Cell Carcinoma: Implications for Tumor Angiogenesis and Metastasis Karen L. Reckamp, Robert A. Figlin, and Robert M. Strieter PI3K/Akt/mTOR Pathway: A Growth and Proliferation Pathway Daniel Cho, James W. Mier, and Micheal B. Atkins EGFR and HER2: Relevance in Renal Cell Carcinoma Eric Jonasch and Cheryl Lyn Walker Proteasome NFkB Signaling Pathway: Relevance in RCC Jorge A. Garcia, Susan A.J. Vaziri, and Ram Ganapathi The Role of Hepatocyte Growth Factor Pathway Signaling in Renal Cell Carcinoma Benedetta Peruzzi, Jean-Baptiste Lattouf, and Donald P. Bottaro Smac/DIABLO: A Proapoptotic Molecular Target in Renal Cell Cancer Yoichi Mizutani, Akihiro Kawauchi, Benjamin Bonavida, and Tsuneharu Miki EphA2: A Novel Target in Renal Cell Carcinoma Mayumi Kawabe, Christopher J. Herrem, James H. Finke, and Walter J. Storkus Restoring Host Antitumoral Immunity: How Coregulatory Molecules Are Changing the Approach to the Management of Renal Cell Carcinoma Brant A. Inman, Xavier Frigola, Haidong Dong, James C. Yang, and Eugene D. Kwon

7 Contents xi The Role of Gangliosides in Renal Cell Carcinoma Philip E. Shaheen, Ronald M. Bukowski, and James H. Finke Tumour Necrosis Factor Misnomer and Therapeutic Target Marina Parton, Tanya Das, Gaurisankar Sa, James Finke, Tim Eisen, and Charles Tannenbaum Molecular Markers for Predicting Prognosis of Renal Cell Carcinoma Mark Nogueira and Hyung L. Kim Adjuvant Therapy for Renal Cell Carcinoma: Targeted Approaches A. Karim Kadar and Christopher G. Wood Index

8 Molecular Genetics in Inherited Renal Cell Carcinoma: Identification of Targets in the Hereditary Syndromes Nadeem Dhanani, Cathy Vocke, Gennady Bratslavsky, and W. Marston Linehan Abstract Kidney cancer affects 51,000 in the United States each year and is responsible for nearly 13,000 deaths annually. Kidney cancer is not a single disease; it is made up of a number of different types of cancer that occur in the kidney. These distinct forms of kidney cancer each have a different histologic type, a different clinical course, respond differently to therapy, and are caused by different genes. The VHL gene is the gene for the inherited form of clear cell kidney cancer associated with von Hippel Lindau as well as for the common form of sporadic, noninherited clear cell kidney cancer. The product of the VHL gene forms a complex with other proteins and this complex targets the hypoxia-inducible factors (HIF) for ubiquitin-mediated degradation. A number of novel agents which target the VHL-HIF pathway have recently been approved by the FDA for treatment of patients with advanced kidney cancer. The MET gene is the gene for the inherited form of papillary kidney cancer associated with hereditary papillary renal carcinoma (HPRC) and has been found mutated in a subset of tumors from patients with sporadic, type I papillary kidney cancer. Clinical trials are currently underway evaluating the role of agents which target the MET pathway in patients affected with HPRC as well as sporadic papillary kidney cancer. The BHD gene is the gene for the inherited form of chromophobe kidney cancer associated with Birt Hogg Dubé (BHD). Biochemical studies have revealed that the BHD pathway interacts with the MTOR pathway and agents which block this pathway are currently being evaluated in preclinical models as a potential approach for the treatment of BHD-associated as well as sporadic chromophobe kidney cancer. The Krebs cycle enzyme, fumarate hydratase, is the gene for the inherited form of type II papillary kidney cancer associated with hereditary leiomyomatosis renal cell carcinoma (HLRCC). In vitro and in vivo studies are currently underway evaluating novel approaches for targeting of this kidney cancer pathway. It is hoped that understanding the genes that cause cancer of the kidney will provide the foundation for the development of effective forms of therapy for patients with this malignancy. W.M. Linehan ( ) Urologic Oncology Branch, National Cancer Institute, 10 Center Drive MSC 1107, Bldg 10 CRC Room1-5940, Bethesda, MD WML@nih.gov R.M. Bukowski et al. (eds.), Renal Cell Carcinoma, DOI: / _2, 13 Humana Press, a part of Springer Science + Business Media, LLC 2009

9 14 N. Dhanani et al. Keywords Kidney neoplasms VHL von Hippel Lindau BHD Birt Hogg Dubé Met Fumarate hydratase 1 Introduction An estimated 36,000 people will be diagnosed with kidney cancer this year, and close to 30% of these patients will die of their disease (1). On the basis of the data from the National Cancer Institute s SEER Cancer Statistics Review, 1 out of every 82 men and women will be diagnosed with cancer of the kidney over the course of their lifetime, and the incidence continues to rise. Men are affected almost twice as often as women, and the incidence among blacks is slightly higher than that of whites. With the increased use of axial abdominal imaging, kidney tumors are being diagnosed at earlier stages, often times incidentally while the patient is still asymptomatic. Nonetheless, at the time of presentation, 40% of these tumors will no longer be confined to the kidney, either through local extension or distant metastatic spread (2). While extirpative surgery is most often curative in tumors restricted to the kidney, treatment of metastatic disease has proven to be a formidable challenge with the traditional therapies currently available. As our understanding of the genetic basis of kidney cancer increases, exciting advances in molecular therapeutics offer novel approaches to treatment for these patients. 2 Identification of the VHL Gene As recently as 30 years ago, very little was known about the contribution of genetic mutations to the development of renal tumors. Like cancer of the colon, breast, and prostate, kidney cancer was known to occur in both sporadic and familial forms. On the basis of the earlier work by Knudson and Strong (3, 4), the concept of tumor suppressor gene inactivation was recognized as an etiological factor in Wilms tumor and retinoblastoma. In keeping with this hypothesis, when compared to the sporadic form, familial kidney cancers were more often multifocal, bilateral, and had earlier onset. Still, there was no gene identified that could be implicated in renal cancer. In 1979, Cohen et al. (5) noted a chromosomal translocation between the short arm of chromosome 3 and the long arm of chromosome 8 in eight of ten affected members of a family known to have heritable kidney cancer. This was followed by several additional reports characterizing chromosomal abnormalities in different families affected with renal cell carcinoma (RCC). In each of these lineages, chromosome 3 was involved, particularly the 3p13-3p14 region. Important insight into the link between the hereditary and sporadic forms of RCC was provided through work by Zbar and colleagues (6) when they reported loss of alleles at loci

10 Molecular Genetics in Inherited Renal Cell Carcinoma 15 on the short arm of chromosome 3 in 11 of 11 evaluable patients with sporadic renal cancers. Shortly thereafter, it was postulated that the genetic mutation responsible for von Hippel Lindau (VHL) was located in a region on chromosome 3p, distinct from the human homologue of RAF1, but apparently linked to it (7). All evidence was pointing to the existence of a tumor suppressor gene encoded on the short arm of chromosome 3, a mutation of which resulted in renal cell cancer. Unfortunately, gene localization was still not feasible because the region of interest was too large for the cloning techniques available at the time. Research efforts were thus shifted to the heritable form of renal cancer, with VHL being the model for investigation. 2.1 von Hippel Lindau VHL is transmitted in an autosomal-dominant pattern with an estimated incidence of 1 in 36,000 live births (8, 9). With a penetrance of over 95% by the age of 65 (10), affected individuals develop neoplastic tumors in multiple organ systems. Central nervous system lesions include retinal hemangioblastomas, endolymphatic sac tumors of the inner ear, and craniospinal hemangioblastomas in the cerebellum, brainstem, spinal cord, lumbosacral nerve roots, and supratentorial lesions. The pancreas may also be affected in these patients, developing cysts, cystadenomas, and neuroendocrine tumors. Benign epididymal papillary cystadenomas occur with increased frequency than the general population and can be bilateral. Rarely, women can have analogous lesions with papillary cystadenomas in the broad ligament. Tumors found in the kidney are solid renal cell cancers, simple cysts, and combinations thereof. In the setting of VHL, it has been estimated that a kidney may contain 600 microscopic tumors and over 1,000 cysts before the age of 40 years (11). Although simple cysts in these patients rarely transform to solid masses (12), complex cysts are known to contain malignant elements and may progress if left untreated. Adrenal lesions found in VHL are pheochromocytomas. Like the kidney tumors, these are frequently multiple and bilateral. Extra-adrenal paragangliomas are also known to occur in these patients, arising in periaortic tissues, the carotid body, and the glomus jugulare (9). In searching for the gene responsible for VHL, researchers explored the applicability of Knudson s two-hit hypothesis to tumor behavior in VHL patients with kidney cancer. Tory et al. (13) evaluated tissue from patients with multiple kidney tumors, and for each patient compared chromosome 3 from one tumor to another. They found that each patient had loss of the same allele of chromosome 3p in all of their tumors. Further analysis of haplotypes revealed that the lost allele was always from the wild-type chromosome, the contribution of the nonaffected parent. This provided strong support for the notion that alteration of a tumor suppressor gene was the causative factor in VHL, and in accordance with Knudson s theory, an

11 16 N. Dhanani et al. individual with a germline mutation was at risk for VHL if they incurred a second hit at the same locus thus inactivating the wild-type allele. Expanding upon earlier work by Seizinger et al. (14), Lerman s group (15) isolated and mapped 2,000 single copy DNA fragments of chromosome 3 from humans, thus generating vital tools which would be used for the future cloning of the VHL gene. With these reagents newly available, Hosoe and colleagues (16) performed further multipoint linkage analysis to localize the VHL gene to an interval between RAF1 and a polymorphic DNA marker, D3S18. Finally, in 1993, researchers at the National Cancer Institute reported identification of the VHL gene through cloning studies and described its role in RCC (17). This small gene, with 854 coding nucleotides on three exons, was found to be located on the short arm of chromosome 3 and responsible for encoding the VHL protein. The gene is evolutionarily conserved and its product shares homology with only a small region of a surface membrane protein of Trypanosoma brucei. Once the causative gene for VHL had been identified, clinicians were eager to find screening methods to identify patients with genetic mutations. Early laboratory studies generated germline mutation detection rates of 39 75% (18, 19). Mutation analyses showed that the type (e.g., insertion, deletion, missense, or nonsense) and location (e.g., codon position) of mutation correlated well with phenotype, thus allowing health care providers to predict the extent of involvement of the various organ systems for any given VHL family. In a study of 469 VHL families from North America, Europe, and Japan, researchers compared the effects of identical VHL germline mutations on different families. On the basis of their findings, VHL was broken down into three distinct phenotypes: pheochromocytoma along with RCC, pheochromocytoma alone, and RCC alone (20). Later studies correlated the relationship between length and location of germline mutations and the incidence of RCCs in VHL patients. A retrospective review of 123 patients from 55 families revealed that individuals harboring a partial deletion suffered a significantly higher rate of RCC when compared to those with complete gene deletions. Moreover, deletion mapping demonstrated the presence of a 30-kb gene on the short arm of chromosome 3, directly adjacent to the VHL gene which, when preserved, may promote the development of RCC (21). Further advances were made when Stolle s group (22) developed a new technique which improved germline mutation detection, accurately identifying a mutation in 93 out of 93 (100%) VHL families tested. The method involved a combination of tests that each demonstrated high sensitivity for the various types of mutations implicated in VHL. Qualitative Southern blotting to detect gene rearrangements and quantitative Southern blotting for the detection of entire gene deletions were the newly added components responsible for the dramatic increase in sensitivity. In addition, fluorescence in situ hybridization (FISH) and full gene sequencing completed the battery of tests. The 100% sensitivity of the new technique lent support to the notion that VHL is genetically homogeneous, and clinically allowed providers to counsel patients with reasonable certainty that a family member found to lack the gene mutation with the new test combination was unlikely to have VHL.

12 Molecular Genetics in Inherited Renal Cell Carcinoma Sporadic RCC Discovery of the VHL gene in the setting of familial RCC allowed scientists to then investigate its role in sporadic tumors. Gnarra et al. (23) used PCR amplification of the three exons of the VHL genes of 108 patients with sporadic RCC in order to analyze the entire coding region in each gene. They identified somatic mutations in the VHL gene in 57% of these patients, and nearly all (98%) were found to have loss of heterozygosity. It was clear that the VHL gene played a role in the development of sporadic RCC in a majority of patients; however, questions arose as to why gene mutations were not demonstrable in all renal cell cancers. One explanation is offered by an important mechanism for VHL gene inactivation as described by Herman and colleagues (24). They discovered hypermethylation of a CpG island in the 5 region of the VHL gene, a region which is normally unmethylated, in nearly 20% of VHL patients with RCC. No other mutation of the VHL gene could be demonstrated in 80% of these patients, and VHL gene expression was absent in all. Furthermore, when treated with 5-aza-2 deoxycytidine, a hypomethylating agent, the VHL gene was once again expressed. Additionally, one has to consider the limitations of current investigative techniques. There are still regions of the VHL gene which have not yet been thoroughly examined and this may hinder our ability to fully detect genetic variation. Furthermore, there is always the possibility of normal tissue interspersed with cancerous cells within a given tumor, thus confounding laboratory findings (25) Cystic Lesions in VHL In addition to solid RCCs, patients with VHL are also frequently found to have cystic lesions within their kidneys (Fig. 1). These lesions range from simple benign cysts, as characterized by radiographic imaging, to complex cystic masses suspicious for malignancy. In this patient population which can be expected to develop numerous multifocal and bilateral lesions requiring surgical extirpation, maximal nephron preservation relies upon the clinician s ability to predict the malignant potential of a cyst or mass, and the likelihood that treatment of that lesion will improve survival. In order to better characterize the relationship between cysts and solid renal masses, Lubensky et al. (26) analyzed 26 renal lesions from two VHL patients for loss of heterozygosity at the VHL region. They found loss of a VHL allele in 25 out of the 26 lesions, thereby demonstrating both benign and malignant lesions to share similar genetic aberration. In both sets of lesions, the mutated gene remained while the normal copy was the one that was lost, thus keeping with Knudson s two-hit hypothesis. Further evidence to support the theory that renal cysts potentially represent precursors to malignant RCC in VHL was provided by the work of Lee et al. (27) when they showed the consistent coexpression of erythropoietin and erythropoietin receptor in RCC as well as many renal cysts. Knowing that simple cysts harbored the same genetic abnormality as solid malignant lesions, clinicians were then faced with the dilemma of when to act on cysts found in the kidneys of VHL patients. If left untreated, simple cysts

13 18 N. Dhanani et al. Central nervous system Retina Cerebellum Brainstem Spinal cord Endolymphatic sac Visceral organs Kidneys Adrenal glands Pancreas Broad ligament (female) Testes (male) Fig. 1 Phenotypic manifestations of VHL. Renal masses are common in VHL patients. a CT scan of a VHL patient demonstrating characteristic bilateral multifocal renal lesions consisting of simple and complex cysts as well as enhancing solid masses. b Gross specimen removed from a VHL patient showing classic multiple golden-yellow tumors. c H&E stain of a classic clear cell renal carcinoma found in patients with VHL. d In addition to renal manifestations, VHL affects organs systems throughout the body. From Linehan et al. (76) (See Color Plates) may develop malignancy over time, and a plan of observation may prove fatal if progression to metastatic disease ensued. On the other hand, unnecessarily operating on benign lesions could lead to a dramatic increase in morbidity for VHL patients, including the perioperative risks of surgery as well as the subsequent renal insufficiency from loss of parenchyma. Thus, investigators focused on determining the natural history of cystic lesions in the VHL population (12). Two hundred and twenty-eight renal lesions from 28 patients were observed for a mean of 2.4 years with serial computed tomography scans. Overall, 74% of the cysts remained stable with respect to size, with an additional 9% actually decreasing in size. Only 2 patients were found to have malignant transformation of their simple cysts based on radiographic criteria. These results supported the practice of conservative management of simple cysts in the VHL population.

14 Molecular Genetics in Inherited Renal Cell Carcinoma Function of the VHL Gene Once the putative gene for RCC was identified, there was an effort to better define the function of the VHL protein, with the hope that this would eventually uncover potential therapeutic targets. One method of determining the function of a protein is to find out what other proteins it complexes with in order to reveal its role in a cellular pathway. In 1995, Duan and colleagues (28) localized the VHL gene product to the cytosol and the nucleus, indicating common translocation of the protein. They were also able to identify two additional proteins of 16 and 9 kda which formed a heterotrimeric complex with VHL. When certain missense mutations of the VHL gene were investigated, the complex did not form. Subsequent studies (29) offered a more detailed description of the protein complex. They explained the function of a transcription elongation factor, Elongin (SIII), made up of three distinct protein subunits, Elongins A, B, and C, which serves to prevent transient pauses of RNA polymerase II (Pol II) during transcription. Although VHL protein was shown to displace Elongin A and compete for binding with Elongins B and C in vitro, there was no evidence of such function in vivo. Iliopoulos et al. (30) demonstrated the effects of VHL protein on certain hypoxia-inducible genes. Under normoxic conditions, intact VHL was shown to downregulate vascular endothelial growth factor (VEGF), platelet-derived growth factor B (PDGF-B), and the glucose transporter GLUT1 by destabilizing their respective mrnas. Thus, presumably, with a VHL mutation there was unregulated expression of these proteins, a finding which was congruent with the known hypervascular characteristics of VHL-associated RCCs. In a search for proteins that interact with the VHL B C complex, Pause and colleagues (31) identified Hs-CUL-2, a newly described gene involved in cell cycle regulation of yeast and Caenorhabditis elegans. They observed that in the presence of a VHL gene mutation, the VHL- B-C-Hs-CUL-2 interaction was markedly diminished, suggesting a tumor suppressor role for this new protein. It was known that VEGF, GLUT1, and PDGF are all targets of hypoxia- inducible factor (HIF) and also that the clear cells of RCC express higher levels of these proteins than nonmalignant cells (32). The role of VHL was further elucidated when researchers showed that the previously described protein complex of VHL- B-C-CUL functioned as a ubiquitin ligase that targets HIF1α and HIF2α for degradation under normoxic conditions (33). Upon hydroxylation by oxygen-dependent prolyl hydroxylases, HIF1α binds to VHL and is subsequently degraded (34). If the hydroxylation does not occur, however, VHL binding is inhibited and ubiquitination of HIF1α fails (35). Transcription of HIF-dependent genes ensues leading to overexpression of VEGF and ultimately increased vascularity. Lending support to this pathway, Maranchie et al. (36) used a competitive inhibitor of the VHL-HIF1α binding site to assess functional outcomes. In preventing this interaction, they found accumulation of cellular HIF1α in normoxia and a conversion to the VHLnegative phenotype.

15 20 N. Dhanani et al. 3 Hereditary Papillary Renal Carcinoma While advances were being made in the genetic basis of RCC resulting from VHL mutations, in 1994 clinicians were uncovering a distinct familial syndrome which was also manifest by renal tumors. Zbar and colleagues (37) reported on a family in which renal tumors had developed in three generations, and whose tumors were multifocal and bilateral. Pathologically these tumors were papillary variants of RCC, as opposed to the conventional type associated with VHL, and they showed no abnormalities in chromosome 3. This new syndrome, termed hereditary papillary renal carcinoma (HPRC), appeared to have an autosomal dominant mode of inheritance with incomplete penetrance. Further analysis of 10 families with HPRC suggested renal cancers occur in both sexes, with a male:female ratio of 2.2:1, have Fig. 2 Manifestations and genetics of HPRC. Patients with HPRC primarily develop bilateral multifocal renal masses. a Abdominal CT demonstrates HPRC tumors with characteristic poor enhancement on contrasted study that may frequently be mistaken for simple cysts. The tumors are best seen on late phase images of a contrast CT. b Low and c high power H&E stain of type I papillary RCC seen in patients with HPRC. d Fluorescence in situ hybridization (FISH) using a MET probe demonstrating trisomy of chromosome 7 (red signal) in papillary type I RCC compared with chromosome 11 serving as control (green signal). From Schmidt et al. (42) (See Color Plates)

16 Molecular Genetics in Inherited Renal Cell Carcinoma 21 a late age of onset (50 70 years), are bilateral and multifocal in nature (38). A later study evaluated 88 surgical pathology slides of grossly normal areas of 12 kidneys from patients with HPRC. More than half of these samples were found to contain microscopic papillary renal cancers, thereby predicting the presence of 1,100 3,400 microscopic tumors in a single kidney of a patient with HPRC (39). Histologically these tumors display a distinct phenotype, with a majority of the architecture in a papillary/tubulopapillary pattern and a chromophil basophilic staining, consistent with a type I papillary renal carcinoma phenotype (40). Radiographically in stark contrast to the hypervascular tumors of VHL, tumors of HPRC display poor contrast enhancement and are markedly hypovascular (41, 42) (Fig. 2). 3.1 Identification of the Gene for HPRC Three years after describing the disease, researchers reported identification of the gene responsible for HPRC (43). Findings of chromosomal trisomy in malignant papillary renal carcinomas raised suspicions of proto-oncogene gene dysfunction and the defect was mapped to the long arm of chromosome 7. Missense mutations in the tyrosine kinase domain of the MET gene ultimately proved responsible for constitutive activation of the MET protein and interference with autoinhibitory mechanisms, resulting in papillary renal cancers. The MET transmembrane protein was found to be a receptor site for hepatocyte growth factor (HGF) also termed Scatter factor (SF) (44). Upon activation by HGF, MET tyrosine phosphorylation induces a host of signaling cascades responsible for embryonic development, cell branching, and invasion (45). 4 Birt Hogg Dubé In 1977, three physicians described a familial syndrome in which affected individuals developed multiple small skin-colored papules on the face, neck, and back (46). Histologically these lesions were found to be fibrofolliculomas, trichodiscomas, and acrochordons, and they were transmitted in an autosomal dominant pattern. Some patients with this constellation of findings, termed Birt Hogg Dubé (BHD), were also known to have concurrent visceral tumors, including thyroid carcinoma, colonic polyps, and one case of a renal tumor. In 1999, a group of clinicians noted that a significant number of their renal mass patients had these distinctive skin lesions that had previously been described in dermatologic literature. They therefore set out to evaluate a large cohort of patients with known familial renal tumors and assess the presence of cutaneous findings. As a result, Toro and colleagues (47) found three extended families in whom there appeared to be common segregation of renal tumors and the cutaneous lesions of BHD. They concluded that BHD seemed to be associated with renal tumors, both transmitted in an autosomal dominant manner.

17 22 N. Dhanani et al. As BHD began to attract more attention and closer scrutiny, numerous additional disease processes were identified in BHD patients. Spontaneous pneumothoraces, parotid oncocytomas, multiple lipomas, angiolipomas, parathyroid adenomas, and colonic polyposis were all postulated to have some connection with BHD (48 51). In order to better define the spectrum of disease processes associated with BHD, Zbar and colleagues (52) solicited participation from patients who were under the care of dermatologists from across the United States and Canada for classic BHD skin lesions. The patients were evaluated for concomitant health problems, particularly kidney, lung, and colon manifestations. The group eventually found no correlation between BHD and colon cancer or polyps. There was, however, a strong link Fig. 3 Phenotypic manifestations of BHD. Classic findings in BHD include (a) characteristic cutaneous fibrofolliculomas, (b) pulmonary cysts that result in a 30-fold increased incidence of spontaneous pneumothoraces, and (c) renal tumors that are usually multifocal and can vary in pathologic subtype, from (d) chromophobe RCC (most common) to oncocytoma, hybrid tumors, or clear cell carcinoma. From Zbar et al. (53) (See Color Plates)

18 Molecular Genetics in Inherited Renal Cell Carcinoma 23 Fig. 4 Phenotypic manifestations of HLRCC. a Classic cutaneous leiomyomatas presenting as multiple firm and erythematous macules and papules that are frequently painful. b Abdominal CT scan showing multiple uterine leiomyomas. This often leads to hysterectomy in HLRCC-affected women in their 20s or 30s. c CT abdomen demonstrating anterior upper pole mass in the left kidney. The renal lesions of HLRCC patients may present early and frequently have an aggressive clinical course. From Toro et al. (62) (See Color Plates) between BHD and renal tumors, as previously suspected, as well as spontaneous pneumothoraces. On multivariate analysis, patients with BHD had an odds ratio of ~9.0 for developing renal tumors, and a risk of developing spontaneous pneumothoraces 32 times higher than the general population (Fig. 3). In order to better characterize the renal neoplasms associated with BHD, researchers examined the pathologic findings of 130 renal tumors from 30 BHD patients from 19 different families (53). Close to 35% of the tumors were pure chromophobe variants of RCC, with an additional 50% being a hybrid of chromophobe RCC and oncocytoma. Less than 10% of the entire cohort had elements of clear cell (conventional) RCC. When present, the clear cell RCC were larger, with a mean diameter of 4.7 cm, versus the chromophobe tumors which averaged 3.0 cm, or the hybrid tumors with a mean diameter of 2.2 cm. Furthermore, analysis of grossly normal appearing surrounding renal parenchyma revealed multifocal oncocytosis throughout a majority of the specimens (Fig. 4). 4.1 Identification of the BHD Gene Knowledge of the genetic basis for BHD came largely in part from work by Schmidt and colleagues (54). Linkage analysis was used to localize the BHD gene to a locus on the short arm of chromosome 17 from a screen of the genome of a large BHD kindred. Further work by Nickerson et al. (55) utilized recombination mapping to localize the gene to a region of 17p11.2. A novel gene in this region was determined to exhibit mutations in the germlines of affected patients. The gene product, folliculin, was truncated as a result of insertions, deletions, or nonsense mutations. The frequency with which BHD is inactivated as a result of genetic

19 24 N. Dhanani et al. mutations suggested a tumor suppressor function. Vocke and coworkers (56) found support for this theory when they sequenced the DNA of 77 renal tumors from 12 patients with germline BHD mutations. They demonstrated a high frequency of mutations in the wild-type BHD allele, thus providing the second inactivating hit. The 579 amino acid protein, named for the hallmark dermatologic findings of the syndrome, has no known functional domains, but is highly preserved across species. BHD mrna expression as measured by FISH has been demonstrated in 17 human tissues, including the kidney, lung, skin, and brain (57). 5 Hereditary Leiomyomatosis Renal Cell Carcinoma A fourth familial syndrome of renal cancer was recently described by Launonen et al. (58). They noted cosegregation of cutaneous leiomyomas and type II papillary renal cell carcinoma in two familial lines (Fig. 4). This syndrome, termed hereditary leiomyomatosis renal cell carcinoma (HLRCC), was mapped to a 14-cM region on the long arm of chromosome 1 (59). Fumarate hydratase, the product of the putative gene for this syndrome, is a catalyst for the conversion of fumarate to malate in the 2C Acety 1 CoA NADH+H + Oxaloacetate 4C 6C Citric acid Malic acid 4C Fumaric acid 4C FH CO 2 6C Isocitric acid NADH+H + FADH 2 GTP P CO 2 5C a -Ketoglutaric acid NADH+H + 4C Succinic acid Shift towards glycolysis as an energy source Upregulation of HIF and HIF-dependent pathways Fig. 5 In HLRCC, mutation of the FH gene leads to dysfunctional fumarate hydratase, one of the key regulatory enzymes in the Kreb s cycle, necessary for mitochondrial respiration and oxidative energy production. This in turn leads to accumulation of fumarate but more importantly to preferential energy production from glycolysis, a phenomenon observed in other malignancies as well. It may also lead to upregulation of HIF and HIF-dependent pathways

20 Molecular Genetics in Inherited Renal Cell Carcinoma 25 Krebs cycle, and its activity is diminished in leiomyomatous tumors (60). The loss of FH function and impediment of the Krebs cycle creates reliance upon glycolytic metabolism and upregulation of HIF and HIF-inducible transcripts (61) (Fig. 5). The resultant environment is ideal for tumor cell survival and proliferation. The largest reported series of HLRCC patients revealed a 93% germline FH mutation detection rate in families suspected of harboring disease with an autosomal dominant inheritance pattern (62, 63). Details of the molecular mechanisms involved in the downstream pathway of this gene are still under investigation, but the renal cancers associated with it appear to be aggressive and lethal if allowed to progress. 6 Treatment 6.1 Localized Disease As our understanding of renal malignancies has evolved, so have our treatment strategies. In 1869, Gustav Simon performed the first planned nephrectomy in the treatment of a ureterovaginal fistula. A century later, Robson and colleagues (64) described refined techniques for radical nephrectomy for renal malignancies. Surgical extirpation remains the gold standard for treatment of localized renal cell carcinoma, although the surgical techniques have become more sophisticated. Since the first laparoscopic radical nephrectomy performed by Clayman (65), great strides have been made in minimally invasive approaches to removing kidneys. Laparoscopy offers patients decreased morbidity as compared with the historical open surgical procedures while not appearing to compromise cancer control. In the setting of localized renal tumors, nephron sparing surgery is becoming more common. In 1890, Czerny performed the first partial nephrectomy for malignancy. Since that time, the scope has increased with surgeons proposing a wide range of acceptable size limits for nephron sparing surgery, with general consensus around 4 cm in diameter (66, 67). Here, too, minimally invasive approaches are being employed and laparoscopic partial nephrectomies are now being performed at specialized centers across the country. Preservation of renal function and maximal sparing of nephrons during therapy is of paramount importance when treating patients with familial syndromes who are at risk for developing multiple, recurrent, bilateral tumors, and may require numerous therapeutic interventions over their lifetime. Nonetheless, cancer control cannot be compromised. In order to minimize the morbidities associated with renal replacement therapy while maintaining vigilance in the containment of cancer, a threshold of 3 cm has been employed whereby tumors are observed until they reach this size criterion (68). In determining the safety of this guideline, researchers found no patients developed metastatic disease nor did they require dialysis when the 3 cm rule was adhered to. In addition to surgical extirpation, ablative techniques have also been employed for the treatment of renal tumors. Thermal tissue ablation with radiofrequency

21 26 N. Dhanani et al. energy can be performed either percutaneously or laparoscopically. With higher wattage generators results for radiofrequency ablation appear promising. Hwang et al. (69) reported favorable outcomes for 23 out of 24 patients treated with RFA at a mean follow-up of 1 year. Nonetheless, this is still considered an experimental technique and further studies will need to be conducted with longer follow-up and validation of post-rfa imaging criteria. 6.2 Metastatic Disease Despite high success rates with treatment of localized renal cancers, the prognosis for patients with metastatic disease is far grimmer. Although immunotherapy has been used, with interleukin-2 being the standard treatment modality, overall response rates are only in the range of 15 22% (70). It is obvious that new strategies are needed for the successful treatment of these patients, and molecular therapeutics seem to hold the key. The success of the tyrosine kinase inhibitor STI-571 in combating gastrointestinal stromal tumors and chronic myelogenous leukemia has fueled enthusiasm for further investigation into the molecular mechanisms of oncogenesis and potential pharmacologic disruption of these pathways (71, 72). In the paradigm of renal cancers, molecular therapeutics can be thought of in two broad categories: those that seek to interrupt specific pathways of tumorigenesis and the individual proteins involved, and those that affect the cancer cell s adaptability. Given the variability of each distinct type of renal cancer, it should not be surprising that this heterogeneous group of diseases offers a wide range of unique molecular targets. Our understandings of the mechanisms involved in the familial syndromes greatly impact our ability to direct therapies at their sporadic counterparts. 6.3 Targeting VHL The VHL pathway offers a variety of targets for intervention. In VHL negative cells, the protein complex responsible for promoting HIF degradation is nonfunctional, resulting in the overabundance of HIF in a normoxic state. One therapeutic approach was demonstrated by Rapisarda et al. (73) when they used a small molecule inhibitor of the HIF-1 pathway, topotecan, to block the transcriptional activity of HIF-1. Although effective in reducing the accumulation of HIF-1α in hypoxic environments, the efficacy of topotecan for VHL remains to be determined since in vitro and in vivo studies in human VHL models suggest HIF-2 to be the major factor in oncogenic pathways (74, 75). Efforts are currently under way to better target HIF-2 function (76, 77). Several components of the downstream pathways in HIF have also been targeted (Fig. 6). Failure to adequately inactivate HIF leads to unregulated expression of

22 Fig. 6 VHL gene mutation, downstream effects, and molecular targeting of the VHL pathway. a. With a VHL gene mutation, the VHL complex is disrupted and allows for accumulation of HIF with subsequent activation of downstream pathways for angiogenesis, glucose transport, and growth. b. Inhibition of overaccumulated HIF and prevention of downstream activation with a small molecule is one of the strategies for molecular targeting of the VHL/HIF pathway. c. New tyrosine kinase inhibitors as well as direct VEGF and PDGF receptor blockers are examples of downstream targeting. From Linehan et al. (76)

23 28 N. Dhanani et al. gene products such as VEGF, PDGF, EGF, TGFα, and GLUT1. Pharmacotherapies inhibiting these pathways may offer a systemic modality to combat metastatic disease. Bevacizumab, a monoclonal antibody to VEGF, has been shown to decrease angiogenesis in renal cell carcinoma (78). Another drug, BAY , inhibits signal transduction and subsequent cell proliferation by antagonizing the tyrosine kinase receptors for VEGF and PDGF (79). The receptor of EGF can be blocked individually through the function of either ZD1839 or erlotinib, or in combination with the VEGF receptor by ZD6474 (80 82). 6.4 Altering the c-met Pathway Type I papillary RCC in HPRC has been shown to result from activating mutations in the cell surface tyrosine kinase receptor for HGF, c-met. Upon activation, the c-met receptor is autophosphorylated, thus recruiting multiple signaling molecules to its cytoplasmic domain and activating intra- and extracellular cascades which ultimately contribute to cellular proliferation, scattering, and invasion (45). On the basis of this knowledge, several therapeutic strategies have been proposed: inhibition of autophosphorylation by the prevention of ATP binding, inhibition of the interaction between HGF and its receptor, and suppression of the downstream signaling cascade of activated c-met (76). 6.5 HSP-90 Inhibition An alternative strategy in the molecular targeting of tumorigenesis is to affect the mechanisms used by the cancer cells to adapt and thrive in surrounding environments. One such group of targets is molecular chaperones, termed heat shock proteins (HSPs), which maintain appropriate protein conformation, assist in protein transport, and play a role in antigen presentation. Out of the entire family of molecular chaperones, heat shock protein 90 (HSP-90) has drawn attention for its active role in renal. HSP-90 is part of a complex that stabilizes and promotes the activity of HIF and the receptor tyrosine kinases MET and KIT (76). An inhibitor of HSP-90, 17-allylamino-17-desmethoxygeldanamycin (17-AAG), has been shown to disrupt the function of this complex, thus leading to rapid inactivation and degradation of its client proteins (83). As a result, HIF-dependent transcriptional activity is impaired, thus decreasing the downstream gene products in the HIF pathway. HSP-90 has also been shown to play a role in chromophobe and papillary RCC through its effects on KIT and MET and their downstream pathways (84, 85). In addition to direct inhibition of KIT, HSP-90 inhibitors also function on AKT and RAF, transcription promoters which are stimulated by KIT but are also themselves client proteins of HSP-90. Finally, with respect to MET, HSP-90 inhibitors may have a potential role as an adjunct to angiogenesis inhibitors. Hypoxia has been shown