|Home | About | Journals | Submit | Contact Us | Français|
The basic biology underlying the development of clear-cell renal cell carcinoma (ccRCC) is critically dependent on the von Hippel–Lindau gene (VHL), whose protein product is important in the cell’s normal response to hypoxia. Aberrations in VHL’s function, either through mutation or promoter hypermethylation, lead to accumulation of the transcriptional regulatory molecule, hypoxia-inducible factor alpha (HIFα). HIFα can then dimerize with HIFβ and translocate to the nucleus, where it will transcriptionally upregulate a series of hypoxia-responsive genes, including vascular endothelial growth factor (VEGF), platelet-derived growth factor (PDGF), and others. Binding of these ligands to their cognate receptors activates a series of kinase–dependent signaling pathways, including the RAF–MEK–ERK and phosphatidylinositol-3 kinase–AKT–mTOR pathways. Targeted agents developed and now approved for use in advanced ccRCC include humanized monoclonal antibodies against VEGF, small-molecule tyrosine kinase inhibitors, and inhibitors of mTOR. Understanding the biology of ccRCC is critical in understanding the current therapy for the disease and in developing novel therapeutics in the future. This review will provide an overview of the genetics of ccRCC, with an emphasis on how this has informed the development of the targeted therapeutics for this disease.
There were an estimated 54,390 newly diagnosed cases of kidney cancer in 2008 and over 13,000 deaths.1 For many patients, the disease is localized when it is found, and for them, surgical excision remains the standard of care and offers excellent long-term recurrence-free survival. However, for those patients who either present with metastatic disease or develop distant relapse after their surgery, the scenario is far less sanguine. Renal-cell carcinoma (RCC) is notoriously resistant to traditional cytotoxic chemotherapy, and for several decades the mainstay of therapy was based on immunotherapy, typically through either high-dose inter-leukin-2 (IL-2) or interferon alpha (IFNα). Neither of these approaches was particularly satisfactory. High-dose IL-2 can, rarely, produce long-term complete remission, but this is generally seen in considerably less than 5% of the time and the overall response rates are on the order of 20%.2 This rate also comes at the cost of very substantial toxicity. Although IFNα is less toxic than high-dose IL-2, it is arguably less effective in that the overall response rates are generally lower, and it almost never results in long-term remission. The way forward in treating RCC, at least the most common clear-cell histological variant, had to await a better understanding of the biology of RCC and the subsequent development of the so-called targeted therapies to attack the components of that biology. The purpose of this review will be to provide an overview of the genetics of clear-cell RCC (ccRCC), and in particular, the central role played by the von Hippel–Lindau gene (VHL). It will also link those principles to the targeted therapies, which have become central to the contemporary management of ccRCC. This story represents a satisfying example of how basic science and molecular biology (the bench) has moved into the clinical setting (the bedside), a theme that will likely be repeated across a range of other diseases over the next decade.
Understanding the biology of ccRCC starts with the discovery and characterization of the VHL gene. VHL disease is an inherited neoplasia syndrome, whose hallmark is the development of benign and malignant tumors across several organ systems.3,4 Its incidence is one in 36,000 births, and is inherited as an autosomal dominant trait with over 90% penetrance by the age 65 years in affected individuals.5–7 The characteristic tumors include hemangioblastomas of the central nervous system, retinal hemangioblastomas, pheochromocytomas, and renal neoplasms, including renal cysts and ccRCC.3 ccRCC, in particular, has a profound influence on families with the disease, as the leading cause of death in this group is metastatic ccRCC.7
The VHL gene’s discovery grew at first from the observation that the ccRCC in VHL disease was identical to the sporadic, nonfamilial form in every respect, except that it tended to be multifocal and occured at a younger age. These characteristics fit the profile of a familial tumor syndrome that was secondary to the loss/mutation of a tumor suppressor gene. Earlier studies looking at different kindreds with a familial tendency toward developing sporadic ccRCC had consistently shown aberrations of the short arm of chromosome 3 (3p).8–10 Subsequent studies of both ccRCC tumors and cell lines confirmed abnormalities of chromosome 3p as a unifying theme.11–15 These alterations in chromosome 3p were not present in the corresponding normal tissues and were not present in other histological variants, such as papillary RCC.16,17
Putting together these observations in sporadic ccRCC’s, with the same tumor type being present in VHL disease, suggested that the abnormality on chromosome 3p fit the profile of a tumor suppressor gene. A series of elegant and ground breaking studies of multiple different kindreds with VHL disease localized the VHL gene to a relatively small region on chromosome 3p,18–20 followed in 1993 by the identification of the VHL gene in a seminal article by Latif et al.21 Studies since then have shown aberrations of VHL in the majority of sporadic ccRCCs.22–28 Roughly, half the VHL mutations found in sporadic clear RCC are frameshift mutations, whereas the second most common is a missense mutation.22 Even in cases in which genetic mutations of VHL have not been identified, other aberrations affecting its function were noted, such as abnormal promoter hyper-methylation of VHL, leading to low or absent protein levels.29 For a more in-depth analysis of specific VHL mutations and their potential relationship to disease biology, the interested reader is referred to two other previously published in-depth reviews.30,31
Normally, VHL’s predominant function is to regulate the cell’s response to oxygen availability in the local microenvironment.32–37 VHL exists in the cell cytoplasm in complex with a series of other proteins, specifically elongin B, elongin C, cullin2, and Rbx, as part of an E3 ligase complex.38–44 This complex can ubiquitinate proteins and thereby mark them for subsequent degradation by the cell’s proteasomal machinery.45,46 In the presence of normal local oxygen levels, a regulatory molecule, termed hypoxia-inducible factor alpha (HIFα), is hydroxylated by a series of prolyl hydroxylases. The presence of a hydroxyl group at these proline residues permits HIFα to bind to the E3 ligase enzyme complex, mediated predominantly by VHL protein (see Figure 1).47,48 The binding of HIFα to VHL and to the E3 ligase complex causes HIFα to be ubiquitinated and marked for degradation by the cell’s proteosomal complex.49–54 Therefore, in the normal circumstance with normal local oxygen availability, HIFα levels are kept low in the cell. In contrast, during hypoxia, HIFα is not hydroxylated, and hence does not bind to VHL protein, and consequently is not degraded. Therefore, as a normal physiological response to hypoxia, HIFα levels rise in the cell, allowing it to bind with a similar molecule that is constitutively present, namely, HIFβ (see Figure 1). The HIFα/β heterocomplex can then translocate to the nucleus and bind to specific hypoxia response elements in the promoters of genes that are important in the cell’s response to hypoxia. Binding of the regulatory HIFα/β complex to the hypoxia response elements in the promoter of these hypoxia responsive genes in turn transcriptionally upregulates their mRNA and subsequent protein levels. The critical genes upregulated by HIFα include vascular endothelial growth factor (VEGF), platelet-derived growth factor (PDGF), transforming growth factor alpha (TGFα), carbonic anhydrase IX (CA-IX), erythropoietin, glucose transporter 1 (GLUT-1), and others.
When VHL protein either cannot function because of a mutation or is abnormally low/absent in the cell (for example, because of promoter hypermethylation), then irrespective of the oxygen levels in the cell, HIFα cannot be bound to the E3 ligase, cannot be degraded, and hence is constitutively present at a higher level in the cell (see Figure 1). High levels of HIFα in turn mean that the HIFα/β complex will interact with the hypoxia response elements in the nucleus and the genes normally regulated by HIF, such as VEGF, PDGF, and TGFα, will be abnormally high and lead to the development of ccRCC.
Of the many proteins upregulated by HIFα, perhaps the one, which has generated the most interest, is VEGF. This is because VEGF has a central role in angiogenesis, and it is now well recognized that this is critical to malignant tumor progression across a variety of tumors and there is the clinical observation that ccRCC’s are generally hypervascular tumors.55–57 The VEGF protein family includes several subtypes, including VEGF-A, -B, -C, -D, -E, and placenta growth factor-1.58–61 These proteins can bind to one or more of at least three VEGF receptors (VEGFRs) at the cell surface, namely, VEGFR-1 (Flt-1), VEGFR-2 (KDR/Flk-1), and VEGFR-3 (Flt-4).58–61 Of the three, VEGFR-1 and -2 are thought to be more important for angiogenesis, whereas VEGFR-3 is thought to be more important for lymphangiogenesis.61
All three members of the VEGFR family are cell-surface membrane-associated tyrosine kinases. Binding of the ligand (VEGF) induces a conformational change in the receptor allowing specific tyrosine amino acid residues in the protein to be phosphorylated. This phosphorylation in turn leads to a downstream cascade of signaling events (see Figure 2). There are at least two main pathways that these cascades follow. One is through the RAF–MEK–ERK series of kinases and the other is through the phosphatidylinositol-3 kinase (PI3K)–protein kinase B (AKT)– mammalian target of rapamycin (mTOR) pathway. It is the activation of these and other pathways that in turn leads to endothelial cell activation, proliferation, migration, and cell survival.58–62 The complex interplay of these events, probably along with similar events from the other HIF target genes, ultimately leads to carcinogenesis through a process that has not yet been fully explained in all its complexity.
A key therapeutic point for the RAF–MEK–ERK and the PI3K–AKT–mTOR pathways is that both also involve kinases to regulate their activity. Another important observation relates to the mTOR, which is downstream of VEGF, but also acts to increase the starting cellular levels of HIFα.63 In principle, then, the abnormal VHL function can set up a vicious positive feedback loop in which HIFα levels rise, leading to abnormally high levels of VEGF, which binds to VEGFR resulting in activation of the phosphatidylinositol-3 kinase–AKT pathway to activate mTOR. This can in turn lead to even higher levels of HIFα. In principle, this could lead to a vicious positive feedback loop exacerbating the defect started by abnormal VHL function.
The fundamental biology underlying the VHL–HIF–VEGF cascade described above is important as it is the members of this biological cascade that form the therapeutic targets for managing advanced ccRCC. By targeting the components of the pathway, the process of carcinogenesis and tumor progression should be reversible. This concept underlies the so-called ‘targeted therapies’, which have now become the standard of care for advanced ccRCC.
The most direct way, conceptually, to block the elevated VEGF in ccRCC would be to inhibit the molecule itself (see Figure 2). Although several such approaches have been explored, the best studied so far has been that of the humanized monoclonal antibody to VEGF, namely, Bevacizumab (Avastin, Genentech, San Francisco, CA, USA).64 This novel, targeted agent was tested for use in ccRCC in a randomized phase II trial and was found to lengthen the time to progression of advanced RCC when compared with placebo.64 Subsequently, it was tested in a large-scale phase III trial in combination with IFNα for men with previously untreated advanced RCC.65 The progression-free survival (PFS) in the combination regimen was 10.2 months, significantly longer than the 5.4 months in the IFN-alone arm. On the basis of these studies, Bevacizumab in combination with IFNα is one of several targeted agents routinely used in advanced RCC.
An alternate to directly blocking VEGF is to target downstream signaling from the VEGFR (see Figure 2). As discussed previously, the VEGFR (as well as PDGF and TNFα) are tyrosine kinases. Downstream of these tyrosine kinase receptors are at least two pathways, the RAF–ME-K–ERK and the PI3K–AKT–mTOR pathways, many of whose members also rely on tyrosine kinase activity. Targeting the tyrosine kinases is, therefore, a logical next step in developing novel therapeutics for ccRCC. The earliest attempts at developing these tyrosine kinase inhibitors (TKIs) were relatively specific for the VEGFR itself.66,67 As the results were generally disappointing, these agents have largely been abandoned. What became apparent over the course of those studies was that less specific TKIs that could affect several different signaling molecules simultaneously were more effective. Presumably this is because of their ability to interrupt multiple signaling cascades simultaneously at multiple different levels using one agent. This conceptual framework has lead to the development, and now approval, of several agents in this drug class. There is a seemingly endless and rapidly expanding pool of potentially active drugs of this type (for example, AG-013736, GW572016, PTK787/ZK222584, plus others), but for this review, we will focus on the two such drugs with large-scale, published level I data confirming their clinical utility and which are now both approved for use in metastatic RCC. (For a more in-depth review of the other agents see Lane et al.62; Shaheen and Bukowski68 and Amato69.)
Sunitinib is an orally bioavailable multitargeted TKI. Pre-clinical studies have shown that it can block downstream signaling from several receptors important in ccRCC, including the receptors for VEGF and PDGF.70,71 Promising phase I studies in the setting of advanced RCC72 led to the initiation of two phase II studies in patients who had failed earlier systemic cytokine therapy (second-line setting). In these two trials, the partial response rates were 34–40% and the median time to progression was 8.3–8.7 months.73,74 Both these data represented substantial improvements over historical controls, and hence, a large-scale, international, multicenter, prospective, randomized, phase III trial was initiated and completed that enrolled 750 patients with ccRCC, who had not received earlier systemic therapy (front-line setting).75 The randomization was between sunitinib and IFNα, and the primary end point of the trial was PFS. The partial response rate for sunitinib (31%) was significantly better than that for IFNα (6%). The median PFS was also significantly better in patients’ who received sunitinib (11 months) compared with those in the IFNα arm (5 months). Overall toxicity was manageable. The clear superiority of the oral TKI sunitinib over IFNα has now led to its approval for use in the front-line setting for the treatment of advanced RCC.
Another orally bioavailable multitargeted TKI is sorafenib. This was originally developed as an inhibitor of Raf-1, another member of the RAF/MEK/ERK pathway downstream of such critical receptors as VEGFR and PDGFR.76 Subsequent studies showed that it was also able to block downstream signaling from a variety of other tyrosine kinases, including VEGFR, PDGFR, as well as others. As with sunitinib, the initial phase II studies with sorafenib showed substantial improvements in PFS, although the partial response rates were overall much lower.47,77 These phase II studies led to a large-scale, multicenter, international, randomized, prospective trial of 903 patients with ccRCC who had failed at least one earlier systemic therapy (second-line therapy).78 The randomization was between oral sorafenib and placebo. PFS was significantly better in the sorafenib arm compared with that in placebo, and it was generally well tolerated, although there were rare cases of significant hypertension and cardiac ischemia. As a consequence, sorafenib is also approved for use in advanced ccRCC, although it is generally used in a second-line setting.
As the aberrations in VHL that underlie the process of carcinogenesis in ccRCC seem to act predominantly through the accumulation of HIFα in the cell, another potential way to target ccRCC is to target HIFα’s starting levels. Although there are several important pathways that coordinate to control HIFα expression, the most important from a therapeutic standpoint is the Akt/mTOR pathway. Activation of mTOR leads to increased expression of HIFα mRNA and increased protein expression. In ccRCC, elevation of HIFα leads to increased signaling along the Akt–mTOR pathway, which can then lead to transcriptional upregulation of HIFα. The elevated HIF cannot be targeted for degradation owing to aberrations in VHL function. This vicious positive feedback loop, as discussed previously, is thought to be important in ccRCC and, therefore, a number of agents have been developed as inhibitors of mTOR. To date, a number of these compounds have been shown to decrease HIFα levels, including rapamycin, temsirolimus, and everolimus. (See also Cho et al.63; Boulay et al.79; Reddy et al.80.) Of these, the two that now have level I evidence supporting their use in the setting of ccRCC are temsirolimus and everolimus.
Temsirolimus (derived as a water-soluble ester of sirolimus) is able to inhibit mTOR’s kinase activity and lead to cell cycle arrest. Several phase II trials showed its efficacy, both as a single agent and in combination with IFNα, in cytokine-refractory advanced RCC in improving PFS when compared with historical controls.81,82 A large-scale, prospective, randomized, phase III trial was completed of patients with high-risk metastatic RCC (based on the Motzer criteria) randomized to receive temsirolimus alone, IFNα alone, or both agents.83 Temsirolimus as monotherapy improved both PFS and overall survival compared with either IFNα or combination therapies. Overall, the toxicity noted in the temsirolimus monotherapy arm was manageable. As a result of this study, temsirolimus is also an approved agent for use in advanced RCC and is generally the preferred front-line option in patients with high-risk metastatic ccRCC.
More recently everolimus, another mTOR inhibitor, was also tested in a large-scale, prospective randomized placebo-controlled phase III trial, specifically in patients who had failed earlier targeted therapies, such as the TKI’s described previously.84 The patients in the everolimus arm had better PFS when compared with patients in the placebo arm. It is anticipated that this will also be approved for use in advanced ccRCC and is likely to be used in the second-line setting given the way the randomized trial showing its efficacy was structured.
The biology of ccRCC is a gratifying example of how discoveries at the bench have directly informed significant advances at the bedside. The broad principles of this biology are reasonably well established, including the pathway from aberrations in VHL, to dysregulated HIFα, downstream changes in hypoxia responsive genes, such as VEGF, PDGF, CA-IX, and others, and then further downstream signaling events through pathways, including RAF–MEK–ERK and PI3K–AKT–mTOR. Clinically, there is ongoing work to target other aspects of this cascade of events. At the more basic biological level, there are innumerable questions still to be answered. Examples include: Which of the many hypoxia responsive genes are the most critical for oncogenesis? What other, as yet understudied, genes are also involved? How do the downstream signaling cascades interact, and why do the normal feedback regulatory mechanisms in the cell fail to stop these cascades of events? Why are these events more critical for cancer in the kidney and not in another organ system such as the lung or bowel? Over time, the answers to some of these questions will hopefully open up even more therapeutic avenues for advanced RCC.
This work was supported, in part, by award number K08 CA113452 from the National Institutes of Health. The content is solely the responsibility of the author and does not necessarily represent the official views of the National Institutes of Health.
The author declared no competing interests.