Targeted agents for the treatment of advanced renal cell carcinoma

Authors

  • Walter M. Stadler M.D.

    Corresponding author
    1. Division of Genitourinary Oncology, Section of Hematology/Oncology, Department of Medicine and Cancer Research Center, University of Chicago, Chicago, Illinois
    • Division of Genitourinary Oncology, Section of Hematology/Oncology, Department of Medicine and Cancer Research Center, University of Chicago, Chicago, IL 60637
    Search for more papers by this author
    • Dr. Stadler has received honoraria and research report from Bayer Pharmaceutical Corporation and has also acted as a consultant.

    • Fax: (773) 702-3163


Abstract

Metastatic renal cell carcinoma (RCC) is currently one of the most treatment-resistant malignancies. However, the elucidation of the molecular mechanisms underlying RCC development has led to the identification of promising targets for novel therapeutic agents. The involvement of the Von Hippel–Lindau protein pathway in clear cell RCC suggests that downstream targets of this pathway, namely, signaling through vascular endothelial growth factor (VEGF) in endothelial cells, platelet-derived growth factor (PDGF) in endothelial cells and pericytes, and the epidermal growth factor receptor (EGFR) pathway in tumor cells are all reasonable and rational therapeutic targets. A number of agents are in development that target VEGF (bevacizumab, a recombinant, humanized monoclonal antibody) or its receptor, VEGFR (PTK787, SU011248, and BAY 43-9006, all of which are small molecule inhibitors). Agents targeting EGFR also are being investigated clinically (gefitinib, cetuximab, erlotinib, and ABX-EGF). The Raf/MEK/ERK pathway is an important downstream convergence point for signaling through VEGFR, platelet-derived growth factor receptor (PDGFR), and EGFR (all have receptor tyrosine kinase activity) and also has important antiapoptotic effects, thereby providing an attractive target for intervention. In addition to inhibiting VEGFR and PDGFR-mediated angiogenic pathways, BAY 43-9006 has been shown to inhibit the Raf/MEK/ERK pathway at the level of Raf kinase. MEK-directed therapeutic approaches are also in development. Given that multiple molecular pathways are implicated in tumor cell growth, antitumor activity may be increased by using individual agents that target multiple pathways, or by combining different agents to allow vertical or horizontal inhibition of relevant pathways. Cancer 2005. © 2005 American Cancer Society.

One-third of patients diagnosed with renal cell carcinoma (RCC) present with or will develop metastatic lesions during the course of the disease.1 Metastatic RCC is currently one of the most treatment-resistant malignancies. Classic cytotoxic chemotherapy is reported to have little antitumor activity in RCC.2 Vinblastine, for example, had a long history of reported activity and still is used occasionally, but in a multiinstitutional study using rigorous response assessment, only 1 objective response was observed in 80 patients.3 Multiple reports also suggest that 5-fluorouracil (5-FU) and related compounds have activity in RCC. However, in a multiinstitutional study, an objective response rate of only 5% was reported.4 Although the combination of gemcitabine and 5-FU may be a little more active, the objective response rate achieved with this combination is still < 20%, and no complete responses (CR) or survival benefit has been demonstrated.5, 6

Immunotherapy with interferon-alpha (IFN-α) or interleukin-2 (IL-2) is generally considered the standard of care in RCC. Objective response rates with IFN-α range from 10% to 15%, and in Phase III studies, improvements in median survival of only 3–7 months compared with placebo-equivalent therapy have been reported.7 The major benefit from IL-2 therapy is the occasional durable CR that is observed. However, this is reported to occur in only 3–8% of highly selected patients believed to be appropriate for this treatment, and many studies suggest that the use of high doses of IL-2—with its attendant toxicities and limited applicability—is necessary to realize the full potential of this cytokine. As a result, the prognosis for patients with advanced RCC is extremely poor. The median survival period is approximately 10 months,8 and the 5-year survival rate is < 10%.9

Despite this rather bleak history, a number of recent discoveries in both the laboratory and the clinic raise the possibility that more effective therapy for RCC is possible. Specifically, significant advances have been made in the elucidation of the molecular mechanisms underlying the development of RCC, and these have led to the identification of promising targets for the treatment of RCC. This has resulted in a shift in drug development from a random screening approach to a more rational, mechanistic approach, and several promising agents currently are in development. In the current review, we will discuss the molecular mechanisms implicated in the development of RCC and the agents currently in clinical development targeting these mechanisms, and provide a summary of the supportive clinical data.

RCC SUBTYPES

RCC can be classified into a number of subtypes according to histology. Clear cell carcinoma is the most common form of renal tumor, and accounts for 70–80% of all cases of RCC.10 Other histologic types are papillary, chromophobe, and collecting duct carcinomas. Sarcomatoid RCC is a poorly differentiated version of the above subtypes that cannot be subtyped more accurately. Because the pathophysiology of clear cell RCC is best understood, and it is the most common subtype, it will be the focus of the current review.

MOLECULAR TARGETS IN RCC—FAMILIAL RCC AS A MODEL

Von Hippel–Lindau (VHL) syndrome is a familial form of RCC that occurs at a young age and is characterized by multiple neoplasms, including RCC, renal cysts, retinal hemangiomas, hemangioblastomas of the cerebellum and spinal cord, pheochromocytomas, and pancreatic carcinomas or cysts. The familial kindreds with VHL allowed investigators to localize and identify the VHL gene on the short arm of chromosome 3. As with other classic tumor suppressor genes that follow the Knudson “two-hit” mechanism,11 affected patients inherit one VHL mutation in their germline and then accumulate a mutation or deletion of the second allele in the susceptible target organ(s). This suggested that the VHL gene could also be implicated in sporadic cases of RCC, in which both copies of the VHL gene are mutated or inactivated in the target organ.12 In fact, both copies of the VHL gene are inactivated by mutation, methylation, or deletion in approximately 40–60% of human sporadic renal clear cell carcinomas.13–16 Further support for this concept was gathered by demonstrating consistent overexpression of genes typically repressed by normal VHL protein (pVHL) function, such as carbonic anhydrase IX (CAIX), in sporadic clear cell carcinomas, including those in which VHL gene itself was not altered.

VHL BIOLOGY AND THE OPPORTUNITY FOR THERAPEUTIC INTERVENTION

The recognition that pVHL is frequently inactivated in RCC, and the elucidation of the pathways regulated by this protein, were important steps in identifying suitable targets for therapeutic intervention. pVHL forms a multiprotein complex (VEC) with elongin B, elongin C, Cullin 2, and Rbx1. This complex binds to the α subunits of hypoxia-inducible factor (HIFα), a transcription factor for genes that allows cells to grow and survive in hypoxic conditions.17 Under normoxic conditions, HIFα undergoes hydroxylation at specific prolyl and aspariginyl residues. This, in turn, leads to pVHL binding and subsequent targeting of HIF for ubiquination and proteasomal degradation. Under hypoxic conditions, HIFα hydroxylation is decreased and, thus, no longer binds effectively to pVHL and can instead bind to its ubiquitously expressed partner HIFβ (also known as aryl hydrocarbon receptor nuclear translocator) to manifest its transcriptional activity. Proteins induced by transcriptionally active HIFα/HIFβ include vascular endothelial growth factor (VEGF), platelet-derived growth factor (PDGF), transforming growth factor-α (TGF-α), GLUT-1, CAIX (G250), and others (Figs. 1 and 2).18, 19VHL mutations inhibit binding of HIFα, even under normoxic conditions and, therefore, HIF activity and up-regulation of these same proteins are pathognomonic for clear cell RCC.

Figure 1.

The VEC multiprotein complex. The VEC complex consists of elongin C (C), Cullin 2 (Cul 2), Rbx1 (R), and NEDD8 (N). pVHL: Von Hippel–Lindau protein; HIF: hypoxia-inducible factor; HRE: hypoxia response element; ECM: extracellular matrix. Adapted from Leung SK, and Ohh, M. Playing tag with HIF: the VHL story. J Biomed Biotechnol. 2002;2:131–-135.

Figure 2.

Sites of action of targeted agents in the development of the treatment of advanced renal cell carcinoma. EGF: epidermal growth factor; HIF: hypoxia-inducible factor; mTOR: mammalian target of rapamycin; P13K: phosphatidylinositol-3 kinase; PDGF: platelet-derived growth factor; pVHL: Von Hippel–Lindau protein; PLC2: phospholipase C-2; VEGF: vascular endothelial growth factor.

The effects of pVHL and HIFα on angiogenic factors such as VEGF and PDGF probably account for the finding that RCC tumors are among the most vascularized of all solid tumors.20 Angiogenesis is essential for tumor growth, and Folkman21 demonstrated that neovascularization is required for tumors to grow beyond 1–2 mm in greatest dimension, and for metastasis to occur. Angiogenesis is a complex multistage process involving many stimulatory and inhibitory factors, the most established of which is the VEGF/vascular endothelial growth factor receptor (VEGFR) system. Multiple isomers and alternatively spliced variants of the VEGF ligand as well as the VEGFR exist, each of which may have important differential effects in the regulation of angiogenesis as well as lymphangiogenesis (see Hicklin and Ellis22 for a review). In RCC, serum levels of the most common circulating forms of VEGF have been shown to predict metastatic potential in patients with organ-confined clear cell RCC.23 PDGF also mediates angiogenic effects and, importantly, may play a role in the later stages of blood vessel formation via an ability to recruit and support the growth of pericytes.24 TGF-α, another HIFα-inducible factor, is a major ligand for the epidermal growth factor receptor (EGFR), and, upon binding, induces important cellular responses (proliferation, survival, differentiation, migration, adhesion). The overexpression of TGF-α probably creates an autocrine loop in RCC that may be important for its malignant phenotype. There is evidence to suggest that activated EGFR can independently up-regulate VEGF production, thereby providing an additional mitogenic signal from the malignant cell to the supporting blood vessels.25

Many of the receptors for HIFα-inducible factors, for example, the VEGFR and platelet-derived growth factor receptor (PDGFR) on endothelial cells, and the EGFR on the tumor cells, exhibit tyrosine kinase activity and, upon ligand binding, activate downstream signaling pathways, of which the Raf/MEK/ERK pathway is the best characterized.26 This is a ubiquitous signaling module that forms a link between extracellular signals and effector molecules in the cytoplasm and nucleus that contribute to cell differentiation, proliferation, and survival.27 Constitutive activation of Raf/MEK/ERK in RCC was demonstrated in a study evaluating primary tumor samples and normal kidney tissue samples obtained from 25 patients with RCC, the majority of whom had clear cell histology.28 That study suggested that activation of mitogen-activated protein kinases involved in Raf/MEK/ERK signaling was common in RCC, and that this was dependent on sequential activation of Raf-1 and MEK.28

Raf/MEK/ERK is not only a key convergent pathway downstream of receptors bound by HIFα-inducible factors (VEGFR, EGFR, and PDGFR), but this pathway, and in fact Raf alone, have also been shown to mediate important survival signals via interaction with members of the Bcl-2 protein family, and with other regulators of apoptosis, including the proapoptotic molecule apoptosis signal-regulating kinase (ASK-1).29–33 These antiapoptotic effects of Raf are relevant to both the tumor cell and the endothelial cell. A recent in vitro study showed that after activation of basic fibroblast growth factor (bFGF) and VEGF, Raf was pivotal in regulating endothelial cell survival during angiogenesis via effects on intrinsic and extrinsic apoptosis pathways.34 Therefore, Raf is a key regulator of tumor cell and endothelial cell proliferation and survival, as is common to many interlinking cellular pathways.33

In addition to the VHL pathway, other pathways are implicated in the regulation of HIFα and hypoxia-induced genes. The phosphatidylinositol-3 kinase (PI3K)/AKT (protein kinase B) pathway and its downstream effector mTOR, which is the mammalian target of rapamycin, have also been shown to increase levels of HIF-1α gene expression.35 mTOR is a central modulator that relays proliferative and anabolic signals to downstream transcriptional and translational machinery, thereby regulating the transition from the G1 phase to the S-phase of the cell cycle.36 Amplifications/mutations in the genes for P13K, AKT, and the PTEN tumor suppressor gene, also believed to be an upstream effector of mTOR, have been found in various cancers.36 Specifically, PTEN gene expression is often down-regulated in RCC.37

There are a number of strategies for targeting the VHL pathway. Agents that directly inhibit HIF currently are in development, although none are yet available in the clinic. Alternatively, individual ligands that are up-regulated in response to HIF can be targeted, or the receptors that bind these ligands, such as the receptor tyrosine kinases VEGFR and PDGFR. Finally, downstream signaling pathways (Raf/MEK/ERK), which commonly relay signals from upstream receptor tyrosine kinases, provide an attractive target for intervention, as this would enable disruption of signaling through multiple growth factors.

TARGETING RECEPTOR TYROSINE KINASES

The pathophysiology of metastatic RCC, in particular, the involvement of HIF in clear cell RCC, suggests that the VEGFR pathway in endothelial cells, the PDGFR pathway in pericytes, and the EGFR pathway in the tumor cells are all reasonable and rational therapeutic targets (Table 1).

Table 1. Targeted Agents in Development for the Treatment of RCC
Drug nameBrand nameTargetDescriptionDelivery methodClinical development stage
  1. RCC: renal cell carcinoma; VEGF: vascular endothelial growth factor; i.v.: intravenous; 5-FU: 5-fluorouracil; VEGFR: vascular endothelial growth factor receptor; PDGFR: platelet-derived growth factor receptor; TKI: tyrosine kinase inhibitor; EGFR: epidermal growth factor receptor; NSCLC: nonsmall cell lung carcinoma CRC: colorectal carcinoma.

BevacizumabAvastin™ (Genentech, South San Francisco, CA)VEGFRecombinant, humanized, monoclonal antibodyi.v. infusion every other wkPhase III for RCC Approved for advanced carcinoma of the colon or rectum in combination with 5-FU
BAY 43-9006 VEGFR, PDGFR, Raf kinaseOral multi-kinase inhibitorDaily oral administrationPhase III
PTK787 VEGFR-1, VEGFR-2, VEGFR-3Oral selective inhibitorDaily oral administrationPhase I/II in RCC
SU011248 VEGFR, PDGFR, FLT3, KITOral selective TKIDaily oral administration (4 wks on/2 wks off)Phase III in RCC
GefitinibIressa® (AstraZeneca, Wilmington, DE)EGFRSmall molecule TKIDaily oral administrationPhase II in RCC (no responses) Approved for third-line use in NSCLC
CetuximabErbitux™ (ImClone Systems, Inc., New York, NY)EGFRChimeric mouse/human monoclonal antibodyWeekly i.v. infusionPhase II in RCC (no responses) Approved for advanced CRC
ErlotinibTarceva® (Genentech)EGFRSmall molecule EGFR inhibitorDaily oral administrationPhase II in RCC
ABX-EGF EGFRHuman monoclonal antibodyWeekly i.v. infusionPhase I/II in RCC
ISIS-5132 Raf-1Antisense therapy2-hr i.v. infusion 3 times weeklyPhase I/II in RCC
CI-1040 MEKOral MEK inhibitorOral administrationPhase I/II in RCC
CCI-779 MTOREster of rapamycin30-min i.v. infusion once weeklyPhase III in RCC

Vascular Endothelial Growth Factor Receptor

Given its pivotal role in tumor angiogenesis, and high specificity of binding, VEGF and its receptor have become much-investigated targets for novel drug development. Bevacizumab (Avastin™, Genentech, South San Francisco, CA) is a recombinant, humanized monoclonal antibody (MoAb) that depletes soluble VEGF from plasma, depriving VEGFRs of their ligand and thereby inhibiting angiogenesis. Phase I studies demonstrated that bevacizumab could be safely administered alone or in combination with conventional chemotherapy.38, 39 In a randomized, placebo-controlled, crossover, Phase II study in patients with advanced RCC, a high dose of bevacizumab (10 mg/kg intravenously [i.v.] every 2 wks) led to significantly prolonged time to disease progression compared with placebo. In addition, bevacizumab-treated patients were more likely to be disease progression free at 4 months versus placebo-treated patients. No significant effects on survival were observed. However, this small trial in which patients on the placebo arm were also crossed over to bevacizumab after disease progression was not designed with survival as a clinical end point. Hypertension and asymptomatic proteinuria were the most common adverse events reported.40 A Phase III trial of IFN-α with or without bevacizumab for the treatment of advanced RCC has completed enrollment, but results are pending. Importantly, IFN-α has also been shown to have antiangiogenic effects in human carcinomas, possibly mediated by down-regulation of bFGF, a proangiogenic factor.41

Bevacizumab was approved in the U.S. in early 2004 for the treatment of patients with metastatic carcinoma of the colon or rectum, in combination with intravenous 5-FU–based chemotherapy. In a recent study of 813 patients with previously untreated metastatic colorectal carcinoma (CRC), a median duration of disease progression-free survival of 10.6 months was reported in patients receiving bevacizumab combined with irinotecan, 5-FU, and leucovorin (IFL) chemotherapy, compared with 7.1 months in patients receiving IFL plus placebo. This improvement in survival was considered statistically significant and clinically meaningful.42

Rather than targeting the VEGF ligand, PTK787/ZK 222584 (PTK787) is an oral, selective inhibitor of VEGFR-1, VEGFR-2, and VEGFR-3 tyrosine kinase receptors. Oral dosing in mouse models led to the inhibition of tumor growth of a range of human carcinomas, in association with inhibition of microvessel formation in the interior of the tumor. This is consistent with antiangiogenic effects.43 In an open-label, Phase I study of 37 patients with metastatic RCC, 1 partial response (PR) and 6 minor responses were reported, and stable disease (SD) was observed in 46% of patients. The overall survival rate at 1 year was 63.7%.44 The antiangiogenic effects of PTK787 were assessed using dynamic contrast-enhanced magnetic resonance imaging scans.45 PTK787 is also in Phase III clinical development for the treatment of CRC in combination with first-line and second-line chemotherapeutic regimens in the CONFIRM 1 and CONFIRM 2 trials, after encouraging Phase I/II results were obtained.46

SU011248 is a more broad-spectrum tyrosine kinase inhibitor (TKI), which targets VEGFR, PDGFR, FLT3, and KIT, and is administered orally, once daily, in a 4-weeks-on/2-weeks-off cycle. In a Phase I dose escalation study in 28 patients with solid tumors not amenable to conventional therapies (including 4 patients with RCC), tumor responses were observed in 6 of 23 evaluable patients, and good oral bioavailability with modest interpatient variability was reported. The most common adverse events were fatigue and general weakness, although central tumor necrosis and organ perforation were reported at high doses of SU011248.47 A single-arm, multicenter, Phase II trial of SU011248 for second-line treatment of metastatic RCC has recently been completed in 63 patients. PRs were observed in 21 (33%) patients, and SD was achieved in an additional 23 (37%) patients. The median time to disease progression was 8.3 months, and in a recent update (June 2004), it was reported that 32 patients were still receiving treatment with continued PRs or SD.48 A Phase III trial of SU011248 versus interferon-alpha for first-line treatment of RCC patients has been initiated.

BAY 43-9006 is an oral multi-kinase inhibitor that has been shown preclinically to inhibit a number of kinases, including VEGFR-2, VEGFR-3, PDGFR-β, and Raf at nanomolar concentrations. Studies in xenograft models demonstrated broad-spectrum antitumor activity and significant antiangiogenic effects, as shown by decreased microvessel density and area in tumor sections.49 Phase I trials of BAY 43-9006 monotherapy have been conducted to investigate 4 different oral dosing schedules (28 days on/7 days off, 21 days on/7 days off, 7 days on/7 days off, and continuous dosing). Data from 118 patients with various solid tumors, including RCC, demonstrated 2 confirmed PRs (1 in RCC), tumor shrinkage in 4 patients, and disease stabilization in 47 patients. In addition, BAY 43-9006 was well tolerated, and drug-related adverse events were mild to moderate. The most frequent events were fatigue, diarrhea, and hand-foot skin reaction (each reported to occur in 35% of patients).50 Additional Phase I studies in patients with advanced-stage solid tumors have shown that BAY 43-9006 can be safely combined with gemcitabine, doxorubicin, oxaliplatin, or carboplatin/paclitaxel, resulting in antitumor activity.51–54

A Phase II study of BAY 43-9006 using a randomized discontinuation trial design has been performed in 484 patients with a range of solid tumors. This encompassed a 12-week, open-label treatment phase, after which patients with SD were randomized to receive either BAY 43-9006 or placebo for an additional 12 weeks. This design is noteworthy because it enables the differentiation of drug activity from the natural course of the disease, and allows a reduction in the number of randomized patients compared with a more classic randomized, controlled trial. Preliminary data for 106 patients with RCC have shown that after the 12-week runin phase, 37 patients had a ≥ 25% tumor shrinkage and continued to receive open-label BAY 43-9006; 38 patients had a < 25% change in tumor size from baseline (SD) and were entered into the randomized phase of the trial; 14 patients had tumor growth ≥ 25% and were discontinued from the study; and 17 patients had missing data, early disease progression, or were discontinued for reasons other than disease progression. Hand-foot skin reaction and hypertension were the most commonly reported Grade 3/4 adverse events.55 More recently, it was announced that a significantly higher percentage of patients randomized to receive continued therapy compared with those randomized to receive placebo remained free of disease progression 12 weeks after randomization.56 Phase III trials with BAY 43-9006 versus placebo for second-line therapy in patients with RCC and Phase II trials in other tumor types currently are ongoing.

Epidermal Growth Factor Receptor

Two main strategies have been developed to block EGFR: MoAbs directed against the external domain of the receptor, or small molecules that compete for binding to the receptor kinase pocket. Gefitinib (Iressa®, AstraZeneca, Wilmington, DE), a small-molecule TKI that is relatively advanced in its clinical development, inhibits EGFR at nanomolar concentrations. In 2 Phase II studies (IDEAL 1 and IDEAL 2) in pretreated patients with nonsmall cell lung carcinoma (NSCLC), treatment with gefitinib monotherapy produced objective response rates of 18.5% and 10%, respectively.57, 58 Gefitinib is now approved in several countries for third-line treatment of advanced NSCLC after the failure of chemotherapeutic regimens. Despite this progress in the treatment of lung carcinoma, the investigation of gefitinib in a small Phase II study of 16 patients with advanced RCC failed to demonstrate any objective responses, and 75% of patients had developed disease progression at 4 months.59 Similarly, a Phase II study of the alternative EGFR TKI erlotinib also revealed only 1 objective response among 40 patients.60 Although these results would suggest a lack of activity of EGFR-targeted agents in RCC, the trials were not designed to adequately test for the possibility of a disease-stabilizing effect. This and the evidence of synergy between EGFR and VEGFR-targeted approaches in preclinical studies have led to combination trials in patients with noteworthy preliminary results.

Cetuximab (Erbitux™, ImClone Systems Inc., New York, NY) is a chimeric mouse/human MoAb that binds EGFR, thereby disturbing cell cycling and inhibiting tumor-induced angiogenesis. Although in vivo studies in mouse xenograft models of human RCC appeared promising,61 no CRs or PRs were observed in a single-arm, Phase II trial of cetuximab in 55 patients with metastatic RCC.62 This effectively halted further investigation of cetuximab as a single agent for the treatment of RCC, although cetuximab is approved for the treatment of advanced-stage CRC. ABX-epidermal growth factor (EGF) is a high-affinity, human MoAb that binds to EGFR in the extracellular domain and prevents binding of both EGF and TGF-α to the receptor. Potent antitumor activity has been demonstrated in vivo with ABX-EGF as a single agent and in combination with chemotherapeutic agents in a range of tumor types including pancreatic carcinoma, prostate carcinoma, breast carcinoma, head and neck carcinoma, and RCC.63 In a Phase I/II study in 88 evaluable patients with metastatic RCC, ABX-EGF monotherapy (weekly i.v. infusion over 8 wks) resulted in 5 tumor responses and disease stabilization in 44 patients.64 A Phase II study of continued ABX-EGF treatment in patients with RCC with a previous response to the agent has completed enrollment.

TARGETING THE RAF/MEK/ERK SIGNALING PATHWAY

The Raf/MEK/ERK signaling pathway is central to the regulation of cell proliferation and survival, both in the tumor cell and the endothelial cell. It represents an important downstream convergence point for signaling through the receptor tyrosine kinases VEGFR, PDGFR, and EGFR. A number of strategies currently are under investigation that target different stages of the Raf/MEK/ERK signal transduction pathway.27

In addition to inhibiting VEGFR and PDGFR-mediated angiogenic pathways in endothelial cells and pericytes, BAY 43-9006 also targets the Raf/MEK/ERK pathway at the level of Raf kinase. Potent inhibition of Raf-1 has been demonstrated in vitro, and inhibition of Raf/MEK/ERK signaling has been shown in colon, pancreatic, and breast tumor cell lines with Ras or B-raf mutations.65, 66 Although the majority of research on Raf and Ras has focused on tumor cells, more recent data also demonstrated that the proliferative and survival responses of endothelial cells to VEGF is dependent on Raf activation.34 Therefore, the exact target or targets for the clinical activity observed with BAY 43-9006 remain(s) to be elucidated. Because of its effects on Raf, it is important to note that BAY 43-9006 can potentially have a downstream effect on pathways in which a direct effect at the level of the tyrosine kinase is not noted, such as EGFR signaling.

MEK-directed therapeutic approaches also are in development. One oral agent —CI-1040—prevents MEK phosphorylation and transmission of proliferative signals through the Raf/MEK/ERK pathway. A Phase I dose escalation study showed that a single dose of CI-1040 was well tolerated at all dose levels tested, with SD observed in 30% of patients.67 Target therapeutic plasma concentrations based on in vivo models were reached after a single dose of 800 mg. A second-generation MEK inhibitor, ARRY-142886, has shown tumor suppressive activity in multiple rodent models of human cancer, including melanoma, pancreatic carcinoma, colon carcinoma, lung carcinoma, and breast carcinoma.

TARGETING THE MOLECULAR TARGET OF RAPAMYCIN

CCI-779, an ester of rapamycin, is a novel mTOR inhibitor that has been shown to lead to G1 cell cycle arrest in preclinical studies.68 In a randomized Phase II study in patients with advanced, refractory RCC, CCI-779 induced an objective response rate of 7% (1 CR and 7 partial responses), a median time to tumor progression of 5.8 months, and a median survival of 15 months. CCI-779 was generally well tolerated at the 3 dose levels tested (25 mg, 75 mg, and 250 mg weekly administration as a 30-min i.v. infusion), and the most frequently occurring Grade 3/4 toxicities were hyperglycemia, hyperphosphatemia, anemia, and hypertriglyceridemia. In addition, six patients were reported to have possible nonspecific pneumonitis, two of whom were withdrawn from the study.69 Additional analyses categorized patients by risk based on previously described prognostic factors (Karnofsky performance status, lactate dehydrogenase level, corrected serum calcium level, serum hemoglobin level, and time from diagnosis to first treatment initiation)8 and demonstrated that patients with two or fewer of these risk factor groups had a twofold to threefold longer median survival than those with three or more risk factors. More importantly, the current analysis also raised the hypothesis that patients with poorer prognostic features had better survival after CCI-779 treatment than historically controlled patients treated with cytotoxic therapy.

CCI-779 has also been tested clinically in combination with the immunotherapeutic agent, IFN-α, in an open-label, single-arm, dose escalation study in a total of 104 patients with advanced-stage RCC. Based on dose-limiting toxicities, the maximum tolerated doses (MTD) for use in future studies were established as 15 mg CCI-779 once weekly and 6 MU IFN-α 3 times weekly. The most frequent Grade 3/4 treatment-related toxicities were leukopenia, hyperlipidemia, and asthenia, and assessment of tumor responses in 55 patients found 7 confirmed PRs and 39 cases of SD.70 A Phase III trial investigating the combination of IFN-α plus CCI-779 compared with either agent alone as a first-line therapy has been initiated in patients with RCC with poor prognosis. Other mTOR inhibitors are in early clinical development, and their potential activity and role in RCC is awaited.

COMBINATION TARGETING

Given that multiple molecular pathways are implicated in tumor cell growth, and the high likelihood for crosstalk between the components of these pathways, single-target inhibition may be insufficient to induce durable antitumor effects. Targeted agents can be combined to allow vertical or horizontal inhibition of relevant pathways. Vertical inhibition targets multiple points within the same molecular pathway, whereas horizontal inhibition acts at points across multiple pathways.

The efficacy of horizontal inhibition in advanced-stage RCC was investigated by combining the anti-EGFR agent, erlotinib, with the anti-VEGF, bevacizumab. Such an approach has preclinical support in a number of models,71 and this Phase II study found an objective response rate of 21%.72 Whether this represents a true benefit over the observed 10% objective response rate with bevacizumab alone remains to be determined. However, given the known biology of the EGFR and VEGFR pathways in clear cell RCC, the combined inhibition of both pathways remains an important area for clinical research and a randomized Phase II trial of bevacizumab with or without erlotinib currently is in progress.

BAY 43-9006 enables potential horizontal and vertical targeting of multiple pathways in 1 single agent. BAY 43-9006 horizontally targets different receptor tyrosine kinases (VEGFR and PDGFR), and through an additional activity against Raf, promotes vertical inhibition of the downstream Raf/MEK/ERK signaling pathway.

FUTURE CLINICAL TRIAL DESIGN

The question of whether traditional clinical research methods are appropriate for the development of novel, targeted, anticancer agents has been the subject of discussions.73 Early dose-finding studies currently are based on toxicity, following the paradigm for classic cytotoxic therapies, for which the therapeutic window is narrow. Whether a similarly narrow window exists for these newer agents remains unknown, but preclinical studies as well as the early clinical results suggest that doses with minimal toxicity can still be therapeutically active. Therefore, the need to fully define the MTD and administering the agent at or close to this dose has been questioned. There has been some hope that the use of pharmacodynamic and predictive markers could help identify a “minimally active dose” for further studies. However, to our knowledge to date, difficulties in standardizing and validating such biomarkers have made practical implementation of these concepts difficult. Only further studies will determine whether the promise of a wider therapeutic window with agents targeted at known biologic abnormalities in RCC is realized.

Assessing antitumor activity with some of these agents in Phase II trials has also been challenging. In preclinical models, many agents induce growth arrest or inhibition rather than tumor shrinkage. Such an effect is most likely clinically significant. For example, IFN-α leads to only 10% objective responses, and yet imparts a survival advantage in randomized Phase III trials. This observation is possible only if patients other than those who experience an objective response are benefiting from treatment. A similar argument holds for bevacizumab, for which an objective response rate of 10% was observed, but an improvement in time to disease progression was demonstrated. A comparison of individual tumor sizes in patients assigned to placebo versus those assigned to bevacizumab in the current study clearly demonstrates that growth inhibition is the major clinical effect of this drug. The observation of significant numbers of patients in the BAY 43-9006 and SU11248 trials with tumor shrinkages that do not meet the usual criteria for objective response suggests that growth inhibition is a relevant effect of these drugs as well. These observations emphasize the point that the 30% unidimensional shrinkage by standard Response Evaluation Criteria in Solid Tumors (RECIST) is an arbitrary standard that does not necessarily have biologic or clinical significance.

Nevertheless, the assessment of disease stabilization in a heterogeneous patient population is difficult. The expected time to disease progression, or fraction of patients without disease progression in an arbitrary small RCC population, is highly variable and not predictable (i.e., the confidence intervals are large). Therefore, time to disease progression or SD endpoints in single-arm uncontrolled studies are difficult to interpret. The use of prognostic markers to better define expected time to disease progression and narrow confidence intervals can help, but may not be sufficient. It therefore is likely that increasing the use of randomized Phase II trials will become necessary to appropriately define activity. The use of novel designs such as the randomized discontinuation design may help to increase the efficiency of such trials.

Conclusions

Significant advances in understanding the molecular mechanisms underlying RCC have led to the development of rationally designed therapies, which are now being tested clinically. Many of these agents target the pathways involved in signal transduction and angiogenesis, which are key to the development of RCC. Until recently, highly specific agents targeting a theoretically critical member of the signaling pathway were sought. However, more recent clinical data suggest that agents that target multiple critical pathways may be more promising. These studies also raise the issue of whether targeting both the malignant tumor as well as the supporting vasculature and stroma may be necessary for effective therapy.

To date, the VEGFR pathway has been the most promising target, and the EGFR and mTOR pathways also have shown potential. As these agents are developed, oncologists will have to learn how to manage a series of toxicities different from those that have plagued the classic cytotoxic therapies. Although these new toxicities—including fatigue, rash, diarrhea, hypertension, and proteinuria— generally are not serious, the potential for using these drugs on a more chronic basis raises certain challenges in symptom management. Finally, and perhaps most important, these new agents, although promising, have not led to CRs in the advanced metastatic setting. Whether they have a role in the adjuvant setting is therefore an important question for future studies. In addition, further work needs to be done to accomplish the goal of eradicating or curing RCC.

Ancillary