Novel agents for advanced bladder cancer

Abstract

Conventional front-line platinum-based combination chemotherapy yields high response rates but suboptimal long-term outcomes for advanced transitional cell carcinoma. Salvage therapy is an unmet need with disappointing outcomes. The emergence of novel biologic agents offers the promise of improved outcomes. Neoadjuvant therapy preceding cystectomy for muscle-invasive bladder cancer provides an important paradigm and an interesting approach in developing novel agents. Patients who are not candidates for cisplatin require special attention. A multidisciplinary approach and collaboration among laboratory scientists, oncologists, urologists and radiation oncologists is necessary to make therapeutic advances. Recent and ongoing trials of novel chemotherapeutic and biologic agents are reviewed.

Introduction

While cisplatin-based combination chemotherapy is associated with improved outcomes in metastatic transitional cell carcinoma (TCC) compared to single-agent or noncisplatin chemotherapy, most patients relapse and die of progressive disease. Several multi-agent cisplatin-based frontline chemotherapy regimens appear to have similar efficacy for metastatic disease, including M-VAC (methotrexate, vinblastine, doxorubicin, cisplatin), dose-dense (DD) M-VAC or GC (gemcitabine, cisplatin) [Sternberg et al. 2006; Von Der Maase et al. 2005; Saxman et al. 1997]. Despite initial high response rates (RR) of 40–70% in metastatic disease, chemotherapy is generally not curative and overall 5-year overall survival (OS) is a suboptimal 5–20%. The median OS and progression-free survival (PFS) are approximately 15 months and 8 months, respectively. GC is employed most commonly due to better tolerability. Addition of other agents to GC has not yielded a significant improvement in outcomes. The recently reported European Organization for the Research and Treatment of Cancer (EORTC) randomized trial did not demonstrate a statistically improved OS with the addition of paclitaxel to GC [Bellmunt et al. 2007a].

The use of neoadjuvant cisplatin-based combination chemotherapy preceding radical cystectomy for muscle-invasive localized or locally advanced TCC of the bladder modestly improves cure rates. Unfortunately, recurrence still occurs in approximately 50% of patients [Grossman et al. 2003].

Salvage chemotherapy for metastatic TCC with conventional chemotherapeutic agents (taxanes, gemcitabine) following one or more prior chemotherapeutic regimens yields generally poor response rates of 10–20% and a median survival of 6–9 months; these responses do not always appear to correlate with survival [Albers et al. 2008; Albers et al. 2002; Vaughn et al. 2002; Sternberg et al. 2001; Lorusso et al. 2000; Mccaffrey et al. 1997; Dreicer et al. 1996]. Therefore, the salvage setting for chemotherapy refractory patients is clearly an unmet need, and these patients are candidates for clinical trials.

Renal dysfunction (usually defined as a creatinine clearance of <60 ml/minute), poor performance status and advanced age are relatively common and preclude cisplatin chemotherapy. Carboplatin-based combination regimens are feasible in such patients, but appear to be suboptimal compared to cisplatin-based regimens [Dogliotti et al. 2007; Bellmunt et al. 1997; Petrioli et al. 1996]. Nonplatinum taxane-gemcitabine regimens also appear to be reasonable alternatives in patients with renal dysfunction [Ardavanis et al. 2005; Li et al. 2005; Dreicer et al. 2003; Meluch et al. 2001]. Randomized trials are specifically evaluating regimens in this population (Table 1). The development of novel and tolerable agents for TCC is clearly warranted. This review will describe novel agents targeting relevant oncogenic molecules and in development for the therapy of locally advanced and metastatic TCC (Table 2).

Table 1.

Ongoing randomized trials of systemic therapy for metastatic metastatic transitional cell carcinoma^*.

Institution	Line of therapy	Phase	Eligibility	Group 1	Group 2
CALGB	First	III	Cr Cl ≥ 60	GC	GC – Bevacizumab
Industry	First	III	Cr Cl ≥ 60	GC	Larotaxel – Cisplatin
Multinational	First	III	Cr Cl 20-60 or CHF	Gemcitabine	Vinflunine Gemcitabine
EORTC	First	III	PS = 2 or Cr Cl 30-60	Carboplatin Methotrexate Vinblastine	Carboplatin Gemcitabine
European	First	II	Cr Cl ≥ 60	GC	GC – Trastuzumab
NCCN	First	II	Cr Cl ≥ 60	GC	GC – Cetuximab
European	First	II	Cr Cl ≥ 60	GC	GC – Sorafenib
European	First	II	Cr Cl ≥ 60	GC	GC – Gefitinib
MDACC	Second-line	II	Consolidation following response/ stability after frontline therapy	Docetaxel	Docetaxel – Cetuximab
U. Michigan	Second-line	II	Consolidation following response/ stability after frontline therapy	Placebo	Sunitinib
Barts London	Second-line	II	Consolidation following response/ stability after frontline therapy	Placebo	Lapatinib
Fox Chase	Second-line	II	Progression after frontline regimen	Cetuximab	Cetuximab – Paclitaxel
DFCI	Salvage	II	1-3 prior regimens	Docetaxel	Docetaxel – Vandetanib

^*From www.Clinicaltrials.gov accessed April 5, 2009; Cr Cl, creatinine clearance (ml/min); GC, gemcitabine plus cisplatin; CHF, congestive heart failure; PS, WHO performance status.

Table 2.

Molecular targets in transitional cell carcinoma being targeted by new agents in ongoing or planned trials.

Agent	Molecular target
Bevacizumab	VEGF
Aflibercept	VEGF, PlGF
AMG-386	Angiopoietin
Trastuzumab	HER-2
Cetuximab	EGFR
Gefitinib	EGFR
Erlotinib	EGFR
Lapatinib	EGFR, HER-2
Sunitinib	VEGFRs
Sorafenib	VEGFRs
Vandetanib	VEGFR, EGFR
Everolimus	mTOR
Temsirolimus	mTOR
Ipilimumab	CTLA-4
Eribulin	Tubulin
Tamoxifen	ER-β

VEGF, vascular endothelial growth factor; PlGF, placental growth factor; EGFR, epidermal growth factor receptor; mTOR, mammalian target of rapamycin; ER, estrogen receptor.

Prognostic factors

Interpretation of phase II studies in metastatic TCC is fraught with difficulty. Poor prognostic factors (visceral metastasis, Karnofsky Performance Status <80) can significantly impact outcomes independent of therapy. In the analysis of patients treated with M-VAC at Memorial Sloan Kettering Cancer Center (MSKCC), median survival of patients with 0, 1, or 2 risk factors was 33, 13.4, and 9.3 months, respectively [Bajorin et al. 1999]. These prognostic factors have been validated with other regimens [Bellmunt et al. 2007a; Sternberg et al. 2006]. Differences in the distribution of various risk factors in small phase II trials can lead to vastly different outcomes independent of the efficacy of agents and this issue confounds the development of novel agents. In a recent presentation from Memorial Sloan Kettering Cancer Center (MSKCC), a nomogram was constructed that incorporated the following four parameters: hemoglobin, serum albumin, Karnofsky Performance Status (KPS) and visceral metastasis [Bajorin et al. 2007]. However, the nomogram requires validation.

Novel cytotoxic chemotherapeutic agents

Vinflunine

Vinflunine is a bifluorinated derivative of the semisynthetic vinca alkaloid vinorelbine, and acts as a tubulin-targeted cytotoxic agent. Fifty-one patients with recurrent metastatic TCC were treated with vinflunine in a phase II trial, of whom nine responded for an overall RR of 18%, and 67% achieved disease control (response + stability) [Culine et al. 2006]. Salvage therapy with vinflunine plus best supportive care (BSC) was compared with BSC in a multinational randomized phase III trial that accrued 370 patients [Bellmunt Molins et al. 2008]. Patients received vinflunine 320 mg/m² every 3 weeks. Grade 3/4 toxicities for vinflunine were febrile neutropenia (6%), anemia (19%), thrombocytopenia (6%), fatigue (28%), constipation (20%), abdominal pain (9%), vomiting (4%) and peripheral neuropathy (1%). The median OS (6.9 versus 4.6 months) was not statistically better (p = 0.29), but the preplanned multivariate analysis adjusting for prognostic factors showed a statistically significant impact of vinflunine on OS (p = 0.036). In the 357 eligible patients or in the 351 patients treated per protocol, OS was significantly longer for vinflunine. The key secondary endpoints of response rate and PFS were also statistically superior for vinflunine. While vinflunine may improve outcomes of previously treated TCC patients, these benefits are at best modest. Another ongoing randomized trial compares the combination of frontline vinflunine and gemcitabine against gemcitabine alone in patients ineligible for cisplatin (Table 1).

Pemetrexed

Pemetrexed is a novel, multitargeted antifolate agent approved for pleural mesothelioma and non-small cell lung cancer. Early studies demonstrated that concomitant supplementation of vitamin B₁₂ and folate attenuated toxicities without compromising efficacy. Frontline pemetrexed in metastatic TCC (without folic acid and vitamin B₁₂ supplementation) yielded an objective RR of 30% and stable disease (SD) was achieved in 35% of patients[Paz-Ares et al. 2003]. Toxicities included grade 4 neutropenia (35%), grade 3/4 anemia (17%), and grade 3/4 thrombocytopenia (9%). Twenty-two per cent of patients developed febrile neutropenia and two patients died. Forty-seven patients were enrolled in another phase II trial in patients with progressive disease following initial chemotherapy for metastatic disease or within 12 months of perioperative chemotherapy [Sweeney et al. 2006]. Three (6.4%) complete responses (CR) and 10 (21.3%) partial responses (PR) were observed for an overall RR of 27.7%, while 10 patients (21.3%) had SD. The median time to progressive disease was 2.9 months and median OS was 9.6 months. Grade 3 or 4 hematologic events were thrombocytopenia (8.5%), neutropenia (4.3%) and anemia (2.1%). In a second phase II trial of second-line pemetrexed from MSKCC, an objective response was achieved in 1 of 12 evaluable patients for an overall response rate of 8% (90% upper limit 29%) [Galsky et al. 2007]. This level of activity did not meet criteria for full accrual based on the predefined 2-stage design, and the study was closed due to lack of efficacy. Frontline treatment with combination pemetrexed–gemcitabine was evaluated in 62 patients with metastatic TCC, 59% of whom had visceral metastases [Von Der Maase et al. 2006]. The RR was 26.5% and the median OS was 10.1 months. Grade 3/4 toxicities included anemia (13.3%), thrombocytopenia (10.0%), neutropenia (36.7%), febrile neutropenia (18.3%) and neutropenic sepsis (3.3%). Although many patients in this trial had poor risk disease, these results do not suggest this combination is promising for future development. An ongoing phase II trial is evaluating combination cisplatin and pemetrexed as frontline therapy (Table 3).

Table 3.

Ongoing and planned phase II trials of frontline therapy for metastatic transitional cell carcinoma.

Drug/Regimen	Institution	Eligibility
CaG – bevacizumab	MSKCC	Frontline, renal dysfunction
Eribulin	USC	Frontline, salvage
Cisplatin – pemetrexed	Spanish	Frontline
Oxaliplatin – docetaxel	Stanford	Frontline, Cr <1.8 mg/dl
AG – paclitaxel	MDACC	Frontline, renal dysfunction
GC – sunitinib	US Oncology	Frontline
GC – lapatinib	EORTC	Frontline
DD-MVAC → GC	Hellenic	Frontline
GC – IMC-A12	SWOG	Frontline

CaG, carboplatin, gemcitabine; Cr, creatinine; AG, doxorubicin, gemcitabine; GC, gemcitabine, cisplatin; DD-MVAC, dose-dense MVAC; IMC-A12, IGF-1 R monoclonal antibody.

Ixabepilone

Ixabepilone is a semisynthetic analog of epothilone B, which is a novel promoter of tubulin polymerization. Ixabepilone was evaluated for the second-line therapy of metastatic TCC in a phase II trial of 45 patients, of whom 40% had received a prior taxane [Dreicer et al. 2007]. Five patients attained a PR among the 42 eligible patients for a RR of 11.9%, and the median OS was 8 months. Toxicities were moderate with neutropenia, fatigue, and sensory neuropathy being the most common.

Oxaliplatin

Oxaliplatin is a non-nephrotoxic third-generation platinum analogue. However, peripheral neuropathy is frequently dose limiting. Oxaliplatin 130 mg/m² every 3 weeks was evaluated in 18 evaluable patients with previously treated metastatic TCC [Winquist et al. 2005]. Patients were stratified as ‘cisplatin sensitive’ or ‘cisplatin resistant’ on the basis of prior cisplatin treatment. One PR was observed in 10 cisplatin-sensitive patients, and no responses occurred in 8 cisplatin-resistant patients. The combination of oxaliplatin and gemcitabine has been evaluated in a frontline phase II trial of 30 patients, and a serum creatinine up to 1.5 × ULN (upper limit of normal) was allowed [Theodore et al. 2006]. Three CRs and 11 PRs were observed for an overall RR of 47%. Median survival was 15 months and toxicities were manageable. The combination of oxaliplatin and docetaxel is being evaluated in an ongoing trial of frontline therapy, and patients with a serum creatinine <1.8 mg/dl are eligible (Table 3). Given the frequency of renal dysfunction and probable inferiority of carboplatin-based regimens, oxaliplatin may warrant further development in TCC.

Novel taxanes

Larotaxel (XRP9881, formerly RPR 109 881) is a novel semisynthetic taxoid made from the natural taxane extracted from the needles of yew trees. Larotaxel exhibited preclinical activity against multidrug-resistant tumors and has been administered clinically with manageable toxicities [Gelmon et al. 2000]. An ongoing phase III trial in Europe is comparing GC with the combination of larotaxel and cisplatin (Table 1). Nab-paclitaxel is a novel solvent-free, albumin-bound formulation of paclitaxel designed to avoid solvent-related toxicities and to deliver paclitaxel to tumors via molecular pathways involving an endothelial cell-surface albumin receptor and an albumin-binding protein expressed by tumor cells and secreted into the tumor interstitium (secreted protein acid rich in cysteine [SPARC]) [Nyman et al. 2005]. Nab-paclitaxel is being evaluated for the salvage therapy of progressive TCC following prior chemotherapy as well as a component of combination regimens in the neoadjuvant setting (Tables 3 and 4).

Table 4.

Ongoing and planned phase II trials of salvage therapy for metastatic transitional cell carcinoma.

Drug/regimen	Institution
Nab-paclitaxel	Canadian
AZD-4877	Multicenter
Pralatrexate	Multicenter
Sunitinib	MSKCC
Pazopanib	NCI
Tamoxifen	Baylor
Everolimus	Belgian
Everolimus-paclitaxel	SWOG
Aflibercept	NCI
TKI-258	Multicenter
Vorinostat	California

Nab-paclitaxel, nanoparticle albumin bound paclitaxel; AZD-4877, antitubulin kinesin spindle inhibitor; Pralatrexate, antifolate; TKI-258, multitargeted TKI against VEGF and FGF receptors.

Other novel chemotherapeutic agents and strategies

The activity and feasibility of DD-MVAC is being further evaluated in the neoadjuvant setting at the Dana Farber Cancer Institute (Table 4). A Hellenic Oncology Research Group trial is evaluating the strategy of sequential DD-MVAC followed by GC as first-line treatment in patients with locally advanced or metastatic bladder cancer (Table 1). In patients ineligible for cisplatin, the EORTC has completed accrual on a trial comparing the combination of carboplatin-gemcitabine with carboplatin-methotrexate-vinblastine, and mature data are awaited (Table 1). Eribulin (E7389) is a synthetic derivative of the marine sponge product halichondrin-B that inhibits tubulin polymerization and is being evaluated as front-line or second-line therapy for metastatic TCC (Table 3). A novel antimitotic agent that inhibits the kinesin spindle protein, AZD-4877, is being evaluated in the second-line setting (Table 4). Pralatrexate, a potent methotrexate analogue is also being evaluated as salvage therapy (Table 4). Irinotecan displayed poor activity in the second-line setting in a recently reported trial [Beer et al. 2008].

EGFR and HER2 targeted therapies

Human TCCs overexpress EGFR (epidermal growth factor receptor), which appears to confer a poor prognosis [Sriplakich et al. 1999]. While EGFR is overexpressed in the majority of TCC cases, targeting EGFR in TCC has not yet proven beneficial. Similarly, many TCC tumors overexpress Her-2/neu, and the role of targeting Her-2/neu is being actively investigated.

Cetuximab

Cetuximab, an EGFR monoclonal antibody, both alone and in combination with paclitaxel inhibited tumor growth and metastasis in a preclinical model [Inoue et al. 2000b]. An ongoing randomized phase II trial conducted by the National Comprehensive Cancer Network (NCCN) is evaluating the combination of cetuximab with front-line GC (Table 1). Another trial conducted at the MDACC is evaluating the strategy of maintenance docetaxel alone or with cetuximab in patients with stable or responding disease after frontline chemotherapy (Table 1). The Fox Chase Cancer center is investigating cetuximab as second-line therapy, either alone or with paclitaxel (Table 1). Cetuximab is associated with the typical toxicities of this class of agents, including skin rash, diarrhea, hypomagnesemia and rare hypersensitivity reactions. Depending on the outcome of these trials, the correlation of rash with clinical benefit may warrant analysis, since rash is associated with improved outcomes in advanced colorectal cancer.

Trastuzumab

Her-2/neu expression is variable in TCC and may be associated with a more aggressive clinical course [Lonn et al. 1995]. Patients with metastatic TCC or squamous cell carcinoma that expressed Her 2/neu (by immunohistochemistry [IHC], serology or fluorescence in situ hybridization [FISH]) in the primary or metastatic site were treated with trastuzumab in combination with paclitaxel, carboplatin and gemcitabine in a phase II trial [Hussain et al. 2007]. Owing to the possibility of trastuzumab-related cardiac toxicity and the lack of historic data on response rates in Her-2/neu expressing TCC, this study was designed with the primary endpoint of assessing cardiac toxicities and a serious cardiac toxicity rate of ≤2% was considered acceptable. Fifty-seven (52.3%) of 109 registered patients were Her-2/neu positive by one or more of the methodologies. Her-2/neu positive patients had more metastatic sites and a higher rate of visceral metastasis than did Her-2/neu negative patients. Forty-four of 57 Her-2/neu–positive patients were treated with the regimen. Overall, 32.6% of patients had previously received peri-operative chemotherapy, and 55% had visceral metastases. The most common grade 3/4 toxicity was myelosuppression with two toxic deaths. Grade three sensory neuropathy occurred in 14% of patients, and 22.7% experienced grade one to three cardiac toxicity. Two patients (4.5%) had grade three toxicity: one left ventricular dysfunction and one sinus tachycardia. Although the overall rate of serious cardiac toxicity was higher than the predefined acceptable threshold, the percentage of symptomatic cardiac dysfunction was low. Thirty-one (70%) of 44 patients responded (five CRs and 26 PRs), and 25 (57%) of 44 were confirmed responses. Median time to progression and survival were 9.3 and 14.1 months, respectively. Given the aggressive course of disease in this high-risk population, these outcomes are considered promising, although the single arm design makes the interpretation of results difficult. A randomized phase II European trial is evaluating GC with or without trastuzumab as frontline therapy for metastatic TCC, which will help to clarify the role of trastuzumab in bladder cancer treatment (Table 2). Trastuzumab is also being evaluated in combination with paclitaxel and radiotherapy for bladder conservation in patients with localized/locally advanced TCC of the bladder.

EGFR and HER2 receptor tyrosine kinase inhibitors (TKIs)

Preclinical antitumor activity of gefitinib correlates with the degree of expression of EGFR [Ciardiello et al. 2000; Sirotnak et al. 2000]. In EGFR-expressing human bladder cancer cell lines, gefitinib inhibited extracellular signal-regulated kinase and Akt/protein kinase B phosphorylation as well as EGFR phosphorylation [Dominguez-Escrig et al. 2004]. Gefitinib demonstrated a PR rate of only 3% in the second-line setting of a broad population with advanced TCC [Petrylak et al. 2003]. A phase II trial by the CALGB (Cancer and Leukemia Group B) combined gefitinib with cisplatin and fixed-dose rate gemcitabine 10 mg/m²/minute [Philips et al. 2008]. Unfortunately, this regimen produced excessive toxicity likely related to the fixed-dose rate gemcitabine. Subsequently, the study was amended to use a conventional 30 minute gemcitabine infusion. However, the conventional GC schedule in combination with daily gefitinib did not demonstrate clearly improved outcomes compared to historical controls, with a RR of 51% and median survival of 14.4 months [Philips et al. 2009]. An ongoing European randomized study is evaluating conventional GC with or without gefitinib (Table 1).

Lapatinib is an oral TKI which targets EGFR and HER2. In a preliminary report of a phase II trial of 59 patients with EGFR and/or HER2 expression, lapatinib had little activity as salvage therapy for metastatic TCC after failure of frontline chemotherapy, with PRs in 3% and clinical benefit (response + stability ≥16 weeks) in 12% of patients [Wulfing et al. 2005]. The median time to progression was 8.6 weeks, although there was a trend towards clinical benefit in those with EGFR or HER2 2+/3+ by immunohistochemistry (IHC). Preliminary analysis suggested that high tumor pHer3, high pErk and both mutant p53 and high pHer3 may predict resistance, while high pAkt and high IGF-1R may predict sensitivity to lapatinib. Key adverse events were diarrhea (39%), rash (32%), nausea (27%), vomiting (22%), asthenia (12%) and fatigue (10%). The primary Grade 3–4 toxicities were vomiting (7%) and diarrhea (3%) and one patient had an asymptomatic Grade 2 decrease in left ventricular ejection fraction. An ongoing phase I/II trial is evaluating the combination of GC and lapatinib for metastatic TCC (Table 3). A randomized trial being conducted in the United Kingdom is evaluating maintenance lapatinib or placebo in patients with EGFR and/or Her2-expressing tumors with stable or responding disease after frontline chemotherapy for metastatic TCC (Table 1). Erlotinib is being studied in the neoadjuvant setting before cystectomy with primarily tumor tissue based correlative and pharmacodynamic endpoints (Table 4).

Anti-angiogenic strategies

Bladder tumors produce high levels of multiple angiogenic stimulatory factors, including VEGF, bFGF and IL-8 [Crew et al. 1997; O’brien et al. 1995; Allen et al. 1993, Nguyen et al. 1993]. Levels of these factors correlate with stage and outcome [Slaton et al. 2004; Izawa et al. 2001]. Microvessel density, a surrogate marker for angiogenic activity, is a predictor of disease progression, vascular invasion, lymph node involvement, tumor recurrence, and poor survival in invasive TCC [Canoglu et al. 2004; Goddard et al., 2003; Dinney et al. 1998a; Bochner et al. 1995; Jaeger et al. 1995; Dickinson et al. 1994] Levels of VEGF and bFGF are inversely associated with prognosis [Beecken et al. 2005]. Based on these findings, it is hypothesized that targeting angiogenesis pathways either alone or in combination with standard chemotherapeutic regimens in TCC of the bladder will lead to improvement in patient outcomes. Preclinical models in bladder cancer suggest that anti-angiogenic therapies alone or in combination with chemotherapy may inhibit progression of bladder cancer, and that VEGF is the primary pro-angiogenic mediator of this progression [Inoue et al. 2000a; Campbell et al. 1998; Dinney et al. 1998b; Tanaka et al. 1997]. Both VEGF mRNA and protein are over-expressed in advanced TCC compared with normal urothelium [Yang et al. 2004; Crew et al. 2000]. In addition to its pro-angiogenic properties, recent in vitro experiments also suggest a role for VEGF signaling as an autocrine and paracrine growth factor to directly promote bladder cancer growth [Wu et al. 2003]. Furthermore, retrospective evaluation of serum VEGF levels in the metastatic setting suggests a correlation of high levels with poor disease-free survival [Bernardini et al. 2001]. Baseline VEGF mRNA expression levels and microvessel density were found to be independent prognostic factors for recurrence and metastasis in 51 patients treated with neoadjuvant MVAC chemotherapy preceding cystectomy [Inoue et al. 2002]. In addition to its pro-angiogenic role, elevated levels of VEGF in tumors lead to abnormal microvasculature. Excessive angiogenic factors recruit endothelial and perivascular cells to form tortuous and dilated blood vessels with poor rheological characteristics, abnormal tumor blood flow and increased vascular permeability [Dvorak et al. 1995]. These changes lead to increased interstitial fluid pressure, which impairs the delivery of chemotherapy to tumor cells due to a decrease in the pressure gradient [Gutmann et al. 1992; Boucher et al. 1991]. By reducing VEGF levels, the aberrant tumor-associated blood vessels are eliminated and the microvasculature also appears to be remodeled, leading to more ‘normal’ blood vessel architecture. This leads to improved transvascular drug delivery directly to tumor cells, which has been demonstrated in other settings [Willett et al., 2004]. Recent evidence demonstrates that VEGFR2 is expressed in urothelial carcinoma and its level of expression correlates with pathologic stage [Wu et al. 2003, Xia et al. 2006]. Targeting VEGFR2 therefore has the potential to suppress both tumor cells and blood vessels.

Bevacizumab

Bevacizumab, a monoclonal antibody targeting VEGF, has proven beneficial when added to chemotherapy in colon and lung cancer. A phase II trial by the HOG (Hoosier Oncology Group) evaluating frontline GC plus bevacizumab for metastatic TCC has completed accrual and the data is maturing. The Cancer and Leukemia Group B (CALGB) will conduct a frontline randomized phase III trial of GC versus GC-bevacizumab (Table 1). Bevacizumab is also being evaluated in a phase II trial in combination with carboplatin plus gemcitabine (CaG) in previously untreated patients ineligible for cisplatin chemotherapy (Table 3). Separate phase II trials are evaluating neoadjuvant GC or DD-MVAC plus bevacizumab followed by radical cystectomy in patients with muscle-invasive and resectable TCC of the bladder (Table 5). While bevacizumab is generally tolerable, it is known to be associated with a small risk of severe toxicities, including cardiovascular events, venous thromboembolism, arterial thrombotic events, bleeding, hypertension, reversible posterior leukoencephalopathy, and proteinuria. Therefore, administration of bevacizumab in combination with chemotherapy for patients with TCC should only be performed in the context of a clinical trial.

Table 5.

Ongoing and planned phase II trials of preoperative therapy for muscle invasive transitional cell carcinoma.

Drug/regimen	Institution
GC – bevacizumab	MUSC
DD-MVAC – bevacizumab	MDACC
DD-MVAC	DFCI
Ifosfamide – cisplatin – Nab paclitaxel	MSKCC
CaG – Nab – paclitaxel	U. Michigan
Erlotinib	UNC
Sunitinib	Cleveland Clinic
GC – sunitinib	Baylor-HOG
Dasatinib	Baylor-HOG
Temsirolimus	UT Southwestern
Ipilimumab	MDACC

GC, gemcitabine, cisplatin; MUSC, Medical University of South Carolina; DD-MVAC, dose-dense MVAC (methotrexate, vinblastine, doxorubicin, cisplatin); CaG, carboplatin, gemcitabine; HOG, Hoosier Oncology Group.

Other monoclonal protein-based anti-angiogenic agents

Aflibercept (VEGF-trap) is a VEGF receptor fusion protein that has greater affinity for VEGF than bevacizumab and also targets placental growth factor (PlGF), and is being evaluated by the NCI in the salvage setting following failure of front-line chemotherapy (Table 4). Novel monoclonal antibodies against VEGF receptors, insulin-like growth factor (IGF)-1 receptor and the angiopoietin-tie2 pathway are emerging and may warrant evaluation for TCC since these targets are expressed [Rochester et al. 2007; Oka et al., 2005]. One patient with metastatic TCC refractory to GC exhibited a CR when receiving the combination of carboplatin-paclitaxel and AMG 386 (angiopoietin neutralizing antibody) in a phase I trial [Mita et al. 2007].

VEGF-receptor TKIs

VEGF signaling primarly occurs via the VEGFR1 (Flt-1) and VEGFR2 (KDR) TKI receptors, both of which are overexpressed in tumor vasculature and represent attractive targets in TCC. A key to success of targeted anti-angiogenic therapy in the future may be the combination of multiple inhibitors against different targets or the use of single inhibitors directed against two or more targets. Sorafenib, a multi-targeted receptor TKI designed as a c-and b-raf kinase inhibitor also inhibits several other receptor tyrosine kinases, among them VEGF receptor (R)-2, PDGFR-β, Flt-3 and c-KIT. Sorafenib did not demonstrate significant activity in the second-line therapy of metastatic TCC following platinum-based chemotherapy [Dreicer et al. 2008]. There were no objective responses and the median survival was only 6.8 months. In the potentially more sensitive setting of first-line therapy with sorafenib as a single agent for metastatic TCC, none of 14 evaluable patients displayed an objective response [Sridhar et al. 2008]. Four patients exhibited stable disease as the best response and the median time to progression was a disappointing 1.8 months. The combination of sorafenib with GC is being evaluated for frontline therapy in a randomized phase II European trial (Tables 1 and 4).

A preclinical study recently demonstrated significant activity for sunitinib against TCC both as a single agent and in combination with cisplatin [Sonpavde et al. 2008]. Preliminarily, modest activity has been demonstrated in phase II trials of sunitinib (50 mg daily, for 4 weeks of every 6-week cycle) as frontline or salvage therapy of metastatic TCC [Bellmunt et al. 2008; Gallagher et al. 2007]. In the salvage setting of a heavily treated population that had received 1–4 chemotherapeutic agents, three of 41 (7%) evaluable patients achieved PR and the clinical benefit rate (responding plus stable disease) was 31% [Gallagher et al. 2007]. Prolonged stable disease was seen in a small proportion of patients (7% for >6 months and 17% for >3 months). The median PFS was 2.4 months and median survival was 6.9 months. Radiographic regression was observed in liver, lung, bone, bladder, soft tissue and lymph node lesions. There were a number of rare but serious Grade 3–4 toxicities including abdominal pain, anorexia, diarrhea, fatigue, hand and foot syndrome, hemorrhage, hypertension, mucositis, skin ulceration, thrombosis and emesis. A different and probably more-tolerable schedule of 37.5 mg daily continuously is being evaluated in the same setting by MSKCC. Additionally, correlative studies are being performed to identify subsets of highly sensitive tumors. In a frontline trial, patients unsuitable for cisplatin with a creatinine clearance between 30 and 60 ml/min and ECOG performance status ≥1 received sunitinib 50 mg daily for 4 weeks of every 6 weeks [Bellmunt et al. 2008]. Of 14 evaluable patients, 2 PRs (14.3%) were obtained, one not confirmed due to brain metastasis progression; 9 patients (64.3%) had SD lasting >3 months. The clinical benefit rate was 78.6% and the median PFS was 6 months. Another trial is evaluating sunitinib compared to placebo in patients that are stable or responding to frontline chemotherapy (Table 1). Sunitinib in combination with GC is being evaluated in separate phase II trials, as preoperative or first-line therapy for metastatic TCC (Tables 3 and 5). The Cleveland Clinic is evaluating neoadjuvant sunitinib alone with primarily correlative studies (Table 5).

Axitinib, a similar multitargeted receptor TKI caused regression of subcutaneous human TCC xenografts and inhibited angiogenesis and phosphorylation of VEGFR-2 and PDGFR-β, and further evaluation may be warranted [Zhu et al. 2006]. An ongoing trial is evaluating pazopanib, a multitargeted TKI, for metastatic TCC in the second-line setting (Table 4). A randomized phase II trial is evaluating salvage docetaxel alone or with vandetanib, a dual EGFR and VEGFR-TKI, in patients that have received up to 3 prior regimens (Table 1).

Other novel agents and strategies

Based on the finding that ER-β expression in TCC increases with increasing stage and grade, and the inhibitory effect of selective estrogen receptor modulators in preclinical models, salvage therapy with oral tamoxifen is being evaluated in a multi-institutional phase II trial of metastatic TCC (Table 4) [Sonpavde et al. 2007; Shen et al. 2006]. Bortezomib, a proteasome inhibitor, displayed poor activity as a single agent in the salvage setting [Rosenberg et al. 2008]. However, based on synergism with chemotherapeutic agents, the evaluation of a combination of bortezomib with chemotherapeutic regimens is ongoing (Table 3). Inhibitors of signaling pathways are being developed premised on preclinical data. Everolimus, a novel orally administered mTOR inhibitor is being evaluated in the salvage setting, as a single agent or in combination with paclitaxel in separate trials (Table 4). Temsirolimus, the mTOR inhibitor approved for renal cell carcinoma, will be evaluated in the neoadjuvant setting with correlative studies as the primary endpoints (Table 5). TKI258, a multitargeted receptor TKI of VEGF and FGF receptors is being evaluated in the salvage setting (Table 4). Other novel avenues of research, including epigenetic therapy (e.g. vorinostat, a histone deacetylase inhibitor) and immune-modulation (ipilimumab to down-regulate CTLA-4 expressing T-regulatory lymphocytes), are being evaluated (Tables 4 and 5). Depsipeptide, another histone deacetylase inhibitor, did not demonstrate activity as salvage therapy for metastatic TCC in a trial conducted by SWOG.

The neoadjuvant therapy paradigm to develop novel agents

The paradigm of neoadjuvant therapy before surgery in localized disease permits rapid in vivo assessment of pathologic response, and may accelerate the development of novel systemic therapies. Pathologic complete remission (pCR) is increased with cisplatin-based combination chemotherapy, and is associated with improved long-term outcomes after cystectomy [Grossman et al. 2003]. Owing to the availability of tissue before and after chemotherapy, it may be possible to determine molecular and biologic characteristics that predict for chemosensitivity and facilitate the development of personalized therapy. The choice of novel agents should be based on the knowledge of potential molecular targets emerging from studies examining TCC biology [Cordon-Cardo, 2008]. If biologic activity can be demonstrated in initial small pilot trials (e.g. 15–25 patients with an appropriate statistical design to detect a biologic effect), additional larger phase II studies of novel agents alone or in combination, potentially using randomized phase II designs may be planned with more stringent efficacy endpoints (pCR). Several ongoing trials are evaluating neoadjuvant regimens and agents with pathological or pharmacodynamic endpoints (Table 5). Testing a regimen in metastatic disease should still be required before embarking on a large randomized trial, since activity in the neoadjuvant setting may not always translate to benefit in the metastatic setting. Since metastatic TCC is uncommon compared to locally advanced resectable disease, efficient clinical trials testing novel agents can help accelerate the development of new TCC treatments.

Predictive factors for response

To guide optimal patient selection, the discovery of factors predictive for response should proceed in concert with the development of novel agents. While cytotoxic chemotherapy is not classically considered targeted therapy, many of these drugs affect specific molecular targets within the cancer cell, and predictors of response may play a role in determining selection for the most appropriate therapy. Levels of DNA-repair genes including ERCC1, RRM1, BRCA1 and caveolin-1 were evaluated in 57 advanced bladder cancer patients treated with cisplatin-based combination chemotherapy [Bellmunt et al. 2007b]. Median survival was significantly higher in patients with low ERCC1 levels (25.4 versus 15.4 months; p = 0.03). A trend towards longer time to progression was observed in patients with tumors expressing low levels of all markers. On multivariate analysis with pretreatment prognostic factors, ERCC1 emerged as an independent predictive factor for survival. Correlation was also found between low/intermediate BRCA1 mRNA levels (which mediates DNA repair) and pCR and long-term outcomes with neoadjuvant cisplatin-based combination chemotherapy in a retrospective study of 49 patients [Font et al. 2008]. Predictors of response to novel agents are critical as well, and will hopefully be defined as studies proceed.

Conclusion

Few patients achieve long-term survival with currently employed regimens for metastatic TCC. Current regimens yield suboptimal outcomes in the frontline setting and there is no proven effective second-line regimen. Therefore, patients with metastatic TCC in both the frontline and salvage chemotherapy settings should be considered candidates for trials. Unfortunately, TCC patients are frequently elderly and have multiple comorbidities. Furthermore, metastatic TCC patients often rapidly progress and experience a decline in performance status, which also renders their participation in trials particularly difficult. Therefore, close attention to tolerability is imperative when developing new treatments.

Disease characteristics of TCC patients are heterogeneous and impact on treatment outcomes. This leads to difficulty assessing the true benefit of an agent in a single-arm phase II trial with objective response as the primary endpoint. Therefore, randomized and appropriately stratified phase II trials with time-to-event endpoints should generally be supported when testing new therapies.

While objective response rates to frontline therapy are generally high, nearly all patients with metastatic TCC will progress. Therefore, therapy to maintain and prolong a response using a tolerable targeted agent following frontline chemotherapy may have value, and is being evaluated with multiple new agents. Consolidation or maintenance of a response appears to be a worthy goal in metastatic TCC, if toxicity is manageable for chronic therapy.

The neoadjuvant paradigm should play an important role in the development of novel agents, as it will allow development and early assessment of biomarkers of response and progression. The neoadjuvant approach to drug development requires close collaboration between medical oncologists, urologists and laboratory scientists. The integration of novel biologic agents with systemic chemotherapy for muscle-invasive and metastatic TCC is required to improve outcomes. GC chemotherapy has been chosen as the platform to further develop combination therapy due to its tolerability and similar efficacy to other cisplatin-based regimens.

While a number of oncogenic molecules are being targeted, a single critically important target has not emerged in TCC (e.g. bcr/abl for chronic myeloid leukemia, Her-2 for breast cancer). Further research into the fundamental biology of TCC may yield more (and possibly better) targets. mTOR inhibition, PI3 kinase/Akt inhibition, FGFR3 inhibition, and Mek inhibition should be tested in TCC once agents are available for phase II testing. A special focus on patients who have recurred following prior chemotherapy or are not candidates for cisplatin is necessary, since these patients currently experience particularly poor outcomes. Factors predictive of response to new and existing agents may facilitate personalized therapy and enable judicious patient selection even in the early stages of drug development. However, novel combinations should only be administered in the context of a clinical trial at this time, since combinations proven in other malignancies may not improve outcomes in TCC.

Conflict of interest statement

Guru Sonpavde: Speakers’ bureau for Sanofi-Aventis, Pfizer, Wyeth and Novartis; Research support from Pfizer, Novartis, BMS, Eli Lilly, AstraZeneca and Cytogen.

Aymen A. Elfiky: None.

Jonathan E. Rosenberg: Speakers bureau for Novartis; Research support from Novartis; Consultant, Genentech.