Health and Prostate

Renal cancer is a relatively rare cancer accounting for approximately 3% of all adult malignancies. There are more than 5000 new cases diagnosed per year in the UK and the incidence is increasing. It most commonly affects people in their fifth to seventh decades with a male-to-female ratio of approximately 2:1.

Renal cell carcinoma (also known as clear cell carcinoma or hyper-nephroma) accounts for 80-85% of all kidney cancers. Papillary renal carcinoma constitutes approximately 10%, with the remainder including chromophobe and collecting duct carcinomas. Transitional cell carcinomas, squamous cell carcinomas and lymphomas can also arise in the kidney.

Risk factors for the development of Renal cell carcinoma include smoking, obesity, adult polycystic kidney disease and long-term renal replacement therapy. Approximately 1% of RCCs are hereditary, the most commonly associated syndrome being the von Hippel-Lindau syndrome. It has recently been demonstrated that most cases of sporadic Renal cell carcinoma have mutations in the von Hippel-Lindau gene, located on the short arm of chromosome three, which normally acts as a tumour suppressor gene.

The most commonly used method of staging renal cancer is the Robson staging system. Stage of disease and prognosis are closely linked. For tumours confined to the kidney (stage I) and with tumours that involve perinephric fat (stage II) the 5-year survival is 50-80% [1]. In patients with distant metastatic disease (stage Intravenous) the median survival is just 8 months with a 5-year survival of less than 10%.

Approximately one quarter of patients present with metastatic disease. A further 30-40% of patients present with apparently localized disease but eventually develop distant metastases . The lung is the most common site of metastasis (50-60%), followed by bone (30-40%), liver (30-40%) and brain (5%).

Whilst radical nephrectomy can be curative in patients presenting with localized disease, the treatment options available to those with metastatic disease remain limited and far from ideal.

Renal cell carcinoma is notoriously resistant to conventional chemotherapy agents and these drugs are not commonly used in disease management. Radiotherapy has a useful palliative role in a limited capacity, used mainly in the treatment of painful bony metastases, localized subcutaneous metastases and brain metastases. Hormonal agents, in particular medroxyprogesterone acetate and tamoxifen (more commonly used in mainland Europe) are often used to palliate the systemic effects of cancer. Their anticancer activity however is very limited, with objective response rates of approximately 1%.

Attention has largely focused on the use of biological agents. Interferon-alpha (IFN-a) and interleukin-2 (IL-2) have both been used extensively in Renal cell carcinoma over the past 2 decades, with encouraging results. It is well established that the use of such agents, termed immunotherapy, can induce complete and durable responses in a small number of patients. In the UK, IFN-a is now widely accepted as standard first-line therapy.

Whilst some progress has been made in recent years, the prognosis for patients with metastatic (m) Renal cell carcinoma remains poor. Thus there remains an urgent need for the development of novel, more active agents. A greater understanding of the biology of this and other cancers has lead to the recent development of a number of new and promising strategies.

Principles of therapy

Drugs available

Future developments

Conclusions

Renal cancer remains amongst the more difficult of cancers to treat. Effective therapies for the majority of patients with mRCC have not been found. Traditional chemotherapy drugs are ineffective and are not commonly used.

Currently, immunotherapy offers the best hope for patients. response rates may be increased by nephrectomy and should be considered before commencing systemic treatment. Standard first-line therapy in the UK is single-agent IFN-a, which has a modest overall RR of approximately 15%.

Atzpodien’s triple therapy is a convenient, relatively well tolerated, outpatient-based regime that has shown high response rates in some studies. The Medical Research Council RE04 study will determine whether it confers a real advantage over single-agent interferon.

Beyond this, there are a wealth of new approaches being developed. A better understanding of the biology of cancer has lead to several novel therapies. Ultimately, success is likely to come from using a combination of agents that simultaneously target the multiple processes that drive the cancer cell.

Chemotherapy

Cytotoxic chemotherapy remains the mainstay of systemic treatment for most solid cancers. Renal cell carcinoma however is a characteristically unresponsive tumour and at present, these agents are not routinely used.

A comprehensive review of chemotherapy in Renal cell carcinoma, performed by Yagoda etal., included 4093 patients in 83 trials between 1983 and 1993. The trials included every class of anticancer agent. The overall RR was 6% (1.3% complete response (CR); 4.7% partial response (PR)). Similarly, a review of published literature between 1990 and 1998 showed no survival benefit for any single-agent chemotherapy drug.

Vinblastine is a cytotoxic drug with some activity in Renal cell carcinoma. For a time it was regarded as the best available agent. However, it is clear that response rates are low, typically between 2% and 7%, and Vinblastine is no longer routinely used in the treatment of Renal cell carcinoma.

5-Fluorouracil is a cytotoxic drug that has some single-agent activity (approximately 10% RR) in Renal cell carcinoma. Introduced in the 1950s, it represents one of the first rationally designed anti-tumour agents. 5-Fluorouracil is a fluorinated analogue of uracil, a pyrimidine base essential for nucleic acid metabolism. Through a number of mechanisms it interferes with DNA, RNA and protein synthesis. Myelosuppression and gastrointestinal (gastrointestinal) toxicity are the predominant side-effects. The role of 5-Fluorouracil, in combination with IFN-a and IL-2, is currently being investigated.

Renal cancer cells have an intrinsic resistance to chemotherapy. Classical multi-drug resistance is associated with a membrane-bound protein, P-glycoprotein, that functions as a drug efflux pump. P-glycoprotein, and its encoding gene, multi-drug resistance-1, have been shown to be overexpressed in many tumours, including Renal cell carcinoma.

This is one of a number of mechanisms of drug resistance that has been described although several others have also been implicated. Until our understanding of these pathways improves, standard chemotherapy drugs are likely to remain largely obsolete in the treatment of renal cancer.

Immunotherapy

Scientists have long suspected that there is considerable interaction between tumour cells and the host’s immune system. The small but significant number of spontaneous regressions of tumours, the increased incidence of malignancy in immunosuppressed patients and the presence of lymphoid infiltrates in solid tumours lent early support to this notion.

In recent years, greater understanding of basic immunology and of the host tumour relationship has lead to the development of immunotherapeu-tic approaches to cancer treatment. Biotherapy can be broadly defined as the use of natural substances that, by modifying the host’s biological response to tumour cells, leads to therapeutic benefit. The term immunotherapy can be encompassed within this and relates specifically to approaches aimed at stimulating immune defence mechanisms.

The ability to differentiate tumour cells from normal cells is based on the ability of the immune system to detect tumour cells expressing abnormal proteins or abnormally expressed normal proteins (tumour antigens). Tumour antigens expressed on the cell surface can be detected by antibodies. Expression of intracellular tumour antigens is detected by T cells. Intracel-lular proteins are proteolytically degraded into small peptides, and a proportion of these are assembled into the groove of human leucocyte antigen molecules. This human leucocyte antigen molecule/peptide complex is transported to the cell surface where it can be recognized by the T cell receptor of cytotoxic T cells. The binding of the T cell receptor to the major histocompatibility complex (MHC)/peptide complex triggers T cell activation and release of cytotoxic factors that induce death of the target cell. T cells however require other signals to give full activation, including cytokines and costimulatory signals from specialized antigen-presenting cells such as dendritic cells. Indeed, effective T cell activation involves several steps.

Antigens shed by dying tumour cells are picked up by dendritic cells and processed into human leucocyte antigen molecules. dendritic cells migrate to the lymph nodes where the antigens are presented to T cells along with the appropriate costimulatory signals. These

partially activated T cells can then circulate and when they encounter tumour cells expressing these antigens they are fully activated and kill the tumour cell. This whole process is controlled by a network of cytokines. Cytokines are a broad class of soluble proteins produced primarily by T cells, antigen-presenting cells, mono-cytes and macrophages and act as signals between cells of the immune system.

Tumour antigens have now been identified in many cancers and cytotoxic T cells specific for these can be isolated from patients with cancer. Unfortunately, in the majority of cancer patients there are defects in this immune response, and to date many different mechanisms of immune evasion have been identified. The aim of immunotherapy is to overcome these to allow effective immune stimulation.

Several approaches have been tried. These include specific and non-specific immunotherapies. Non-specific immunotherapies include cytokines such as IFN and IL-2 that enhance the immune response by a range of different mechanisms. Immunotherapies that enhance the immune response to specific antigens include vaccination with antigens in the form of whole cell vaccines, specific protein or specific peptides. Although many approaches have been tried in renal cancer the current mainstay of immunotherapy in renal cancer is the cytokines IFN-a and IL-2.

Surgery

The role of surgery in the management of localized Renal cell carcinoma is well established. Radical nephrectomy is potentially curative for many patients with early stage disease. For patients with metastatic disease, decisions regarding surgery are more complicated.

At one time, nephrectomy was carried out in the hope of inducing a spontaneous regression of metastases. In fact, the frequency of this phenomenon is less than 1% [8] whilst the surgery itself carries a mortality of between 1% and 5%. There is therefore no justification for nephrectomy based solely on this assumption.

More recently however, data from two randomized trials have been published suggesting that nephrectomy in patients with metastatic disease can increase the likelihood of an objective response to immunotherapy at metastatic sites.

The European Organization for Research and Treatment of Cancer (EORTC) 30 947 trial randomized 83 patients with mRCC to either radical nephrectomy plus IFN-based immunotherapy versus IFN-a alone. Time to progression (5 vs 3 months) and median duration of survival (17 vs 7 months) were significantly better in the study patients than in the controls.

The results of a trial by the Southwest Oncology Group (SWOG) were similar. In that study of 245 patients, median survival was 12.5 months in the combined treatment arm compared with 8.1 months in patients who had interferon only and no nephrectomy.

It is important to note that only patients with a WHO performance status of 0-1 were eligible for these trials. In addition, both studies recruited relatively small numbers of patients over a very long time period.

Nevertheless, the current weight of evidence is in favour of nephrectomy before commencing immunotherapy for patients of good performance status. However, this approach is not appropriate or feasible in all cases and requires careful patient selection.

Adjuvant treatment

Adequate surgical excision is the only effective means of curing Renal cell carcinoma. Approximately 70% of patients have nonmetastatic disease at the time of first presentation. However, of those who undergo tumour nephrectomy 2-14% relapse locally and 31-36% develop distant metastatic disease.

The concept of adjuvant therapy is well established in several tumours, such as of the breast and colon. In Renal cell carcinoma however, chemotherapy, hormone therapy and radiotherapy have all failed to show a survival benefit when used in this setting. It remains unclear whether immunotherapy can make an impact in patients with Renal cell carcinoma at high risk of relapse.

Pizzocaro et al. randomized 247 patients with Robson stage II and III Renal cell carcinoma to 6 months of thrice-weekly IFN-a versus observation after radical nephrectomy. Overall, no advantage for adjuvant interferon was demonstrated in terms of overall and event-free survival. Other similar studies using single-agent interferon in this setting have failed to demonstrate a survival advantage.

The EORTC 30955 trial is currently open and is using combination therapy in an adjuvant setting. Eligible patients should have a pT3b, pT3c or pT4 stage tumour; any pT stage and nodal status pNlor 2; or any pT stage and microscopic positive margins or evidence of microscopic vascular invasion. Patients are randomized either to an 8-week regimen of triple therapy with IFN-a, IL-2 and 5-Fluorouracil or to observation.

There is currently no evidence to support the routine use of adjuvant therapies in Renal cell carcinoma and observation remains the standard of care. Suitable patients should however be considered for participation in clinical trials.

Interferons

Interferons are a heterogeneous group of glycoproteins produced by mammalian cells in response to viral infections or other inducers. Three major types have been identified – interferon-a, interferon-β (class I) and inter-feron-8 (class II).

As well as helping to fight viruses, interferons have anti-tumour properties. These may be mediated through a direct cytotoxic effect on tumour cells or through augmentation of the immunogenicity of tumours by upregulation of histocompatibility and tumour-associated antigens, and/ or activation of macrophages, T lymphocytes and natural killer cells [14]. It is also thought that interferons possess anti-angiogenic properties.

The anti-neoplastic activity of IFN-a was first shown in hairy cell leukaemia and Kaposi’s sarcoma. Subsequent studies have documented its activity in chronic mylogenous leukaemia, B and T cell lymphomas and melanoma. IFN-a is the most extensively tested IFN in clinical studies of patients with mRCC.

The use of IFN-a in Renal cell carcinoma first began in the 1980s. It is now widely accepted in the UK as first-line treatment for metastatic renal cancer. A series of phase I and II studies have shown overall objective response rates for single-agent IFN-a in the region of 15%, with CRs in approximately 1%. The best responses have been achieved in patients with a good performance status and limited metastatic disease. Prognostic factors that help to predict response include prior nephrectomy, performance status, sites of metastatic disease and interval from presentation to the development of metastases. However, the modest increases in survival can be offset by treatment-related side-effects and the fact that responses are often not durable. Two large randomized phase III trials have been published demonstrating a clear survival advantage with the use of IFN-a.

Pyrohnen et al. prospectively randomized 160 patients with either locally advanced Renal cell carcinoma or mRCC to receive either Vinblastine alone or IFN-a plus Vinblastine for 12 months or until disease progression. At the time, many considered Vinblastine to be standard treatment. In both groups, intravenous Vinblastine was given at 0.1 mg/ kg every 3 weeks, and in the combination arm IFN-a was administered subcutaneously (s.c.) at 3 mega units three times a week for 1 week and 18 mega units three times a week thereafter. The median survival was 67.6 weeks in the combination arm and 37.8 weeks in the Vinblastine arm (p < 0.0049). Overall response rates were 16.5% in the combination arm and 2.5% in the Vinblastine arm (p < 0.0025). Survival rates for patients treated with the combination or with Vinblastine alone were 55.7% and 38.3%, respectively, after 1 year, and 11.7% and 5.1%, respectively, after 3 years. Survival was prolonged in the overall patient population, even in the subset of patients who did not have objective tumour responses, and was not decreased in patients who required a reduction of their IFN-a dose from 18 to 9 mega units to improve tolerability.

A subsequent multicentre randomized trial by the Medical Research Council   Renal   Cancer   Collaborators   (Medical Research Council   RE01)   randomized   335 patients with mRCC to receive either s.c. IFN-a (three doses – 5, 5, 10 mega units – for week 1, then 10 mega units three times per week for a further 11 weeks) or MPA 300 mg once a day for 12 weeks. A survival advantage of IFN-a over MPA was demonstrated, with an absolute improvement in 1-year survival of 12% (43% vs 31%). The median survival time was 8.5 months for patients treated with IFN and 6 months for patients treated with MPA. The time to relapse, however, was short, and progression-free survival at 2 years was 5% or less for both groups. Side-effects were also more marked in the IFN-a group, notably tiredness, anorexia, loss of energy and nausea. The difference was most significant at 4 weeks, persistent at 12 weeks but had resolved by 6 months

Treatment-related toxicity can be marked and, whilst rarely life-threatening, does affect quality of life. Acutely, patients suffer from flu-like symptoms, namely fevers, rigors and myalgia that can be largely alleviated by coadministration of paracetamol. These symptoms usually resolve within approximately 2 weeks of treatment. Lethargy is common and tends to accumulate with repeated injections. In addition, anorexia, weight loss, depression and loss of libido can occur. At higher doses, hepatotoxicity and myelosuppression may be seen.

Several dose schedules have been investigated. Overall, the differences in terms of response rates have been modest. The most commonly used regimen in the UK is 9-10 mega units s.c, three times per week. Patients typically self-administer their interferon on an outpatient basis. Injecting at night means that patients can sleep through the worst of the acute side-effects.

Single-agent IFN-a produces a small but consistent response in renal cancer. Efforts to increase the proportion of patients who respond and to make responses more durable have largely focused around using interferon in combination with other active agents.

Interleukin-2

IL-2 is a 15000-Da glycoprotein secreted predominantly by T helper-1 (Thl) lymphocytes. It is a potent growth factor that causes activation and proliferation of cytotoxic T lymphocytes, natural killer cells and macro-phages. The first clinical trials with IL-2 were carried out in 1984. Promising early results were reported by Rosenberg et <xl., who used high-dose, intravenous bolus IL-2. A 22% overall RR was reported in 54 patients with mRCC treated with IL-2 .

A subsequent multicentre study by Fyfe et <xl. reported an overall RR of 14%. Two hundred fifty-five patients were treated with high-dose bolus IL-2, 600 000-720 000 IU/kg, administered intravenously for 14 consecutive doses over 5 days, and repeated after a 5- to 9-day rest period. A long-term follow-up showed that 15% of patients achieved a CR or a PR. Of those patients achieving a CR, the median duration of response exceeded 80 months and 60% remained in CR at 10-year follow-up. The associated toxicity was severe and treatment-related death was 4%. This was despite the study population consisting of a relatively young group of patients (median age 52 years) with a good performance status.

The side-effects of IL-2 therapy are dose-dependent and at high dose can be formidable. Toxicity is manifest by a vascular leak syndrome as a result of increased capillary permeability. This leads to fluid retention, pulmonary oedema, decreased peripheral vascular resistance, hypotension, tachycardia and oliguria. These in turn can lead to respiratory failure, cardiovascular collapse and renal impairment. Patients often require inotropic support to maintain adequate blood pressure, and occasionally may need transfer to the intensive care unit. As experience with the use of high-dose IL-2 has increased, treatment-related mortality has reduced considerably and is now less than 1%. Careful patient selection is important and all patients should have an adequate assessment of their cardiac, respiratory, renal and hepatic function before commencing treatment.

Since many patients are unsuitable for high-dose treatment, alternative lower-dose regimens have been developed. Continuous infusional IL-2 has been shown to have efficacy comparable to bolus regimens and, by avoiding high peak levels, carries less toxicity.

The use of daily, self-administered subcutaneous IL-2 in patients with mRCC has been widely tested. The convenience and reduced toxicity of this regimen has made it an attractive alternative; it is the commonest route of administration in use in Europe today.

Many phase II studies have been published using subcutaneous IL-2. Essentially these studies show that subcutaneous IL-2 monotherapy produces response rates of 15-20%, similar to those achieved with high-dose therapy. In addition, it is well tolerated. Typical side-effects include fever, malaise, nausea, vomiting and diarrhoea. Rises in serum creatinine, especially in those with pre-existing renal impairment, and thyroid dysfunction may also occur. These toxicities are generally not severe and do not require hospitalization.

It remains uncertain whether high-dose IL-2 confers any benefit over lower-dose regimens in terms of clinical outcome. Whilst response rates appear to be similar, it is not known whether response duration and overall survival are increased using high-dose treatments. IL-2 has been approved by the US Food and Drug Administration for use in patients with mRCC on the strength of studies showing that, in small numbers of patients, IL-2 can induce complete, durable responses, sometimes lasting more than 10 years. The challenge of identifying which patients will respond and of defining the optimal dose and route of administration remains.

Combined immunotherapy

After establishing the individual activity of IFN-a and IL-2 in mRCC, investigators have looked at using them in combination. Studies in in vitro and in vivo animal models were encouraging and showed synergism between the two cytokines. Clinical trials have been disappointing however. A series of small randomized studies have suggested no advantage to combination therapy over IL-2 alone.

However, in a large phase III trial, Negrier et al. randomized 425 patients to receive high-dose intravenous IL-2 alone, subcutaneous IFN-a alone or a combination of both agents. The RR was better for the combination arm (IFN-a 7.5%; IL-2 6.5%; combination 18.6%; p < 0.01) and event-free survival at 1 year also improved (IFN-a 12%; IL-2 15%; combination 20%; p = 0.01). This did not however translate into a significant increase in overall survival.

Biochemotherapy

The concept of combining immunotherapy with standard chemotherapy agents is one that has generated much interest over recent years. The theoretical rationale behind such a combination is that chemotherapy may enhance immunogenicity by causing cellular damage and release of tumour cell antigens that are processed by IFN-a-stimulated antigen-presenting cells and, in turn, activate IL-2-stimulated cellular effectors. This theory, however, has yet to be proven.

One of the more promising regimens in mRCC was first used by Atzpo-dien. Patients with mRCC were treated with a combination of IFN-a, IL-2 and 5-Fluorouracil. Treatment was given in an outpatient setting with 8 weeks of IFN-a (6-9MU/m² s.c, 1-3 times/week), IL-2 for 4 weeks (5-20 mega units/m² s.c, 3 times/week, weeks 1—4) and intravenous 5-Fluorouracil 750 mg/m² weekly for weeks 5-8. The regimen was well tolerated. For the first 35 patients treated, an overall RR of 48.6% was reported.

Since 1993, there have been several phase II trials using this triple-therapy regime with response rates in the range of 15-45%.

In a randomized trial comparing the triple-agent regimen against single-agent tamoxifen, enrolling a total of 78 patients, those patients receiving triple therapy were shown to have an overall RR of 39% compared with 0% in the tamoxifen-only arm. The overall survival (24 vs 13 months) was significantly increased in the triple-therapy arm.

However, one phase III study has been published showing no advantage for triple therapy over interferon plus IL-2 alone. The study used different drug schedules, which may explain the difference.

RE04 is an Medical Research Council randomized controlled phase III clinical trial that is aimed at addressing these issues. The study, currently recruiting, will randomize 670 patients with mRCC to either triple therapy or standard single-agent IFN-a.

Medroxyprogesterone acetate

Medroxyprogesterone acetate (Provera) is a 17-OH progesterone derivative. The rationale for its use in mRCC is derived from the fact that a proportion of renal cancers express oestrogen and progesterone receptors. The exact mechanism of its anti-tumour effect is unclear.

At the time of its introduction there were few alternatives and hence Provera became widely used for patients with mRCC. The reality is that overall response rates with Provera are low (-2%) and short in duration.

At the standard dose of 300 mg daily, Provera is a well-tolerated drug. Its main side-effect is appetite stimulation, which may be beneficial in patients with tumour cachexia or anorexia. Nausea can be a problem early on but tends to improve with continued use. Fluid retention can occur but rarely requires intervention.

Today, Provera can be considered for use in those patients who have either progressed on, or are unsuitable for, first-line immunotherapy. For these patients it may provide a modest palliative benefit.

Anti-angiogenesis agents

Angiogenesis refers to the formation of new blood vessels and is an important step in tumour growth. It is estimated that most tumours need to trigger angiogenesis in order to grow beyond 2 mm in diameter. The initiation and promotion of angiogenesis is under the control of a variety of cytokines and hormones, termed angiogenic factors, and include vascular endothelial growth factor, basic fibroblast growth factor, IL-8 and a tumour necrosis factor. Renal tumours are highly vascular and this makes them an attractive target for inhibitors of angiogenesis.

The development of anti-angiogenesis agents is the focus of much research, both in Renal cell carcinoma and in other solid malignancies. These novel approaches have burgeoned over recent years. Two such agents are discussed below.

Thalidomide is infamous as a drug that was prescribed to pregnant women in the late 1950s to combat morning sickness, which resulted in birth defects in thousands of children. We now know that thalidomide possesses anti-angiogenic properties and in utero this impaired limb development, with tragic consequences.

One mechanism through which thalidomide is thought to work is through the increased degradation of tumour necrosis factor mRNA. tumour necrosis factor plays a role in promoting new blood vessel formation and is also thought to account for many of the systemic effects of malignancy such as cachexia and fever.

There have been several small non-randomized phase II trials using thalidomide in patients with mRCC. Eisen et al. used thalidomide at 100 mg daily in 18 patients with advanced Renal cell carcinoma. Three patients achieved a PR and three showed stable disease for up to 6 months.

In a study of 40 heavily pretreated patients with mRCC, 11 patients had at least stable disease at 6 months (2 out of 11 achieved PR). Thalidomide was given at 400 mg/day and increased to 800 mg/day or 1200 mg/day if patients progressed. Motzer et al. recently published data on 26 patients (15 of whom had had prior systemic treatment) with mRCC treated with 200 mg/day of thalidomide, which was increased every 2 weeks to a maximum of 800 mg/day. No responses were seen but 32% of patients had stable disease at 6 months.

Thalidomide is a relatively well tolerated drug, particularly at a low dose. The most common side-effects include lethargy, sedation, constipation and skin rash. At higher doses peripheral neuropathy and venous thromboem-bolism are occasionally seen.

The monoclonal antibody (mAb) bevacizumab (Avastin) is an anti-vascular endothelial growth factor antibody that is currently in phase II trials in Renal cell carcinoma. In a randomized placebo-controlled trial, 110 patients with mRCC were treated with either high-(10 mg/kg) or low-(3 mg/kg) dose bevacizumab, or placebo. The time to progression was significantly increased in the high-dose group compared with the placebo group (hazards ratio = 2.3, p = 0.001). Only three PRs were observed, all in the high-dose arm. Toxicity at both doses of antibody was minimal.

It is too early to know whether such agents will become a recognized treatment option in Renal cell carcinoma. The use of thalidomide in combination with other agents, such as IFN, is being investigated. It appears that inhibiting angio-genesis may result in stabilization of disease over long periods. If this is true then these drugs may represent a useful option in adjuvant therapy.

Dendritic cell vaccines

Harnessing the immune system to specifically target and destroy cancer cells has long been the goal of tumour immunologists. Approaches using systemic IFN-a and IL-2 lead to non-specific immune activation and, whilst useful responses are achieved, there remains scope for improvement.

Recently, with the identification of tumour-associated antigens), and an improved understanding of the processes underlying antigen acquisition and presentation, the potential for a more rational application of immune-based therapies is becoming possible.

dendritic cells are ‘professional’ antigen-presenting cells. They play a pivotal role in the host tumour response by presenting tumour-associated antigens to cytotoxic T cells. Advances in laboratory techniques have enabled investigators to culture dendritic cells in vitro, ‘load’ them with tumour-associated antigens, and re-infuse these cells back into patients. Antigen binding is achieved by pulsing dendritic cells with known antigen peptide epitopes, whole-antigen proteins or even whole tumour cells or cell lysates.

Holtl et al. performed a pilot study of antigen-loaded dendritic cells in 12 patients with metastatic Renal cell carcinoma. dendritic cells were loaded with cell lysate from cultured autologous tumour cells, then activated with a combination of tumour necrosis factor and prostaglandin E2. The vaccine was administered by three intravenous infusions at monthly intervals. Potent immunological responses against cell lysate could be measured in vitro after the vaccinations, suggesting that a DC vaccine can induce antigen-specific immunity in patients with metastatic Renal cell carcinoma.

DC vaccines seem to be well tolerated and, to date, there have been no significant adverse events. This makes them attractive, particularly for use in less-well patients and combination therapies. More work is needed and clinical studies using DC vaccines are being published in increasing numbers.

Mini allogeneic stem cell transplants

Stem cell transplants have, for many years, been used to salvage the bone marrow of patients who have undergone high-dose, myeloablative chemotherapy. They are most commonly used in haematological malignancies, where they form part of potentially curative treatment. The term allogeneic transplant refers to the use of donor stem cells, usually from human leucocyte antigen-matched siblings.

For a long time investigators suspected that the donor cell population itself may be capable of exerting a direct anti-neoplastic effect. Much work has since been carried out in patients with leukaemia, which has helped to confirm this. It is now clear that immunocompetent donor T cells within the SCT can exert an anti-tumour response, called the graft-versus-tumour response.

Allogeneic Stem cell transplants are now an accepted treatment modality in many haematological malignancies that are refractory to standard chemotherapy. Infusions of donor lymphocytes alone are capable of inducing remissions in patients with relapsed leukaemias.

A similar approach has more recently been tested in patients with solid malignancies, with the most promising results achieved in patients with Renal cell carcinoma. Since high-dose chemotherapy is not beneficial in Renal cell carcinoma, less toxic non-myeloablative conditioning regimens (hence the term ‘mini’) have been developed. This significantly reduces the morbidity and mortality of the procedure.

In a phase II study of 19 patients with mRCC resistant to IL-2, treated with mini allogeneic SCT, a 53% RR (CR: three patients; PR: seven patients) was reported. Patients with a CR remained in remission over 2 years after transplant. There were two treatment-related deaths. The time to regression of metastases was delayed, occurring on average 5 months after transplant.

More recently Childs presented data on 47 patients treated to date. A 47% overall RR was achieved (CR: 4 patients; PR: 18 patients). The majority of the patients had failed standard immunotherapies. There were four treatment-related deaths (9%), two due to infection and two due to graft- versus-host disease. Again, the average time to response was delayed at a median of 6.5 months.

These response rates and their durability are impressive, especially in a study population that has failed conventional treatments. However, there are limitations. Firstly, patients usually require an human leucocyte antigen-matched or single-antigen mismatched sibling. The use of mismatched donors is possible but this increases the risks of an already hazardous procedure. Even in ideal circumstances the procedure carries a significant morbidity and mortality. Patients also need a life expectancy of at least 6 months to make the procedure worthwhile, due to the delayed time to response. Nevertheless, this is a promising new therapeutic option in the treatment of Renal cell carcinoma.

Monoclonal antibodies

The concept of using mAbs in cancer treatment, the so-called ‘magic bullet’ approach, is in itself not new. For many years, investigators have been using monoclonals, alone or in combination with cytotoxic drugs, radioactive isotopes and toxins. To date, two mAbs have achieved broad market success, namely Rituximab in non-Hodgkin’s lymphoma and Herceptin in Her-2-positive breast cancer.

mAbs remain under investigation in a number of tumour types, including renal cell cancer. Clinical experience with use of mAbs in Renal cell carcinoma is most extensive with mAb G250. This antibody recognizes carbonic anhydrase IX, whose expression is suppressed in normal renal epithelium by pVHL. In Renal cell carcinoma, loss of the von Hippel-Lindau gene product leads to expression of the target antigen and, therefore reactivity with mAb G250. Phase I and II clinical trials with unlabelled or 131 I-labelled radioactive G250 antibody have shown some stable disease responses.

ABR-214936 is an mAb fused to a staphylococcal enterotoxin that is currently being tested in a phase II trial in patients with mRCC. The antibody is targeted to an oncofetal antigen, 5T4, which is expressed in non-small cell lung cancer, carcinomas of the kidney, pancreas and breast, but with limited normal tissue expression.

Another antibody in early development is ABX-EGF, which targets epidermal grown factor receptor (epidermal growth factor receptor), a receptor tyrosine kinase. epidermal growth factor receptor is overexpressed in many cancers, including Renal cell carcinoma. In a phase II trial of 31 mRCC patients, treated with 8 weeks of ABX-EGF, 58% of patients showed minor response/stable disease whilst 36% progressed.

A germ cell tumour of the testis is a rare disease although it is the most common tumour in men aged 20-35 years. The incidence of testicular cancer is about 4-5 per 100000 men per year, but there is a geographical and racial variation. Most patients present themselves with a painless lump in the testicle. Sometimes the first symptoms are related to retroperitoneal lymph node metastasis (back pain) or to lung metastasis (haemoptysis or breathlessness). A few patients present with gynaecomastia as a result of an elevated level of the tumour marker human chorionic gonadotrophin.

The diagnosis is established after an inguinal orchiectomy, and germ cell tumours are distinguished into seminomas and non-seminomas, each accounting for about 50% of the total. Staging includes, next to physical examination, computed tomographic scanning of the chest, the abdomen and the pelvis and determination of the serum levels of lactodehydrogenase, alpha-fetoprotein and human chorionic gonadotrophin. The Royal Marsden staging system is widely used [1]. In stage I there is no evidence of metastatic disease and the tumour is confined to the testicle. In stage II there is abdominal metastasis and in stage III supradiaphragmatic metastasis. In stage Intravenous extralymphatic metastasis are present. Furthermore, this staging system also quantifies the volume of metastatic disease. In 1997, the International Germ Cell Cancer Collaborative Group published a prognostic classification system for patients with disseminated disease based on histology (seminoma vs non-seminoma), origin of the primary (gonadal vs extra-gonadal), place of metastases and the level of serum tumour markers. Patients can be divided into three prognostic groups: good, intermediate and poor.

Because the histopathology of testicular cancer is complex, as is the treatment, referral to a specialist centre is frequently advised, especially because survival of patients with testicular cancer appears to be related to the experience of the treating institution and because of the need for long-term medical care of survivors.

Table Staging and classification of testicular cancer

Seminoma stage I

Radiation to the para-aortic and ipsilateral lymph nodes is the standard treatment for stage I seminoma, the so-called ‘dogleg’ field. Doses of 25-30 Gy are given and provide excellent local control. In order to reduce haematologic, gastrointestinal and gonadal toxicity and to maintain efficacy, a recent randomized prospective trial compared the conventional ‘dogleg’ field with a field limited to the para-aortic region alone. It was found that the limited field produced statistically less significant morbidity while the 3-year relapse-free survival was identical (96%), as was the overall survival (99-100%). The more limited field radiation, however, had a nonsignificant increased risk of pelvic recurrences (1.8% vs 0%).

Concerns regarding acute and chronic toxicity of radiation have resulted in interest in surveillance for stage I seminoma. Several large surveillance series showed a recurrence rate in the range of 15-20% with a median follow-up of 4-6 years, but nearly all patients with recurrent disease can be cured by radiation therapy or cisplatin-containing chemotherapy, leading to a survival rate of more than 99%. The risk of recurrence in patients on surveillance seems to be related to several adverse prognostic factors, such as tumour size, tumour invasion of small vessels and age at presentation, but in a multivariate analysis tumour size (more than 4 cm) and invasion of the rete testis were the only important factors. Nevertheless, surveillance is not yet an accepted alternative to radiation.

The effectiveness of one or two courses of adjuvant chemotherapy with the single-agent carboplatin in stage I seminoma has been studied in 160 patients [8]. Although only two patients developed a recurrence, the use of adjuvant carboplatin should be considered as investigational until the results of randomized trials are available.

Standard treatment options

For patients with stage I: Radical inguinal orchiectomy, followed by radiation to para-aortic and ipsilateral lymph nodes, ‘dogleg’ field, with a dose range from 25 to 30 Gy.

Seminoma stage IIA/B with lymph node metastasis < 5 cm

In stage II seminomatous testicular cancer, retroperitoneal or para-aortic lymph nodes are usually present in the region of the kidney. Retroperitoneal involvement should be further characterized by the size of the involved nodes. For treatment planning and estimation of prognosis, stage II seminoma is divided into bulky (IIC) and nonbulky (IIA/B) disease. Radiotherapy alone is standard in seminoma stage IIA/B. The risk of recurrence after radiotherapy is increased if more than five nodes are involved, or if the maximal size of the lymph node metastasis is greater than 5 cm in diameter. Stage IIA/B disease has a cure rate of more than 90% with radiation alone and chemotherapy cures more than 90% of patients who have a relapse after radiation therapy.

Bulky stage II (IIC) disease describes patients with extensive retroperitoneal nodes (< 5 cm) who require primary chemotherapy and who have a less favourable prognosis.

Standard treatment options

For patients with stage IIA/B: Radical inguinal orchiectomy followed by radiation to the retroperitoneal and ipsilateral pelvic lymph nodes. (Radiation to inguinal nodes is not standard unless there has been some damage to the scrotum, putting inguinal lymph nodes at risk.)

Seminoma advanced disease: stage IIC with lymph node metastasis > 5 cm to stage Intravenous

Patients with stage IIC seminomatous testicular cancer have metastatic tumours greater than 5 cm on a computerized tomography scan. Historically, radiotherapy was used for all stages of seminoma; however, the success of the radiation is inversely related to the bulk of the disease. These studies suggest that radiotherapy had a high failure rate if the abdominal mass was more than 5 cm in diameter. Higher risk of relapse can amount to 35% for abdominal masses larger than 10 cm in diameter. Another problem with the use of radiotherapy in the initial management of patients with an abdominal mass of more than 5 cm in diameter is the extent of the kidney within the radiation field. These considerations lead most authors to recommend primary chemotherapy for patients with bulky disease (IIC). Combination chemotherapy with cisplatin is an effective therapy in patients with stage IIC seminomas, leading to a probability of progression-free survival of ±90%. Patients with stage III and Intravenous disease are also treated primarily with combination chemotherapy. Chemotherapy combinations include BEP (bleomycin + etoposide + cisplatin) for four courses and EP (etoposide + cisplatin) for four courses in good-prognosis patients only [13]. Other regimens, such as VIP (etoposide + ifosfamide + cisplatin), appear to produce similar survival outcomes but are less commonly used. A randomized study comparing four courses of BEP with four courses of VIP showed equivalent overall survival and time-to-treatment failure for the two regimens in patients with advanced disseminated germ cell tumours who had not received prior chemotherapy. Haematologic toxic effects, however, were substantially worse with the VIP regimen. A recent study in patients with good-risk germ cell cancer (including seminoma) showed equivalence of three versus four courses of BEP chemotherapy. Four hundred and six patients were compared in both arms; 23% of the randomized patients had seminoma in both arms. The projected 2-year progression-free survival was 90.4% for three cycles and 89.4% for four cycles of BEP. Because of the toxicity of these cisplatin regimens in relatively old patients with seminoma, there is a need for less toxic treatments. Although monotherapy with carboplatin has a relatively high failure rate of about 23% , combinations of carboplatin-based chemotherapy have been propagated as active.

Residual radiologic abnormalities are common at the completion of chemotherapy, but many gradually regress over a period of months. Some clinicians advocate empiric radiation of persistent residual abnormalities or attempt to resect residual masses if the diameter is 3 cm or more. Either approach is controversial. In a combined retrospective series of 174 seminoma patients with postchemotherapy residual disease treated in ten centres, empiric radiation was not associated with any significant improvement in progression-free survival after completion of the chemotherapy. In some other series, surgical resection of specific masses has yielded a significant number with residual seminoma that requires additional therapy. Nevertheless, other reports indicate that size of the residual mass does not correlate well with active residual disease; most residual masses do not grow and frequent marker and computerized tomography scan evaluation is a viable option even when the residual mass is 3 cm or more in diameter.

Table Common chemotherapy regimens for patients with disseminated non-seminomatous testicular cancer in different prognostic groups

Standard treatment options

For seminoma patients with stage IIC-Intravenous: Radical inguinal orchiectomy followed by combination chemotherapy (with a cisplatin-based regimen). Chemotherapy combinations include BEP for three or four courses in good-or intermediate-prognosis patients or EP for four courses in good-prognosis patients. There is controversy whether any residual masses present at the completion of chemotherapy should be empirically irradiated, or whether masses greater than 3 cm should be resected.

Non-seminoma stage I

Non-seminoma stage II-Intravenous

Surgery after chemotherapy

If patients do not achieve a complete radiological response after chemotherapy, surgical removal of residual masses should be performed. The timing of such surgery requires clinical judgement, but occurs most often after three or four cycles of combination chemotherapy and after normalization of serum marker levels. The probability of finding residual teratoma or carcinoma after chemotherapy may depend on the histology of the primary tumour. However, others have reported that irrespective of initial histology there is a significant risk of teratoma or carcinoma in residual masses. Moreover, neither size of the initial metastasis nor degree of shrinkage while on therapy appears to accurately identify patients with residual teratoma or carcinoma. This has led some authors to recommend surgery with resection of all residual masses apparent on scans in patients who have normal or normalized markers after chemotherapy. The presence of persistent viable tumour cells in the resected specimen seems to be an indication for additional chemotherapy, although this strategy may not improve overall survival. Surgical removal of residual masses is also necessary to prevent regrowth of teratomas and growth of non-germ cell elements present in some of these masses. Some patients may have discordant pathological findings (fibrosis/necrosis, teratoma or carcinoma) in residual masses in the abdomen versus the chest; some medical centres therefore perform simultaneous retroperitoneal and thoracic operations. However, most centres do not perform simultaneous retroperitoneal and thoracic resections. Although the agreement among the histologies of residual masses above, versus below, the diaphragm is only moderate, there is some evidence that if retroperitoneal resection is performed first, results can be used to guide decisions about whether to perform a thoracotomy. In a multi-institutional case series of surgery to remove postchemotherapy residual masses in 159 patients, only necrosis was found at thoracotomy in about 90% of patients who had also only necrosis in their retroperitoneal masses. This figure was about 95% if the original testicular primary tumour did not contain teratomatous elements. Conversely, the histology of residual masses at thoracotomy was not nearly as good a predictor of the histology of retroperitoneal masses.

The standard treatment options

If patients with disseminated non-seminomatous testicular cancer do not achieve a complete radiological response on chemotherapy, surgical removal of residual masses should be performed. The timing of such surgery should be done after three or four cycles of combination chemotherapy and after normalization of serum tumour marker levels.

Treatment of recurrent disease

Follow-up

The aim of follow-up care in patients treated for testicular cancer is to detect a relapse at a stage where salvage treatment has the best chance of being effective, to monitor and treat treatment-related toxicity, to detect cancers in the contralateral testicle and to offer support and counselling about issues such as fertility and employment. Recently minimal recommendations have been published by the European Society for Medical Oncology (ESMO), but the optimal timing of clinical, biochemical and radiological follow-up is still under investigation.

Early detection of recurrence of testicular cancer after successful treatment with cisplatin combination chemotherapy is beneficial if there is a chance of achieving another durable remission with salvage treatment. The possibility of early recognition of recurrence and subsequent treatment prolonging survival will increase with more effective salvage therapies. However, the optimal regimen of physical examination, tumour marker estimations and chest X-rays for use in the follow-up of patients after initial treatment has not been determined. The widely used follow-up strategies come from large multi-institutional chemotherapy trials that defined the optimal chemotherapy combination for disseminated non-seminomatous testicular cancer during the last 2 decades. However, the primary focus of this particular follow-up was to define the efficacy of the first-line treatment regimen and not to evaluate the value of follow-up examinations. Furthermore, there are few data in the medical literature concerning the effectiveness of these follow-up regimens. In daily practice, the aim of follow-up after successful chemotherapy is to detect a tumour relapse in time without unnecessary procedures. Recent data suggest that routine chest X-rays have limited or no additional value in the detection of a relapse during follow-up in patients who have a complete biochemical response and no residual masses. The value of chest X-rays in follow-up of clinical stage I patients with non-seminomatous testicular cancer also does not show additional value in detection of disease recurrence. So tumour marker measurements, medical history and physical examination seem to be of key value.

From an oncologic point of view, recent data and recommendations suggest that it is reasonable to discharge patients with stage I non-seminomatous testicular tumours and all stages of seminoma from follow-up after 5 years. Metastatic non-seminomatous testicular cancers seem to have a continuing annual relapse rate of 1-2% even after 10 years, suggesting that life-long follow-up might be needed. However, an important part of the long-term follow-up is surveillance of long-term toxicity of administered treatment. Since most of the cured patients are men in their twenties or early thirties, long-term treatment-related toxicity is of growing importance.

Treatment toxicity and long-term side-effects

Chemotherapy with cisplatin causes significant side-effects both in the short and the long term. Acute side-effects include nausea and vomiting, alopecia, bone marrow suppression with risk for neutropenic fever, fatigue, renal toxicity and acute cardiovascular toxicity. A particular complication of the BEP combination chemotherapy is lung toxicity associated with bleomycin; in most studies, 0.5-1% developed fatal bleomycin-induced pneumonitis. Bleomycin combined with cisplatin is also associated with the risk of developing Raynaud’s phenomenon. Cisplatin may also cause damage to both peripheral and auditory sensory nerves. This resolves in most patients over 6-12 months but long-term studies suggest persistent damage in a proportion of patients.

Infertility is one of the most distressing adverse effects of cancer therapy. Patients with germ cell tumour may have azoospermia related to the disease itself or to the sterilizing effects of chemotherapy. Fertility is an important predictor of long-term health-related quality of life in testicular cancer survivors. Testicular patients undergoing chemotherapy are usually counselled about the risks of infertility and offered the opportunity for sperm banking before commencing therapy. For the azoospermic germ cell cancer survivor, donor insemination and adoption have historically been the main reproductive options. A recent report by Damani et al. explores the possibilities of testis sperm extraction in testicular cancer survivors. This assisted reproductive technology, initially developed for conditions such as congenital absence of the vas deferens, resulted in successful retrieval of sperm in approximately two-thirds of the patients. Rather, it should be considered a reproductive option for the azoospermic cancer survivor without banked sperm. This technique represents the development of an effective intervention for an established treatment-related adverse effect, with the potential to improve the long-term well-being of the cancer survivor. For many other physiologic adverse effects of cancer treatment, the situation is not so clear.

The prevalence and time course for development of certain other late effects have not been well defined. The main concerns relate to the increased risk of second malignancies that can occur after treatment with chemotherapy or radiotherapy or of cardiovascular events in long-term survivors.

Cisplatin-containing chemotherapy for germ cell cancer has been associated with Raynaud’s phenomenon, and serious vascular complications, including myocardial infarction, stroke and thromboembolic disease, have been reported. Some, but not all, studies have suggested that after cisplatin-containing chemotherapy, patients may be at increased risk for the premature development of hypertension and lipid abnormalities, well-known major cardiovascular risk factors. Testicular cancer survivors develop a metabolic syndrome or syndrome X-like state after chemotherapy, which makes them more prone to cardiovascular events. However, does cisplatin combination chemotherapy result in an increased risk for early cardiovascular events? In one study of testicular cancer patients treated with surgery or surgery plus chemotherapy, no increase in cardiovascular events was noted in the chemotherapy group at a median follow-up of 5 years. A more recent study reported an increased risk of cardiovascular events for testicular cancer survivors younger than 50 years of age who had received chemotherapy and were in remission for 10 or more years. Further studies are needed to better define the actual risk, if any, of early cardiovascular events in these patients. What do these data tell us regarding the education and counselling of testicular cancer survivors concerning cardiovascular risk? Are there rational early intervention possibilities?

Who will be following the cancer survivor when these adverse effects become manifest? While the oncologist might be the most knowledgeable about the potential late adverse effects of cancer treatment, many survivors may not regularly see an oncologist once the risk of tumour recurrence is unlikely. Probably many of these patients are followed by primary care physicians, who may not be fully aware of the details of the patient’s oncologic history and may not be familiar with the long-term sequelae of cancer and its treatment. Other patients may exit the health care system altogether. For uncommon cancers such as germ cell tumours, few centres have enough patients to define a large enough long-term cohort for studies. Our preference is to undertake the long-term follow-up at a cancer centre to allow the build-up of well-documented databases on the well-being and actual health status of testicular cancer survivors facilitating cancer survivor research.

For future well-defined health care problems of testicular cancer survivors, either primary care physicians with knowledge of testicular cancer and treatment sequelae or a cancer specialist with knowledge of general internal medicine should take care of treatment sequelae or risk factors for disease.

The cure rate for patients with non-seminomatous tumours in clinical stage I exceeds 95%. About 20% of patients with stage I disease without lymphatic or vascular invasion or without invasion into the tunica albuginea, spermatic cord or scrotum are discovered to have regional lymph node metastases at surgery. Nerve-sparing retroperitoneal lymph node dissection and surveillance are both standard treatment options.

Nerve-sparing retroperitoneal lymph node dissection

A primary retroperitoneal lymph node dissection after orchiectomy allows careful pathological staging, while at the same time offering a therapeutic benefit if the retroperitoneal lymph nodes are positive. A nerve-sparing retroperitoneal lymphadenectomy that preserves ejaculation in virtually every clinical stage I patient appears to be as effective as the standard retroperitoneal lymphadenectomy. Despite the improved accuracy of clinical staging methods about 20% of patients with clinical stage I have pathological stage II disease at retroperitoneal lymphadenectomy. Many of these patients are cured surgically without subsequent chemotherapy. However, approximately 80% of clinical stage I patients who undergo primary retroperitoneal lymphadenectomy are found to have pathological stage I disease and do not benefit from the surgical procedure.

In case of a pathological stage I after retroperitoneal lymphadenectomy, patients can go into follow-up without additional treatment. In a large study, 15% of patients with a negative lymph node dissection experienced recurrence, usually pulmonary and usually within 18 months. The overall survival rate of patients with pathological stage I is about 99%.

In case of tumour in the resected lymph nodes in patients with a clinical stage I, a pathological stage II is documented. The relapse rate of a pathological stage II not treated with adjuvant chemotherapy is related to the volume of retroperitoneal disease up to 30%. These patients are therefore further treated with two courses of adjuvant cisplatin-combination chemotherapy.

Prognostic factors for patients with stage I disease that may predict the likelihood of occult metastases are the presence of lymphatic or venous invasion in the primary tumour, the presence of embryonal cell carcinoma and the absence of yolk sac elements in the primary tumour. A more sophisticated way to stain proliferating tumour cells in testicular tumours with a monoclonal antibody MIB-1 against Ki-67 in combination with the volume of embryonal cell carcinoma and the transaxial diameter of retro-peritoneal lymph nodes in the predicted landing zone allows a low-risk clinical stage I classification. However, none of these strategies reliably predicts the presence of occult metastases in clinical stage I disease.

If performed, surgery should be followed by monthly determination of serum markers and chest X-rays for the first year and 1- to 2-month determinations for the second year, every 6 months in years 3 to 5, and follow-up is then indicated yearly thereafter.

Surveillance

Approximately 75-80% of patients with clinical stage I disease who undergo retroperitoneal lymphadenectomy have negative lymph nodes. The rational for surveillance is to avoid surgical ‘overtreatment’ of patients with clinical stage I disease. In this strategy, radical inguinal orchiectomy without retroperitoneal node dissection is followed by regular follow-up (e.g. every 1-2 months) consisting of history, physical examination, determination of serum tumour markers, and during the first year, abdominal computerized tomography scans. Intervals for abdominal computerized tomography scans have varied from every 2 months to scans only at 3 and 12 months post-orchiectomy, with apparently similar outcomes. Disease recurrence is rarely detected by chest X-ray alone, so chest X-ray may play little or no role in routine surveillance. In a Medical Research Council (Medical Research Council) surveillance study of non-seminomatous germ cell tumours (NSGCTs), 396 patients with a median follow-up of 4 years had a 25% recurrence rate and a mortality rate of less than 2%. Long-term follow-up is important, since relapses have been reported more than 5 years after the orchiectomy in patients who did not undergo a retroperitoneal dissection.

Surveillance should be considered only if:

1   computerized tomography scan and serum markers are normal;

2  the patient and the physician accept the need for repeating computerized tomography scans as necessary to continue the periodic monitoring of the retroperitoneal lymph nodes up to 24 months;

3  the patient diligently follows a programme of regular check-ups, which includes history, physical examination, radiology and determination of serum markers;

4 the physician accepts responsibility for seeing that a follow-up schedule is maintained as noted for 2 years and then periodically beyond 2 years.

Data suggest that relapse rates are higher in patients with histological evidence of lymphatic or venous invasion and lower when the primary tumour contains mature teratomas. Some investigators have reported higher relapse rates in patients with embryonal cell histology and recommend retroperitoneal lymphadenectomy for such patients. Other investigators have not found a higher relapse rate for this subgroup. Additionally, some investigators recommend retroperitoneal lymphadenectomy in patients with a normal pre-orchiectomy alpha-fetoprotein, because they feel this marker cannot be used as an indicator of relapse during follow-up. However, since marker-negative patients may be marker-positive at relapse and marker-positive patients may be marker-negative at relapse, other investigators do not view a normal pre-orchiecomy alpha-fetoprotein as a contraindication to a surveillance policy.

Adjuvant therapy consisting of two courses of BEP has been administered to patients with clinical stage I disease who were considered at high risk of relapse (about 50% predicted relapse rate based on presence of vascular invasion and histologic type). In 114 such patients, the relapse-free survival at 2 years was 98%. Another study of high-risk clinical stage I patients treated with two adjuvant courses of BEP reported a relapse rate of less than 5%, while in historical series of high-risk patients followed without adjuvant chemotherapy the relapse rate was 50%. However, in the historical series, cure rates have also been 95% and greater after chemotherapy is given for relapse. Given the present criteria, high-risk patients will relapse, at most, around 50% of the time, and thus approximately 50% of patients who would not have relapsed would receive chemotherapy ‘unnecessarily’.

It is unclear which approach is superior in outcome. The adjuvant chemotherapy series are too small to draw definite conclusions about ultimate efficacy and about the risk of chemotherapy-induced long-term toxicity, secondary malignancies, impact on fertility or risk of late relapse.

Standard treatment options

For patients with non-seminoma stage I: Radical inguinal orchiectomy followed by either retroperiotoneal lymph node dissection (in case of pathological stage I: follow-up, in case of pathological stage II: two adjuvant courses of BEP) or surveillance.

Disseminated non-seminoma is highly curable. In most patients, an orchiectomy  is   performed   before   starting  chemotherapy.   However,   if  the diagnosis has been made by biopsy of a metastatic site and chemotherapy initiated, subsequent orchiectomy is generally performed due to the fact that chemotherapy may not eradicate the primary cancer. This is illustrated by case reports in which viable tumour was found on postche-motherapy orchiectomy despite complete response of metastatic lesions.

After the introduction of cisplatin, vinblastine and bleomycin (PVB) combination chemotherapy, consisting of a remission-induction part and a maintenance part, the strategy for treatment outcome improvement had focused on less toxicity with similar efficacy. It was shown that the dosage of vinblastine could be reduced (0.3 mg/kg vs 0.4 mg/kg) and that maintenance chemotherapy does not prevent relapses but adds significantly to the toxicity. Later on vinblastine has been replaced by etoposide; based on the efficacy of etoposide in salvage therapy, and based on the results of a randomized study with BEP, this combination became the new standard.

Other centres have developed their own combinations such as the Memorial Sloan-Kettering Cancer Center, New York, using the so-called VAB schemes, or Charing Cross Hospital, London, using the POMB-ACE combination, but most often the BEP regimen is used.

The success of treatment of disseminated testicular cancer has led to refinements in treatment, with a greater importance on prognostic factors. Several groups have devised schema for stratifying patients into prognostic groups. Although each prognostic system has advantages and disadvantages, several characteristics are common. On the other hand, substantial differences occur between the various classifications in their use of prognostic variables and in their ability to separate patients into good-and poor-prognostic groups. This means that the description of a good prognosis differs, depending on the prognostic system used. To achieve more uniformity in classifying the prognosis of patients with metastatic disease, the International Germ Cell Cancer Collaborative Group (IGCCCG) recently developed a prognostic classification for germ cell tumours based on a large analysis of more than 5000 patients who were treated in prospective studies in North America, Europe, New Zealand and Australia. Primary tumour site, degree of elevation of serum tumour markers (alpha-fetoprotein, human chorionic gonadotrophin and lactodehydrogenase) and the presence or absence of non-pulmonary visceral metastases were identified as the most important independent prognostic variables. Integration of these prognostic factors produced three groupings of testicular cancer patients with good, intermediate and poor prognosis, with 5-year overall survival rates of 92%, 80% and 48%, respectively. Since then this system is used by all collaborative groups.

Good-risk patients with metastatic non-seminomatous testicular cancer

The strategy for treatment outcome improvement in ‘good-risk patients’ has focused on less toxicity with the same efficacy compared with the standard treatment of BEP. Attempts to improve the toxicity profile have focused on the role of bleomycin (especially because of bleomycin-induced pulmonary fibrosis). The European Organization of Research and Treatment of Cancer (EORTC) compared four courses of BEP with four courses of EP in patients with a ‘good prognosis’. The total dose of etoposide per course, however, was 360 mg/m compared with 500 mg/m , as is the US standard. In total, 419 patients were randomized. In the EP arm 87% of the patients achieved a complete response, if necessary followed by surgical resection of residual disease; in the BEP arm this was the case in 95% of the patients. This difference is significant. Due to the low number of relapses (4%) no difference in progression-free survival was found. An Eastern Cooperative Oncology Group (ECOG) study compared three cycles of BEP with three cycles of EP. In the BEP arm 94% of the patients achieved a complete response versus 88% in the EP arm. The progression-free survival of the BEP group was significantly higher than that of the EP group (86% vs 68% after 5 years). So bleomycin is an essential part of the standard BEP regimen.

To address the question whether cisplatin can be replaced by the less-toxic analogue carboplatin, an Medical Research Council-EORTC study was performed. Almost 600 patients with a ‘good prognosis’ were randomized between four courses of BEP and four courses of carboplatin-etoposide-bleomycin. Significantly less patients in the carboplatin-etoposide-bleomycin arm achieved a complete response (94% vs 88%). After 1 year, the progression-free survival was significantly lower in the carboplatin-etoposide-bleomycin arm compared with the BEP arm. These data demonstrate that cisplatin cannot be substituted by carboplatin. To assess the optimal number of courses of BEP a study was performed in which three courses of BEP have been shown to be equivalent to four courses in patients with minimal or moderate extent of disseminated germ cell tumours [40]. To estimate equivalence of three and four courses of BEP, an EORTC-Medical Research Council study was performed randomizing three courses of BEP with three courses of BEP plus one course of EP. The median 2-year progression-free survival was 90.4% versus 89.4%. Therefore it can be concluded that for good-risk patients based on the IGCCCG criteria these regimens are equivalent.

One question remaining is whether in good-risk patients three courses of BEP are equivalent to four courses of EP. Probably this question will never be answered and the choice is based on personal preferences.

The standard treatment options for ‘good-prognosis’ patients

Radical inguinal orchiectomy followed by chemotherapy. Chemotherapy regimens include BEP for three courses or EP for four courses. If these patients do not achieve a complete radiological response on chemotherapy, surgical removal of all residual masses should be performed.

‘Intermediate- and poor-risk’ patients with metastatic non-seminomatous testicular cancer

Compared with good-risk non-seminoma patients, patients with intermediate or poor risk have a worse prognosis. This is a strong argument for treating patients as soon as possible after being diagnosed as having metastatic disease. IGCCCG data show that intermediate prognosis accounts for ± 28% of the non-seminomatous testicular cancer patients and the 5-year survival of this group is 80%. The non-seminomatous testicular cancer patients with a poor prognosis (± 16% of the patients) have a 5-year survival of 48%. The patients who are not cured with standard chemotherapy usually have widespread visceral metastases, high tumour marker levels or mediastinal primary tumours at presentation. Some retrospective data suggest that the experience of the treating institution may impact the outcome of non-seminoma. Data from 380 patients treated from 1990 to 1994 on the same study protocol at 49 institutions were analysed. Overall, 2-year survival for the patients treated at institutions that entered less than five patients onto the protocol was 62% (95% CI = 48-75%) versus 77% (95% CI = 72-81%) in the institutions that entered at least five patients. As in any nonrandomized study design, patient selection factors and factors leading patients to choose treatment at one centre over another can make interpretation of results difficult.

Although the standard treatment for patients with an intermediate or poor prognosis has been four courses of BEP chemotherapy, the strategy for treatment outcome improvement has focused on non-cross-resistant chemotherapy combinations, and dose escalation or intensification. A study in which 244 patients were randomized between four courses of BEP and four alternating courses of PVB and BEP showed no significant differences in complete remission numbers: 72% versus 76%, respectively. The progression-free survival was 80% in both groups. Because of its activity in second-line treatments, ifosfamide was incorporated into first-line treatments. A randomized study comparing four courses of BEP with four courses of VIP showed equivalent overall survival (83% vs 85%, respectively) and time-to-treatment failure for the two regimens in patients with advanced disseminated germ cell tumours who had not received prior chemotherapy. Haematologic toxic effects were substantially worse with the VIP regimen.

In patients with poor-risk germ cell tumours, the standard dose cisplatin regimen has been shown to be equivalent to high-dose cisplatin (40 mg/m daily x 5 per course) in terms of complete response, cure rates and survival; moreover, patients in the high-dose cisplatin regimen experienced significantly more toxic effects. A randomized comparison of an intensive induction-sequential chemotherapy schedule BOP/VIP-B (bleomycin, vincristine, cisplatin, etoposide, ifosfamide) with BEP in patients with poor-prognosis non-seminomatous testicular cancer showed more toxicity without evidence of an improved response rate or survival for the BOP/VIP-B regimen.

Based on its activity in patients with a relapsed or refractory germ cell tumour, paclitaxel is an interesting drug to add to the first-line regimen in patients with intermediate- or poor-prognosis disease [44]. The EORTC is currently performing a study in which intermediate-risk patients are treated with standard BEP versus BEP plus paclitaxel (T-BEP).

More intensive approaches are explored in several studies, including high-dose chemotherapy with peripheral stem cell transplantation. This approach has been fuelled by results from small studies in patients who failed second- or third-line cisplatin-containing regimens. Long disease-free periods were established in 10-20% of patients who were treated with high-dose chemotherapy and peripheral stem cell rescue. This approach has also been used in a French study in which patients with poor prognostic factors were randomized between conventional dose chemotherapy and conventional dose combined with high-dose chemotherapy as first-line treatment. The 2-year survival rate was not different in both treatment arms; however, the trial was inconclusive because the dose of cisplatin was lower in the experimental arm compared with the standard arm. A dose-intense regimen using the VIP combination has been exploited by a German study group . The dose intensity of etoposide was three times higher and that of ifosfamide two times higher compared with standard VIP. The EORTC is currently performing a randomized study in poor-prognosis testicular cancer patients, comparing standard BEP with high-dose VIP and peripheral stem cell rescue. Patients who present with brain metastases as a poor prognostic factor should be treated with chemotherapy and simultaneous whole-brain irradiation (5000 cGy/25 fractions).

The standard treatment options for ‘intermediate- and poor-prognosis’ patients

Radical inguinal orchiectomy followed by chemotherapy with postche-motherapy surgery for removal of residual masses (if present). Chemotherapy regimens include BEP for four courses and VIP for four courses.

Deciding on further treatment in case of recurrent testicular cancer depends on many factors, including the histology, prior treatment, site of recurrence, as well as individual patient considerations. Salvage regimens consisting of ifosfamide, cisplatin and either etoposide or vinblastine can induce long-term complete responses in about one quarter of patients with disease that has persisted or recurred following first-line cisplatin-based regimens. Patients who have had an initial complete response to first-line chemotherapy and those without extensive disease have the most favourable outcome. The VIP regimen is now the standard initial salvage regimen. However, few, if any, patients with recurrent NSGCTs of an extragonadal origin achieve long-term disease-free survival using VIP if their disease recurs. High-dose chemotherapy with autologous stem cell transplantation has also been used with some success in the setting of refractory disease. Durable complete remissions may be achievable in 10-20% of patients. The durable complete remission rate may even exceed 50% in selected patients if high-dose chemotherapy is used as salvage chemotherapy at the first relapse of primary testicular cancer. In general, patients with progressive tumours during frontline or after salvage treatment and those with refractory mediastinal germ cell tumours do not appear to benefit as much from high-dose chemotherapy with autologous stem cell transplantation as those who relapse after an initial response. In some highly selected patients with chemorefractory disease confined to a single site, surgical resection may yield long-term disease-free survival. The choice of salvage surgery versus high-dose chemotherapy with autologous stem cell transplantation for refractory disease is based on resectability, the number of sites of metastatic disease and the degree to which the tumour is refractory to cisplatin.

A special case of late relapse may be patients who relapse more than 2 years after achieving complete remission; this population represents less than 5% of patients who are in complete remission after 2 years. Results with chemotherapy are poor and surgical treatment appears to be superior, if technically feasible. This may be because mature teratoma may be amenable to surgery at relapse and also has a better prognosis than carcinoma. Mature teratoma is a relatively resistant histologic subtype, so chemotherapy may not be appropriate.

Clinical trials are appropriate and should be considered whenever possible, including phase I and II studies for those patients not achieving a complete remission with induction therapy or not achieving a complete remission following salvage treatment for their first relapse or for patients who have a second relapse.

Patients who relapse with brain metastases after a complete initial response to chemotherapy require further chemotherapy, with simultaneous whole-brain irradiation and consideration of surgical excision of solitary lesions.

The standard treatment options

Patients with recurrent non-seminomatous testicular cancer can be treated with a salvage VIP regimen. However, since only few of these relapsed patients achieve a long-term disease-free survival, high-dose chemotherapy with autologous stem cell transplantation can also be used. Participation of these patients in clinical trials should be considered whenever possible.

Despite radical treatments with curative intent, the chance of long-term survival for patients with muscle-invasive bladder cancer remains disappointing. In a large series of more than 1000 patients with apparently organ-confined disease, 5-year overall survival was only 47% for all muscle-invasive tumours, ranging from 72% for T2 tumours to only 33% for T4 disease. Death from bladder cancer following radical primary treatment is largely due to occult systemic disease, present at the time of surgery, which is below the limits of resolution of currently available cross-sectional imaging.

This post reviews systemic therapy for muscle-invasive cancer of the urothelium and describes the common drugs that are used, the various clinical scenarios when they may be indicated and the selection of patients for treatment.

Drugs in use

Orthodox cytotoxic agents

Transitional cell carcinoma (transitional cell carcinoma) is sensitive to a variety of drugs as shown by complete or partial response in patients with measurable metastatic or locally advanced disease. Of the older cytotoxic drugs, cisplatin and meth-otrexate are the most active agents, with significant activity also seen with carboplatin, doxorubicin, vinblastine, cyclophosphamide, ifosfamide and 5-fluorouracil. Newer agents with significant activity against bladder cancers include gemcitabine and paclitaxel.

CISPLATIN

Cisplatin is a derivative of the heavy metal platinum. As a single agent, cisplatin produces objective response rates between 12% and 35% in bladder tumours. Its activity as a DNA-modifying cytotoxic drug is as a bifunctional alkylating agent, which results in intra-strand cross-linking of DNA, thereby interfering with synthesis of new DNA during replication. It has a broad range of anti-tumour activity, being particularly effective in tumours that are defective in DNA repair.

Table   Efficacy   of  cytotoxic agents in transitional cell carcinoma

Cisplatin is administered intravenously. Typical doses range from 40mg/m² to 100mg/m², either in a single dose, or divided over several days. Intravenous administration is almost invariably as a slow infusion over 2-4 h. When given as an Intravenous bolus, nephrotoxicity is dose-limiting. Pre-and post-hydration with Intravenous crystalloids is routinely employed to minimize nephrotoxicity. This usually means that the patient is admitted overnight, although recently, in selected patients, outpatient schedules have been explored. Cisplatin is not usually administered to patients with glomerular filtration rate (GFR) less than 40 ml/min.

Cisplatin is highly emetogenic; however, the introduction of effective anti-emetic regimens, based around combinations of corticosteroids such as dex-amethasone and 5-hydroxytryptamine-3 antagonists such as ondansetron or granisetron, have dramatically reduced acute nausea and vomiting, although ‘delayed’ effects 4-6 days after administration can still be troublesome.

Haematological toxicity (leukopenia, more than thrombocytopenia or anaemia) is not usually dose-limiting, and recovery from myelosuppression is usually sufficiently rapid to permit 3-week dosing. Alopecia, a side-effect often associated with chemotherapy, is rarely encountered with cisplatin.

Significant neurotoxicity is a common feature of treatment with cisplatin, being dose-limiting for a significant proportion of patients. Peripheral neuropathy producing numbness and tingling in a glove-and-stocking distribution is seen in up to 40% of long-term survivors, although it rarely severely compromises performance status. Autonomic neuropathy may result in postural hypotension. Central nervous system toxicity is also seen, with tinnitus and high-frequency hearing loss reported in up to 50% of patients, particularly in older age groups. Neurotoxicities may be permanent, and even when they are not, only resolve very slowly, over a period of many months.

Most patients treated with cisplatin have at least temporary reduction in fertility, but sperm counts back to pretreatment levels occur in 40% of younger patients treated with cisplatin, within 2 years. There is no convincing evidence of an increased incidence of second malignancies in long-term survivors following cisplatin-based chemotherapy. A frequently overlooked long-term side-effect of cisplatin is an increase in risk of thromboembolic events. These can often be attributed to comorbidity but careful series have shown a significant increase in myocardial infarction, pulmonary embolism and stroke after cisplatin chemotherapy.

METHOTREXATE

Until relatively recently, methotrexate was the only cytotoxic with single-agent activity comparable with cisplatin, with objective response rates in small studies ranging from 26% to 56%.

It is an antimetabolite cytotoxic that acts as specifically during the S phase (DNA synthesis) of the cell cycle as a folate antagonist via reversible inhibition of the enzyme dihydrofolate reductase (DHFR), the enzyme that is required to regenerate tetrahydrofolate for thymidine synthesis. By producing a metabolic block, methotrexate produces an accumulation of dihydrofolate, depletion of intracellular pools of reduced folate and reduced synthesis of both the pyrimidine nucleoside thymidine and the purine nucleotides. Methotrexate also inhibits another key nucleic acid-metabolizing enzyme, thymidine synthase, via the formation of polyglutamate derivatives.

Methotrexate can be administered either orally, as an Intravenous solution, via intrathecal injection or, rarely, as an intramuscular injection. When used in chemotherapy for advanced bladder cancer it is usually given as an Intravenous infusion, usually at 30 mg/m² bolus weekly for 2 out of a 3-week or 3 out of a 4-week cycle.

Clearance of methotrexate is predominantly renal, and elimination is inhibited by renal dysfunction but increased toxicity may also be seen in patients with liver dysfunction. Particularly at higher doses, adequate renal function is required for administration of the drug. Renal dysfunction results in slower clearance and prolonged exposure of normal tissues, resulting in increased toxicity. Free distribution into ‘third-space’ fluid compartments such as pleural fluid or ascites also prolongs drug exposure and increases toxicity. Methotrexate is readily absorbed by the gut. In bladder cancer patients who have had urinary diversion procedures involving the formation of reservoirs from segments of the gastrointestinal tract, such as an ileal conduit, reabsorption of the drug from the urine may also result in increased toxicity.

At high doses, good renal function and adequate hydration are required, with alkalinization of the urine. Unless this occurs, crystallization of the drug in the urine can cause nephrotoxicity. In contrast to many cytotoxic agents, methotrexate is not particularly myelotoxic and as a result can be successfully combined with other agents without a requirement for a reduction in dose.

Inhibition of DHFR in normal tissues as well as in tumours also accounts for much of the toxicity of methotrexate. gastrointestinal tract toxicity is common, particularly stomatitis, oral mucositis, nausea and diarrhoea. Acute, reversible hepatitis and pneumonitis are also seen, although more rarely.

The metabolic block produced by methotrexate can be reversed by providing exogenous reduced folates. Folinic acid (leucovorin) ‘rescue’ can be administered to minimize stomatitis and gastrointestinal tract toxicity. For lower doses of methotrexate, a ‘flat’ dose of folinic acid at 15-30 mg QDS for 4-6 doses is usually sufficient, but at higher doses, dose and duration of folinic acid rescue are based on plasma methotrexate levels. Despite some concern, there is no evidence that folinic acid rescue abrogates the therapeutic effect of methotrexate.

VINBLASTINE

Single-agent response rates for vinblastine range from 4% to 28%. Vinblastine is a polycyclic organic base that belongs to the vinca alkaloid class of cytotoxic drugs, derived from the periwinkle plant, Vinca rosea. Like its formylated analogue vincristine, its mechanism of action is via inhibition of the polymerization of tubulin, thereby inhibiting the formation of microtubules, including the mitotic spindle required for cell division, although effects on microtubules may also produce other toxic effects in non-proliferating cells. However, its spectra of activity and of toxicity are significantly different from those of vincristine. It is usually administered intravenously as a bolus injection into a fast-running drip at a dose of 3-6 mg/m², with doses being given weekly in some regimens.

Vinblastine is extensively bound to serum proteins, and to platelets and red blood cells. Metabolism is predominantly hepatic, and it is excreted in the stools via the biliary tract, and, to a lesser extent, is also eliminated in the urine. Care must therefore be taken in administering the drug to patients with an obstructive pattern of liver function tests, and consideration given to dose reduction. Dose modification to take account of renal dysfunction is not usually required.

Extreme caution must be exercised to avoid extravasation, as vinblastine is a vesicant, capable of causing severe tissue necrosis if not promptly and appropriately treated. Principal toxicities of vinblastine are myelosuppression

and gastrointestinal disturbances. Mucositis and stomatitis are more frequently seen with vinblastine than with vincristine, whilst neuropathic effects such as sensory peripheral neuropathy and autonomic neuropathy, resulting in urinary retention or ileus, are seen less often than with vincristine. Like vincristine, vinblastine may cause hyponatraemia, secondary to a syndrome of inappropriate ADH secretion. Hypertension and Raynaud’s phenomenon may also be induced, but alopecia is rare and usually only mild.

DOXORUBICIN

Doxorubicin belongs to the anthracycline class of microbially derived anti-tumour antibiotic compounds. It is composed of a tetracyclic chromophore, adriamycinone (which gives it its characteristic ruby colour), covalently linked to an aminosugar, daunosamine. It has multiple effects on cancer cells, via both the formation of free radicals and intercalation between the stacked base pairs of DNA. In addition to its effects on DNA, and subsequent DNA replication and transcription, it causes damage to mitochondria (where interaction of doxorubicin with the electron transport chain is important in the formation of free radicals), to cytoplasmic structures and to the cell membrane.

Administration of doxorubicin is usually via the Intravenous route, most commonly as a bolus injection into a fast-running Intravenous drip. Intravenous dosing is usually on a 3-week basis, most commonly at doses in the range 45-90 mg/m². It has also been given intravesically as a treatment for superficial bladder cancer.

The major route of elimination is via the biliary tract, with no more than 10% being excreted in the urine. Consequently, dose adjustments are required in the case of abnormal hepatic biochemistry, but not in the case of impaired renal function.

The dose-limiting acute toxicities for the administration of doxorubicin are myelosuppression and mucositis/stomatitis. Nausea and vomiting can usually be well controlled with modern anti-emetic regimens. Alopecia is usually seen, although scalp cooling can been used to limit this. However, there is also a chronic dose-limiting toxicity of cardiomyopathy. This is dose-related and clinically significant in about 10% of cases above a cumulative dose of 550mg/m². Transient acute dysrhythmias may also occur during the administration of the drug, which are not usually clinically significant. Like vinblastine, doxorubicin is vesicant. Another notable skin toxicity seen with doxorubicin is the so-called ‘radiation recall phenomenon’, observed 4-7 days after administration of doxorubicin in patients who have had previous external-beam radiotherapy. Patients experience an erythematous dermatitis, similar in appearance to radiation dermatitis, in the anatomical distribution of previous radiation exposure. The severity of the dermatitis is variable, from mild warmth and erythema to severe burning pain, occasionally accompanied by vesicle formation, ulceration or desquamation.

CARBOPLATIN

Carboplatin is an analogue of cisplatin that has shown higher efficacy when compared with cisplatin in some tumour types. Although carboplatin acts via a cell cycle-independent mechanism similar to cisplatin, it has a very different toxicity profile. Carboplatin is generally considered easier for frail or elderly patients to tolerate and can also be given to patients with impaired renal function. Like cisplatin it is predominantly cleared via the renal route, although with slower clearance than cisplatin and rarer nephro- or neurotoxi-city. Myelosuppression, particularly thrombocytopenia, is the dose-limiting toxicity. The nadir is late compared with most other cytotoxics, with thrombocytopenia being maximal at around 21 days after administration. Carboplatin-based chemotherapy may therefore be given on a 4-week schedule. Most non-haematological toxicities of carboplatin are mild in comparison with cisplatin, for example, it is less emetogenic. Nevertheless, significant hypersensitivity reactions are seen in a proportion of patients, requiring cover with corticosteroids and antihistamines for subsequent chemotherapy cycles, and occasionally discontinuation of the use of carboplatin entirely.

Like cisplatin it is almost invariably administered via the Intravenous route. Its reduced nephrotoxicity means that it can be administered as a short infusion, without pre- or post-hydration, and can thereby almost always be administered on a day-case basis, rather than requiring admission. Unlike most cytotoxic drugs, dosing of carboplatin is not based on body surface area, but upon renal function, which reliably predicts tissue exposure to a given dose of drug.

GEMCITABINE

Gemcitabine is a nucleoside analogue of deoxycytidine that belongs to the anti-metabolite class of cytotoxic drugs. The active metabolite acts as a chain terminator when DNA strands elongate during replication. Its action is, therefore, cell cycle phase-specific, acting predominantly in the S phase, but also blocking progression through the Gl/S phase ‘check-point’. It also competitively inhibits the DNA synthesis enzymes DNA polymerase and ribonucleotide reductase.

Like other cytotoxics used in the treatment of metastatic bladder cancer, gemcitabine is administered intravenously, as a short infusion. Typically, dosing is at 1000-1250 mg/m² on a weekly or 2 weeks in 3 schedule. Metabolism involves inactivation via deamination, forming difluorodeox-yuridine, which, along with the parent compound is excreted, predominantly (up to 98%) in the urine.

The dose-limiting toxicity of gemcitabine is myelosuppression, particularly neutropenia and thrombocytopenia. Non-haematological toxicities are usually mild. Commonly observed non-haematological adverse effects including nausea and vomiting, rash and flu-like symptoms. Toxicity is greater when longer infusion times are employed; this may be because shorter infusions swamp the capacity for activation and there may be a ceiling effect at about 10mg/m²/min.

Gemcitabine has shown single-agent response rates of 23-29% in patients pretreated with cisplatin and 28-36% in previously untreated patients. Complete responses have been reported between 4% and 13%.

PACLITAXEL

The taxanes (paclitaxel and docetaxel) have shown impressive activity both as single agents and in combination with other agents. Indeed, paclitaxel is the most active drug yet investigated as single-agent first-line therapy.

Paclitaxel is a semi-synthetic derivative of a naturally occurring anti-tumour agent, which is extracted from the bark and needles of the yew tree, Taxus baccata. Like vinblastine, its effect is on microtubules, but its mechanism of action is entirely different, causing stabilization of microtubules by preventing depolymerization, with consequent disruption of the intracellular cytoskeleton and of the mitotic spindle.

Paclitaxel is administered intravenously, in a variety of infusional schedules, including 1-, 3-, 6- and 24-h infusions. One of the commonest schedules involves administering at a dose of 175 mg/m² as a 3-h infusion, once every 3 weeks, but weekly schedules are also under investigation.

Hypersensitivity or anaphylactic reactions to the Intravenous administration of paclitaxel were seen in 2% of patients receiving paclitaxel in early clinical trials. They are unrelated to dose or schedule of the drug, consistent with the hypothesis that they are due to the presence of polyoxyethylated castor oil (Cremophor® EL) required to improve the solubility in aqueous solution of the highly lipophilic paclitaxel molecule. Consequently, all patients receiving paclitaxel now receive premedication, which includes oral or Intravenous corti-costeroids and antihistamines. Paclitaxel is metabolized by cytochrome P450 enzymes in the liver and eliminated predominantly in the faeces, via biliary excretion. Very little of the drug or its metabolites is found in the urine. Consequently, dose modification is required for patients with an obstructive pattern of hepatic dysfunction, but is not necessary for patients with renal impairment. Paclitaxel is therefore an attractive drug for the treatment of advanced bladder cancer, where impaired renal function is commonly seen.

In addition to the acute hypersensitivity reactions discussed above, the common toxicities of paclitaxel are myelosuppression, alopecia, transient arthralgia or myalgia, nausea/vomiting, diarrhoea, mucositis and peripheral neuropathy, with neutropenia being the major dose-limiting toxicity. Neu-rotoxicity is seen in approximately 50% of patients without previous symptoms. It is severe in 2% and results in discontinuation of the drug in 1%. This neurotoxicity is thought to result from effects on microtubules involved in axonal transport of neurotransmitters. Interestingly, when used in combination with carboplatin, paclitaxel seems to interact to produce a platelet-sparing effect, with reduced incidence and severity of thrombocytopenia compared with carboplatin alone.

Relevant clinical trial data

Combination versus single-agent therapy

Cytotoxic chemotherapy with a single drug usually fails to achieve complete cure due to the development of drug-resistant clones. Combining a number of different agents results in an attack on the cancer cell on multiple fronts in the hope that at least one of these attacks will inflict a lethal injury to the largest possible number of cells. With respect to bladder cancer, this rationale has been supported in clinical trials, showing that single-agent regimens are inferior to combination regimens. Importantly, combination chemotherapy containing cisplatin is superior to the same combination without cisplatin, and cisplatin is the foundation of all of the most effective regimens in systemic treatment of urothelial tumours.

Comparison of well-known combination regimens

Principles of therapy

Future developments

Conclusion

Combination chemotherapy is now well established as a useful treatment modality for advanced bladder cancer, which is moderately sensitive to cytotoxic drugs. For patients with good performance status and adequate renal function, such systemic chemotherapy should be cisplatin-based, with two alternative regimens as treatment of choice: MVAC and gemcitabine/ cisplatin. Many patients are unable to tolerate such relatively aggressive chemotherapy, and particularly in patients with renal impairment, combinations such as carboplatin, methotrexate and vinblastine can be valuable.

Newer cytotoxic drugs such as gemcitabine and paclitaxel may improve the therapeutic index of chemotherapies for metastatic TCCs, and are currently the subject of much clinical research activity.

The success of such chemotherapy regimens in bladder cancer raises the question of whether, in an analogous situation to the treatment of breast and bladder cancers, perioperative chemotherapy might improve the disappointing survival figures seen in muscle-invasive bladder cancer, treated with curative intent. Although a recent meta-analysis of randomized controlled trial data suggests that preoperative ‘neoadjuvant’ chemotherapy may become a new standard of care, adjuvant chemotherapy cannot currently be recommended outside the context of a clinical trial.

Despite these advances in treatment, survival rates for both advanced and muscle-invasive bladder tumours remain disappointing, and novel therapies are required. Potential future areas of promise include molecularly targeted therapies, targeting tumour phenotypes such as overexpression of epidermal growth factor receptor at the cell surface or tumour suppressor gene mutations intracellularly. Therapies with orally bioavailable tyrosine kinase inhibitors and/or targeted cancer gene therapies are promising areas of development.

CMV and MVAC

The CMV regimen  is a combination of cisplatin, methotrexate and, vinblastine. In the context of a randomized controlled phase III clinical trial, treatment with CMV resulted in an objective response rate of 46%, with a median survival of 7 months.

MVAC combines cisplatin with methotrexate, vinblastine and doxorubicin (Adriamycin®). In randomized controlled trials, this regimen has been found superior to single-agent cisplatin, combination chemotherapy with cisplatin, cyclophosphamide and doxorubicin (CISCA) and also methotrexate, carboplatin and vinblastine (MCAVI). For these phase III studies, overall response rates have varied between 39% and 65%, with median survivals up to 16 months. Dose escalation of MVAC has been attempted using recombinant haematopoietic growth factors. Better objective response rates have been reported with the more dose-intense regimen; this did not translate into improved time to disease progression or overall survival in a large randomized controlled trial. However, toxicity was significantly less in the accelerated arm, suggesting that the accelerated regimen may be a reasonable alternative to the classical MVAC .

Until recently, CMV has been regarded as a standard combination chemotherapy for advanced bladder cancer in the UK while MVAC has been more widely used in the USA. Unfortunately, these two regimens have never been compared head-to-head in a randomized controlled trial. MVAC is probably associated with more toxicity than CMV. The most significant side-effects (grade 3 and 4) include bone marrow suppression (causing anaemia, neutropoenia and thrombocytopenia), mucositis, alopecia, nausea, vomiting, infection and diarrhoea. As a result of these toxicities, only 37% of cycles of MVAC were given without a dose adjustment in a recent phase III study. Furthermore, treatment-related mortality is 3-4% in most studies.

GEMCITABINE AND  CISPLATIN

The therapeutic index of MVAC is therefore suboptimal. There has therefore been a search for novel cytotoxic agents with superior cost benefit ratio to MVAC. Several new cytotoxics, including gemcitabine and paclitaxel, are of particular interest, as they have single-agent response rates as high as, or higher, than those of cisplatin itself.

The combination of gemcitabine and cisplatin (GC) has been shown to be safe and effective in 119 patients between three phase II trials. In a direct randomized comparison with MVAC, the combination was shown to have a superior toxicity profile and efficacy appeared equivalent although the trial was not powered to confirm equivalence. As a result of this, GC has now been adopted as a standard of care in many institutions.

New doublet and triplet combinations

The introduction of several new agents into the portfolio of drugs with activity in bladder cancer has given rise to a confusing plethora of small phase II trials testing new doublet and triplet (and even quadruplet) combinations. In many cases these have been tested in the second line and the results can only be interpreted as screening tests for activity. Not surprisingly, given the inherent chemosensitivity of transitional cell carcinoma, most of these series confirm activity of the new combinations. Unfortunately, few large phase III trials have been performed to test whether any of these newer combinations are superior to those already described above. However, randomized comparison of taxane platinum doublets has shown inferiority in comparison to MVAC although these trials were small. In a rare assessment of cisplatin versus carboplatin, a randomization of Gem Cis versus Gem Carbo suggested inferior results for the carboplatin-containing regimen.

Sequential therapy

The disadvantage of combination treatment is that it may be necessary to reduce the dose of each component of the regimen, because of overlapping toxicity (usually myelosuppression) with other components. The more the components in a chemotherapy regimen, the more this is likely to be the case. An alternative to combination regimens with many agents is to treat sequentially with single-agent or doublet combinations using the highest possible dose to ensure the highest possible fractional cell kill of the clones that are sensitive to each agent. This approach is being explored at Memorial Sloan-Kettering, with encouraging results. In this treatment patients receive six cycles of a combination of doxorubicin and gemcitabine every 2 weeks followed by six cycles of a combination of ifosfamide, paclitaxel and cisplatin every 3 weeks. All cycles of this treatment required support with granulocyte colony-stimulating factor.

Table  New  doublet and triplet cytotoxic combination chemotherapies for transitional cell carcinoma

Clinical scenarios when systemic chemotherapy is indicated

FIRST-LINE TREATMENT FOR PATIENTS WITH METASTATIC DISEASE

It is generally assumed that the confirmation of metastatic disease implies the patient is incurable and survival is short. Chemotherapy used in this setting may improve quality of life and survival by causing tumour deposits to shrink and thereby reducing both the local symptoms of cancer (e.g. pain) and the systemic symptoms (e.g. fatigue). Chemotherapy in this setting needs careful management as the burden of chemotherapy in terms of toxicity and hospital visits can easily outweigh the benefits. For some patients however, the rewards are definitely worthwhile.

However, in a large retrospective review of the patients treated at the memorial with the MVAC regimen, a small number of patients were alive and disease-free at 5 years. It should be remembered therefore that selected patients, treated aggressively, often with a combination of chemotherapy and local therapy, can be cured.

SECOND-LINE TREATMENT FOR METASTATIC  BLADDER CANCER

Given that even when chemotherapy produces a response in a metastatic tumour, it is usually partial and temporary, selection of patients for a second-line treatment at disease progression is difficult. This is an area with a weak evidence base and wide variation of practice around the world. Generally speaking, responses to second-line chemotherapy are less common than to first-line treatment and for that reason patients who have had a good response to previous treatment are usually selected, lasting a significant period of time (6 months). Re-challenge with cisplatin is an option for patients having a treatment-free interval of more than 6 months or a year. Another approach is to use drugs the patient has not previously received in the hope that mechanisms of drug resistance that developed during the previous course of chemotherapy will not apply to a different agent. Finally, a patient in this situation should be offered entry to clinical trials of new agents.

PATIENTS WITH CLINICALLY ORGAN-CONFINED  DISEASE

Systemic treatment given after radical local therapy in an attempt to eradicate occult micro-metastatic disease and prevent treatment failure is called adjuvant therapy, while the same treatment given before local therapy is called neoadjuvant therapy. These two approaches have a number of similarities and differences, as shown in Table Neoadjuvant versus adjuvant chemotherapy for locally advanced bladder cancer. These differences mean that each approach must be evaluated separately in clinical trials and results from adjuvant chemotherapy should not be extrapolated to neoadjuvant and vice versa.

ADJUVANT THERAPY

There is a good reason for thinking that adjuvant chemotherapy should improve the outcome for patients with bladder cancer since the risk of metastatic relapse is high, the disease is chemosensitive, but cures are rare in the setting of overt metastatic disease. Unfortunately, there is an absence of good clinical trial data to demonstrate benefits because of a lack of well-powered, well-designed trials.

Table Neoadjuvant versus adjuvant chemotherapy for locally advanced bladder cancer

To date, four randomized studies have assessed the role of adjuvant chemotherapy following cystectomy and two following radical radiotherapy, against an observation control arm. None of the trials individually had sufficient statistical power to conclusively demonstrate a significant survival advantage. Indeed, the combined accrual to all the published trials includes only a few hundred patients. Just considering the patients with high-risk disease whose 5-year survival is 35% would require more than 1300 patients to generate sufficient events to observe a 7% improvement in overall survival at a power of 80% with 95% confidence. Of the four published trials randomizing patients between surgery alone and surgery plus postoperative chemotherapy, three employed cisplatin-based chemotherapy and reported improved disease-free survival. Significant improvements in overall survival were reported in only one. However, the design and/or analysis of each of these trials has been questioned.

There is thus, to date, insufficient evidence to advocate the routine use of adjuvant chemotherapy for muscle-invasive bladder cancer outside of the context of a clinical trial. Nevertheless, the existing data can be interpreted as  giving  an  encouraging  suggestion  towards  improved  survival  with chemotherapy, and a definitive trial with appropriate statistical power to answer the question of adjuvant chemotherapy for bladder cancer is therefore now vital. The European Organization for Research and Treatment of Cancer (EORTC) is currently coordinating an international intergroup trial and we look forward to the results of this in due course.

Table Previous randomized studies of adjuvant chemotherapy after cystectomy for bladder cancer

NEOADJUVANT THERAPY

In view of the difficulties of delivering chemotherapy after radical local therapy to the bladder, there are advantages to neoadjuvant therapy. Furthermore, this approach minimizes the delay before the introduction of a therapy that may have an effect on systemic disease and may initiate a response in the primary tumour, both reducing the chance of metastatic spread at surgery and maximizing the possibility of organ-sparing surgery. Clearly, there is a risk that if the chemotherapy is ineffective for a particular patient, the tumour may continue to progress and/or delay definitive treatment and allow an organ-confined cancer to progress. For this reason, more intense monitoring of the tumour is required when giving neoadjuvant chemotherapy so that, if there is evidence of continued progression, immediate salvage local therapy can be given.

As for adjuvant therapy, none of the individual trials of neoadjuvant chemotherapy for muscle-invasive bladder cancer published in the peer-reviewed literature, including the largest single trial of neoadjuvant chemotherapy for bladder cancer, the Medical Research Council BA06 trial, has achieved conventional statistical significance. The BA06 trial randomized 976 patients with locally advanced transitional cell carcinoma (at least T2 G3, but node-negative and free of distant metastases) either to radical treatment alone (surgery or radiotherapy) or to local therapy plus three cycles of neoadjuvant CMV combination chemotherapy, demonstrated trends towards the neoadjuvant chemotherapy arm both for median survival (increased from 35.5 months to 44 months) and for 3-year survival from 50.5% to 55.5%. However, the trial did not reach statistical significance at a median follow-up of 4 years. It has been suggested that this failure to reach statistical significance may be due to the use of CMV, rather than MVAC chemotherapy and/or the fact that 34% of patients were in the relatively good prognosis T2 G3 group, for whom it would be more difficult to demonstrate a convincing treatment benefit. The SWOG 8710 trial provided similar support for neoadjuvant chemotherapy, with an improvement in median survival from 3.8 years to 6.2 years and a 26% reduction in mortality in 624 node-negative T2-T4a patients treated with three cycles of preoperative MVAC, compared with patients who had surgery only. However, these data were obtained only over a recruitment period of 11 years, and randomized only a small proportion of the patients eligible for entry. Furthermore, they only reached statistical significance when analysed using a one-sided t test, whilst re-analysis of the SWOG data using two-sided tests, to permit comparison with the Medical Research Council BA06 trial data, showed that the SWOG 8710 data no longer achieved statistical analysis in their own right.

Recently, an updated meta-analysis of individual patient data on 2688 patients treated in ten published randomized controlled trials of neo-adjuvant chemotherapy for muscle-invasive bladder cancer has been published. This showed that in patients treated with cisplatin-based combination chemotherapy (already demonstrated to be superior to single-agent cisplatin chemotherapy and to chemotherapy lacking cisplatin, as above) there was an absolute benefit in overall survival at 5 years of 5% in favour of neoadjuvant chemotherapy, compared with treatment with surgery alone. This has led to the suggestion that neoadjuvant chemotherapy be considered a new standard of care for muscle-invasive bladder cancer.