The term cancer is used to describe a multitude of diseases, all of which are linked by loss of control of normal growth and replication of cells. It is a major cause of death worldwide, although it is proportionately more common in developed countries. Globally, approximately 10 per cent of deaths are caused by malignant diseases, but in the United Kingdom and the United States of America, these diseases account for approximately 20 per cent of deaths. Approximately 6 million new cases of cancer develop each year, causing some 4 million deaths annually. As populations become older this figure is likely to increase as the treatment of infections and cardiac diseases improves.

There is great variability in the incidence of different types of cancer between countries, and also between ethnic populations within those countries. Recently there has been an increase in the incidence of carcinoma of the prostate and cervix, with a steady decline in the incidence of other cancers such as gastric cancer.

In developed countries incidence of lung cancer is still increasing in women, but appears to be stable in men; this is closely correlated with smoking habits. However, in third-world countries the incidence continues to increase.

Despite wide variations in cancer incidence, it is still difficult to identify local environmental factors important in the aetiology of these differences, largely due to the long latency associated with many forms of cancer.

If diagnosed early many cancers, particularly tumours such as skin cancers, can be cured by local treatment such as surgery or radiotherapy. However, the majority of solid malignant tumours are not curable by local measures alone. This is due to either local invasion or dissemination of the tumour via the lymphatics or blood. Dissemination via the bloodstream can give rise to patterns of metastatic disease which are diagnostic of particular tumour types.

The mechanisms underlying variability in metastatic patterns are poorly understood at present, although multiple factors are probably involved. The pattern of venous drainage is important: e.g. dissemination of colorectal carcinoma via the portal veins leads to a high incidence of liver metastases. Alternatively, tumour cell types show differences in the expression of surface molecules such as intercellular adhesion molecule - 1; this important factor affects metastatic potential in malignant melanoma. The local extracellular millieu at the site of metastases may also be important, with detectable differences in expression of many growth factors produced by normal and malignant cells. Differences in expression of these growth factors may be important for the establishment of metastases and for subsequent stimulation of tumour cell growth.

The successful treatment of a patient with cancer involves close co-operation between surgical oncologists, radiation oncologists, medical oncologists, general practitioners, nurses, and support care workers, including clinical psychologists.

The role of the surgeon is central to this. Obtaining tissue for adequate histological analysis and identification of patients who can be cured by resection is a major component of management. Apart from curative resection of primary tumours, excision of secondary deposits can also offer long-term disease control. This paradigm is exemplified in the management of patients with colorectal cancer, in whom resection of isolated hepatic or pulmonary metastases is of proven value in a subgroup of patients. A similar situation arises with local recurrences or pulmonary metastases from soft tissue sarcomas, resection of which can result in long-term disease control.

Appropriate debulking improves the outlook in women with ovarian carcinoma: increased response rates to chemotherapy and prolonged response duration are achieved following debulking to less than 2 cm (or preferably less than 1 cm) in diameter. Likewise, adequate debulking of residual masses following induction chemotherapy for metastatic germ cell malignancy has led to improvements in overall survival in patients with high bulk disease. Surgery is also a valuable option for the palliative treatment of many patients with advanced cancer. Defunctioning colostomy or bypass operations in patients with potentially obstructive lesions of the bowel, or the palliative resection of fungating lesions eroding through the skin, can be of great benefit.

In order to increase our knowledge of the biology of malignant diseases, close links must be developed between surgeons, medical oncologists, and research scientists. High-quality fresh normal tissue and tumour tissue are necessary for laboratory analysis and research. Ultimately differences between normal and malignant tissue could be identified which may lead to the development of novel and more selective anticancer agents.

A number of systemic approaches are used to treat advanced cancer. These include chemotherapy and hormone therapy, and there has been a recent resurgence of interest in immunomodulation with cytokines such as interferons and interleukins.

The decision to treat any individual patient with systemic therapy is based on a number of factors.

Detailed histological evaluation of every tumour is mandatory and gives valuable information relevant to the choice of treatment and the overall prognosis. The response rate of different tumours to chemotherapy varies widely (Table 2) 677. Some cancers have high response rates, and cure is frequently possible (such as germ cell malignancies and Hodgkin's lymphoma); others, such as colorectal carcinoma and melanoma are highly refractory to chemotherapy, with low response rates, few remissions, and no cures. However, good palliation can sometimes be achieved in the absence of an objective response.

Histological grade is important; high-grade or anaplastic tumours have a higher proportion of cells in cycle, and when untreated they have a poorer prognosis than slowly growing tumours. Conversely, high-grade tumours frequently respond better to chemotherapy. Patients with low grade non-Hodgkin's lymphoma exhibit a continuing relapse pattern, only 20 to 25 per cent being disease free at 5 years. In contrast, aggressive combination chemotherapy regimens such as MACOP-B and ProMACE-CytaBOM have been reported to achieve complete response rates in the region of 80 per cent in patients with high-grade disease. Many of these patients remain in remission for more than 2 years.

It is important to establish the site and extent of the tumour. Staging investigations are helpful in determining the most appropriate treatment plan and also give valuable information on prognosis. Tumour staging is normally expressed using the TNM system, as described in the UICC classification of malignant tumours. Tumour bulk itself may affect response to treatment. In patients with testicular teratoma, response rates to combination chemotherapy of between 80 and 90 per cent can be achieved overall, but the response rate of patients with bulky disease is reduced, with less than 60 per cent achieving complete remission.

Age is important in the choice of systematic treatment, particularly in view of the increased likelihood of drug toxicity occurring in elderly patients. The general condition of the patient is also a major factor: there is decreased overall benefit and increased toxicity of chemotherapy in patients in poor general condition. A number of scales are used to assess the general condition of patients: the UICC version is simple and easy.

Many cytotoxic drugs are metabolized via cytochrome p-450 dependent-reactions in the liver, and disorders of hepatic function can result in either loss of efficacy or enhanced toxicity. Cyclophosphamide is inert in its parental form, but it is metabolized by the liver to the active metabolite 4-hydroxycyclophosphamide. The liver is also important in the detoxification of drugs, hyperbilirubinaemia leading to increased anthracycline toxicity due to impaired detoxification. Abnormalities in renal function are also important, causing delayed excretion of drugs and their major metabolites, and resulting in increased toxicity, as is seen with methotrexate and carboplatin.

Abnormalities of the blood count can indicate bone marrow infiltration. Chemotherapy in patients with bone marrow infiltration can induce profound bone marrow depression, which could result in life-threatening sepsis. Modification of the doses of cytotoxic drugs may be appropriate in this setting.

As chemotherapy is frequently associated with unpleasant side-effects, social factors and the psychological status of the patient are important considerations in the overall treatment plan.

One of the most important decisions in the management of a patient with cancer is to identify the ultimate aim of the treatment. When the aim is cure, the approach is completely different to the situation where palliation or short-term extension of survival is possible. When attempting cure, more intensive chemotherapy is used and significantly more toxicity is acceptable than when palliative treatment is being given.

Continued evaluation of the benefits of cytotoxic chemotherapy is essential to ensure that patients do not suffer unnecessary toxicity when no clinical benefit is being achieved. There are a number of criteria by which response is assessed. It should be stressed that patients with poorly responsive tumours may still benefit from chemotherapy to relieve specific symptoms, even in the absence of any effect on overall survival.

Cytotoxic drug administration is frequently associated with side-effects. These can be classified into non-specific toxicities, which are frequently observed with many different classes of cytotoxic drugs, and organ-specific toxicities, which are frequently associated with individual drugs. Non-specific side-effects of chemotherapy include myelosuppression, nausea and vomiting, alopecia, tiredness, and loss of concentration. Marked differences in their severity are observed between and within different classes of cytotoxic drugs.

Approximately 50 cytotoxic drugs are currently licensed for clinical use. These drugs vary in their mode of action and indications for use. They can be grouped according to their predominant mode of action.

Many alkylating agents are commonly used in clinical practice. The majority of these drugs are bifunctional, having two distinct alkyl groups that can link covalently to DNA or protein. They have a predilection for alkylation on the N7 position of guanine in DNA: this can result in abnormal base pairing of the guanine with thymidine or the formation of inter- or intrastrand cross-links between two guanine bases, inhibiting DNA replication. Alkylating agents have single agent activity against a range of human cancers. Common major toxic effects of alkylating agents include myelosuppression and gastrointestinal toxicity. Alkylating agents can also cause infertility and have been implicated in the development of second neoplasms. It is therefore appropriate, if at all possible, to limit the use of alkylating agents in the treatment of curable malignancies. The second neoplasms are predominantly drug-resistant leukaemias that occur 5 to 7 years following drug exposure. There is also an increased risk of solid tumours developing, with a peak incidence at 6 to 8 years, although the overall incidence continues to increase with time following treatment. Cisplatin, which also alkylates and cross-links DNA, is significantly less likely to cause infertility, and as yet has not been implicated in the development of second malignancies.

Cyclophosphamide and ifosfamide are active against lung cancer, breast cancer, ovarian cancer, sarcomas, and lymphomas. Dose-limiting toxicities of cyclophosphamide are myelosuppression and nausea and vomiting.

Chlorambucil is active in lymphomas and ovarian cancer. It is normally well tolerated, with minimal nausea and vomiting. The main problem associated with this drug is delayed myelosuppression associated with severe stem cell damage.

Melphalan has many similar properties to chlorambucil. It causes delayed myelosuppression and minimal gastrointestinal toxicity or alopecia. It is used to treat myeloma, normally in combination with corticosteroids, and has also been used in high doses, combined with autologous bone marrow transplantation, to treat a range of solid tumours. Resistance to melphalan is occasionally associated with resistance to cisplatin; this is thought to be related to a common membrane transport mechanism.

Nitrosoureas have multiple mechanisms of action, one of which is the alkylation of DNA. They are used to treat brain tumours. Their use has, however, been inhibited because of their association with severe gastrointestinal toxicity and prolonged myelosuppression.

Dacarbazine is a methylating agent. It is the most active agent in malignant melanoma with responses of 20 to 25 per cent. It is also used in the treatment of soft tissue sarcoma and lymphomas. Resistance to this compound is associated with increased levels of the enzyme O&sub6;-methylguanine alkyltransferase. It can cause rapid acute and severe gastrointestinal toxicity.

Methotrexate is a dihydrofolate reductase inhibitor, and also partially inhibits thymidylate synthetase. It is active against breast tumours, sarcomas, and lymphomas. Prolonged administration of high doses of methotrexate can lead to profound toxicity in the form of myelosuppression, conjunctivitis, and stomatitis. However, in the doses normally used methotrexate causes few side-effects. The toxicity associated with methotrexate can be alleviated by administration of reduced folates (leucovorin). Increased methotrexate toxicity is seen in patients with hypoalbuminaemia, renal failure, or with large third spaces such as ascites or pleural effusions.

5-Fluorouracil is a well tolerated drug active against gastrointestinal tract cancer and breast cancer. It is activated to 5′FdUMP (5′-fluoro-2′-deoxyuridine-5′-monophosphate which binds to and inhibits the activity of thymidylate synthase. Side-effects include nausea and vomiting, myelosuppression, and occasionally stomatitis. The latter is dose dependent and increased when 5-fluorouracil is combined with folinic acid (leucovorin).

Cytosine arabinoside is an inhibitor of DNA polymerase-a and -d. The drug is used predominantly in the treatment of leukaemia and has very little activity against solid tumour malignancies. Its predominant side-effects include myelosuppression and gastrointestinal toxicity.

Anthracyclines are active against lung cancer, breast cancer, lymphomas, bladder cancer, sarcomas, and gastric cancers. Doxorubicin is the most widely used anthracycline, but although it is very active, it has significant toxicity. Epirubicin appears to have a better toxicity profile, particularly in causing less cardiotoxicity. Toxicities associated with anthracycline therapy include myelosuppression, nausea and vomiting, and cardiac toxicity. Acute cardiac arrhythmias are related to peak plasma levels, and chronic administration of anthracyclines can lead to a dose-dependent cardiomyopathy. Free radical scavengers may be able to protect the heart, and trials are underway to assess this.

The mechanisms of action of anthracyclines are complex. They intercalate DNA, inhibit topoisomerase 2 activity, and cause free radical damage, and there is increasing evidence that they have a direct effect on the tumour cell membrane. The relative role of these different mechanisms of action has not been defined. Anthracyclines can cause severe local tissue damage if extravasation occurs.

Mitoxantrone is an anthracenedione active against lymphomas and breast cancer. It is used predominantly in elderly breast cancer patients in view of its relative lack of toxicity compared to doxorubicin. Its main dose-limiting toxicity is myelosuppression.

Mitomycin C has activity against breast cancer, gastric cancer, and cervical cancer. It is activated in hypoxic conditions via cytochrome p-450 reductase. Dose-limiting toxicity is myelosuppression, although renal toxicity and pulmonary fibrosis can occur, and local extravasation can cause ulceration.

There are two epipodophyllotoxins in clinical use, etoposide (VP16) and teniposide (VM26). The predominant mode of action is inhibition of topoisomerase 2, but they also produce free radicals. Etoposide is active against germ cell malignancies, lymphomas, and lung cancer. The predominant dose-limiting toxicity of this class of compounds is myelosuppression, but nausea and vomiting and alopecia are also common.

Cisplatin is highly active against many tumours, including germ cell tumours, ovarian carcinoma, bladder carcinoma, and lung cancer. Although very active, cisplatin can cause significant toxicity: the risk of dose-limiting nephrotoxicity can be reduced by adequate hydration. Nausea and vomiting is a major side-effect, although this has been alleviated to a degree with the introduction of new antiemetics. Other toxicities include hypomagnesaemia and peripheral neuropathy. Carboplatin is very active against germ cell malignancies, and also solid tumours such as ovarian carcinoma. Renal toxicity and ototoxicity are not a problem, but renal impairment can lead to increased myelosuppression, which is the dose-limiting toxicity. Problems associated with myelosuppression make the drug less suited to combination chemotherapy than cisplatin, and it is significantly more expensive.

The vinca alkaloids are mitotic spindle poisons. Vincristine is active against Hodgkin's disease and breast cancer, the predominant dose-limiting toxicity being a peripheral, predominantly sensory, neuropathy, although an autonomic neuropathy can occur. Vinblastine is active against lymphomas and testicular tumours. Myelosuppression is the dose-limiting toxicity, but neuropathy is not a problem.

Most cytotoxic drugs are administered intravenously as a bolus, or as short intravenous infusions. This overcomes the problems of unpredictable patient compliance and drug-induced nausea and vomiting associated with oral formulations.

In the majority of instances, in particular when the overall aim is cure, cytotoxic drugs are administered in combination. Each drug should ideally show a different mode of action, have proven activity as a single agent, and a different toxicity pattern from the other agents. Cytotoxic drugs are normally administered in cycles given every 3 to 4 weeks, to allow bone marrow recovery.

The apparent lack of efficacy of cytotoxic chemotherapy against the majority of common solid tumours has resulted in the exploration of different schedules of drug administration. One approach has been to use more frequent administration of compounds to which the tumours do not exhibit cross-resistance. The use of drugs that cause minimal myelosuppression allows the administration of large doses of cytotoxic drugs on a more frequent basis, hopefully inhibiting potential regrowth of tumour between cycles of chemotherapy. Scheduling may be particularly important for some cytotoxic drugs. Etoposide efficacy has been shown to be strongly dependent on scheduling: divided dose schedules are significantly superior to bolus doses and likewise continuous infusions of certain drugs such as 5FU may result in enhanced antitumour activity.

Lung cancer is the most common cancer in the Western world, accounting for approximately one-third of all male cancer deaths. Many patients die within 1 year of diagnosis with the majority of lung cancers not resectable.

Lung cancer can be divided into a number of histological subtypes. There has been recent increase in the incidence of adenocarcinoma, and presentation of a mixed cell type is not infrequent. Many histological diagnoses are made on limited biopsy material obtained during flexible fibreoptic bronchoscopy. It is likely, therefore, that the number of mixed tumours is underestimated.

Cough, breathlessness, and haemoptysis are the most common presenting symptoms in patients with lung cancer. Other common symptoms include tiredness, malaise, anorexia, weight loss, and chest pains which can occur in up to 50 per cent of patients. Symptoms suggestive of extensive mediastinal infiltration of the tumour or of distal metastatic spread are relatively rare at presentation, except in those with small cell carcinoma, in whom early dissemination of disease is common.

Management varies for the different histological subtypes. It is, therefore, important to make a firm histological diagnosis and assessment of disease spread at presentation. Patients should be discussed jointly between chest physicians, thoracic surgeons, and oncologists. Confirmation of malignancy can occasionally be made on sputum cytology, but fibreoptic bronchoscopy is frequently required to obtain tissue. Occasionally transbronchial biopsy, percutaneous lung biopsy, mediastinoscopy, or open lung biopsy is necessary to confirm the diagnosis.

Common sites of metastatic disease include liver, bone, and brain, although imaging of these areas is not indicated in the absence of any other clinical indication. All patients should undergo routine haematological and biochemical tests. A normochromic anaemia is common, although diffuse bone marrow infiltration is occasionally indicated by a leucoerythroblastic blood picture. Abnormal biochemical indices are particularly common in small cell carcinoma with this tumour known to produce multiple hormones. Hyponatraemia can indicate ectopic antidiuretic hormone secretion or hypoadrenalism from adrenal metastasis. Abnormalities of liver function are important: they may indicate hepatic metastases, which may significantly affect the body's handling of cytotoxic drugs.

The treatment of choice for patients with non-small cell lung cancer and sufficient respiratory reserve is surgery. Unfortunately only approximately 30 per cent of patients have tumours that are resectable at diagnosis. Among those whose tumour is totally resected, 5-year survival is 30 per cent. Five-year survival of patients with squamous carcinoma is 25 per cent, compared with 13 per cent in those with other histological subtypes. Significantly higher survival rates are seen in patients who do not have mediastinal lymph node involvement at operation.

Radical radiotherapy can be considered in patients who are unwilling to undergo surgical resection or in whom resection is impossible because of the site of the tumour. One-year survival rates following radical radiotherapy range from 30 to 60 per cent; less than 10 per cent of patients survive 5 years. An interesting recent approach has been the use of hyperfractionated and accelerated radiotherapy (CHART). Response rates using this radiotherapy regimen are higher, although it is too early to comment on long-term survival rates. A potential problem associated with this approach is the increased possibility of spinal cord damage and this is being closely monitored.

Radiotherapy is frequently used to relieve superior vena caval obstruction, stridor, and haemoptysis: symptomatic benefit is obtained in about 75 per cent of patients. Bone pain is frequently well controlled, and irradiation is useful for palliation of cerebral metastases or spinal cord compression, although neurosurgical decompression may be necessary for the latter.

Historically, non-small cell lung cancer has been considered resistant to chemotherapy. However, with the advent of platinum combinations, initially vindesine and cis-platinum, response rates of 30 to 40 per cent have been achieved. This has not been associated with improved long-term survival, however.

Three of the drugs most active as single agents are cis-platinum, ifosfamide, and mitomycin C (Table 7) 682. Their use in combination, as the MIC regimen, has produced response rates of 56 per cent and 47 per cent in two separate studies, with 10 to 20 per cent of patients achieving complete remission. Previous studies with platinum have been limited by severe toxicity to normal tissue, but this particular regimen appears to be relatively well tolerated: its further evaluation is indicated. Median survival in patients receiving chemotherapy for non-small cell lung cancer is, however, only 6 to 8 months.

A major challenge is the identification of patients with limited non-small cell lung cancer who relapse early and have a poor prognosis. Greater understanding of the biology of the disease may help identify these patients, who may be suitable for more aggressive treatment, and may also define targets for future therapeutic development. For example, non-small cell lung cancers express an increased number of epidermal growth factor receptors on the cell surface, although whether this affects prognosis remains unanswered. The epidermal growth factor receptor itself may prove to be a therapeutic target for imaging or therapy, with the development of monoclonal antibodies to this receptor. These monoclonal antibodies may inhibit growth via inhibition of growth factor binding, or by toxicity due to conjugation with toxins or cytotoxic drugs.

Small cell lung cancer is totally different to non-small cell lung cancer. It is highly metastatic, and the majority of patients have widely disseminated disease at presentation. The prognosis is poor: untreated patients have a median survival of 3 to 4 months.

Small cell lung cancers contain large numbers of neurosecretory granules on electron microscopy. These cells frequently secrete a range of potential tumour markers such as antidiuretic hormone, ACTH, and calcitonin, although secretion of parathyroid hormone is normally associated with squamous carcinomas. They exhibit high levels of neuroendocrine markers such as bombesin (gastrin releasing peptide), neurone specific enolase, the BB isoenzyme of creatinine kinase, and L-dopa decarboxylase. The more aggressive, so-called ‘variant phenotype’ has lower levels of L-dopa decarboxylase and bombesin and is more frequently associated with amplification of the c- myc oncogene.

A recessive deletion of a fragment of the short arm of chromosome 3 (3p14 - 23) has been identified in small cell carcinomas. This deletion is present in all small cell lung cancers, and is likely to involve loss of more than 1 tumour suppressor gene although the genes affected have not been identified.

In view of its early dissemination, the mainstay of treatment is chemotherapy. A large number of cytotoxic drugs have caused tumour regression in patients with small cell lung cancer, but when used as single agents they have minimal effect on survival. Combination chemotherapy is significantly superior to single agent chemotherapy in terms of response rate, response duration, and overall survival. Response rates approaching 90 per cent can now be achieved, with 2-year survival rates ranging from 10 to 20 per cent. However, less than 5 per cent of patients are cured or long-term survivors, and the median survival is approximately 12 to 14 months.

The role of radiotherapy in the treatment of small cell lung cancer has not been clarified. Patients with limited disease who achieve a complete response following chemotherapy may derive additional benefit from thoracic radiotherapy, although increased morbidity in the form of oesophagitis and pneumonitis may result. In 1985, Arriagada reported a 37 per cent 2-year relapse-free survival rate in patients treated with chemotherapy and intercollated radiotherapy. However the toxicity of this regimen was severe, and it is currently being re-evaluated.

Cerebral metastases are found in about 20 per cent of patients with small cell lung cancer. Prophylactic cranial irradiation can reduce the incidence of this to less than 10 per cent, although it may be associated with significant toxicity to the central nervous system. The role of prophylactic cranial irradiation is currently being addressed in a Medical Research Council trial in the United Kingdom, and at this time, this treatment should be limited to patients with the best prognosis.

The role of surgery in small cell carcinoma is ill defined. In general, surgery is not an appropriate approach. However, a number of patients with peripheral nodules, subsequently found to be small cell carcinomas, have exhibited survival similar to that of patients with non-small cell lung cancer. Patients with resectable small cell carcinoma of the lung may benefit from primary surgery followed by adjuvant combination chemotherapy.

Carcinoma of the breast is the most common malignant cause of death in females in the United Kingdom.

There is an increased incidence of breast cancer in patients who have a first-degree relative with breast cancer. Molecular genetic studies have revealed a locus on chromosome 17, associated with familial breast cancer, and studies are well advanced towards the isolation and sequencing of the gene involved (BrCa1). Patients with breast cancer, particularly of lobular histology, have an increased incidence of malignancy in the contralateral breast. The age at which a woman has her child is important: there is a decreased incidence of breast cancer in women whose first child is born before 18 years of age. Women with early menarche or a late menopause are at higher risk of breast cancer.

Beatson first described regression of breast carcinomas following oophorectomy in 1896. Subsequent studies have confirmed the oestrogen dependency of this tumour.

Oestrogen is the most important factor in the development and growth control of breast cancer. This hormone stimulates cellular proliferation, transcription of several mRNAs, and the production of a number of growth factors which may act in an autocrine or paracrine fashion to stimulate breast cancer proliferation. Over 60 per cent of breast cancers express oestrogen receptors; this number increases with age. Breast cancer tissue also expresses receptors for progesterone, prolactin, androgen, and glucocorticoid.

Increased expression of epidermal growth factor receptors (EGFR) is observed in one-third of breast carcinomas, and is associated with high-grade and hormone-independent cancer. Other proteins with homology to EGFR have been identified. C- erb B2 (HER - 2/neu) encodes a putative growth factor receptor whose expression is amplified in 17 per cent of breast cancers. A structurally similar receptor C- erb B3 has also recently been described. Its ligand is amphiregulin, which has also been shown to bind to the epidermal growth factor receptor, although the predominant ligands for the latter are epidermal growth factor itself and transforming growth factor-&agr;.

Although axillary node disease is the most important prognostic indicator in breast cancer, the expression of a number of growth factors and their receptors have been associated with poor prognosis independent of oestrogen receptor status. Patients with stage II breast carcinoma which expresses epidermal growth factor receptors have poorer relapse-free and overall survival rates than those with cancers that do not express the receptor. A similar correlation is observed with expression of pro-cathepsin D. In stage 1 disease the role of expression of these receptors on prognosis remains unclear: in these patients the high S-phase fraction appears to be the most reliable indicator of poor prognosis although recent studies suggest that the degree of angiogenesis in the primary tumour may be very important.

In recent years there has been a switch to more conservative surgical treatment of the primary breast tumour, where possible. Axillary nodes should always be biopsied in view of the prognostic implications of stage II disease, to select patients for adjuvant therapy.

Adjuvant systemic treatment should be considered in all patients with stage II breast cancer. Improvement in relapse-free and overall survival rates has been seen following chemotherapy in premenopausal patients. In postmenopausal patients, improvements in relapse-free and overall survival rates can be achieved with the anti-oestrogen tamoxifen. The role of adjuvant chemotherapy has been evaluated widely. Initial studies were based on 12 months adjuvant chemotherapy, but, recent study has shown that 6 months of treatment with cyclophosphamide, methotrexate, and 5-fluorouracil is as effective. A recent overview of adjuvant trials in breast cancer reviewed over 20 000 patients treated with adjuvant chemotherapy or endocrine therapy entered in controlled trials. This confirmed that combination chemotherapy significantly reduced early mortality in premenopausal patients and that adjuvant hormone therapy significantly reduced mortality in postmenopausal patients.

Adjuvant treatment of patients with node negative disease is controversial. At present, there is no general indication for treatment of this group as a whole, although a small number of these patients suffer early relapse and die of metastatic disease. A number of studies are currently addressing the problem of identification of this subset of patients, and randomizing them between adjuvant chemotherapy or to a control group. Other studies are evaluating whether more aggressive chemotherapy schedules are better in high-risk patients, and early results appear to confirm this. A recent study looked at the effects of chemotherapy on relapse-free and overall survival in node-negative patients with oestrogen receptor-negative tumours or primary tumours of at least 3 cm in diameter. Five-year disease-free interval in treated patients was 83 per cent to 61 per cent in the control group. It is too early to assess whether overall survival is affected by chemotherapy in this study. Tamoxifen has been shown to be of benefit, in terms of overall survival, in both node-negative and node-positive postmenopausal breast cancer patients.

Although high response rates can be achieved with systemic therapy in advanced disease, it is not curative. It is very effective for symptomatic control, can produce remission in a subgroup of patients, and may extend survival in some patients, although this remains controversial. Current studies are addressing whether high dose chemotherapy, including bone marrow transplantation, is more effective for long-term control, and results are eagerly awaited.

The likelihood of response to endocrine therapy is indicated by the presence of steroid hormone receptors. Patients who express high levels of oestrogen receptor have response rates of 50 per cent, whereas in those who have few or no oestrogen receptors the likelihood of response is 5 per cent.

Alteration of the endocrine profile in patients, either with hormones or antihormones, or by ablation of endocrine glands, is associated with an overall response rate of 25 to 35 per cent; symptomatic improvement is seen in a further 20 per cent of patients. The median duration of response to hormone therapy is 18 months. Response can be delayed, and patients with life-threatening metastases, such as lymphangitis or liver metastases, should be considered for systemic chemotherapy in the first instance.

Tamoxifen exhibits anti-oestrogen properties. It inhibits breast cancer growth by binding to the oestrogen receptor, preventing binding of oestradiol, and subsequently leading to the secretion of the growth inhibiting protein, transforming growth factor -&ggr;. Response rates are virtually identical to those seen following oophorectomy, and toxicity is minimal, particularly in postmenopausal patients.

Alternative endocrine therapies are now available and are effective in advanced disease. Aromatase is an important enzyme involved in the peripheral production of oestrogen from androgens. Aminoglutethimide is a potent inhibitor of this enzyme: doses of 125 mg twice a day produce a response rate of 19 per cent. Although steroid supplementation is not necessary at this dosage, the addition of hydrocortisone further reduces oestradiol levels and increased the response rate to 27 per cent.

Luteinizing hormone releasing hormone agonists are also effective, particularly in premenopausal patients. Depot injections of compounds such as zoladex are given every 4 weeks. This results in decreased oestradiol levels, although there is an initial stimulation of oestradiol production, occasionally associated with a tumour flare response. Responses have also been documented in postmenopausal patients, where a direct antitumour effect is postulated.

Many cytotoxic drugs are effective in breast cancer. The most effective agents so far tested have been the anthracyclines, such as doxorubicin and epirubicin. Response rates range from 20 to 50 per cent with single agent chemotherapy. Higher response rates have been achieved with anthracyclines where bone marrow toxicity was controlled by administration of bone marrow growth factors. Combination chemotherapy yields response rates of up to 70 per cent but survival rates are not significantly better than those achieved with effective single agent anthracycline therapy. Complete response rates range from 10 to 20 per cent and responses last for 8 to 12 months. Median survival of patients following chemotherapy ranges from 12 to 24 months,

In view of its limited effect on survival, it would appear appropriate to limit the duration of chemotherapy. A recent study compared mitoxantrone given for four courses only with prolonged administration of the same drug and no significant difference was observed.

Germ cell malignancy is one area where major therapeutic advances have been achieved in the systemic management of a solid tumour in the last 10 to 15 years. There are many reasons for this, including the introduction of platinum compounds. In addition, identification of residual disease has been improved both by CT scanning and by measurement of serological markers such as &agr;-fetoprotein and &bgr;-human chorionic gonadotrophin.

Germ cell tumours are the most common malignancy in young men, and the incidence is increasing. Diagnosis of testicular teratoma at inguinal orchidectomy should be followed by detailed screening, including CT scans of the abdomen and thorax and repeated estimations of tumour markers to identify residual or recurrent disease. Levels of either &agr;-fetoprotein or &bgr;-human chorionic gonadotrophin, or both, are elevated in 80 to 90 per cent of patients with teratoma. These markers are generally negative in seminoma.

Poor prognostic features include bulky disease, para-aortic or mediastinal lymphadenopathy greater than 10 cm in diameter, high levels of tumour markers (&agr;-fetoprotein > 500 i.u./l; &bgr;-human chorionic gonadotrophin 1000 i.u./l). Metastatic spread to liver, bone, or brain are bad prognostic features, and primary extragonadal germ cell tumours have a poor prognosis.

There are a number of unresolved questions in the management of patients with seminoma. In patients with stage I disease, orchidectomy and locoregional radiotherapy produces a cure rate approaching 100 per cent. However, 80 per cent of those patients would be cured by surgery alone. It would seem logical, therefore, to undertake close surveillance to detect the 20 per cent of patients who are going to relapse at an early stage, avoiding administration of radiotherapy in 80 per cent of patients. However, such monitoring is costly, and there is no reliable screening test for seminoma, making this policy difficult to follow in routine clinical practice. Stage 2A disease is treated with radiotherapy.

The management of stage 2B disease is controversial. Relapse rates following radiotherapy approach 20 per cent, and chemotherapy would appear to be a logical option, although many groups still use radiotherapy. Stage 2C and above should be treated with chemotherapy using platinum either alone or in combination chemotherapy. The role of additional radiotherapy in these patients is not resolved.

The recommended approach to stage I disease is observation, with measurement of serum markers every month and CT scans every 3 months over the first year. Combination chemotherapy is the standard treatment for disease of stages 2A to 4. Platinum-based combinations such as bleomycin, etoposide, and cisplatinum produce cure rates of 80 to 90 per cent in patients with small tumours; new approaches are aimed at reducing treatment/related toxicity. The EORTC is currently evaluating standard bleomycin, etoposide, and cisplatinum chemotherapy compared with etoposide and cisplatinum alone. Early reports show no significant difference between the two treatment groups, although slightly more relapses were observed in patients receiving only two drugs.

The response of patients with large tumours is significantly inferior: less than 60 per cent achieve a complete remission. Newlands et al. reported on the administration of cisplatinum, vincristine, methotrexate, and bleomycin, alternating with actinomycin D, cyclophosphamide, and etoposide every 10 to 14 days. Patients received two cycles of chemotherapy following the achievement of complete remission. Complete remission was achieved in more than 80 per cent of patients with bulky abdominal or mediastinal disease using this regimen.

Debulking following chemotherapy is now standard treatment for residual disease. In one study, surgically removed tissue was found to contain differentiated teratoma (44 per cent), fibrotic/necrotic material (34 per cent), and residual malignancy (23 per cent). The last finding is an indication for continuation of chemotherapy. Such surgery, together with additional chemotherapy where indicated, has significantly increased the survival of patients with bulky germ cell tumours.

Colorectal cancer is the second most common cancer in the United Kingdom, after lung cancer in males.

The last few years has seen a tremendous increase in our understanding of the progression of malignancy in bowel tumours. Primarily, this stemmed from molecular genetics and our understanding of the mechanisms involved in allelic loss, and its relationship to the development of human malignancy. This was most clearly demonstrated in retinoblastoma. Cytogenetic and linkage studies mapped the retinoblastoma locus to 13q14. Retinoblastoma is believed to occur when an individual who has inherited one mutant retinoblastoma gene loses the normal allele by one of a number of mechanisms, including mitotic non-junction or recombination. Abnormalities at this locus are also found in other human cancers, such as small cell lung cancer.

Similar studies into the mechanism of malignant progression have been undertaken in patients with polyposis coli. Although only a small proportion of colorectal cancers are found in patients with polyposis, between 10 and 20 per cent of all colorectal carcinomas probably have an underlying genetic predisposition. The gene responsible for familial adenomatous polyposis has been mapped to chromosome 5q: this region is deleted in approximately 30 per cent of sporadic adenomas. Frequent genetic changes involving tumour suppressor genes have been identified in colorectal tumours: 17p and 18q deletions are observed in 75 per cent of colorectal carcinomas. It is thought that the deletion at 17p causes loss of a p53 allele, leaving a single mutated p53 allele. An 18q deletion has also frequently been detected in colorectal carcinomas. There is a candidate tumour suppressor gene in this area termed the DCC gene. The true function of the protein product of this gene has not been identified.

It has been suggested that there are a number of additional abnormalities in colorectal cancer, and these appear to develop sequentially. Hypomethylation is present early, as are mutations in the K- ras oncogene, together with deletion of alleles on chromosome 5q. Deletions involving 18q and 17p appear to be late events. Further evaluation of this process may help in our understanding of the development of other solid tumours and also may identify valuable targets for drug design.

Surgery remains the mainstay of treatment for primary, and occasionally metastatic colorectal carcinoma. In addition, palliative surgery may occasionally be indicated to improve the quality of life in patients with unresectable colorectal carcinoma.

Systemic chemotherapy of colorectal carcinoma has been extremely disappointing. Response rates to many cytotoxic drugs are less than 20 per cent, complete remission is almost never seen. 5-Fluorouracil has been the drug of choice, with response rates approaching 20 per cent; however, the median duration of this response is relatively short and is normally less than 6 months.

Recently there has been great interest in the modulation of 5-fluorouracil cytotoxicity using folinic acid. Response rates range from 30 to 40 per cent obtained—a significant improvement over those obtained with 5-fluorouracil alone.

New drugs that are more active than the commonly used agents are required. The development of monoclonal antibodies to specific antigens on the surface of colorectal carcinomas may prove to be of value. In the longer term, development of new compounds based on our understanding of the pathogenesis of the tumour may be more successful, such as targeting to the abnormal p53 protein product produced in this tumour type.

Although progress has been made in the treatment of certain malignant conditions, cure remains elusive for the majority of disseminated malignancies affecting adults. Resistance to the currently used cytotoxic drugs is common; this may occur for many reasons (Table 10) 685. At the cellular biochemical level resistance mechanisms are multiple and tend to vary with the class of compound to which resistance has developed.

Mechanisms of resistance to antimetabolites vary with different compounds. Methotrexate resistance has been widely studied and a number of potential mechanisms of resistance have been proposed.

Cells made resistant to single cytotoxic drugs such as doxorubicin or vinblastine in vitro frequently exhibit cross-resistance to a range of other structurally unrelated cytotoxic drugs derived from natural products. This form of resistance is associated with an alteration in drug transport, leading to decreased intracellular cytotoxic drug levels. Increased drug efflux is energy dependent and is associated with the presence of a 170 kDa membrane glycoprotein, p-glycoprotein. The gene encoding p-glycoprotein has since been cloned, and called mdr1. Multiple drug resistance genes have been conserved through many species. In man there are two genes, mdr1 (associated with the multiple drug resistance phenotype), and mdr2 for which no function has as yet been identified. Transfection of the mdr1 gene into recipient cells leads to expression of the resistance phenotype, indicating that this resistance is associated with over-expression of a single gene.

In normal tissues, expression of mdr1 is increased in the liver, colon, kidney, and adrenal gland, while low levels are found in the bone marrow and in lung. p-Glycoprotein is found on the luminal surface of a number of organs such as the bile canaliculi in the liver, the luminal surface of the proximal renal tubule, and on the luminal surface of the colon. Although the precise function of this glycoprotein in normal cells is not well defined, this distribution suggests that it has a major role in drug transport and drug excretion.

High levels of mdr1 mRNA have been found in tumours derived from tissues that normally expresses high levels of mdr1 mRNA. These tumours are highly resistant to cytotoxic chemotherapy at presentation suggesting a potential role for p-glycoprotein in intrinsic drug resistance. Cancers such as acute lymphoblastic leukaemia and myeloma, in contrast, have low levels of mdr1 RNA at presentation and are relatively sensitive to cytotoxic drugs. On relapse, mdr1 levels are increased, and drug resistance is frequently observed, indicating a possible role for this gene in acquired drug resistance.

Cytotoxic drugs and their metabolic products are frequently highly reactive, and there are many natural mechanisms within the cell designed to detoxify reactive molecules. Glutathione, a tripeptide (&ggr;-glutamyl cysteinyl glycine) which is present in millimolar concentrations in all cells of the body, protects the cell against reactive molecules using the free sulphhydryl group on the cysteine. This can detoxify a range of metabolites via conjugation reactions: the reaction rate of these reactions is dramatically increased if they are catalysed by one of the glutathione transferase isoenzymes. Glutathione is also an important detoxification mechanism for the handling of free radicals: Glutathione forms conjugates with may cytotoxic drugs, and elevated glutathione levels have been associated with resistance to cytotoxic drugs both in vitro and in vivo. Exposure of cells in vitro to buthionine sulphoximine causes a progressive depletion in cellular glutathione levels, resulting in enhanced sensitivity to a range of cytotoxic drugs. However, one of the problems of using this approach as a therapeutic strategy has been its relative lack of selectivity for tumour tissue.

Glutathione levels are generally higher in tumour cells than in normal tissue. This fact could prove to be valuable clinically, using glutathione to activate cytotoxic drugs, with neocarzinostatin, which is inactive in its parental form but is reduced by glutathione to the active metabolite. Toxicity to normal tissue precludes the clinical use of neocarzinostatin, but better tolerated analogues may be developed.

The glutathione transferases comprise a multigene family of isoenzymes with a wide range of substrate specificity. They are conserved through many species and form an important cellular protective mechanism against a wide range of xenobiotics and carcinogens. The nomenclature of these enzymes is complex. However, the enzymes can be loosely grouped into three main cytosolic classes of glutathione transferases, which are divided according to their separation by isoelectric focusing (Fig. 4) 2779.

Acidic transferases are expressed at higher levels than the other classes of glutathione transferases in all solid tumours studied. It has been suggested that the acidic transferases may serve as a serological tumour marker, but this remains unsubstantiated. Increased expression of the acidic transferase has been described in drug-resistant cell lines, such as MCF7 breast cancer cells made resistant to doxorubicin. However, resistance to alkylating agents has also been associated with elevated expression of other glutathione transferases including basic, neutral, and microsomal glutathione transferases.

A number of glutathione transferase genes have now been cloned and expressed in cellular models. Two studies found that transfection of an acidic transferase was not associated with doxorubicin resistance, whereas two other studies observed increase resistance to alkylating agents and doxorubicin following transfection of acidic and basic glutathione transferases. The precise role of glutathione transferases in clinical drug resistance therefore remains unclear. Glutathione transferases may provide a target for chemotherapy in the future if it becomes possible to identify specific differences in isoenzyme profile between tumour and normal tissues.

The topoisomerases are nuclear enzymes that are intimately involved in DNA replication, transcription, and recombination. The type 2 topoisomerases, which have been most widely studied, exist in &agr; and &bgr; forms, which appear to have different cellular functions. These energy-dependent enzymes are essential for cell growth and division. Since they bind covalently to DNA they represent a target for a number of cytotoxic drugs, including anthracyclines such as doxorubicin, and epipodophyllotoxins such as etoposide. These cytotoxic drugs act by freezing the normal catalytic activity of the enzyme, producing the so-called ‘cleavable complex’ which prevents strand passage and religation of previously nicked DNA. Increased topoisomerase 2 activity is associated with sensitivity to inhibitors of the enzyme, with high levels of enzyme activity identified in sensitive tumour types such as testicular teratoma.

Resistance to topoisomerase 2 inhibitors develops rapidly. The majority of topoisomerase inhibitors used in clinical practice are transported by p-glycoprotein and these drugs form part of the ‘so-called’ multiple drug resistance phenotype. Development of inhibitors that are not transported by this mechanism could prove to be a major therapeutic advance. However, there are other mechanisms by which resistance to topoisomerase 2 inhibitors may develop. Decreased enzyme activity can be caused by modification of gene expression through hypomethylation of DNA, mutations of the gene decreased mRNA expression, or reduced specific activity of the enzyme itself. Alternatively, it is also potential feasible to enhance cellular chemosensitivity by up-regulating topoisomerase 2 expression. This has been achieved using growth factors such as oestrogen in breast cancer cells.

DNA topoisomerase 1 is a monomeric protein that may also prove to be an important target. Camptothecin, which was toxic in early clinical trials, inhibits topoisomerase 1 activity by freezing the cleavable complex, in a similar manner to the action of topoisomerase 2 inhibitors. Recently an analogue of camptothecin has been shown to produce remissions in human colon cancer xenografts, and the development of this and other topoisomerase 1 inhibitors may produce valuable anticancer drugs.

Abnormalities in DNA repair are associated with hypersensitivity to DNA damaging agents and with a number of cancer-prone syndromes, including ataxia telangectasia, xeroderma pigmentosum, Cockayne's syndrome, and Fanconi's anaemia. It is less clear whether increased levels of DNA repair are important in resistance to cytotoxic drugs, although over-expression of O&sub6;-methylguanine alkyltransferase, has been associated with resistance to methylating agents such as dacarbazine.

At present there is a great deal of interest in the role of DNA repair in resistance to cisplatin. Resistance to cisplatin is multifactorial: reduced cellular accumulation of the drug is important in some models, and enhanced cellular detoxification by glutathione or metallothioneins important in others. However, some models associate resistance to cisplatin with enhanced DNA repair. In one model, resistance is associated with enhanced intrastrand cross-link repair. This may be important clinically: the reduced capacity of testicular teratoma to repair platinum adducts in vitro compared with bladder carcinoma cells may, in part, account for their exquisite sensitivity to cisplatin.

In bacteria, DNA repair is inducible by a number of mechanisms. An ‘SOS’ response which can be activated by a wide range of DNA damaging agents results in increased expression of a number of DNA repair genes. Alternatively, alkylation of DNA can lead to an adaptive response which results in increased expression of a limited number of repair enzymes, including O&sub6;-methyl guanine alkyltransferase. Inducibility of DNA repair has not been demonstrated in mammalian cells, although this is likely to occur and may be an important factor in resistance to cytotoxic drugs. The many mechanisms by which tumours may become resistant to cytotoxic chemotherapy at the cellular level are poorly understood at present. More advanced tumours and increasing numbers of cell divisions undergone by the tumour, are correlated with a greater likelihood of resistance developing via a mutation affecting one of these cellular mechanisms. Greater understanding of the mechanisms of drug resistance would help us to devise novel treatment strategies to overcome it.

There are a number of potential ways to circumvent the clinical problem of resistance to cytotoxic drugs.

A number of analogues of commonly used cytotoxic drugs are already in clinical use. These were developed primarily to reduce the toxicity associated with cytotoxic chemotherapy; for example, carboplatin has a totally different pattern of toxicity compared to cisplatin. Likewise, the anthracycline epirubicin causes significantly less gastrointestinal toxicity and decreased cardiac toxicity compared to doxorubicin.

In addition to circumventing dose limiting toxicities of the parent drugs, analogues may also exhibit different patterns of antitumour activity by circumvention of particular mechanisms of drug resistance. For instance novel anthracyclines are being developed that do not appear to exhibit the cross-resistance pattern shown by parent anthracyclines such as doxorubicin. It is hoped that these drugs will enter clinical trials in the near future.

Increased doses of chemotherapy can be used if the common major dose-limiting toxicities of cytotoxic administration can be controlled.

Improved control of chemotherapy-induced emesis has been achieved by greater understanding of the mechanisms involved. Combinations of antiemetics with different modes of action, administered up to 24 h before the commencement of cytotoxic chemotherapy has led to much better control of emesis. Elucidation of the role of 5-hydroxytryptamine receptor in cytotoxic drug induced emesis has led to the development of a range of relatively non-toxic specific receptor antagonists which are very effective. Complete control of nausea and vomiting can be achieved with 5-hydroxytryptamine receptor antagonists in 50 per cent of patients receiving cisplatin. Greater control is possible when these antiemetics are used in combination with steroids.

Myelosuppression is the most common dose-limiting toxicity associated with many cytotoxic drugs. This problem can be reduced by bone marrow or peripheral blood stem cell transplantation or with haematopoietic growth factor support.

Bone marrow transplantation has been of unquestionable benefit in the treatment of many leukaemias, but results using this approach in many solid tumour malignancies have been less successful. In the treatment of small cell carcinoma of the lung, neither early or late intensification of chemotherapy has made a significant impact on overall survival.

A number of haematopoietic growth factors have recently been cloned and will be available for clinical use. This group of glycoproteins include the colony stimulating factors (granulocyte, granulocyte - macrophage, macrophage), erythropoietin, and the interleukins. Recent clinical studies have shown the beneficial effect of granulocyte colony stimulating factor and granulocyte macrophage colony stimulating factor in the prevention of sepsis associated with severe cytotoxic drug-induced neutropenia. Growth factor support reduces the extent and duration of the white blood cell count nadir in these studies, reducing the risk of life-threatening neutropenic sepsis. The introduction of these compounds into clinical practice represents an exciting new development in cancer chemotherapy.

The dose intensity of cytotoxic chemotherapy could be increased if dose-limiting organ specific toxicity could be controlled. For example, haemorrhagic cystitis associated with ifosfamide and cyclophosphamide therapy can be controlled by administration of the sulphhydryl-containing compound mesna. Mesna acts by conjugating acrolein, a highly reactive metabolite of ifosfamide and cyclophosphamide, preventing bladder toxicity.

Many compounds bind to p-glycoprotein, inhibiting its function as an efflux pump and reversing the resistance phenotype. Verapamil, a calcium channel blocker, was the first compound shown to bind to p-glycoprotein and reverse the resistance phenotype. Unfortunately levels of verapamil required to reverse resistance are not clinically achievable without severe cardiotoxicity. New, less toxic modulators are required to enable the achievement of plasma levels equivalent to those shown to be effective in vitro. Most calcium channel blockers are effective, and other compounds such as quinidine, chlorpromazine, chloroquine, reserpine, tamoxifen, and cyclosporin A have been shown partially to reverse drug resistance in vitro. Clinical trials of these potential modulators of drug resistance are currently underway. Cyclosporins and tamoxifen show the greatest promise, as the levels effective in vitro are clinically achievable, with acceptable toxicity. In the future combinations of multidrug resistance modifiers or combinations of modifiers affecting multiple mechanisms of action may be used to increase the therapeutic efficacy of cytotoxic chemotherapy.

Drug detoxification can be modulated experimentally by compounds such as buthionine sulphoximine which deplete intracellular glutathione levels. Such agents are only of value if they are selective against tumour tissue. A more promising approach may be to develop drugs that require activation by glutathione.

A number of drugs used in clinical practice are known to inhibit DNA repair. Hydroxyurea inhibits ribonucleotide reductase at clinically achievable levels, starving the cell of deoxyribonucleotides and inhibiting DNA synthesis. Cytosine arabinoside inhibits DNA polymerase &agr; and &dgr; at clinically achievable levels. These compounds have minimal activity against human solid tumours. Their use to inhibit DNA repair following treatment with cisplatin is currently being studied. However, new and more specific repair inhibitors are required: aphidocolin, one such agent currently under investigation, enhances cisplatin toxicity in the experimental setting.

There is a great need to develop new compounds with novel mechanisms of action. One way to achieve this is by the random screening of a large number of compounds as exemplified by the National Cancer Institute (USA) Drug Discovery Project which tests a large number of new compounds annually against panels of human cell lines derived from a wide range of histological subtypes of cancer.

The second, and possibly more optimistic approach, is to use our increasing knowledge of the cell biology of human cancer to identify potential targets for drug development. The identification of differences between normal and neoplastic cells gives us the opportunity to develop new classes of anticancer agents with greater selectivity than do the present drugs. Recent understanding of the importance of autocrine and paracrine growth factors and growth factor receptors in malignant cell growth identifies valuable targets for potential growth inhibitors. These inhibitors could be monoclonal antibodies linked to cytotoxic drugs, toxins or radionuclides, or, indeed, short peptides.

Increasing understanding of second messenger systems involved in transmitting signals from the cell membrane to the nucleus opens an exciting new field for the development of new classes of anticancer agents, and lead compounds are already undergoing clinical evaluation.

Although improvements have been made in the treatment of solid tumour malignancies over this time there has been minimal impact overall on the mortality associated with these diseases. Over the next decade there is the opportunity for great improvements to occur in the management of solid tumours, in particular, the identification of high risk groups to be selected for adjuvant systemic treatment and also in the development of novel anticancer agents reflecting increasing understanding of the biology of neoplastic cell growth.

Arriagada R, et al. Alternating radiotherapy and chemotherapy schedules in small cell lung cancer limited disease. Int J Radiation Oncology Biol Phys 1985; 11: 1461 - 7.

Batist G, Tupule A, Sinha BK, Katki AG, Meyers CE, Cowan KH. Overexpression of a novel anionic glutathione transferase in multidrug resistant human breast cancer cells. J Biol Chem 1986; 261: 15544 - 9.

Bonadonna G, Valagussa P. Systemic therapy in resectable breast cancer. In: Henderson IC, ed. Haematology/Oncology Clinics of North America. Philadelphia: WB Saunders, 1989: 727.

Breast Cancer Trials Committee, Scottish Cancer Trials Office (MRC). Adjuvant tamoxifen in the management of operable breast cancer: the Scottish trial. Lancet 1987; ii: 171 - 5.

Bronchud MH, et al. Phase I/II study of recombinant human granulocyte colony-stimulating factor in patients receiving intensive chemotherapy for small cell lung cancer. Br J Cancer 1987; 56: 809 - 14.

Bronchud MH, Howell A, Crowther D, Hopwood P, Souza L, Dexter TM. The use of granulocyte colony-stimulating factor to increase the intensity of treatment with doxorubicin in patients with advanced breast and ovarian cancer. Br J Cancer 1989; 60: 121 - 5.

Calvert AH, Carter SK, Egorin MJ, eds. Proceedings: The cell membrance and cell signals as target in cancer chemotherapy, Cancer Chemother Pharmacol 1989; 24 (Supplement 2).

Carmichael J, Cantwell BMJ, Edwards CM, Rapeport WG, Harris AL. The serotonin type 3 receptor antagonist BRL 43694 and nausea and vomiting induced by cisplatin. Br Med J 1988; 297: 110 - 1.

Carney DN, et al. Establishment and identification of small cell lung cancer cell lines having classic and variant features. Cancer Res 1985; 45: 291 - 323.

Cullen MH, Joshi R, Chetiyawardana AD, Woodroffe CM. Mitomycin, ifosfamide and cis-platin in non-small cell lung cancer: Treatment good enough to compare. Br J Cancer 1988; 58: 359 - 61.

Cunningham D, et al. Prevention of emesis in patients receiving cytotoxic drugs by GR 38032F, a selective 5-HT3 receptor antagonist. Lancet 1987; i: 1461 - 3.

Currie DC, et al. Mitomycin, ifosfamide and cisplatin in non-small-cell lung cancer. Cancer Chemother Pharmacol 1990; 25: 380 - 1.

DeVita VT, Hellman S, Rosenberg SA. Cancer Principles & Practice of Oncology, Vol 1 & 2. 2nd edn. Philadelphia: JB Lippincott Company, 1985.

Dische S, Saunders MI. Continuous, hyperfractionated accelerated radiotherapy (Chart). Br J Cancer 1989; 59: 325.

Drake FH, Hofmann GA, Bartus HF, Mattern MR, Crooke ST, Mirabelli CK. Biochemical and pharmacological properties of p170 and p180 forms of topoisomerase II. Biochemistry 1989; 28: 8154 - 60.

Early Breast Cancer Trialists Collaborative Group. The effects of adjuvant tamoxifen and of cytotoxic therapy on mortality in early breast cancer: an overview of 61 randomised trials among 28,896 women. N Engl J Med 1988; 319: 1681.

Eastman A, Schulte N. Enhanced DNA repair as a mechanism of resistance to cis-diamminedichloroplatinum. Biochemistry 1988; 27: 4730.

Einhorn LH, Donohue JP. Cis-diamine-dichloroplatinum, vinblastine and bleomycin combination chemotherapy in disseminated testicular cancer. Ann Intern Med 1977; 87: 293 - 8.

Fisher B, Fisher ER, Redmond C. Ten year results from the NSABP clinical trial evaluating the use of L-phenylalanine mustard (L-PAM) in the management of primary breast cancer. J Clin Oncol 1986; 4: 929.

Giovanella BC, et al. DNA Topoisomerase I-targeted chemotherapy of human colon cancer in xenografts. Science 1989; 246: 1046 - 8.

Hamilton TC, Masuda H, Young RC. Modulation of cisplatin cytotoxicity by inhibition of DNA repair in a cisplatin resistant human ovarian cancer cell line. Proc Am Assoc Cancer Res 1987; 28: 1151.

Harris AL, et al. Comparison of short-term and continuous chemotherapy (Mitozantrone) for advanced breast cancer. Lancet 1990; 335: 186 - 90.

Hayes JD, Wolf CR. Role of glutathione transferase in drug resistance. Glutathione Conjugation—Mechanisms and Biological Significance. London: Academic Press 1988.

Hermanek P, Sobin LH. TNM Classification of Malignant Tumours. 4th edn. UICC, Springer-Verlag, 1987.Kessel D. Resistance to Antineoplastic Drugs. CRC Press Inc., 1989.

Knudson AG. Mutation and Cancer: Statistical study of retinoblastoma. Proc Natl Acad Sci USA 1971; 68: 820 - 3.

Kramer RA, Zakher J, Kim G. Role of the glutathione redox cycle in acquired and de novo multidrug resistance. Science 1988; 241: 694 - 6.

Kraus MH, Issing W, Miki T, Popescu NC, Aaronson SA. Isolation and characterization of ERBB3, a third member of the ERBB/epidermal growth factor receptor family: evidence for overexpression in a subset of human mammary tumours. Proc Natl Acad Sci USA 1989; 86: 9193 - 7.

Mansour EG, et al. Efficacy of adjuvant chemotherapy in high-risk node-negative breast cancer. N Engl J Med 1989; 320: 485 - 90.

Newlands ES, Reynolds KW. The role of surgery in metastatic testicular germ cell tumours (GCT). Br J Cancer 1989; 59: 837 - 9.

Newlands ES, Begent RH, Rustin GJ, Parker D, Bagshawe KD. Further advances in the management of malignant teratoma of the testis and other sites. Lancet 1983; i: 948 - 51.

Nicholson S, Sainsbury JRC, Halcrow P, Chambers P, Farndon JR, Harris AL. Expression of epidermal growth factor receptor associated with lack of response to endocrine therapy in recurrent cancer. Lancet 1989; i: 192 - 185.

Poon MA, et al. Biochemical modulation of fluorouracil: evidence of significant improvement of survival and quality of life in patients with advanced colorectal carcinoma. J Clin Oncol 1989; 7: 1407 - 18.

Prasad US, Naylor AR, Walker WS, Lamb D, Cameron EWJ, Walbaum PR. Long term survival after pulmonary resection for small cell carcinoma of the lung. Thorax 1989; 44: 784 - 7.

Rothenberg M, Ling V. Multidrug resistance: molecular biology and clinical relevance. J Natl Cancer Inst 1989; 81: 907 - 9.

Sikora K, Halnan KE. Treatment of cancer. 2nd edn. London: Chapman and Hall Medical, 1990.

Slevin ML, et al. A randomized trial to evaluate the effect of schedule on the activity of etoposide in small-cell lung cancer. J Clin Oncol 1989; 7: 1333 - 40.

Steward WP, et al. Recombinant human granulocyte macrophage colony stimulating factor (rhGM-CSF) given as daily short infusions—a phase I dose-toxicity study. Br J Cancer 1989; 59: 142 - 5.

Tait D, Peckham MJ, Hendry F, Goldstran P. Post-chemotherapy surgery in advanced non-seminomatous germ cell tumours: the significance of histology with particular reference to differentiated (mature) teratoma. Br J Cancer 1984; 50: 601 - 9.

Tandon AK, Clark GM, Chamness GC, Chirgwin JM, McGuire WL. Cathepsin D and prognosis in breast cancer. N Engl J Med 1990; 322: 297 - 302.

Tsuchida S, Sekine Y, Shineha R, Nishihira T, Sato K. Elevation of the placental glutathione S-transferase form (GST-&pgr;) in tumour tissues and the levels in sera of patients with cancer. Cancer Res 1989; 49: 5225 - 9.

Veale D, Kerr N, Gibson GJ, Harris AL. Characterization of epidermal growth factor receptor in primary human non-small cell lung cancer. Cancer Research 1989; 49: 1313 - 17.

Vogelstein B, et al. Genetic alterations during colorectal-tumour development. N Engl J Med 1988; 319: 525 - 32.

Whang-Peng J, et al. Specific chromosome defect associated with human small cell lung cancer: deletion 3p (14 - 23). Science 1982; 215: 181 - 2.WHO Handbook for Reporting the Results of Cancer Treatment. Geneva: WHO Offset Publication No. 48, 1979.

Wright C, et al. Expression of c-erb B2 oncoprotein: a prognostic indicator in human breast cancer. Cancer Res, 1989; 49: 2087 - 90.

Woll PJ, Rozengurt E. Multiple neuropeptides mobilise calcium in small cell lung cancer: effects of vasopressin, bradykinin, cholecystokinin, galanin and neurotensin. Biochem Biophys Res Comm 1989; 164: 66 - 73.