Neoplastic disease and immunosuppression

Clinical Pharmacology, 11e

Harpreet Wasan

Synopsis

The causes of cancer are multi-factorial. Most cancer incidences are sporadic, with fewer than 5% being familial. In many cases, environmental risk factors are recognised, which include lifestyle choices, e.g. tobacco smoking, diet and exposure to sunlight. The growing number and efficacy of systemic modalities available to treat patients with cancer are significantly improving disease outcomes. Immunosuppressive drugs are described here as they share many characteristics with anticancer drugs.

• Cancer treatments and outcomes.

• Rationale for cytotoxic chemotherapy.¹

• Classes of cytotoxic chemotherapy drugs.

• Chemotherapy in clinical practice.

• Endocrine therapy.

• Immunotherapy.

• Targeted biological therapies.

• Immunosuppression and immunosuppressive drugs.

Neoplastic disease

Cancer treatments and outcomes

Cancers share some common characteristics:

• Growth that is not subject to normal spatial restrictions for that tissue and fails to respond to apoptotic signals (see below) or in which a high proportion of cells are dividing, i.e. there is a high ‘growth fraction’.

• Local invasiveness.

• Tendency to spread to other parts of the body (metastasise).

• Less differentiated cell morphology.

• Tendency to retain some characteristics of the tissue of origin, at least initially.

Cancer treatment employs six established principal modalities:

1. Surgery.

2. Radiotherapy.

3. Cytotoxic chemotherapy.

4. Endocrine therapy.

5. Immunotherapy.

6. Biological (or targeted) therapy.

This account describes the main groups of drugs (see p. 510) but it is important to understand the overall context in which systemic therapy is offered to patients.

Systemic cancer therapy

Cancers originating from different organs of the body differ in their initial behaviour and in their response to treatments (Table 31.1). Primary surgery and/or radiotherapy to a localised cancer offer the best chance of cure for patients. Drug treatments offer cure only for certain types of cancer, often characterised by their high proliferative rate, e.g. lymphoma, testicular cancer, Wilms’ tumour. More often, systemic therapy offers prolongation of life from months to many years and associated improvements in quality of life, even if patients ultimately die from their disease.

Table 31.1 Degree of benefit achieved with systemic therapy for common cancers

Curable: chemosensitive cancers	Improved survival: some degree of chemosensitivity	Equivocal survival benefit: chemoresistant cancers
Teratoma	Colorectal cancer	Sarcoma
Seminoma	Small cell lung cancer	Bladder cancer
High-grade non-Hodgkin's lymphoma	Ovarian cancer	Melanoma
Hodgkin's lymphoma	Breast cancer	Renal cancer Insensitive to cytotoxic chemotherapy but now can be controlled temporarily with oral VEGF2 and mTOR inhibitors
Wilms’ tumour	Cervical cancer	Primary brain cancers
Acute myeloblastic leukaemia	Endometrial cancer	Nasopharyngeal carcinoma
Acute lymphoblastic leukaemia in childhood	Gastro-oesophageal cancer Cholangiocarcinoma and gall bladder cancer	Hepatoma Insensitive to cytotoxic chemotherapy but now can be controlled temporarily with oral VEGF inhibitors
	Myeloma
	Pancreatic cancer
	Low-grade non-Hodgkin's lymphoma
	Non-small cell lung cancer
	Adult acute lymphoblastic leukaemia

Use of drugs as adjuvant therapy attempts to eradicate residual microscopic cancer by treating patients after their primary surgery. This strategy has improved overall survival for patients after surgical resection of primary breast, colorectal and gastric cancer. In some situations, drugs are administered prior to surgery (neoadjuvant therapy), primarily to shrink large, locally advanced disease to subsequently enable surgical resection. Many patients with cancer are not cured by their primary treatment due to the presence of micrometastatic disease; the disease often returns months or years later even though at the time of completing their initial treatment there was no visible evidence of cancer. Clearly, this is a limitation of current standard techniques used to identify residual disease. Currently, radiological techniques cannot clearly visualise lesions smaller than 5 mm in most organs, which equates to over many million cancer cells.

Palliative therapy, offered to patients with advanced, incurable cancer, aims both to increase survival and to improve quality of life by symptom control. Despite significant improvements in cancer outcomes in the last 5–10 years, there remain a number of types of cancer that are poorly responsive to currently available drugs. Patients with chemoresistant cancers who are fit enough and willing may be offered experimental treatments within Phase 1 or 2 clinical trials.

Most treatments currently available are associated with unwanted effects of varying degrees of severity. The risk of causing harm must be weighed against the potential to do good in each individual case. Systemic therapy aims to kill malignant cells or modify their growth but leave the normal cells of the host unharmed or, more usually, temporarily harmed but capable of recovery. When there is realistic expectation of cure or extensive life prolongation, then to risk more severe drug toxicity is justified. For example, the treatment of testicular cancer with potentially life-threatening platinum-based combination chemotherapy regimens offers a greater than 85% chance of cure, even for those with extensive, metastatic disease.

Where expectation is confined to palliation in terms of modest life prolongation of less certain quality, then the benefits and risks of treatment must be judged carefully. Palliative treatments should involve low risk of adverse effects, e.g. 5-fluorouracil-based chemotherapy for advanced colorectal cancer is well tolerated by most patients, while improving survival by around 1–2 years. A modern prerequisite of cancer chemotherapy Phase 3 trials is concomitantly and objectively to assess patient quality of life while on drug therapy. This helps clinicians and nurses to explain the potential benefits and harm of treatment to patients and their families, who may themselves hold strong views about the quality and quantity of life.

Rationale for cytotoxic chemotherapy

The narrow therapeutic index of cytotoxic agents means that escalation of drug doses is constrained by damage to normal cells and the maximum doses that patients can tolerate are often suboptimal to achieve total cancer cell killing. Even so, cytotoxic chemotherapy agents remain the mainstay of systemic anticancer treatment, as an understanding of their pharmacology has enabled clinicians to exploit the benefits of these drugs (see below).

The very real limitations of cytotoxic chemotherapy have forced a concentration of cancer research on trying to understand the carcinogenic process, the aim being to identify specific molecular targets that can be exploited to develop novel therapeutic approaches. So-called targeted therapies are now well-established groups of anticancer drugs.

Classes of cytotoxic chemotherapy drugs

Cytotoxic chemotherapy drugs exert their effect by inhibiting cell proliferation. All proliferating cells, whether normal and malignant, cycle through a series of phases of: synthesis of DNA (S phase), mitosis(M phase) and rest (G₁phase). Non-cycling cells are quiescent in G₀ phase (Fig. 31.1).

Fig. 31.1 The cell cycle. Most cytotoxic drugs inhibit the processes of DNA replication or mitosis.

Cytotoxic drugs interfere with cell division at various points of the cell cycle, in particular G₁/S phase (e.g. synthesis of nucleotides from purines and pyrimidines), S phase (preventing DNA replication) and M phase (e.g. blocking the process of mitosis).

They are thus all potentially mutagenic. Cytotoxic drugs ultimately induce cell death by apoptosis,² a process by which single cells are removed from living tissue by being fragmented into membrane-bound particles and phagocytosed by other cells. This occurs without disturbing the architecture or function of the tissue, or eliciting an inflammatory response. The instructions for apoptosis are built into the cell's genetic material, i.e. ‘programmed cell death’.³

In general, cytotoxics are most effective against actively cycling cells and least effective against resting or quiescent cells. The latter are particularly problematic in that, although inactive, they retain the capacity to proliferate and may start cycling again after a completed course of chemotherapy, often leading later to rapid regrowth of the cancer.

Cytotoxic drugs can be classified as either:

• cell cycle non-specific: these kill cells whether they are resting or actively cycling (as in a low growth fraction cancer such as solid tumours), e.g. alkylating agents, doxorubicin and allied anthracyclines, or

• cell cycle (phase) specific: these kill only cells that are actively cycling, often because their site of action is confined to one phase of the cell cycle, e.g. antimetabolite drugs.

Table 31.2 provides a summary of the key groups of anticancer drugs, their common toxicities and main treatment applications.

Table 31.2 Principal classes of cytotoxic drug, their common toxicities and examples of clinical use

Drug class	Common toxicities	Examples of clinical use
Cytotoxic drugs
Alkylating agents	Nausea and vomiting, bone marrow depression (delayed with carmustine and lomustine), cystitis (cyclophosphamide, ifosfamide), pulmonary fibrosis (especially busulfan). Male infertility and premature menopause may occur. Myelodysplasia and secondary neoplasia	Widely used in the treatment of both haematological and non-haematological cancers, with varying degrees of success
Platinum drugs	Bone marrow depression, nausea and vomiting, allergy reaction (esp. carboplatin), nephrotoxicity, hypomagnesaemia; hypocalcaemia; hypokalaemia; hypophosphataemia; hyperuricaemia (all as a consequence of renal dysfunction, primarily associated with cisplatin); Raynaud's disease; sterility; teratogenesis; ototoxicity (cisplatin); peripheral neuropathy; cold dysaesthesia and pharyngolaryngeal dysaesthesia (oxaliplatin)	Testicular cancers, ovarian cancer; oxaliplatin acts synergistically with 5FU and is licensed in combination with 5FU to treat both advanced and early stages of colorectal cancer
Nucleoside analogues, e.g. cytarabine, gemcitabine, fludarabine	Bone marrow depression, mainly affecting platelets; mild nausea and vomiting; diarrhoea; anaphylaxis; sudden respiratory distress with high doses (cytarabine); rash, fluid retention and oedema; profound immunosuppression with fludarabine	Cytarabine is used in haematological regimens; gemcitabine is used for pancreatic cancer, bladder cancer and some other solid tumours; fludarabine is active in chronic lymphatic leukaemia and lymphoma
Taxanes	Nausea and vomiting, hypersensitivity reactions, bone marrow depression, fluid retention; peripheral neuropathy; alopecia; arthralgias; myalgias; cardiac toxicity; mild GI disturbances; mucositis	Breast and gynaecological cancers; recent evidence that docetaxel improves survival in advanced prostate cancer
Anthracyclines	Nausea and vomiting, bone marrow depression; cardiotoxicity (may be delayed for years); red-coloured urine; severe local tissue damage and necrosis on extravasation; alopecia; stomatitis; anorexia; conjunctivitis; acral (extremities) pigmentation; dermatitis in previously irradiated areas; hyperuricaemia	Common component of many chemotherapy regimens for both haematological and non-haematological malignancies
Antimetabolites, e.g. 5-fluorouracil, methotrexate	Nausea and vomiting; diarrhoea; mucositis, bone marrow depression, neurological defects, usually cerebellar; cardiac arrhythmias; angina pectoris, hyperpigmentation, hand–foot syndrome, conjunctivitis	Commonly used in haematological and non-haematological malignancies
Topoismerase I inhibitors	Nausea and vomiting; cholinergic syndrome; hypersensitivity reactions; bone marrow depression; diarrhoea; colitis; ileus; alopecia; renal impairment; teratogenic	Irinotecan is effective in advanced colorectal cancer; topotecan is used in gynaecological malignancies
Mitotic spindle inhibitors (vinca alkaloids)	Nausea and vomiting; local reaction and phlebitis with extravasation, neuropathy, bone marrow depression; alopecia; stomatitis; loss of deep tendon reflexes; jaw pain; muscle pain; paralytic ileus	Commonly used in haemato-oncology regimens
Hormones
Tamoxifen	Hot flushes; transiently increased bone or tumour pain; vaginal bleeding and discharge; rash; thromboembolism; endometrial cancer	Oestrogen receptor-positive, advanced and early stage breast cancer
Aromatase inhibitors	Nausea; dizziness; rash; bone marrow depression; fever; masculinisation	Equivalence with tamoxifen suggested
Medroxyprogesterone acetate	Menstrual changes; gynaecomastia; hot flushes; oedema, weight gain; hirsutism; insomnia; fatigue; depression; thrombophlebitis and thromboembolism; nausea; urticaria; headache	Third-line therapy for slowly progressive breast cancer in postmenopausal women
Flutamide	Nausea; diarrhoea; gynaecomastia; hepatotoxicity	Prostate cancer
Goserelin	Transient increase in bone pain and urethral obstruction in patients with metastatic prostatic cancer; hot flushes; impotence; testicular atrophy; gynaecomastia	Prostate cancer
Leuprolelin (LHRH analogue)	Transient increase in bone pain and ureteral obstruction in patients with metastatic prostatic cancer; hot flushes, impotence; testicular atrophy; gynaecomastia; peripheral oedema	Prostate cancer
Immunotherapy
BCG (bacille Calmette-Guérin)	Bladder irritation; nausea and vomiting; fever; sepsis, granulomatous pyelonephritis; hepatitis; urethral obstruction; epididymitis; renal abscess	Localised bladder cancer
Interferon-α	Fever; chills; myalgias; fatigue; headache; arthralgias, bone marrow depression; anorexia; confusion; depression; psychiatric disorders; renal toxicity; hepatic toxicity; rash	Renal cancer
Interleukin-2	Fever; fluid retention; hypotension; respiratory distress; rash; anaemia, thrombocytopenia; nausea and vomiting; diarrhoea, capillary leak syndrome, nephrotoxicity; myocardial toxicity; hepatotoxicity; erythema nodosum; neuropsychiatric disorders; hypothyroidism; nephrotic syndrome	Renal cancer
Trastuzumab (Herceptin)	Fever; chills; nausea and vomiting; pain; hypersensitivity and pulmonary reactions, bone marrow depression; cardiomyopathy; ventricular dysfunction; congestive cardiac failure; diarrhoea	Advanced and early stage breast cancer, combined with cytotoxic chemotherapy
Rituximab (MabThera)	Hypersensitivity reaction, bone marrow depression, angioedema, precipitation of angina or arrhythmia with pre-existing heart disease	Non-Hodgkin's lymphoma

Adverse effects of cytotoxic chemotherapy

Principal adverse effects are manifest as, or follow damage to, the following:

Nausea and vomiting

may occur within hours of treatment or be delayed, and last for several days, depending on the agent. As emetogenicity is largely predictable, preventive action can be taken. The most effective drugs are competitive antagonists of serotonin (5-hydroxytryptamine type 3, 5HT₃) receptors, e.g. ondansetron, and corticosteroids such as dexamethasone, which benefit by unknown, multi-factorial mechanisms. Other effective antiemetics include domperidone, metoclopramide, cyclizine and prochlorperazine (see p. 534). Combinations of drugs are frequently used and routes of administration selected as commonsense counsels, e.g. prophylaxis may be oral, but when vomiting occurs the parenteral route and suppositories are available.

Suppression of bone marrow and the lymphoreticular system

Myelosuppression with depression of both antibody- and cell-mediated immunity is the single most important dose-limiting factor with cytotoxic agents, and carries life-threatening consequences. Repeated blood monitoring is essential and transfusion of red cells and platelets may be necessary. Cell growth factors, e.g. the natural granulocyte colony-stimulating factor (filgrastim), are available to protect against or to resolve severe neutropenia.

Opportunistic infection

by Gram-negative bacteria from the patient's own flora, e.g. from the gut that has been damaged by chemotherapy, may occur. Infections with virus (herpes zoster), fungus (candida) and protozoa (pneumocystis) are also increased. Fever in a patient receiving chemotherapy usually requires immediate hospitalisation, collection of samples for microbiological studies and urgent empirical initiation of antibiotic treatment. Where risk of neutropenia is high, antimicrobial prophylaxis may be used. High-dose chemoradiotherapy and allogeneic bone marrow transplant produce profound immunosuppression with significant risk of opportunistic infection and third-party graft-versus-host disease following unirradiated blood transfusion. Live vaccines are contraindicated in these patients.

Diarrhoea and mouth ulcers

usually arise from drug damage to gut epithelium and other mucosal surfaces with a naturally rapid cell turnover.

Alopecia

is due to an effect on the hair bulb but is not invariable; it recovers 2–6 months after ceasing treatment. Scalp cooling may prevent or limit this with certain drugs, e.g. vinca alkaloids.

Urate nephropathy

is due to rapid destruction of malignant cells releasing purines and pyrimidines, which are metabolised to uric acid that may crystallise in and block the renal tubule (urate nephropathy). In practice this occurs only when there is a large cell mass or a tumour is very sensitive to drugs, e.g. acute leukaemias and high-grade lymphomas. High fluid intake, alkalinisation of the urine and use of allopurinol or rasburicas during the early stages of chemotherapy avert this outcome.

Local extravasation

may damage surrounding tissues; it is a problem with certain vesicant cytotoxics, e.g. doxorubicin, dacarbazine. This is a medical emergency and policies for management (which may include debridement by a plastic surgeon) should be in place in every centre.

Hypersensitivity reactions

may occur with susceptible patients. These are more problematic with certain cytotoxic agents, e.g. paclitaxel, carboplatin and the newer hybrid targeted monoclonal antibodies, for which prophylactic corticosteroid and antihistamine are offered routinely.

Specific organ damage

may result, e.g. lung toxicity with bleomycin, cardiotoxicity with anthracyclines, nephrotoxicity with platinum agents.

Delayed wound healing

can be expected. Surgical wounds should be healed prior to commencing chemotherapy, wherever possible.

Germ cells and reproduction

deserve special attention as chemotherapy may cause infertility. In addition, the theoretical mutagenic effects of cytotoxic drugs mean that reproduction is to be avoided during and for several months after therapy (although both men and women have reproduced normally while undergoing chemotherapy). When treatment may cause permanent sterility, men are offered the facility for prior storage of sperm. Cryopreservation of ovarian tissue is now also feasible. Prior contraceptive advice is necessary, as most cytotoxic drugs are teratogenic and are contraindicated during pregnancy.

Carcinogenicity

may result in delayed second malignancies, a potentially serious issue where treatment improves life expectancy. Many cytotoxic drugs are themselves carcinogenic, and a patient may be cured of the primary disease only to succumb to a second, treatment-induced, cancer 5–20 years later. Examples include patients with Hodgkin's lymphoma who are often young with high cure rates. Whether this is due to a mutagenic effect, to chronic immunosuppression, or to both, remains unclear. Alkylating agents and radiotherapy are particularly incriminated, as are some antimetabolites (mercaptopurine) and anthracyclines (doxorubicin). The relative risk can be as high as 10 to 20 times the normal risk. The second cancers caused include leukaemia, lymphoma and squamous carcinoma.

Classes of cytotoxic agents

Alkylating agents

Alkylating agents (nitrogen mustards and ethylenimines) act by transferring alkyl groups to DNA in the N-7 position of guanine during cell division. Normal synthesis is prevented because of either DNA strand breakage or cross-linking of the two strands. Examples include: busulfan, carmustine, chlorambucil, cyclophosphamide, ifosfamide, lomustine, melphalan, mustine (mechlorethamine), thiotepa, treosulfan.

Antimetabolites

Antimetabolites are synthetic analogues of normal metabolites and act by competition to ‘deceive’ or ‘defraud’ bodily processes.

Methotrexate, a folic acid antagonist, competitively inhibits dihydrofolate reductase, preventing the synthesis of tetrahydrofolic acid (the coenzyme that is important in synthesis of amino and nucleic acids). The drug also provides a cogent illustration of the need to exploit every possible means of enhancing selectivity. Where the desire is to maximise the effect of methotrexate, a potentially fatal dose is given, followed 24 h later by a dose of tetrahydrofolic (folinic) acid as calcium folinate (Ca Leucovorin) to bypass and terminate its action. This is called folinic acid ‘rescue’, because, if it is not given, the patient will die. The therapeutic justification for this manoeuvre is the cell kill obtained with very high plasma concentrations of methotrexate, allied to the fact that the bone marrow cells recover better than the tumour cells. The outcome is a useful degree of selectivity.

Pyrimidine antagonists: 5-fluorouracil (5FU) is metabolised intracellularly and its metabolite binds covalently with thymidilate synthase, thereby inhibiting DNA (and RNA) synthesis. 5FU has a short duration of action and addition of folinic acid in 5FU therapy improves its antitumour activity; protracted infusion can achieve the same outcome. Oral prodrugs of 5FU include capecitabine and UFT (a mixture of tegafur and uracil). These prodrugs have a cytotoxic action equivalent to that of 5FU but cause less myelosuppression and stomatitis; the risk of hand–foot syndrome (damage to the palmar and plantar surfaces of the hands and feet causing reddening, soreness and blistering) is considerably higher.

Arabinosides (cytosine arabinoside, gemcitabine) and the purine antagonists (deoxycoformycin, fludarabine, 2-chloroadenisine) azathioprine, mercaptopurine and tioguanine are also converted intracellularly to active metabolites that inhibit DNA synthesis.

Antimetabolites find extensive use in anticancer therapy, either alone or in combination with other drugs. They remain the mainstay of treatment for haematological as well as common solid tumours such as breast and gastrointestinal tract cancers.

Anthracyclines and related compounds

The original anthracyclines were antibiotics produced by microorganisms such as Streptomycetes spp. Daunorubicin and doxorubicin were the first compounds to be isolated and appear to interfere with both DNA and RNA synthesis. Other examples include bleomycin, dactinomycin, epirubicin, mitoxantrone, idarubicin, plicamycin (mithramycin), mitomycin and streptozotocin (most often used to treat the rare islet-cell pancreatic tumours).

Topoisomerase inhibitors

These inhibit enzymes essential for spatial conformation (‘unwinding’) prior to DNA replication, The epipodophyllotoxins (etoposide, teniposide) are major inhibitors of topoisomerase II. Topotecan and irinotecan selectively inhibit topoisomerase I and are effective in relapsed ovarian and colorectal cancer respectively. Bone marrow depression is dose limiting as, in the case of irinotecan, is delayed diarrhoea (which relates to a specific polymorphism in a drug metabolising gene). Administration of irinotecan is often complicated by an acute cholinergic reaction, reversible by prophylactic subcutaneous atropine. Doxorubicin is a non-specific inhibitor of topoisomerase I and II.

Spindle poisons

The plant alkaloids (vincristine, vinblastine, vindesine and vinorelbine) and taxoids (paclitaxel, docetaxel) inhibit microtubule assembly and cause cell cycle arrest in mitosis. They particularly cause bone marrow depression and alopecia. Vincristine causes neuropathy.

Platinum drugs

This family of drugs (which include cisplatin, carboplatin and oxaliplatin) act by cross-linking DNA in a similar manner to alkylating agents. The parent drug, cisplatin, is associated with a variety of adverse effects, including severe emesis, nephrotoxicity and ototoxicity. Renal damage is ameliorated by carefully pre-hydrating patients, and emetogenicity is effectively controlled with 5HT₃ receptor (serotonin) antagonists. Second- (carboplatin) and third- (oxaliplatin; no nephrotoxicity) generation platinum agents have improved toxicity profiles, by small alterations to the basic molecular structure, and offer effective treatment for germ cell, ovarian and colorectal cancers.

Miscellaneous agents

Asparaginase starves tumour cells dependent upon a supply of the amino acid, asparagine (except those able to synthesise it for themselves); its use is largely confined to acute lymphoblastic leukaemia.

Chemotherapy in clinical practice

Drug use and tumour cell kinetics

Evidence from leukaemia in laboratory animals shows that:

• Survival time is inversely related to the initial number of leukaemia cells, or to the number remaining after treatment.

• A single leukaemia cell is capable of multiplying and eventually killing the host.

Cytotoxic drugs

act against all multiplying cells. Bone marrow, mucosal surfaces (gut), hair follicles, reticuloendothelial system and germ cells all divide more rapidly than many cancer cells and are damaged by cytotoxic drugs, leading to the particular adverse effects of chemotherapy. In contrast to haematological cancers, most solid tumours in humans divide slowly and recovery from cytotoxic agents is slow, whereas normal marrow and gut recover rapidly. This speed of recovery of normal tissues is exploited in devising intermittent courses of chemotherapy.

In cancer, the normal feedback mechanisms that mediate cell growth are defective and cell proliferation continues unchecked, cancer cells multiplying, at first exponentially. Cancers with high growth fractions, e.g. acute leukaemias, high-grade lymphomas, may visibly enlarge at an alarming rate, but are frequently highly sensitive to cytotoxic chemotherapy. In later stages, the growth rate of these cancers often slows and the volume-doubling time lengthens due to several factors, most of which conspire to render the advanced cancer less susceptible to drugs, namely:

• Increased cell cycle (division) time.

• Decrease in the number of cells actively dividing, with more in the resting state (decrease in growth fraction), but with the potential to switch back to a fast growth state.

• Increased cell death within the tumour as it ages.

• Overcrowding of cells leading to necrotic, avascular and hypoxic areas that cannot easily be penetrated by drugs. These are fertile areas for clonal selection of the most robust cancer cells.

Selectivity of drugs for cancer cells is generally low compared with the selectivity shown by antimicrobial agents but it can be substantial, e.g. in lymphoma, where tumour cell kill with some drugs is 10 000 times greater than that of marrow cells. Cell destruction by cytotoxic drugs follows first-order kinetics, i.e. a given dose of drug kills a constant fraction of cells (not a constant number) regardless of the number of cells present. Thus a treatment that reduces a cell population from 1 000 000 to 10 000 (a two-log cell kill) will reduce a cell population of 1000 to 10. Furthermore, cell chemosensitivity within a cancer is not homogeneous owing to random mutations (clonal selection) as the tumour grows, the cells remaining after initial doses being more likely to resist further treatment. Therefore, combining several drugs may be more effective than a single agent given repeatedly to the limit of tolerance.

The selection of drugs

in combination chemotherapy is influenced by:

• Choosing drugs that act at different biochemical sites in the cell.

• Using drugs that attack cells at different phases of the growth cycle (see Fig. 31.1). ‘CHOP’ (cyclophosphamide, doxorubicin (previously known as hydroxydoxorubicin) vincristine (previously called oncovin) and prednisolone), is a standard combination chemotherapy regimen for non-Hodgkin's lymphoma. The first three cytotoxic drugs exert their antitumour effect on different aspects of cell proliferation. The antitumour effect of corticosteroid remains unclear.

• The desirability of attaining synchronisation of cell cycling to achieve maximum cell kill. Cells are killed or are arrested in mitosis by vincristine, which is then withdrawn. Cells then enter a new reproductive cycle more or less synchronously, and when most are judged to be in a phase sensitive to a particular phase-specific drug, e.g. methotrexate or cytarabine, it is given.

• Avoidance of cross-resistance (see below) between drugs. In some instances, use of one drug regimen followed by another rather than using them simultaneously in combination avoids drug resistance and improves therapeutic efficacy. For example, epirubicin given for four cycles followed by CMF (concomitant cyclophosphamide, methotrexate and 5-fluorouracil) for four cycles has largely replaced CMF alone as standard adjuvant chemotherapy for breast cancer, because the outcome is better.

• Non-overlapping toxicity profiles. Before establishing a combination regimen, Phase 1 trials (see p. 40) are undertaken, frequently fixing the dose of one drug while escalating the dose of another, in small cohorts of carefully monitored patients, so that toxicity and patient safety can be monitored.

• Empirical evidence of efficacy against a particular tumour type. The antitumour activity of platinum complexes was a chance finding (see below).

• Enhanced cell killing in preclinical models when drugs are combined. Oxaliplatin on its own has limited cytotoxicity against colorectal cancer cell lines in vitro and in mouse xenograft models, but its combination with 5FU confers a more than additive, i.e. synergistic, killing effect on tumour cells.

Considerations of pharmacokinetics in relation to cell kinetics are of great importance, as drug treatment alters the behaviour of both malignant and normal cells.

Drug resistance

Resistance

to a cytotoxic chemotherapy agent may be present at the outset (primary resistance), or may develop with repeated drug exposure (acquired resistance). Increasing dosage is limited by toxicity, e.g. to bone marrow, which does not become tolerant. Combination chemotherapy is a strategy commonly used to address the problems of tumour resistance.

Multiple drug resistance

(MDR) of a cancer is not uncommon. MDR is most frequently due to increased expression of an ATP-dependent membrane efflux pump called P-glycoprotein (Pgp), which is a member of a class of membrane proteins called the ATP-binding cassette superfamily. Pgp is an important protective mechanism possessed by many normal cells against environmental toxins and has broad specificity for hydrophobic compounds. Long-lived cells such as the haemopoietic stem cell, cells on excretory surfaces such as biliary hepatocytes, proximal renal tubule and intestinal cells, and the cells of the blood–brain barrier all have high expression of Pgp. A number of agents including immunosuppressants (ciclosporin) and calcium channel blockers (verapamil and nifedipine) block Pgp in theory.

The MDR phenomenon illustrates how tumour cells adapt and enhance normal cell mechanisms to deal with the effects of chemotherapy, and how repeated cycles of chemotherapy select out a population of cells that have developed adaptive survival mechanisms, e.g. in myeloma where MDR proteins are rare at diagnosis but common at progression.

In those tumours for which cures can be achieved by chemotherapy (childhood acute lymphoblastic leukaemia, Hodgkin's lymphoma, choriocarcinoma) it is essential that optimal doses of initial chemotherapy be administered and dose intensity maintained in order to avoid the emergence of chemoresistance.

Improving efficacy of chemotherapy

Methods that potentially widen the narrow therapeutic index of cytotoxic agents include:

• Regional (as opposed to systemic) administration of drugs: intrathecal, intra-arterial into liver and isolated limb perfusion.

• Selective organ delivery of drug by altered formulation e.g. Caelyx is a formulation comprising high concentrations of doxorubicin encased in liposomes. This also alters toxicity profiles (e.g. less renal but more mucosal toxicity).

• High-dose (bone marrow ablative) chemotherapy is feasible by harvesting stem cells prior to drug exposure, and returning the cells to the patient on completion of treatment. This is a successful strategy in haematological malignancies but failed trials in almost all solid tumours.

• Circadian rhythms exist in cell metabolism and proliferation, and those of leukaemic cells differ from normal leucocytes. The time of day at which therapy is administered can theoretically influence the outcome (chronomodulation).

• In large solid tumours, the fraction of cells multiplying rapidly is often relatively small. In ovarian cancer, for example, patients benefit from debulking surgery (cytoreduction) prior to cytotoxic drug therapy.

Hazards to staff handling cytotoxic agents

Urine from nurses and pharmacists who prepared infusions and injections of anticancer agents revealed drugs in concentrations that were mutagenic to bacteria. When they stopped handling the drugs, the contamination ceased. It can be assumed that absorption of even small amounts of these drugs is harmful (mutagenesis, carcinogenesis), especially repeatedly over long periods. Pregnant staff should not handle these drugs.

A note of caution

Certain chemotherapy regimens require the simultaneous administration of intrathecal methotrexate and intravenous vincristine. In the UK until recently, each drug was presented in similar bolus volumes and the drug-filled syringes appeared very alike except that the syringe for intrathecal administration had a red stopper. Nevertheless, from 1985 in the UK inadvertent intrathecal administration of vincristineoccurred on 14 occasions: 10 patients died and the remainder suffered paralysis.⁴

Interactions of anticancer agents with other drugs

The diverse modes of action of cytotoxic drugs offer ample scope for serious unwanted drug–drug interactions, and by different mechanisms. There is general cause for alertness. Drugs that inhibit enzymes and thus delay normal metabolic breakdown may cause harmful reactions to standard doses of cytotoxics, e.g. allopurinol (xanthine oxidase inhibitor) with mercaptopurine or cyclophosphamide. Enzyme-inducing drugs can reduce the therapeutic efficacy of anticancer drugs by accelerating metabolism. Competition with non-steroidal anti-inflammatory drugs (NSAIDs) reduces the renal tubular excretion of methotrexate, leading to methotrexate toxicity. A combination of cytotoxics causing a dangerous degree of immunosuppression represents an adverse pharmacodynamic reaction.

Therapeutic drug–drug interactions are an essential part of treatment, as witnessed by the many drug combinations used to treat cancer (see Drug use and tumour cell kinetics, p. 515).

Endocrine therapy

Hormonal influence on cancer

The possibility of interfering with cancer other than by surgery, e.g. by endocrine manipulation, was first tested in 1895 when a Scottish surgeon, faced with a woman aged 33 years with advanced breast cancer:

put it to her husband and herself as to whether she should have performed the operation of removal of the [fallopian] tubes and ovaries. Its nature was fully explained to them both, and also that it was a purely experimental one … She readily consented … as she knew and felt her case was hopeless. [Eight months after operation] all vestiges of her previous cancerous disease had disappeared. [The surgeon concluded, after treating two further cases, that there may be ovarian influences in breast cancer and added that] whether [this is] accepted or not, I am sure I shall be acquitted of having acted thoughtlessly or recklessly.⁵

The treatment had logic. The author, observing the weaning of lambs on a local farm, had noted a similarity between the proliferation of epithelial cells of the milk ducts in lactation and in cancer, and had conceived the idea that cancer of the breast might be due to an abnormal ovarian stimulus.

In 1941⁶ it was shown that prostatic cancer with metastases was made worse by androgen and made better by oestrogen (diethylstilbestrol).

Hormonal agents

The growth of some cancers is hormone dependent and is inhibited by surgical removal of gonads, adrenals and/or pituitary. The same effect is achievable, at less cost to the patient, by administering hormones, or hormone antagonists, of oestrogens, androgens or progestogens and inhibitors of hormone synthesis.

Breast cancer

cells may have receptors for oestrogen, progesterone and androgen, and hormonal manipulation benefits some 30% of patients with metastatic disease; when a patient's tumour is oestrogen-receptor positive the response is about 60%, and when negative it is only 10%. After treatment of the primary cancer, endocrine therapy with the anti-oestrogen, tamoxifen, is the adjuvant therapy of choice for postmenopausal women who have disease in the lymph nodes; both the interval before the development of metastases and overall survival are increased. Adjuvant therapy with cytotoxic drugs and/or tamoxifen is recommended for node-negative patients with large tumours or other adverse prognostic factors. Cytotoxic chemotherapy is more useful in younger women, with tamoxifen, increasingly, as adjuvant therapy. The optimal duration of dosing with tamoxifen is not yet established, but is likely to be 5 years or more.

Aromatase inhibitors cause ‘medical adrenalectomy’ in postmenopausal women by blocking conversion of adrenal androgens to oestrogens in peripheral fat by the enzyme aromatase. The first drug in this class, aminoglutethimide, causes significant adverse effects. More selective and less toxic aromatase inhibitors now include anastrozole, letrozole and exemestane, and find use after treatment with tamoxifen fails. Clinical trial data suggest that these drugs rival tamoxifen in efficacy for both advanced and early breast cancer. Progestogens, e.g. megestrol or medroxyprogesterone, are third-line agents in postmenopausal women.

Prostatic cancer

is androgen dependent and metastatic disease can be helped by orchidectomy, or by pituitary suppression of androgen secretion with a gonadorelin (LHRH) analogue, e.g. buserelin, goserelin, leuprorelin or triptorelin (see p. 615). These cause a transient stimulation of luteinising hormone and thus testosterone release, before inhibition occurs; some patients may experience exacerbation of tumour effects, e.g. bone pain, spinal cord compression. Where this can be anticipated, prior orchidectomy or anti-androgen treatment, e.g. with cyproterone or flutamide, is protective.

Benign prostatic hypertrophy

is also androgen dependent, and drug therapy includes use of finasteride, an inhibitor of the enzyme (5α-reductase) that activates testosterone.

Adrenocortical steroids

are used for their action on specific cancers and also to treat some of the complications of cancer, e.g. hypercalcaemia, raised intracranial pressure. In leukaemias, corticosteroid may reduce the incidence of complications, e.g. haemolytic anaemia and thrombocytopenia. A glucocorticoid is preferred, e.g. prednisolone, as doses are high, and the mineralocorticoid actions are not needed and cause fluid retention.

In general,

endocrine therapy carries less acute serious consequences for normal tissues than do cytotoxic agents. In a sense, they represent the first generation of mechanism-driven targeted agents used to treat cancer. As cancer patients live longer, the chronic effects of hormonal therapies are becoming more evident, such as a higher incidence of endometrial cancer with chronic use of tamoxifen in a minority of patients and osteoporosis in breast and prostate cancer patients who have had anti-oestrogens and/or anti-androgens including gonadorelin (LHRH) analogues.

Immunotherapy

Immunotherapy (immunostimulation) derives from an observation in the 19th century that cancer sometimes regressed after acute bacterial infections, i.e. in response to non-specific immunostimulant effect. In general, it appears that the immune response is attenuated in cancer. Strategies to stimulate the host's own immune system to kill cancer cells more effectively are:

• Non-specific stimulation of active immunity with vaccines, e.g. BCG (bacille Calmette–Guérin⁷) instilled into the urinary bladder for bladder cancer. Other approaches involve the injection of tumour cells or tumour cell extracts combined with an immune stimulant such as BCG.

• Specific immunisation strategies, where tumour-specific and tumour-associated antigens have been identified. Melanomas, for example, possess melanoma differentiation antigens (tyrosinase, gp100, MART1) as well as tumour-associated antigens (MAGE, BAGE, GAGE series of major histocompatibility complex (MHC)-associated peptides and a family of lipoproteins known as gangliosides). Both DNA and whole-protein vaccines derived from these antigens are being evaluated in melanoma, but results to date have been clinically disappointing.

Naturally occurring substances are increasingly used to treat cancer. Cytokines are produced in response to various stimuli, such as antigens, e.g. viruses. These peptides regulate cell growth, activation and differentiation, and immune responses and can be synthesised by recombinant DNA technology. Examples include:

• Interleukins that stimulate proliferation of T lymphocytes and activate natural killer cells; interleukin-2 is used in metastatic renal cell carcinoma with marginal effectiveness.

• Interferons. Interferon-α is used for chronic granulocytic leukaemia, hairy cell leukaemia, renal cell carcinoma and Kaposi's sarcoma.

Thalidomide was withdrawn in the 1960s following evidence of its teratogenic effects, including some on fetal limb development (see p. 63). These very effects prompted the notion that suppression of cell proliferation might actually provide benefit. Investigation revealed that thalidomide possessed immunomodulatory properties, anti-inflammatory actions, direct effects on tumour cells and their microenvironment, and actions on angiogenesis (see below). Thalidomide and analogues designed to reduce toxicity (immunomodulatory drugs: lenalidomide) have a therapeutic role in myeloma and are synergistic with dexamethasone and chemotherapeutic agents.

ATRA

All-trans-retinoic acid (ATRA) induces remission in newly diagnosed patients with promyelocytic leukaemia (APL), by leukaemic cell differentiation. APL is due to reciprocal translocation between chromosomes 15 and 17 producing a fusion gene PML–RARa. The fusion protein blocks differentiation but is overcome by ATRA. Subsequent administration of anthracycline improves cure rates.

Development of anticancer drug therapy

In general, anticancer drugs develop from:

• Chance discovery: cisplatin. In the 1960s, scientists studying the effect of an electric current on bacteria cultured in a Petri dish noted that the cells stopped dividing, instead forming long filamentous structures. Further investigation revealed that the inhibitor of cell division was in fact an ion formed in solution from the platinum electrodes used in the experiment. The platinum complex, cis-diammine-dichloroplatinum (II), later known as cisplatin, was isolated and subsequently developed for its potential to kill cancer cells. When given to patients with a variety of different types of cancer, germ cell tumours in particular were found to possess remarkable sensitivity to cisplatin treatment, and this drug remains in use for treating such patients today.⁸

• Analogues. Severe vomiting, renal and nerve damage, and deafness limit the therapeutic efficacy of cisplatin. Carboplatin and oxaliplatin, second- and third-generation compounds derived from cisplatin, combine enhanced toxicity towards cancer cells with improved tolerance.

• Mass screening programmes. See Chapter 3.

• Rational drug design. Academic institutions and commercial biotechnology companies involved in experimental therapeutics study the cancer process to identify key (‘target’) genes or gene products that regulate aspects of carcinogenesis and then try to find ways of blocking the function of these targets (Table 31.3). Unlike conventional cytotoxic chemotherapy, many of these agents are thus more cancer-cell selective and thus cytostatic. In other words, a targeted biological agent may prevent tumour growth or progression and delay recurrence, but may not induce rapid tumour shrinkage, hitherto the key conventional endpoint for evaluating cytotoxic drugs. Some examples follow to illustrate the opportunities created by this type of approach.

Table 31.3 Some novel molecular targets being exploited in anticancer drug development^a

Target	Drug^b	Examples of current clinical use
Her2/neu	Trastuzumab (Herceptin)	Advanced and early stage breast cancer Advanced gastric cancer
CD20	Rituximab (MabThera)	Non-Hodgkin's lymphoma
EGFR	Cetuximab (Erbitux)	Improves survival in advanced colorectal cancer Lung cancer and head and neck cancer (in combination with radiotherapy)
EGFR	Gefitinib (Iressa)	Advanced non-small cell lung cancer
EGFR	Erlotinib (Tarceva)	Advanced non-small cell lung cancer, advanced pancreatic cancer
VEGF	Bevacizumab (Avastin)	Advanced colorectal cancer; trials confirm some efficacy in non-small cell lung cancer and renal cancer
Bcr-abl	Imatinib (Glivec)	Chronic myeloid leukaemia
c-kit	Imatinib (Glivec)	Gastrointestinal stromal tumours
Raf/MAPK	Sorafenib	Hepatocellular and renal cancer Clinical trials ongoing in breast cancer, GIST and melanoma
Cyclin-dependent kinase	Flavopiridol	Undergoing clinical trials
mTor (a key regulator of cell cycle progression)	Everolimus and temsirolimus	Renal Pancreatic neuroendocrine tumours
Proteasome inhibitor	Bortezomib	Active in myeloma; trials of combination therapy in progress

a Drugs at various stages in the process of obtaining a licence for use in the UK.

b The suffix ‘mab’ identifies a monoclonal antibody, whereas ‘nib’ identifies a tyrosine kinase inhibitor.

Targeted biological therapies

Passive immunotherapy using monoclonal antibodies raised against specific tumour-associated antigens on the cell surface

Targeted antibodies have the advantage of high cancer specificity and relatively low host toxicity.

• Rituximab, an anti-CD20 monoclonal antibody, for the treatment of low-grade follicular lymphomas and for use in combination with CHOP (see above) for high-grade lymphoma, as these tumours carry the antigen CD20 on the cell surface.

• Significant over-expression of the Her2/neu (erbB2) cell surface growth factor receptor occurs in approximately 20% of breast cancers and gastric cancers and is associated with a far more aggressive form of breast cancer compared with non-Her2-expressing tumours. Trastuzumab (Herceptin), a humanised monocolonal antibody, binds specifically to the Her2/neu receptor, blocking its function in regulating intracellular processes, including cell proliferation. In combination with conventional cytotoxic chemotherapy, trastuzumab significantly improves the survival of patients with advanced or early breast cancer, compared with cytotoxic chemotherapy alone. A series of key adjuvant trials conducted across Europe and the USA showed that trastuzumab combined with chemotherapy provided the biggest survival gains ever recorded for this disease.⁹^,¹⁰^,¹¹ Potential cumulative cardiac dysfunction with trastuzumab is dose limiting.

• Her2 is a member of the epidermal growth factor receptor (EGFR) family. EGFRs are highly expressed by about 85% of colorectal cancers and are important in regulating cell proliferation. Another monoclonal antibody, cetuximab(Erbitux), blocks EGFR function and is used for selected colorectal cancers. It was subsequently found that this drug had no benefit in approximately 40% of all colorectal cancers with a mutation in the K-RAS oncogene. The mutation constitutively activates the cell pathway downstream of the EGFR receptor, so blocking it with cetuximab has no effect. This is an important example of personalising medicine to the individual cancer type and selecting which agents are suitable in individual cancers. It is also crucial in the design and selection of patients for future evolutions of this therapy in colorectal cancer, as patients with K-RAS mutations would only get the unwanted effects and absolutely no benefit if treated with EGFR inhibitors. Limiting toxicities for this class of drugs are fatigue, rash, mucositis and hypomagnesaemia.

• Vasculoendothelial growth factor (VEGF) is a major angiogenic signal regulator for new blood vessel formation (angiogenesis). Angiogenesis, a process that is common to all cancers, is vital for the growth and establishment of secondary tumours to grow beyond 1–2 mm when diffusion of nutrients becomes insufficient to maintain tumour growth. Blockade of VEGF and its receptor is a successful strategy for treating several types of neoplasm. The monoclonal antibody bevacizumab (Avastin) improves survival or slows tumour growth significantly, when combined with cytotoxic chemotherapy for advanced colorectal, lung and breast cancers. This novel approach evokes a range of adverse drug reactions that differ from those of conventional cytotoxics: hypertension, proteinuria, bleeding, and increased risk of thromboembolic events.

Similar drugs that block effects of VEGF are now used to treat neovascular age-related wet macular degeneration, a condition that causes loss of central vision in 2.3% of people > 65 years.

Radioimmunotherapy

Monoclonal antibodies targeted against epitopes¹² on tumour cells, e.g. rituximab against CD20 in lymphoma, are conjugated to radionuclides such as yttrium-90 (ibritumomab) or iodine-131 (tositumomab) to deliver radiation directly to the cellular target; they produce durable responses in patients resistant to chemotherapy and unconjugated antibody.

Chemo-immunotherapy

Monoclonal antibodies conjugated to toxins deliver high concentrations of agents that are too toxic to give systemically, e.g. CD33 plus calicheamicin (Gemtuzumab ozogamicin) in AML.

Most therapeutic antibodies are monoclonal immunoglobulin (Ig) G antibodies produced in mammalian cell lines by recombinant DNA technology. Genetic engineering alters the molecular structure of key immunogenic portions of the antibody to generate ‘humanised’ chimeric¹³ antibodies that avoid rejection by the human immune system (but hypersensitivity reactions may occur).

Signal transduction inhibitors

Tyrosine kinase activation of cell surface receptors and their downstream proteins is an important mechanism by which messages are translated to the nucleus to affect cell function. A family of small molecules called tyrosine kinase inhibitors (TKIs) is now showing significant promise as anticancer agents. These small molecules are often orally administered. Multi-targeted kinase inhibitors are also attractive, as they may possess a wide spectrum of antitumour activity, but their potential for toxicity is a real concern.

• Imatinib (Glivec, Gleevec) blocks the dysregulated tyrosine kinase hyperactivity produced by the Philadelphia chromosome (bcr-abl) that occurs in chronic myeloid leukaemia and some cases of acute lymphoblastic leukaemia; clinical trials support its therapeutic efficacy. This is an important example of a drug designed precisely to address the biological abnormality that causes a disease. By serendipity, ii imatinib also blocks an oncogenelled c-kit (CD-117) and has now revolutionised treatment of a rare cancer called GIST (gastrointestinal stromal tumour) that hitherto had no known systemic treatment options (both chemo- and radiotherapy primary resistant), and frequently carries a c-kit mutation.

• Tyrosine kinase inhibitors of VEGF receptor and its downstream effector pathways are also in clinical trial (see Fig. 31.2 and Table 31.3).

• The EGFR tyrosine kinase inhibitors, gefitinib (Iressa) and erlotinib (Tarceva), now find use for a variety of EGFR-expressing tumours, specifically in lung cancer, that have a mutation in the EGFR gene, with marked benefit, although cure remains elusive.

Fig. 31.2 Many new drugs target the vasculoendothelial growth factor (VEGF) pathway.

Targeting the cell cycle

Recent advances in molecular biology have shown that the cell cycle is regulated by a series of proteins that include cyclins, cyclin-dependent kinases and cyclin-dependent kinase inhibitors (Fig. 31.3). Aberrations in these proteins are implicated in uncontrolled progression through the cell cycle (and hence in carcinogenesis), but also represent a new set of therapeutic targets for anticancer therapy, e.g. flavopiridol, which has cyclin-dependent kinase inhibitor activity, and rapamycin, which inhibits mTor (mammalian target of rapamycin), another key regulator of cell cycle progression. Arsenic trioxide modulates cell growth and differentiation, and induces remission in relapsed refractory acute promyelocytic leukaemia in part through apoptosis induction and down-regulation of Bcl-2.

Fig. 31.3 The cell cycle is regulated by a series of proteins called cyclins, cyclin-dependent kinases (CDKs) and cyclin-dependent kinase inhibitors (CDKIs). Many CDKIs appear to be tumour-suppressing genes. These moieties are potential targets for anticancer therapy. MAPK, mitogen-activated protein kinase.

Protease inhibition

The ubiquitin–proteasome pathway is an intracellular proteolytic system that degrades cyclins and cyclin-dependent kinase inhibitors which regulate cell cycle progression. Bortezomib inhibits proteasome activity and, in myeloma, prevents degradation of nuclear factor κB inhibitor (IκB), resulting in a directly apoptotic effect, antiangiogenesis and inhibition of myeloma–stromal cell interaction. It has single-agent activity and restores chemosensitivity in resistant cells.

Chemoprevention of cancer

Because many cancers are currently incurable once metastasised, cancer prevention is a logical objective. Individuals can change aspects of their lifestyle significantly to influence their risk of developing particular cancers. Ceasing to smoke tobacco is an obvious health benefit. The individual benefits attributable to other changes are more difficult to quantify. The connection between environment and cancer risk is complex and as yet poorly understood, as genetic susceptibilities may obviate or accentuate certain risks.

Chemical interventions to reduce cancer risk may be an option for the population as a whole, or for groups at high risk of a specific cancer. Retrospective epidemiological and association studies suggest large effects of certain dietary manipulation but prospective interventional studies have rarely been confirmatory and have even shown harm. The best example is supplemental vitamins, derivatives and dietary micronutrients that may inhibit the development of cancers in the laboratory, e.g. β-carotene, isotretinoin, folic acid, ascorbic acid, α-tocopherol. In a trial of antioxidant supplementation with ascorbic acid, vitamin E, β-carotene, selenium and zinc over 7.5 years, the total cancer incidence was lower in men (who had a lower baseline antioxidant status) than in women.¹⁴ Subsequent large-scale randomised trials with supplementation have demonstrated the opposite effect with an unexpectedly higher incidence of new cancers.

The anti-oestrogen tamoxifen, used as an adjuvant therapy in women undergoing surgery for primary breast cancer, reduced the risk of cancer occurring in the contralateral breast. Tamoxifen and anastrozole (see above) are undergoing assessment for chemoprevention in women at high risk of breast cancer.