Cancer is essentially the abnormal, out of control growth of cells within the body. Leukaemia results if the problem is in the bone marrow; elsewhere, it leads to a solid mass of tumour. If the tumours grow and spread, then they are termed malignant and are likely to cause serious problems.

As a tumour grows, it develops a blood supply and lymphatic system, through which tumour cells can be carried around the body. If they are not destroyed by the immune system, then they can begin to metastasise and develop in other organs. Where once cancer chemotherapy meant simply the use of highly toxic substances to kill the cancer cells, hopefully without killing too many healthy cells in the process, more recently many different approaches have been developed, which attack at different points in the tumour's lifecycle.

Cytotoxic drugs act at one point or another in the cell cycle - the process by which cells divide. The cycle lasts 16 to 18 hours, and has four phases. In the G1 phase, cells create the nucleoside building blocks for making DNA. Next, in the S phase, the nucleosides are put together to create an exact copy of the parent DNA. The G2 phase sees other structures, such as microtubules, being formed. Finally, in the M phase, the two sets of chromosomes pull apart, form into nuclei, and the two new cells separate. Some will divide again almost immediately, but most will enter a resting phase, lasting perhaps a few hours, or even indefinitely. Various points in this cycle can be attacked by drugs to prevent the out of control cell division.

Currently available anticancer agents essentially fall into six groups: alkylating agents, antimetabolites, topoisomerase inhibitors, antitumour antibiotics, microtubule inhibitors, and drugs that act through hormone dependent mechanisms.

Some of the earliest anticancer agents are alkylating agents, which bond covalently to a cell's DNA, stopping replication. Older drugs include busulphan (GSK), carmustine (BMS), chlorambucil (GSK), isophosphamide and cyclophosphamide. More recently, a range of platinum-based compounds were introduced, including oxaliplatin (Sanofi-Synthélabo) and carboplatin (BMS).

Newer compounds in this class include bizelesin (Pharmacia/National Cancer Institute), which is in Phase I trials and looks promising against various forms of leukaemias and several different solid tumours. Temozolomide (Schering-Plough) is used in the treatment of primary brain cancers, and is now being investigated in brain metastases and, in combination with other drugs, for various other cancers. Another promising drug, ZD0473 (AstraZeneca) is related to the platins, and is in Phase II trials for resistant ovarian tumours.

The second class of drugs, antimetabolites, interfere with the G1 stage of the cell cycle. Again, many have been in use for years, including 5-fluorouracil (ICN), hydroxyurea (BMS) and methotrexate (Pharmacia). Nucleoside analogues, such as fludarabine (Schering Health Care) and mercaptopurine and thioguanine (GSK) work by mimicking the natural nucleosides used to build DNA. Gemcitabine is another drug with a mimic effect: it is incorporated into the new DNA chain and stops its production.

A newer drug from AstraZeneca, raltitrexed, is used as a last resort in colorectal cancer, and a modified version, ZD9331, has entered Phase II trials in a wide range of solid tumours. A related compound from Lilly, permetrexed, shows promise against a variety of different tumours, and is now in Phase III for the difficult to treat lung cancer mesothelioma.

Topoisomerase inhibitors interfere with the unwinding of DNA strands during the S phase of the cell cycle. The enzymes topoisomerase I and II initiate this process. Topotecan (GSK/Merck) and irinotecan (Pharmacia) inhibit topoisomerase I; two inhibitors of the second enzyme are doxorubicin Schering-Plough) and etoposide (BMS). An investigational drug from Xenova, coded XR11576 and currently undergoing preclinical work, inhibits both enzymes. It appears to have good activity against multiply drug resistant tumours, as well as being orally available.

Microtubules are formed during the G2 and M phases of the cell cycle and, rather than acting directly on the cancer cells' DNA, they prevent the formation of the microtubules that are an essential part of the process of separation of the chromosomes. Several vinca alkaloids, derived from an extract of the periwinkle plant Vinca rosea, have this effect and are on the market, including vinblastine, vincristine and vindesin from Lilly, and Pierre Fabre's vinorelbine. Also plant-derived are the taxoid molecules, which are based on an extract from the Pacific yew tree Taxus brevifolia, and notably include paclitaxel (BMS) and docetaxel (Aventis).

Both classes of drugs are extremely expensive to extract from the natural source, as they are only present in extremely low quantities, so various new derivatives that are easier to make are being investigated. These include two more being looked at by BMS, coded BMS-184476 and BMS-188797, which appear promising against some resistant cell lines. Aventis and Elan are also looking at new taxoids. A newer class of microtubule inhibitor, the epithilones, are being looked at by both BMS and Novartis. BMS's semisynthetic BMS-247550 is now in Phase II trials, having been shown in the laboratory to have twice the activity of paclitaxel in cell cultures.

Another part of the cell cycle that can be targeted is the signalling system that regulates the process. AstraZeneca's ZD1839, for example, which is in Phase III trials, is able to block these pathways in calls with epidermal growth factor on their surfaces, such as non-small cell lung cancer. Another compound with a similar action PKI 166 from Novartis, is in Phase II. Recently launched by Novartis, imatinib (acts to block the signal that cause tumours to develop in chronic myeloid leukaemia. And Lilly's ISIS stops signals that trigger cell division, and is in Phase III.

Because both breast and prostate cancers are dependent on the sex hormones, oestrogen in women and testosterone in men, they can be treated indirectly by blocking the action of the hormone in one way or another. Receptor antagonists - antiandrogens for testosterone receptors in the prostate, and antioestrogens for oestrogen receptors in the breast - are one way of attacking the cancer. Three antiandrogens are currently available: bicalutamide (AstraZeneca), flutamide (Chiron) and cyproterone acetate (Schering Health Care). Bristol-Myers Squibb markets the antioestrogen megestrol acetate. More recent work on bicalutamide suggests that it significantly reduces the chances of early prostate cancer progressing.

Probably the best known treatment for breast cancer, AstraZeneca's tamoxifen, is a selective oestrogen receptor modulator, or SERM. These drugs selectively antagonise the oestrogen receptors on cancer cells, so they can only be used on cancer cells that have the hormone receptors. However, as well as being used as a therapy, tamoxifen has been found to have a preventive effect in women at high risk of developing breast cancer. It has been in use for nearly 30 years now, and a number of companies are looking at second generation SERMs which have less of an oestrogen-like activity, including Lilly's raloxifene and arzoxifene, and lasofoxifene from Pfizer. It is thought these compounds may also have some effect in prostate cancer, but nothing has yet been proved definitively.

There are more indirect ways of blocking the effect of the sex hormones. Oestrogen is made in the ovaries, and testosterone mostly in the testes, and to some extent in the adrenal gland. A range of hormones made in the pituitary gland regulate their release, so inhibiting the production of these release hormones would have a knock on effect on oestrogen and testosterone. The three hormones responsible are follicle stimulating hormone, which acts on the ovaries; luteinising hormone, on the testes; and adrenocorticotropic hormone, which acts on the adrenal gland. Luteinising hormone releasing hormone (LHRH) stimulates the release of these three in the brain, and LHRH agonists block their production. Several are now available, including AstraZeneca's goserelin, triptorelin (Ipsen), leuprorelin (Wyeth) and Shire's buserelin. Of these, only goserelin is licensed for use in women, and it is often administered in combination with antiandrogens in men.

A more recent class of anticancer drugs that act on hormone dependent cancers are the aromatase inhibitors. These directly stop the body from producing the sex hormones, and available drugs include exemestane (Pharmacia), anastrozole (AstraZeneca) and two compounds from Novartis, formestane and letrozole. Several others are in development, including Yamanouchi's YM511. Another investigational approach is to reduce the number of sex hormone cells the cancer cells produce, and one compound with this action, AstraZeneca's ICI 182,780 is in Phase III trials in post-menopausal women whose advanced breast cancer is resistant to tamoxifen.

Cancer is a particularly fruitful area for imaginative new pharmaceutical strategies, as there are so many points in the cancer's growth cycle that can be interrupted, whether directly or by affecting a related process. Farnesyl protein transferase, for example, is involved in the growth of solid tumours, and several companies, notably Janssen-Cilag, Schering-Plough and Bristol-Myers Squibb, are looking at inhibitors of the enzyme. Janssen-Cilag and BMS have compounds in Phase I (BMS-214662 and R115777 respectively), and Schering-Plough has a compound in Phase II.

Sanofi-Synthélabo is developing a drug that, unusually, can attack cells in the resting phase after division. The compound, tirapazamine, is converted by an enzyme within resting cells into an activated form that is able to react with the DNA strands within. It is now in Phase II/III trials in combination with radiotherapy and platinum drugs.

Large tumours have to generate their own blood supply in order to continue growing, by the process of angiogenesis. They use chemical messengers like vascular endothelial growth factor (VEGF) to create offshoots from existing capillaries nearby. Several compounds that affect this process are under development. The endothelin receptors are a potential target, and Abbott's atrasentan and YM 598 from Yamanouchi both act at endothelin receptors, and are being investigated in prostate cancer.

PTK787 from Novartis prevents the release of chemical messengers after the VEGF receptors are stimulated, and trials showed it reduced tumour size and tendency to metastasise It is now in Phase II. Similarly, TAP Pharmaceuticals' TNP-470, also in Phase II, has been shown to shrink small tumours. And Pharmacia has two VEGF inhibitors in clinical trials.

BMS and OxiGene are looking at combretastatin, isolated from the South African bush willow Combretum caffrum, which affects the capillaries around the tumour, and trials have shown it to reduce blood flow by as much as 95%. Aventis and Ajinomoto are developing a related compound.

Matrix metalloprotease enzymes are implicated in the growth of tumours, as they dissolve the network that holds healthy cells together around the outside of the tumour, giving it space to grow. Matrix metalloprotease inhibitors should thus be able to reduce the size of both primary and secondary tumours. British Biotech's infamous marimastat worked in this way and, although this ultimately proved to be a high profile failure, other drugs in this class are still being investigated, including prinomastat at Agouron/Pfizer, and Bristol-Myers Squibb's BMS-275291.

Not all drugs under development for cancer are small molecules. As well as the biologically active proteins like interferons and interleukins, there has been much interest in the potential of monoclonal antibodies. First discovered in 1975, these substances are produced by immune cells, and target single tissue or cell types. This property gives them huge potential as Œmagic bullets' for specific tumour types, and should be able to fire anticancer agents at malignant cells accurately, while leaving the surrounding healthy tissue alone. Several are now available for treating advanced cancers.

For example, Celltech's gemtuzumab, is based on an antibody to the CD33 protein on acute myeloid leukaemia cells. The anticancer agent calicheamicin is attached to the antibody, which can then deliver the drug very accurately to the CD33 protein on malignant cells within the body, while leaving healthy cells without this marker alone. A couple of other monoclonal antibodies use the drug maytansine, and these are under development at British Biotech for treating small cell lung cancer, and GlaxoSmithKline is collaborating with ImmunoGen on the second. Many others are in the pipeline, mostly in collaborations between biotech companies and the large pharma companies.

Four monoclonal antibodies have thus far been given marketing licences. Two are from Genentech and Roche: rituxumab for non-Hodgkin's lymphoma, and trastuzumab for breast cancer. The other two target forms of leukaemia: gemtuzumab, and Millennium/Schering Health Care's alemtuzumab treats chronic lymphocytic leukaemia.

One of the biggest problems facing cancer therapeutics is the development of resistance to drugs within tumours, making compounds that once had an effect in a patient useless, particularly in second courses of chemotherapy. Often, resistance to one drug leads to multiple drug resistance, and more than one class of cancer drugs, never mind single compounds, can become useless. It is estimated that around 90% of people who die of cancer are resistant to at least some chemotherapeutic agents. Much effort has been put in in recent years to the search for compounds that could prevent, or even reverse, resistance to anticancer agents.

Several mechanisms have thus far been discovered, but perhaps the most important is the overproduction of P-glycoprotein by cancer cells. This substance pumps out unwanted substances from within the cells, preventing the drugs from accumulating inside them. Several P-glycoprotein pump blockers are being investigated. Xenova's XR9576 has been shown to restore sensitivity to a number of therapeutic agents in different call lines, and Phase II trials on its interaction with doxorubicin, vinorelbine and paclitaxel have been completed. LY335979 showed promise in Phase II trials in women with locally recurrent or metastatic breast cancer. And Vertex Pharmaceuticals' biricodar is now in Phase II trials in lung, ovarian and prostate cancers.

Many other ideas for cancer drugs are also being investigated, ranging from vaccines that stimulate the immune system to recognise and destroy the tumours, through gene therapy, to novel ways of interrupting the cell growth cycle. Cancer is such a complex range of conditions, with different forms of tumour responding in different ways to drug therapy, that there will never be a single pharmaceutical answer to all tumours. So the more different strategies there are for interrupting the spread of cancers, the more likely it is that a successful cure for most - or even all - different tumours can be identified.