New cancer drugs are first tested in the laboratory before they are given to people in trials. If it seems that they may be helpful in the treatment of a particular cancer they are tested in phase 1 trials. If these are successful the drug is entered into phase 2 trials and then phase 3.

Manufacturers of drugs that have gone through phase 3 trials and have been shown to be safe and effective can apply for the drug to be granted a licence (also known as a marketing authorisation). Licensed drugs are then available to be used in healthcare. Most licenses are granted by the European Medicines Evaluation Agency as the license covers all countries in the European Union. The Medicines and Healthcare products Regulatory Agency can grant a license for a drug to be used just in the UK. Drugs that are licensed may be further researched in phase 4 trials.

People who run trials involving patients (trial researchers) have to offer a treatment that they believe is at least as good as, and possibly better, than the best available treatment. The trials have to be conducted following strict rules and guidelines.

Phase 1 trials of new cancer treatments use drugs that have been tested on cancer cells in the laboratory and which have had an effect. Phase 1 is the first stage in which the drugs are used to treat people. These studies involve small numbers of patients. Phase 1 drugs are offered to people whose cancer has come back or spread and there is no other standard treatment that may help them. It is not expected that the treatment will cure the cancer; rather, the trial aims to find out:

• how much of the treatment can be given without causing serious side effects
• if the drug has an effect on the body
• what side effects it causes
• whether the drug has any effect on the cancer.

In a phase 1 trial, the first group of 3–6 patients are given a small dose that is expected to be safe. If none of the patients has any side effects the next group will be given a higher dose. In this way the dose of the drug is increased until some people do have side effects. When side effects are seen that make it unsafe to increase the dose further, it is known as the maximum tolerated dose and the dose is not increased any more.

It is impossible to predict whether a new treatment will cause harmful side effects and people in this type of trial need to be monitored very carefully. Phase 1 trials are usually carried out in clinical research units at specialist hospitals, rather than in local hospitals. Sometimes this may involve a lot of travelling to and from the hospital.

If the drug has an effect on the cancer, and a safe dose has been found, the drug will then be tested in phase 2 trials.

By the time a treatment reaches phase 2, researchers will know quite a lot about it. The phase 1 trials will have shown what most of the major side effects are and whether the drug actually has some effect on particular types of cancer. The aim of phase 2 trials is to find out:

• if the treatment works well enough to be tested in a larger, phase 3 trial
• which types of cancer it might be best used to treat
• more about the side effects
• more about the best dose to use.

Researchers will know a lot about the side effects that are likely to occur, so that they can deal with these straight away if necessary.

Phase 2 studies have more patients, perhaps a few dozen. These patients will be closely monitored to see whether their cancer is getting smaller. If the cancer shrinks, it is known as a response to the treatment. The trial will also look at any side effects caused by the treatment.

Sometimes, phase 2 studies include different ways of giving the new treatment; for example, patients may be given the same drug by injection, or as tablets. These trials may be randomised.

These studies follow phase 2 and the main aim is usually to compare the effectiveness of newer treatments with that of standard treatments. They also give more information about the side effects that the new treatment may cause.

Sometimes a trial may test whether a new treatment is as good as the standard treatment but cause fewer side effects. An example is testing a new way of giving radiotherapy and comparing it to the standard radiotherapy treatment.

Phase 3 trials are large and may include hundreds, or sometimes thousands, of patients from many different hospitals, often from several countries. Researchers can measure various aspects, such as:

• how long patients stay free of cancer (known as disease-free survival)
• the number of people who are alive, with or without signs of cancer (known as overall survival)
• whether the cancer grows more slowly
• how the treatment affects patients' quality of life.

Phase 3 trials always involve randomisation.

Phase 4 trials are carried out after a drug has been shown to be effective and has been granted a license. These trials aim to find out:

• how well the drug works when it is used more widely than in clinical trials
• the long-term risks and benefits of the drug
• more about the possible rare side effects and safety of the drug.

These bring together the results of a large number of similar studies. In doing this, doctors and researchers can try to get more accurate answers than from individual trials. They may help to show more about the value of a treatment and may also help to identify other important questions that need to be answered. Overviews and meta-analyses do not involve further patients: they analyse the information gained from previous trials.