Drug testing

More on drug testing

A new drug goes through a series of stages before it can be sold:

Large numbers of animals are still experimented on, both to discover effective drugs and to test any that may be useful. Currently drugs, by law, still have to be tested on animals before they can go on to clinical trials. Increasingly, though, in vitro (test tube) tests are also being used. Tissue culture technology is continuing to develop and will be used more and more in future.

In clinical trials, human volunteers try out the new drugs. If the trials are successful and the drugs work with minimum side effects, then they can be registered by health authorities and go on the market.

There are two main problems with this process. Firstly, many animals not only die, but also suffer the effects of induced diseases (eg cancer, arthritis, etc) and the effects of drug overdoses (eg convulsions, internal bleeding, nausea/vomiting, etc). Secondly, the process does not guarantee human safety.

Drug discovery/design

Traditionally diseases have been produced in animals, and drugs that might be useful for these diseases are then tested on the animals. Clearly this approach is cruel. Animals suffer just as much as humans when they have diseases like cancer, and at the end of the experiment they are killed. There are better ways to discover drugs useful to humans.

A good example is the National Cancer Institute (NCI) in the US. It has over a period of 35 years injected 400,000 chemicals into mice deliberately bred to develop leukemia. The aim was to find general tumour-fighting chemicals. A few drugs against leukemia have been found in this way, but not against other forms of cancer. There isn't a lot to show for 35 years' work, 400,000 chemicals, and millions of mice!

To become more effective, the NCI has stopped using mice. They now use 60 human tumour cell lines from 7 main areas: colon, lung, melanoma, kidney, ovary, brain and blood. By using cell cultures researchers are more efficient: they can test 300 chemicals a week. They are more biologically accurate: they are using cells from the right species (human) and a specific type of cancer (eg lung cancer). Some of the chemicals previously tried on mice are being tested again. Researchers think the cell culture test will more accurately show tumour-fighting potential (1).

Pharmagene is a company which tests drugs exclusively on human tissue. It started in 1996. The company studies human tissue to find the genetic basis of disease. It looks for genes that express particular proteins involved in disease states, and then designs drugs to alter the function of these proteins. Visit the Pharmagene web site .

More on cancer

Various species of animals of given cancer, but most often mice are used. They are either bred to develop a particular type of cancer, they are given chemicals to produce cancer, or human cancers are grafted into mouse strains with defective immune systems. Fortunately cell culture tests are becoming more and more common in cancer research. Some examples:

Cell cultures can also be used to investigate differences between people. For example, 4 drugs were tested on ovarian tumours from 100 people. There was a big difference in response to all the drugs. So, cell culture tests could be very useful to determine the best drug for each individual patient (6).

People also vary in terms of how much damage cancer drugs do to healthy cells. Human white blood cells (lymphocytes) were assessed for DNA damage after being exposed to chemotherapy drugs. There were big differences in DNA repair. With this knowledge, doses of drugs can be adjusted according to a patient's capacity to repair damage (7).

HIV/AIDS

Chimpanzees, monkeys, cats, and mice with defective immune systems have been used to study immunodeficiency viruses. However, HIV grows in cell cultures, and many studies have used human cells to test drugs. For example, white blood cells and blood corpuscles that digest bacteria (phagocytes) were infected with 3 strains of HIV. The drug being tested limited virus replication (8). Other studies have also used white blood cells (T4-lymphocytes) to test antiviral effectiveness (9).

Other virus infections

Drugs against other viruses, such as hepatitis and herpes, can be tested in cell cultures. Infecting animals with these diseases causes unnecessary suffering. For example, guinea pigs infected with genital herpes developed ulcers and scabs, no doubt with pain, itching, and general ill-health. One in six of the animals suffered paralysis and death (10).

As an alternative, 13 chemicals were tested on human cells infected with herpes. One was effective in limiting virus replication, especially when combined with another antiviral drug (11). Similarly, chemicals can be tested on human liver cell lines infected with hepatitis B (9). In some cases drugs already used for other purposes have been effective against infected cells, decreasing the amount of virus in the cell culture dramatically (12).

Epilepsy

Brain slices can be made to produce discharges similar to those in epilepsy. Drugs are then added to the tissue culture to see if they will reduce or block these discharges (13-14). The alternative is to produce seizures in live animals. Unfortunately it is usually rat brain slices that are used in studies, although there is no reason in principle why human tissue could not be used. Human brain tissue is removed during operations for intractable epilepsy and tumours, and could be used for research.

Cell cultures can also be used to tailor treatment to the patient. Some people have adverse reactions to anticonvulsant drugs, but a culture of their white blood cells (lymphocytes) can be used to test drugs and find out which ones are least likely to have a negative effect (15).

Arthritis and other inflammation

To produce inflammation in animals, a chemical injected into the paws or the leg joints. This is clearly very painful. However, human tissue can be used to study the process of inflammation and its control. For example, one study used samples of human synovial membrane from people with rheumatoid arthritis to test 2 drugs (16). Other studies have used epithelial cells from a vein in the umbilical cord (17), epithelial cells from nose surgery (18), or blood corpuscles (19). Each study could demonstrate anti-inflammatory action of the drugs tested.

In vitro (test tube) methods are being used to fight other diseases as well; the studies here are only a few examples. As these methods continue to develop, they will be used more and more in future.

Another way to "discover" new drugs is to design them by computer. A drug has an effect by binding to a particular receptor in the body. This arrangement is like a lock and key; the arrangement of the molecules in the drug fits exactly into the receptor.

Many such receptors have now been precisely described. With this knowledge, computers can be programmed to design molecules to fit specific receptors and have specific effects. This approach is called "rational drug design", and will continue to expand in the future (20). A few drugs now in use have been designed in this way, for example, some HIV drugs (21).

Drug testing: toxicity

Before a new drug can go on to clinical trials it has to go through short-term and long-term animal tests. An LD50 is recorded, which is the dose of the drug it takes to kill half the animals of a particular species. Animals are also fed daily doses, usually for 90 days, to see if there are any long-term toxic effects. Pregnant animals are fed the drug to see if it will cause birth defects. Tests of up to 2 years are done to see if the drug causes tumours. For more details of these tests see:

Toxicity testing
Tests for chemicals that cause cancer or birth defects

Once the animal studies are finished, the drug goes on to clinical trials with human volunteers. In spite of the animal tests, only 1 or 2 out of every 20 drugs in clinical trials go on to the market; the majority either don't work well or have too many negative effects (20) (22).

Drugs are not necessarily safe, even after the clinical trials. Some are later withdrawn from sale, or restricted, because of their adverse effects, including deaths. Very often these effects were not predicted by the animal tests. Here are a few examples:

Isoprenaline was in the past used in inhalers for asthmatics. It is thought to have killed at least 3500 people due to heart failure (22). However, researchers found:
" Intensive toxicologic studies with rats, guinea pigs, dogs and monkeys at dosage levels far in excess of current commercial metered dose vials ... have not elicited similar adverse responses " (23). Cats could tolerate 1mg/kg, which is 175 times the dose that could kill a 70 kg human (0.4mg/kg) (24).

Clioquinol (Entero-Vioform, Mexaform) was an anti-diarrhoea drug. It was withdrawn after it caused a nervous system disorder (subacute myelo-optic neuropathy) in Japan. At least 10,000 people suffered symptoms such as walking difficulty, numbness in the feet, paralysis and eye problems, including blindness (25-26). However, there was no sign of neurotoxicity in rats, cats, dogs and rabbits given the drug for 90 days or more. Even in the LD50 tests there was no sign of eye damage (27).

Benoxaprofen (Opren) was an arthritis drug that was withdrawn after less than 2 years. There were 3500 adverse reactions, including 61 deaths in the UK alone. Deaths were caused by liver damage in the elderly. The drug also caused severe photosensitivity; people developed a rash in the sun (28). However, monkeys fed 7 times the maximum human dose daily for a year showed no signs of toxicity, and there was no sign of photosensitive skin (22).

A number of other anti-inflammatory drugs have been withdrawn over the years due to adverse reactions and deaths. These include Zomax, Flosint, Alclofenac, Ibufenac and Osmosin . Here too animal tests did not predict the danger (22). With Flosint, for example, dogs and monkeys fed high doses daily for 1 year showed no signs of toxicity (29).

Zimelidine (Zelmid) was an antidepressant withdrawn after causing 300 adverse reactions in the UK, including liver damage, convulsions, and 7 deaths (22). Long-term studies with rats and dogs did not warn of these problems. The worst effect on dogs at the highest dose was slight gastrointestinal irritation (30).

Practolol (Eraldin) was a heart drug withdrawn after it caused severe eye problems, including blindness, and 23 deaths in the UK (31). Eventually 1000 people were compensated for injuries by the drug company (22). This drug was thoroughly tested in animal studies, which did not predict the dangers (32).

A number of antimicrobial drugs have caused problems not seen in animal tests. Clindamycin caused a serious intestinal disease (pseudomembraneous colitis) and 36 deaths in the UK (31). This problem was not seen in dogs and rats fed much higher doses daily for 6 months and 1 year (33). Chloramphenicol caused serious blood disorders and 42 deaths in the UK (31). Ketoconazole (Nizoral) produced 82 reports of liver toxicity and 5 deaths in the UK, although due to under-reporting no exact figure could be given (33). Apart from liver damage, the drug also produced nausea, headache, itching and dizziness. These effects were not predicted by animal tests (35).

Adverse drug effects continue to be a problem, even after lengthy animal tests. According to the US General Accounting Office, 51.5% of the 198 drugs approved by the US Food and Drug Administration from 1976-1985 caused side effects in humans. These effects were serious enough to cause withdrawal of the drug or, more often, relabelling with more warnings. Adverse effects included heart failure, shock, breathing difficulties, seizures, kidney failure, liver failure and death (36).

In 1995 there were 130,950 adverse reports made to the US Food and Drug Adminis-tration. Some of the drug effects were life-threatening, involved hospitalisation, and caused death. Visit the FDA web site. Such reports are only the tip of the iceberg. It has been estimated that only 10% of all adverse effects are reported (37).

In 1997 there have been reports in medical journals about diet pills and a hay fever drug. The drugs dexfenfluramine (Redux) and fenfluramine (Pondimin) have produced dangerous pulmonary hypertension and heart disease, including deaths. The New England Journal of Medicine states:
" The mechanism by which these drugs cause pulmonary hypertension is not known, in part because it has not been possible to produce the disease in animals " (38).

The hay fever drug terfenadine (Seldane) has been withdrawn in some countries. It is now restricted in the UK after it produced 33 serious cardiac arrhythmias and 14 deaths there (39). Part of the problem is that terfenadine interacts with other chemicals, including grapefruit juice!

Animal tests are no guarantee of safety. There is also another problem: the adverse effects seen in animals may not apply to humans, and so possibly valuable drugs will not be investigated further.

These problems are highlighted in 2 studies. In one study, there was a total of 53 adverse effects in humans. Of these, 23 were seen only in humans, not in animals. Rats showed only 18 of the 53 effects seen in humans. A further 19 adverse effects were seen only in rats, not in humans. Dogs showed only 29 of the 53 effects seen in humans. A further 24 adverse effects were seen in dogs only, not in humans (40).

In the other study, 45 newly licensed drugs were examined. Some toxic effects were seen only or mainly in animals, including convulsions, sedation, liver, kidney and blood toxicity, breathing problems, salivation, and central nervous system problems. Other toxic effects were seen only or mainly in humans, including nausea, dizziness, headache, dry mouth, sweating, cramps, low blood pressure and skin problems. Vomiting and gastrointestinal problems were similar in humans and animals. Only about 25% of the toxic effects seen in animals were also seen in humans (41).

Understanding drug effects and individual variation

Human cell cultures can be used to test how toxic a drug is. However, they can do much more than that. They can also show exactly how a drug is broken down in the body, and why there are such large differences between people in their response to drugs. Animal tests cannot explain human variability.

In vitro tests are now being developed to investigate the sequence of processes a drug goes through:

The speed with which enzymes in the liver break down drugs is a major difference between species. This speed determines how long a drug is in the body, and how much toxic effect it is likely to have. Animal tests do not necessarily show how long it will take humans to break down a drug. For example, the anti-inflammatory drugs phenylbutazone and oxyphenbutazone produced at least 512 deaths in the UK alone due to blood disorders and gastric bleeding (31). The drug company admitted to 1182 deaths worldwide, although estimates have ranged as high as 10,000 (42).

Humans break down these drugs very slowly compared to other animals. It takes 72 hours to break down phenylbutazone in humans, compared to 8 hours in monkeys, 6 hours in dogs and rats, and 3 hours in rabbits (22). Dogs take only 1/2 hour to break down oxyphenbutazone, compared to 72 hours in humans (40).

In vitro tests can show these kinds of differences before drugs go on the market and cause injury. Many researchers now acknowledge the importance of in vitro methods in drug testing (43-46). Cell culture methods can be used to investigate:

Human liver cells (hepatocytes) are an excellent tool for studying these aspects of drug metabolism. They can be obtained from operations or organs for transplantation. A disadvantage is that the complete set of liver enzymes only works for a few days in fresh samples. In future, better cryopreservation (a special freezing technique) may provide a steady supply of good quality hepatocytes for research. However, parts of these cells or even individual enzymes can already be stored for years, and they are also very useful for drug studies. Fifteen of the important cytochrome P450 enzymes have now been identified and described.

Variations in response to drugs can be predicted when it is known:

For example, it is known that some chemicals interfere with the action of certain enzymes. If such a chemical is taken together with a drug that relies on one of these enzymes for its metabolism, then there will be a drug interaction. There are likely to be adverse effects. The interaction of the hay fever drug terfenadine with several other drugs, described earlier, could have been predicted in this way (46-47).

Human cell cultures can be used to calculate how long it will take to break down a drug (49). This information would avoid disasters like the one with phenylbutazone and oxyphenbutazone, described earlier.

It is known that some people are "poor metabolisers". Some of their enzymes do not work efficiently, and they are more likely to suffer adverse drug effects. These effects can be predicted when it is known which enzymes break down a particular drug.

Other factors influence the response to drugs, and can be investigated in vitro. For example, how quickly do enzymes from old people break down a drug? Does it make a difference if the person is a smoker or drinks a lot of alcohol? Does it make a difference if the person has a particular disease? These factors could all influence whether or not someone has an adverse drug effect.

Finally, it is very difficult to predict human allergic reactions to a drug in animal studies. There is a test which uses human leukocytes, blood corpuscles involved in an immune response. A test drug, as well as liver enzymes to break it down, are added to the leukocytes to study the reaction. Variations between people can be investigated in this way (50)

Liver cells are very important in drug metabolism, but other cell cultures are also being investigated for other aspects of drug activity. For example, a cancer cell line has been used to effectively simulate absorption through the small intestine (51). A kidney cell line has been used to investigate the blood-brain barrier, and whether drugs are likely to disrupt this barrier to have a toxic effect on the brain (52).

I would like to see References for this document on drug testing.