What is the response of the governmental agencies, whose mission should include the education of the public? Let the industries self-regulate the junk they put on the market, undermine any attempt to promote other lifestyles, limit other ways to use resources and ridicule or ignore those who seek alternatives. In chapter 8 of Sacred cows and Golden Cows: The Human Cost of Animal Experimentation, the authors have proposed that one of the reasons why the battle against cancer is far from being won, is the abusive extrapolation from animal results and the stubborn emphasis on studying human-like cancer in animals to find drugs.

Some scientists stressed that our understanding of human cancer cannot be gleaned from animal studies because genetic changes and control seem different. The majority of human cancers results from sporadic mutations of tumor suppressor genes or protooncogenes mainly due to environmental factors, while mice carry their mutations since early development. Furthermore, most modern mouse models involved one single genetic change. Others believed that despite a large amount of literature, no evidence for induction of malignant tumors in animals could be observed.(1) Certainly, one of the most alarming comments came from Dr. Richard Klausner of the National Cancer Institute: "The history of cancer research has been a history of curing cancer in mouse. We have cured mice of cancer for decades, and yet it simply didn't work."(2) Until now, the history of cancer research in animals is a succession of gossips spread through the media machine, altering further years after years the hope of desperate cancer patients.

Short-lived achievements such as the interferon and taxol have been echoed and praised to work in animals so well just to show later that cross-species speculations had little validity in practice. In addition, what we know about chemotherapy has emerged from the use of human tissue and in vitro studies. Malignant human tissue is in theory available in virtually unlimited amount. Epidemiological studies have led to a better understanding of the causes of some of the cancers and earlier prevention and diagnosis. Gene therapy is also a perfect example of a field of research where few animal experiments made it to the clinical trials, despite the involvement of over 2000 laboratories working worldwide now and the $350 million received in grants from the NIH. However, based on an increasing knowledge of the human genome and depending on the discovery of critical genes, therapeutics become customized to the patient's genetic makeup to create a personal treatment.

Mindful consumers should know now the benefits of a healthy life style, including fruits, vegetables, grains and nuts as source of antioxidants in a diet low in animal fat, daily physical exercise. No smoking and little alcohol, at best, and a life in a low stress environment free from pollutants would seem ideal. Certainly, one of the best ways to prevent cancer lies in a long-term commitment to a healthy life-style that should start as early as possible in life. This is precisely what good doctors, based also on recent findings, are trying to pass on to their patients. Obviously, a commitment to a better lifestyle does not please everyone. While the tabacco, liquor, food and petroleum industry lobby groups promote unhealthy products and pressure politicians, the pharmaceutical industry promises the most extravagant and costly drugs and promote the idea that the patients can get away with years of bad habits by just taking their magic pills.

This section is about the use of animal models in cancer research. We know that cancers are multifaceted diseases, they are influenced by genetic, environmental factors and also influenced by gender, age and psychological conditions. One or another combination of these various factors in one individual at one point in life will be enough to initiate a tumorigenic process leading to cancer. Why and how humans develop cancers so greatly is still not clear and it seems that despite the battery of therapeutics being used-surgery, immunotherapy, chemotherapy and radiotherapy-and increasing availability of tests for early detection, cancer is on the rise and continues to kill yearly tens of thousands of Canadians. According to 2001 Canadian Cancer Statistics, an estimated 134,100 new cases of cancer and 65,300 deaths from cancer occurred in Canada in 2001 with men outnumbering women for both new cases and deaths, by 4.8% for incidence and 12.7% for mortality. Only three types of cancer account for at least 50% of the new cases in each sex including prostate, lung, and colorectal cancers in males, and breast, lung, and colorectal cancers in females. Also, nearly one-third of the cancer deaths in men and almost one-quarter in women are due to lung cancer alone. Lung cancer was the leading cause of cancer death among Canadian women in 2001, accounting for an estimated 7,400 deaths, as compared with the 5,500 deaths expected for breast cancer. Prostate cancer will continue as the leading form of cancer incidence, with an estimated 17,800 newly diagnosed cases, as compared with 12,100 lung cancers among Canadian men in 2001.

In 1999-2000, according to the figures given by the Canadian Institutes for Health Research, research in the fields of cancer therapy, brain tumors and therapy, retinal cancer, breast cancer and therapy, and carcinogens cost a total amount over $35 million. Of course, many studies involve the use of animal models and it is difficult to know from the information provided by the funding agencies what kind of animal models are being used in one specific study. Actually, this kind of information is readily omitted for political reason since animal experimentation in Canada is a sensitive issue. The government and the biomedical community are not willing to advertise the feats of their vivisectionists. For many years, imperfect animal models, which are living systems presenting one or several features that "approximate" the human condition, have been created. Rats and mice are incontestably the most manipulated animal species and if they are susceptible to both spontaneous and induced tumorogenesis, results vary between species and show sex difference as well.

All the chemical agents known to be carcinogenic to humans do not produce tumors in other animal species, like chromium, which is carcinogenic in humans but not in animals. Also, some chemical agents can cause cancer in two species even though the tissues involved are different. Another example, benzedrine causes bladder tumors in humans, and middle ear tumors in rats. Furthermore, all the chemical agents known to be carcinogenic to animals do not necessarily produce tumors in humans. Then, it is obvious that no general rules can be drawn from so many discrepancies to predict the action of carcinogens, which limits the validity of the animals as models of specific human neoplasms. Yet, animals develop spontaneous neoplasms mostly in older animals and brain tumors are rare in animals.(3)

When it comes to non-human primates, our genetically closest cousins on the phylogenetic tree of life, the data showed that leukemia could be induced by ways of procarbazine, neutron and proton irradiation; neoplastic growth may resemble that of the human condition in terms of metastasis, lethality and cell phenotypes. Then, if there would be one species scientifically valid as a model of human cancer, investigators would logically choose primitive simians such as rhesus monkeys or even higher primates such as apes to experiment on. Ironically, the favorite laboratory species were and continue to be rodents. Mice and rats are being used not because they are scientifically proven to be satisfactory models but because they are cheap and convenient to use. As well, the NIH requires the use of rodents as a mandatory test for virtually all drugs. In a cost-efficient setting, rodents are perfect animals to get students their Ph.D. thesis and the grants but they are not small 'furry humans' from which can be predicted the remedies for cancer, as past research demonstrated.

Strikingly, there is the great variety existing among the cytological and histological features but also various frequencies of occurrence of neoplasms vary as well. Many animals including the common and well-studied domestic species develop tumors. Mammary tumors of cows are rare; nephroblastomas of the urinary system are present in pigs but not in dogs, cats or horses. Leukosis is found in the lymphatic system of oxes and pigs but not in dogs or cats. Undoubtedly, animals provide good models for veterinarians to study animal cancers and find therapies but stretching the study of these animal models to understand the human pathology, in the light of present-day knowledge is counterproductive. Certainly, on a fundamental level, the study of such neoplastic tissue can lead to a better understanding of the cell and molecular basis of tumorogenesis. Nevertheless, biochemical reactions, metabolisms and genetics are different among species and what apply to one species may not apply to another, even between two apparently related species like rats and mice. Anyone can easily understand the even greater discrepancies between humans and rodents. Historically, the emphasis on using rodents in cancer research started when a geneticist noticed the course and distribution of neoplasms in mice.(4) Then, inbred strains that develop neoplastic processes have been selected. In contrast to genetically transformed rats and mice, the animals have inherited naturally a genetic defect that is expressed at an incidence rate varying between 20 to 100%, depending on the strain, sex, age and means of induction.

Animals have been extensively used as a test to screen potential drugs in an attempt to develop efficacious and safe therapy. They are based on three important prerequisites: the animal system must be sensitive enough to prevent the occurrence of false-negatives and false-positives, the test must be reproducible and pharmacological and toxicological studies must lead to further development of drugs and a better understanding of cancer. In the 1960s, in vitro testing with various test systems, including neoplastic tissues or cells from mouse, hamster and rat was the rule. The use of human tissue was restricted to the HEP-2 (i.e. human epidermoid carcinoma), KB (i.e. human epidermoid carcinoma of nasopharynx) and human sarcoma HS1 systems but most systems were derived from animals. These tests were designed to give a relative indication of the cytotoxicity of the tested drugs. In contrast, in vivo tests aimed at studying the drug in a more complex environment through which the animals could develop a tumor by tissue or cell transplantation.

Generally, a primary screening is used to obtain a dose-response curve and prove the repeatability of the tests. Then, a secondary screening will emphasize on the routes of administration, comparison with other test systems, drug activity, and its correlation with the human condition. If we looked at the animals selected for their neoplasms, the literature is abundant and many animal models were claimed to resemble the human cancer. However, it is interesting to note that there are no good models for human stomach cancer. All laboratory animals but rats are not subject to this form of neoplasm. As well, many animals develop experimental hepatomas in the liver including mice, but their expression of alpha-globulin cannot be compared with its expression in humans. Furthermore, the occurrence of large bowel cancer in laboratory animals is a very uncommon event and among the animals the most used in research, rats and mice, they do not show spontaneous or induced tumors of intestines.

Experimental tumors could not be induced in some animals even though the use of chemicals or radiation or any other means was massive and several hundred times beyond the doses that any person can ever be exposed to accidentally. As a result, these treatments killed the animals rather than creating the expected tumors. In addition, there is no animal model of prostate cancer because there is no ability for the carcinoma to metastasize. As for the practical applications of the studies of animal models of breast cancer, they are possible but unlikely to glean effective treatments. A good model of breast cancer would include morphological changes at adolescence, influences of endocrine and sexual activities and susceptibility of accelerating agents like radiation, immunosupression and drug exposure. Of course, the murine model does not meet these criteria from which will depend the development, course and treatments of cancer. Other models of lymphoma exist, such as the Burkitts-like lymphoma produced by viral infection in the squirrel and spider monkeys, but their incidence and histopathological profiles in humans are not the same in monkeys, because monkeys are not the natural host of those viruses.

In short, animals are not human copies, lack of validity and consequently should not be claimed models of human diseases. Despite the billions of dollars allocated to our scientists, the emphasis on studying the biology of rodents and their resourcefulness to create animal models of human cancers, the causes of cancer remain unclear, the therapies treat the symptoms and cure but cancer continues to grow. In the case of cancer, genetic and environmental factors, viruses, chemicals, radiation, immunological factors, trauma can initiate tumorogenesis. As we face with so much gray, it seems that the keyword in medicine is making its way: detection and prevention. After much delusion over the fight against cancer, finally some medical doctors and scientists are tuning down their optimism and resort to the old adage that prevention is better than the cure.

There are an estimated one million men with prostate cancer in Canada and according to the figures provided by the Cancer Research Foundation of Canada only 20% were diagnosed. The animal based-research community will switch to a more rational and effective approach that is based on science? Prevention has always been far more cost-efficient than the cure and new techniques of early diagnosis should help medical staffs in the fight against cancer. For example, trials have been under way to see whether selenium and vitamin E do prevent or slow down the growth of the tumors. The U.S. National Cancer Institute (NCI) has put some $187 million to fund a 12-year study of 32,000 men without prostate cancer in 430 cities across Canada. Some results showed that after 6 years there was a 40% reduction in mortality rate in the group that received vitamin E. Another study showed that after 10 years of treating the patients with selenium, there was a 70% drop of prostate cancer occurrence among those of the patients who received the treatment versus placebo.(5) Vitamin E and selenium are found in some varieties of nuts and seeds. Without the long-term commitment of men and women in trials such results would be impossible to obtain. Again, it is a chance to be at the cutting edge of medical research, to benefit from the state of art of medical technology and to work with the leaders in their fields. More importantly, guinea pigs would not be as helpful.

Scientists have traditionally used mice and rats not only because they are cheap but also because using human tissue is an administrative, legal and societal hardship. Unfortunately, no single model can recapitulate all aspects of the human pathology. Nevertheless, this has little importance for the competitive and productive researchers because they will pick up the system that best suits the question that they want to address. Transgenic animals seem just to create that. Unfortunately, mouse models have invaded for too long the field of cancer research before in vitro research on human tissue was preferred to study cancer and test drugs. Now, molecular biology makes it possible to engineer a mouse in such a way that a foreign gene is expressed (knock-in) or a host gene is repressed (knock-out). This avenue has become popular because some genes are thought to predispose to cancers and participate in cell cycle regulation and DNA repair. One important protein is encoded by the p53 gene, which is involved in many forms of cancer. Its primary role is to maintain the integrity of the genome. There are many criteria that a mouse has to fulfill before deserving the title of "animal model" including biologic, genetic, etiologic and therapeutic as summarized by Byron Hann and Allan Balmain at the Mount Zion Comprehensive Cancer Center, University of California. Biologically, the tumor initiates in a single cell, progress in a population of other cells and histology and pathology must be similar to those of the human disease.(6) As well, some specific features must facilitate the study of cell death, cell cycle, vascularization or angiogenesis. Genetically, mutation of genes in common with human genes must be found, common biochemical pathways must be evidenced and gene expression in mice and human must be similar in cancer cells. Etiologically, exposure to the same chemicals, or physical factors or other environmental conditions must induce the same tumorogenic response in both animals and humans. Finally, the model is predictive of human clinical results.

The ideal mouse does not exist but some models have been used to study the role of genes known in cancer patients such as K-ras, one of the most commonly mutated genes in the lung cancer or p53, a tumor-suppressor gene controlling cell cycle and thought to be mutated in half of human cancers. However, such studies recapitulated what was already known clinically. Other models exist based on a similar expression or deletion of a gene in a whole organism and the subsequent induction of a tumor type affecting the skin or the pancreas. Once a tumor is created, scientists have to assume that there is continuity between the naturally occurring cancer and the induced tumor in animals. Modifying genes in animals may not just affect the biochemical pathway that is of interest to the researcher but possibly many of them since genes can be at the corner stone of different but interdependent pathways, thus making interpretation more difficult. There are numerous events between a cause and its effect(s), and most of them remain unidentified.

Much data is gathered from studies on mice, which have raised more questions than answers. For instance, mice lacking T-cells, a population of cells of the immune system, are more susceptible to skin tumors after treatment of carcinogens. What would be the role of the immune system in terms of regulating tumor progression? The loss or alteration of the gene encoding another protein called p16 and thought to be implicated as a tumor-suppressor protein in the patients suffering from melanoma, causes an increased incidence of the carcinogen-induced tumors in mice, although the spontaneous occurrence of those tumors in mice is low. This means that there is a need of gamma irradiation or exposure to chemicals to cause a tumor in a strain already deficient in a gene that would have a protective role. Would this mouse gene have the same role in people?

We know that mice are not naturally prone to spontaneous cancer in contrast to humans and such data substantiates this well-known fact among experts. Certainly, there are many other proteins that have abnormal expression during the onset and evolution of cancer and it would be interesting to know what they are, how they regulate or are regulated with regard to cancer. So far, transgenic animals have received much attention, prompted much excitement, generated a whole industry of animal suppliers, but little credit can be given to them when compared to more traditional and simpler approaches already developed in the last 50 years. It is likely that mice will continue to contribute little or nothing to our understanding and treatment of human cancers. Instead, as it is stipulated in the law, as it is required by the NIH mice will continue to serve as poor models or tasters of the new anti-tumor cocktails synthesized by pharmaceutical or biotechnology companies. Genetic manipulations are the state of art of modern biology but beyond the feat of the procedure more importantly is the validity of the method to study human diseases. In contrast, genetic studies can be made through epidemiological studies to identify individuals at risk, genetic inheritance in family, race, sex or age variability and etiologic factors. Then, genes can be identified, and biopsies and tissue cultures are performed for further analysis. In the battle against cancer, there is a need to explore the toxicity of some chemicals to cancer cells.

Recently, some compounds known as metalloprotease inhibitors, and studied in a mouse model, made it to clinical trials to be stopped later because of a lack of efficacy but also because an increased tumor progression in human patients was seen. This kind of failure happens as it happens for investigators to be lucky and ultimately find a drug that is efficient and that does not kill the host receiving it. The history of cancer research took another turn the day when nitrogen-containing mustards, initially during World War I as warfare, were shown to reduce the size of tumors in the patients. From the satisfactory results gained in clinical studies of these compounds in the patients suffering from Hodgkin's lymphoma, a massive search of anti-tumor molecules was launched. The modern era of chemotherapy was born.(7)

Cyclophosphamide was derived from nitrogen-containing mustards. Later in 1965, Professor Barnett Rosenberg who was a physicist established the action of in cell cycle and his research led to the development of the compound known as the trade name Neoplatin, which was effective in testicular and ovarian cancers. Antimetabolites such as methotrexate-an antagonist of folic acid that is the precursor of dihydrofolate and tetrahydrofolate and both necessary to the synthesis of thymine, one of the essential components of DNA-used against leukemia since 1952, were derived from the studies of the effects of brewer's yeast, rich in folic acid, on mice bearing a breast tumor. The results showed that 43% responded to the treatment by complete regression of the tumors even though another group could not repeat the results. After the war, animal experiments continued and clinical trials appeared to correlate with animal results.

Since the 1950s, other drugs, including the immunosuppressant azathioprine, the anti-viral agent acyclovir, and the inhibitor of reverse transcriptase azidothymine (AZT) that were first synthesized as potential anti-cancer drugs proved inefficient in this role. Many of these compounds were derived from natural compounds and serendipity played an important role in the discovery of unparalleled drugs such as antileukemic vinblastine, quinine used against malaria, the painkiller morphine and so many other derivatives. Also, camptothecin, isolated from a Chinese tree interferes with the topoisomerase I, an important enzyme involved in DNA replication. Another drug Taxol, which stabilizes microtubules, was derived from the bark of the pacific yew, and tests in animals quickly led to clinical trials and great excitation. But unfortunately, the drug proved to be toxic to bone marrow cells and is now being replaced by another natural product and a synthesized analogous molecule. Not only substances isolated from plants and trees, tested in animals and later tried in humans, demonstrated anti-tumor properties but other compounds obtained from bacteria and fungi have largely contributed to help humankind in its quest for cures by providing hundreds of useful molecules, such as antibiotics and other anti-cancer molecules such as mitomycin C isolated in 1958 from Streptomyces verticillatus. Later a massive search of anti-tumour molecules was launched in the USA during the Nixon years. Of 500,000 compound tested in animals, 12 were found to have a substantial role in human cancers. They were derivatives of already known molecules, whose efficacy could have been deducted by their chemical structure.

There are volumes of data on the mechanisms of cancerogenesis at the cell and molecular levels derived from animal studies. That seems great but what does it tell us about cancer in humans? Leading clinicians will tell you that anti-cancer chemicals are tested in vitro on human tissue because the rodent tissue is not a reliable one. In cancer research, some researchers assume that artificially created neoplasms in animals are models of spontaneous cancer in people. They also assume that environmental factors, which according to some experts account for 80% of human cancer, have no impact in cancerogenesis and assume that mice are like humans of course in terms of molecular pathways and gene regulation. These are assumptions one must believe in order to continue animal experimentation.

The list of the chemicals that Nature has engineered during the evolution of species may continue to be part of our pharmacy. Unfortunately, the natural habitats harboring these interesting, unknown and potentially very useful organisms are disappearing every year as a result of human encroachment. Humankind should be grateful for the gifts it received in legacy, but as time passes, the natural bounties of our small planet vanish.

• 1. Herzig M. and Christofori G. "Recent advances in cancer research: mouse models of tumorogenesis." Biochimica et Biophysica 1602, 97-113 (2002)
• 2. Richard Klausner as quoted in Los Angeles Times Wednesday, May 6 (1998)
• 3. Mitruka B. et al. Animal Models of Cancer Research. In: "Animals for Medical Research: Models for the study of Human Disease." John Wiley & Sons Medical Publication (1976)
• 4. Slye M.J. Cancer Research 7: 107 (1922)
• 5. National Post. "Prostate Cancer: A joint Venture." Thursday, September 27 (2001)
• 6. Byron Hann and Allan Balmain. "Building 'validating' mouse models of human cancer." Current Opinion in Cell Biology 13: 778-784 (2001)
• 7. John Mann. "The elusive magic bullet." Oxford University Press (1999)