ILAR Journal Online, Volume 35(1) 1993: Models of Type I Diabetes

Online Issues

<< All Back-issues

<< This Issue's Table of Contents

ILAR Journal V35(1) 1993 [FORMERLY ILAR NEWS]
Models of Type I Diabetes - Part One

Issues for Institutional Animal Care and Use Committees (IACUCs)

Ethical Issues Involved in the Development of Animal Models for Type I Diabetes
Frederick E. Sieher and Richard J. Traystman

Frederick E. Sieber, M.D. is an associate professor and Richard J. Traystman is a distinguished research professor and vice chairman for research in the Department of Anesthesiology and Critical Care Medicine, Johns Hopkins Medical Institutions, Baltimore, Maryland.

Numerous animal models have been developed to mimic human disease states. The underlying assumption in using these animal models in medical research is that they will provide additional knowledge about and insight into disease processes and, hopefully, better methods for treatment or prevention of diseases in humans. Mounting pressure from both the scientific and lay communities makes it necessary for researchers to clearly articulate these assumptions in all animal research. This is most challenging when designing and testing new animal models. All scientific research involving the use of animals should begin with an ethical focus by examining the risk-benefit ratio (i.e., morbidity or mortality to the animal versus the potential importance of the knowledge acquired). Simply stated, if animals are to be used as models of human diseases a clear understanding of the ethical standards that apply should be established. The aim of this paper is to discuss these ethical aspects within the context of research on type I diabetes, using scientific evidence as a framework for the decision-making process.

Wessler defines an animal model as "a living organism with an inherited, naturally acquired, or induced pathological process that in one or more respects closely resembles the same phenomenon occurring in man" (Wessler, 1976). He emphasizes that animal models offer only approximations of human disease and can never actually duplicate the same process. In attempting to approximate type I diabetes in animals, both spontaneous and induced models have been developed. The means by which the disease is induced or how it develops in an animal is important in determining how the model is to be studied.

The first step in the scientific method is a clear statement of the hypothesis, which in medical research originates from the human side of the equation. This hypothesis may involve a clinical question or pathophysiologic mechanism that cannot be tested ethically or appropriately using human subjects or alternative methods. The hypothesis is brought to the laboratory, where an appropriate animal model is chosen. As outlined in the following section, several animal models of type I diabetes are currently available. One must evaluate the appropriate use of these models to answer the hypothesis based on our original contention that clinical relevance and the risk-benefit of expanded knowledge and animal welfare is weighed carefully. It is our contention that these ethical issues must be addressed before the project begins and adhered to throughout the project.

DIABETIC ANIMAL MODELS AND THEIR USES

Induced Diabetes

Diabetes can be induced by pharmacologic or surgical means. The classic surgical model of type I diabetes is total pancreatectomy, and the animal of choice is the dog, because total pancreatectomy is technically easiest in this species (Sarr, 1988). The advantages of this model include large size, which enables taking multiple tissue samples, frequent blood sampling, and more easily managing insulin regimens; and ease of handling, thereby making long-term care easier (although more expensive). The diabetes obtained is highly reproducible, and in several species, microangiopathy similar to that observed in human type I diabetes develops in both the eyes and kidneys. This model was used by Banting and Best in their studies leading to the discovery of insulin and the mechanism of diabetes. It is also of benefit in the study of diabetic retinopathy (Gelatt et al., 1979) and in chronic studies on tight versus loose control of blood glucose level and the effects of diabetes on end-organ vascular disease. This model has been used to validate various insulin delivery systems, as well as to test various insulin preparations (Goriya et al., 1978) and combinations thereof. This model has validated the use of the artificial pancreas and led to clinical trials of the artificial pancreas in humans (Grau and Saudek, 1987). In addition, various types of artificial pancreatic implants and the biohybrid artificial implantable pancreas have been tested using this model (Sullivan et al., 1991).

There are, however, some disadvantages to total pan-createctomy, including the following: 1) the procedure requires major surgery, adequate postoperative analgesia, and good postoperative care with administration of antibiotics; 2) following pancreatectomy, the animal has little if any counterregulatory response to hypoglycemia; and 3) chronic pancreatic enzyme supplementation is required. These factors must be taken into account in designing a study using this model.

Type I diabetes may be pharmacologically induced via a number of agents that selectively destroy pancreatic ]~ cells. Streptozotocin and alloxan are the most commonly used drugs. In contrast to total pancreatectomy, the use of these chemical agents leaves the remainder of pancreatic function intact. Pharmacologic induction of diabetes has been performed in many species including nonhuman primates, dogs, cats, rabbits, and rodents. The advantages of this model include the following: 1) the use of these agents allows the investigator to study a reproducible type of diabetes in a variety of species; 2) investigators are able to use smaller animal models for economy of care; 3) many of the end-organ effects of diabetes occur with this model. Disadvantages of pharmacologic induction of diabetes are that the drugs used can be toxic to other organ systems and the response to the drugs can be variable. Models using streptozotocin-induced type I diabetes have been useful in studying the effects of tight versus loose control of blood glucose level on the development of diabetic retinopathy (Engerman and Kern, 1987), examining changes in kidney basement membrane associated with the development of diabetic nephropathy (Bloodworth and Engerman, 1980), the effects of diabetic glucose control on the vas-culature and endothelium (Engerman et al., 1977), and validating pancreatic transplant techniques (Barker et al., 1982). Pharmacologically induced models of type I diabetes have been invaluable in advancing knowledge about the pathology of diabetes (e.g., gaining basic information about the impairment of endothelial-dependent dilation of arterioles), and treating this disease (e.g., development of surgical techniques for placement of programmable insulin pumps and the development of therapeutic modalities, such as supplementation of the diet with myo-inositol, for treatment of neuropathy).

Spontaneous Type I Diabetes

There are several spontaneous models of type I diabetes, two of which have been extensively studied: the BBDP rat and the NOD mouse. The diabetic syndrome in BBDP rats is highly reproducible (>50 percent incidence) and closely resembles insulin-dependent, ketosis-prone type I diabetes mellitus in humans. As in human type I diabetes, the syndrome in BBDP rats is characterized by hyperglycemia, lymphocytic insulitis, and the presence of antibodies to islet cell-surface molecules. A major distinguishing feature of diabetic BBDP rats is a genetically transmitted, lifelong lymphopenia characterized by elevated numbers of natural killer cells and depressed numbers of T-helper and T-suppressor cells. The onset of diabetes in BBDP rats is from 60-120 days of age and is associated with lymphocytic insulitis and destruction of pancreatic ]J cells. It is believed that the development of diabetes in the model is secondary to a cell-mediated autoimmune process and may have implications for the pathophysiology of type I diabetes in humans. In addition to diabetes, some BBDP-related lines develop an autoimmune thyroidiris. This model has been useful in determining several genetic markers of diabetes, studying the immunologic mechanisms of diabetes, and developing cyclosporin therapy for use in early type I diabetes in humans (Stiller et al., 1984). In addition, BBDP rats are prone to several of the long-term complications of diabetes; in particular, the neuropathy and retinopathy that occur in this model are very similar to complications in human diabetics (Marliss et al., 1982).

NOD mice develop an autoimmune lesion involving lymphocytic infiltration and destruction of the pancreatic /~ cells, which leads to hypoinsulinemia, hyperglycemia, ketoacidosis, and death. These mice are particularly well suited for genetic and immunologic studies, as well as for research on environmental factors that influence the expression of diabetes. Important insights into transplantation techniques (Ikehara et al., 1985; Lum et al., 1991) and the role of lymphotrophic viral infection in preventing the development of diabetes have been achieved using this model (Oldstone, 1988). A more complete discussion of the NOD mouse is found elsewhere in this journal (pp. 15).

ETHICS OF DEVELOPING AND UTILIZING ANIMAL MODELS OF TYPE I DIABETES

Animal models of type I diabetes have clearly provided benefits to humans, including the discovery of insulin, better treatment of ocular and vascular complications, development of pancreatic transplantation techniques, and increased understanding of the immune basis of juvenile onset diabetes (Council on Scientific Affairs, 1989). There are many other areas of investigation, such as those examining diabetic-induced changes in cellular polyol metabolism or Na-K⁺ATPase activity, in which the results cannot be directly applied to therapeutic interventions in humans. However, advances in biomedical sciences often come from combining results of experimentation at all levels, from the molecular to the clinical. The model selected should depend on the hypothesis and the technical requirements. For instance, larger animal models may be needed for studies requiring numerous procedures, including blood collection, tissue biopsy, and surgical implantation of catheters and mini-pumps. Smaller animal models that have high rates of reproduction might be helpful in studies on the effects of genetics and the environment on the expression of diabetes.

In the development of type I diabetic animal models it is important to justify the research objectives by carefully answering the five vital questions posed by Lane-Petter (1972), as delineated below. The principles discussed in these five questions have been incorporated into guidelines for animal protocol review and ethical assessments of animal experiments (Prentice et al., 1990; Porter, 1992) as well as into policy statements of the American Diabetes Association on the responsible use of animals (ADA, 1985). Others have discussed these concepts within the context of the "three R's": replacement, reduction, and refinement (Rowan, 1979). The five questions for the researcher to consider are:

1. Is an animal the best experimental system for the problem? It is important to carefully evaluate the possibilities and ethics of answering the research question, either in humans or by alternative means. With type I diabetes, we are studying a disease and its end-organ effects, a situation that virtually eliminates alternative methods of research. The possibilities afforded by clinical research depend on the questions addressed. On the other hand, no model of diabetes can accurately and completely reproduce the human syndrome. It must be determined if the particular animal model in question will provide new insight into the problem studied. Therefore, one must assess an animal model on the basis of species appropriateness, ability to accurately reproduce the lesion to be studied, its reproducibility from one animal to another, ability of the animal to survive, and simplicity and versatility of the model (Held, 1983).

2. Must the animal be conscious at any time during the experiment? The study and development of models of type I diabetes are chronic experiments that require conscious animals.

3. Can pain or discomfort of the procedure be lessened or alleviated? Both animals and humans share certain capacities for pain and suffering. The investigator who uses animals has both a legal and moral obligation to ensure that the pain and suffering of experimental subjects is minimized and that the experimental methods are continually refined and improved. In chronic experiments, such as those involving diabetes, anesthesia is not an option. In developing models of diabetes, one is inducing a serious metabolic disease in an organism. If the animal is allowed to remain diabetic for a protracted period, the end-organ effects of the disease may occur, including retinopathy, microangiopathy, and neuropathy. Investigators must evaluate the experimental needs and endpoint of the study to eliminate as much of the suffering incurred with this disease state as possible. Investigators must closely monitor the animals and treat their metabolic disease as is appropriate to the research goals of the protocol. The effects of stress on the animal caused by continual blood glucose measurements or administration of insulin is undoubtedly species-dependent and must be assessed. In addition, investigators must ask themselves if progression of the disease is absolutely required to meet the research objectives and, if so, how far it is necessary to allow the disease to progress.

4. Could the number of animals used be reduced? Chronic animal studies should be designed for maximum efficiency to obtain the highest possible rate of success and the maximum amount of information from each experimental subject. New investigators to the field should seek advice and counsel from both the institutional animal care and use committee and experienced researchers in the field.

5. Is the problem worth solving? Of all the ethical questions proposed by Lane-Petter, it is our opinion that this is the most important and one that has been asked many times in the lay press and by animal-rights activists. If the hypothesis is clearly articulated as a relevant problem, the benefit of developing such a model is confirmed. The risk to the animal must then be minimized to obtain the best model to test the hypothesis.

Hoff (1980) asserts that animals should not be used in experiments when substantial benefits, defined as the saving of human lives or the provision of a substantial contribution to human welfare, are not expected to result. This definition, however, is too narrow. Biomedical research in which animals are used does not always provide an immediate benefit to humans. Merely seeking to understand the mechanisms by which biological systems operate is important and may eventually lead to better treatment of human ailments. This is often impossible to predict when developing animal models of human disease. However, if the research hypothesis has met the relevant criterion, the development of an animal model to test the hypothesis is justified. What are the relevant, unanswered questions in diabetes research? Significant morbidity and mortality are the hallmarks of this chronic disease that affects six percent of the American population. For instance, diabetics have twice the risk of developing heart disease and hypertension as nondiabetics. Diabetes is the second leading cause of blindness and one of the leading causes of renal insufficiency, unquestionably providing a rationale for using animal models to study the disease. The research questions that arise from this chronic illness encompass such fields as genetics, immunology, nephrology, ophthalmology, cardiovascular disease and stroke. There is no doubt that the development of animal models of type I diabetes has improved treatment and provided a better understanding of this disease.

Animal research and the development of animal models of human disease will continue to make important contributions to the treatment of type I diabetes in both animals and humans. Scientists today are compelled to address both their peers and their critics by formally answering the questions posed by Lane-Petter (1972) and strictly adhering to the scientific method in formulating and testing a research hypothesis. Clearly, the development and use of many different animal models to study type I diabetes can stand up to both scientific and lay scrutiny.

REFERENCES

ADA (American Diabetes Association). 1985. Responsible use of animals in research. Diabetes 34:714.

Barker, C. F., A. Nail, L. J. Perioff, D. C. Dafoe, and S. Bartlett. 1982. Animal models of diabetes and immunological problems with islet allografts. Trans. Am. Soc. Artif. Intern. Organs 28:691-699.

Bloodworth, J. M. B., and R. L. Engerman. 1980. Experimental diabetic glomerulosclerosis. Pp. 521-540 in Secondary Diabetes, S. Podolski and M. Viswanathan, eds. New York: Raven Press.

Council on Scientific Affairs. 1989. Animals in research. J. Am. Med. Assoc. 261:3602-3606.

Engerman, R. L., J. M. B. Bloodworth., and S. Nelson. 1977. Relationship of microvascular disease in diabetes to metabolic control. Diabetes 26:760-769.

Engerman, R. L., and T. S. Kern. 1987. Progression of incipient diabetic retinopathy during good glycemic control. Diabetes 36:808-812.

Gelati, K. N., R. L. Peiffer, Jr., and L. W. Williams. 1979. Diabetic relinopathy. Pp. 177-178 in Spontaneous Animal Models of Human Diseases, vol. I, E. J. Andrews, B. C. Ward, and W. H. Airman, eds. New York: Academic Press.

Goriya, Y., R. Kawamori, M. Shichiri, M. Kikucki, Y. Yamasaki, Y. Shigeta, and H. Abe. 1978. Validation of i.v. small-dose insulin infusion therapy in diabetic ketoacidosis of depancreatized dogs. Acta Diabetol. lat. 15:236-242.

Grau, U., and C. D. Saudek. 1987. Stable insulin preparation for implanted insulin pumps: Laboratory and animal trials. Diabetes 36:1453-1459.

Held, J. R.. 1983. Appropriate animal models. Ann. N.Y. Acad. Sci. 406:13-19.

Hoff, C. 1980. Immoral and moral uses of animals. N. Engl. J. Med. 302:115-118.

Ikehara, S., H. Ohtsuki, R. A. Good, H. Asamoto, T. Nakamura, K. Sekita, E. Muso, Y. Tochino, T. Ida, H. Kuzuya, H. Imura, and Y. Hamashima. 1985. Prevention of type I diabetes in nonobese diabetic mice by allogeneic bone marrow transplantation. Proc. Natl. Acad. Sci. 82:7743-7747.

Lane-Petter, W. 1972. The place and importance of the experimental animal in medicine today. Proc. Roy. Soc. Med. 65:343-344.

Lum, Z., I. T. Tai, M. Krestow, J. Norton, I. Vacek, and A. M. Sun. 1991. Prolonged reversal of diabetic state in NOD mice by xe-nografts of microencapsulated rat islets. Diabetes 40:1511-1516.

Marliss, E. B., A. F. Nakhooda, P. Poussier, and A. A. Sima. 1982. The diabetic syndrome of the "BB" Wistar rat: Possible relevance to type I (insulin-dependent) diabetes in man. Diabetologia 22:225-232.

Oldstone, M.B.A. 1988. Prevention of type I diabetes in nonobese diabetic mice by virus infection. Science 239:500-502.

Porter, D.G. 1992. Ethical scores for animal experiments. Nature 356:101-102.

Prentice, E., D. Crouse, and R. Rings. 1990. Approaches to increasing the ethical consistency of prior review of animal research. Invest. Radiol. 25:271-274.

Rowan, A.N. 1979. The concept of the three R's: An introduction. Dev. Biol. Stand. 45:175-180.

Sarr, M. 1988. Pancreas. Pp. 204-21 in Experimental Surgery and Physiology: Induced Animal Models of Human Disease. M.M. Swindle and R. J. Adams, eds. Baltimore, Md.: Williams & Wilkins.

Stiller, C. R., J. Dupre6, M. Gent, M. R. Jenner, P. A. Keown, A. Laupacis, R. Martell, N. W. Rodger, B. von Graffenried, and B. M. J. Wolfe. 1984. Effects of cyclosporine immunosuppression in insulin-dependent diabetes mellitus of recent onset. Science 223:1362-1367.

Sullivan, S. J., T. Maki, K. M. Borland, M. D. Mahoney, B. A. Solomon, T. E. Muller, A. P. Monaco, and W. L. Chick. 1991. Biohybrid artificial pancreas: Long-term implantation studies in diabetic, pancreatectomized dogs. Science 252:718-721.

Wessler, S. 1976. Introduction: What is a model? Pp. xi-xvi in Animal Models of Thrombosis and Hemorrhagic Diseases. Proceedings of the Workshop on Animal Models of Thrombosis and Hemorrhagic Diseases, which was organized by the Institute of Laboratory Animal Resources and held March 12-13, 1975, in Washington, D.C. Washington, D.C.: U.S. Department of Health, Education, and Welfare.