ILAR Journal Online, Volume 31(3) 1989: Perspectives on Animal Use

Online Issues

<< All Back-issues

<< This Issue's Table of Contents

ILAR Journal V31(3) 1989
Perspectives on Animal Use

UKCCCR Guidelines for the Welfare of Animals in Experimental Neoplasia
UKCCCR Ad Hoc Committee

Dr. P. Workman (MRC Clinical Oncology Unit, Cambridge, Chairman); Dr. A. Balmain (Beatson Institute for Cancer Research, Glasgow); Dr. J. A. Hickman (CRC Experimental Cancer Chemotherapy Research Group, Aston); Dr. N. J. McNally (deputy, Dr. A. M. Rohas, both CRC Gray Laboratory, Northwood); Prof. N. A. Mitchison (University College, London); Dr. C. G. Pierrepoint (Tenovus Institute, Cardiff); Mr. R. Raymond (JCRF, London); Dr. C. Rowlatt (ICRF, London); Dr. T. C. Stephens (ICI Pharmaceuticals, Alderley Park, Macclesfield); and Mr. J. Wallace (Institute of Cancer Research, London).

Background and Scope

Animals with local or disseminated tumours are likely to experience pain and/or distress, thus justifying special care and attention from both licensees and others involved in their welfare. Associated techniques including surgical preparation, irradiation, and drug administration may increase the severity of an experimental procedure. Recognizing this, the United Kingdom Coordinating Committee on Cancer Research (UKCCCR),¹representing the main cancer charities and the MRC, has prepared the following guidelines for research workers using animals in experimental neoplasia. Particular emphasis is focused on the prediction and recognition of adverse effects and the implementation of humane end points. The majority of work in this area utilizes small laboratory animals, particularly rodents. Consequently, we have drawn largely on available expertise with these species. However, the general principles are applicable to all species of animal.

While we recognize and encourage the development of alternative research techniques which do not involve animals, we consider that there are many questions which can be answered only by the study of tumours growing in vivo. The general welfare of laboratory animals and the performance of regulated procedures upon them are both covered by the Animals (Scientific Procedures) Act (1986) effective from 1 January 1987. Under this Act all scientific procedures on living vertebrates which may have the effect of causing pain, suffering, distress, or lasting harm are controlled by the Home Office and require specific authority through Personal and Project Licences. Recommendations for the housing and care of laboratory animals are specified in the Royal Society/UFAW Guidelines (Part 1, 1987). In addition, the following references are recommended for advice on general animal husbandry and experimental techniques: Gay (1965), Fowler (1978), Tuffery (1987), and the Institute of Animal Technology (in press).

We welcome the new Act and the Royal Society/ UFAW Guidelines, and look forward to the publication of further guidelines from expert sources. We envisage that the present guidelines will be of general value to workers carrying out experiments which involve the growth of tumours in experimental animals, whether these arise spontaneously, are produced by transplantation (including passage and hybridomas), or are induced by carcinogenic agents or genetic manipulation. They may be especially helpful in the completion of Project Licence Applications, in particular section 19b (v and vi) which requires the applicants list the possible adverse effects and their likely incidence as well as the proposed methods of controlling severity, e.g. the use of analgesia, regional or local anaesthesia and sedation, and the implementation of humane end points.

It is an important feature of the present guidelines that the procedures practised upon animals in cancer research, and particularly the humane end points used, should be subject to a continuous process of refinement. The guidelines will therefore be modified and updated as appropriate. The guidelines are not mandatory. The term "should" is used to encourage attainment of desirable standards; the term "must" is used only where legal obligations apply.

The recommendations are divided into two parts: the general recommendations are applicable to all regulated procedures, and the specific recommendations are more directly targeted to the particular problems of experimental neoplasia. It is important to emphasize that procedural guidelines, especially with respect to implementation of humane end points, must be tailored to the precise nature of each individual experimental neoplasia model. To illustrate this, the Appendix gives some examples of criteria for particular tumour systems.

Recommendations

General Recommendations

1. The following recommendations are based on the premise that for each study those involved in the procedures will weigh the likely adverse effects on the animals used against the benefits likely to accrue from the work. The potential benefits of cancer research are clear. Nevertheless, the feasibility of using alternative methods not involving live animals should be considered. In vitro cell lines may be appropriate in many instances.

2. Where animals must be used, the degree of pain and distress must be minimized by judicious use of anaesthetics and analgesics, the refinement of experimental techniques and the early implementation of humane end points. Licensees must know the severity band for each regulated procedure (i.e. mild, moderate, substantial, or unclassified). The severity band will have been arrived at by agreement between the applicant and the Home Office and takes into consideration details of the procedure itself, the nature and incidence of any likely adverse effects and any practical measure which will be used to minimize severity. The severity condition of Personal and Project Licences requires a Personal Licensee to notify the Project Licence holder if one or more severity bands may have been or are likely to be exceeded. The Project Licence holder must notify the Home Office Inspectorate of this at the earliest possible opportunity. In addition, there is an inviolable termination condition in every Personal Licence, which requires the Personal Licensee to ensure the immediate killing (by an approved painless method) of any animal in severe pain or severe distress which cannot be alleviated.

3. Where certain procedures cause particular concern, these must be noted specifically in the Project Licence application. A more detailed justification and definition of severity limits will be needed. Such procedures may be subject to additional conditions in the Personal Licence to control numbers of animals and/or severity. In addition, the Home Office may require particular reports on them.

4. It is important that pilot experiments should be undertaken on small numbers of animals before new procedures are carried out on a larger scale. The pilot experiments should identify particular problems, define the time scale of critical events, and help to refine the appropriate end point. In all experiments the numbers of animals used should be restricted to the minimum consistent with the design and purpose of the experiment.

5. All involved staff should be aware of their individual responsibilities, and a clear chain of consultation should be established. The decision-making process should be designed so that under all circumstances, appropriate action is taken promptly to deal with any problems which may arise, for example, if the clinical condition of a tumour-bearing animal deteriorates unexpectedly or if the individual effects of tumour and therapeutic treatment are difficult to distinguish (see section 3.5).

6. Ail involved staff should receive appropriate training and supervision. Where research workers are using unfamiliar procedures, information and guidance should be obtained from experienced colleagues, as well as from the scientific literature.

Specific Recommendations

1. Assessment of Severity

1.1 Before assessing the severity of any regulated procedure on the well-being of an animal, it is essential that the observer is familiar with the normally accepted behaviour, anatomy, physiology, and environmental requirements of the species used, for example, growth rate, dietary intake, and microbial status.

1.2 Particular attention should be paid to those body systems most likely to be affected by the procedure. With solid tumours this will include ulceration, distension of covering tissues, and cachexia. In the case of ascitic tumours, abdominal distension, anaemia, and cachexia will be important. Lymphatic involvement from lymphoma and neurological disturbance from intracerebral tumours are examples of special complications arising in specific situations.

1.3 Certain deviations from normal well-being may be difficult to observe, for example, induction of anaemia or the development of metastases, and special investigations may be required to detect them.

1.4 Appropriate control animals should always be included, so that the individual effects of the tumour and of any treatment can be distinguished.

2. Biology of Tumours

2.1 Due consideration should be given to the known biology of the tumour. For spontaneous and transplanted tumours, important features will include growth rate, invasion, distension, ulceration, metastases, site, and production of cachectic factors.

2.2 In the case of tumours induced by carcinogens, viruses, or genetic manipulation, factors such as method of induction may affect the nature and location of resulting tumours.

2.3 Contamination of tumour cell lines with viruses and other microorganisms may compromise experimental results, as well as causing an outbreak of disease among laboratory animals. Screening of cell lines for rodent viruses is strongly recommended. For example, Sendai virus is often used to induce cell fusion in vitro and is pathogenic to mice and rats. A potential hazard exists for research workers from immune-compromised animals receiving human tumour xenografts which may be contaminated with human pathogens. In such cases, special facilities should be considered for both tissue preparation and animal containment (e.g. flexible film isolators) (see UKCCCR Guidelines for the Xenografting of Human Tumours, 1980).

3. Humane End Points

3.1 Considerable care should be given to the judicious choice of end point for tumour growth. This should take into account predictable indications of pain, distress, or significant deviation from normal behaviour. Unless specified otherwise on the Project Licence, animals should be killed before:
(I) predictable death occurs;
(ii) they get into poor condition;
(iii) the tumour mass becomes over-large, likely to ulcerate or unacceptably limits normal behaviour.

3.2 In the case of local solid tumours, the required information on response to therapy may be obtained by tumour regrowth delay or clonogenic assay, rather than by tumour weight at a given time. Difficulties may arise with this last method because optimum shrinkage of treated tumours may not be achieved before control tumours become excessively large and/or distressing to the host animal. Where such an assay has to be used, the tumour burden should be regulated as indicated in section 3.1.

3.3 The choice of site for solid tumours also requires considerable care, and particular attention should be given to avoidance of sites involving the special senses or where the capacity for tumour growth is limited. Subcutaneous or intradermal growth on the back or in the flank are considered to cause the least distress, while tumours implanted in the footpad, tail, brain, and eye will require much greater justification. Distension of musculature is generally painful and this should be considered with intramuscular implants. Extra attention must be paid if multiple sites are used.

3.4 The survival end point should be avoided wherever possible, and its use will require special justification. Where this end point has to be used for animals with ascites or disseminated tumours, particular care should be taken. It is, in general, unacceptable to wait for predictable death. Animals expected shortly to become moribund should be killed, unless specified otherwise in the Project Licence.

3.5 Difficulties may occur where the effects of anti-cancer agent's on tumour growth are being evaluated. The high toxicity of these agents may combine with the adverse effects of the tumour, but this might be justified by the prospect of a therapeutic remission (as occurs in man). Thus, the outcome of such experiments may be uncertain, and uncritical culling would frustrate the purpose of the study. However, where the outcome can be reasonably predicted, animals about to become moribund should be killed.

3.6 No precise quantitative guide can be given as to the acceptable upper limit of tumour burden, since the adverse effects on the host will depend on the biology of the tumour, the site and mode of growth, and the nature of associated treatments. However, tumour burden should not usually exceed 10% of the host animal's normal body weight. It should be emphasized, however, that problems may arise with much smaller tumour burdens.

3.7 With ascitic tumours, including hybridomas, care should be taken to ensure that the volume of ascitic fluid does not become excessive, causing gross abdominal distension, and that solid deposits and cachexia are not allowed to become clinically significant. Ascitic volumes should not usually exceed 20% of normal body weight in mice and rats. Retired breeders are advantageous for monoclonal antibody production, since their abdominal musculature more readily allows larger ascites volumes to be tolerated without discomfort. Ascitic tumours should normally be drained only once. This will minimize the development of solid tumour deposits, bleeding into the peritoneal cavity, and cachexia, as well as reducing the risk of introducing an infection. General anaesthesia should always be employed.

3.8 In tumour therapy experiments with adult rodents, it is recommended that weight loss should not normally exceed 20% of the host body weight at the commencement of the experiment. For younger animals, failure to maintain the weight gain seen in untreated control animals should be considered as an indication of toxicity.

3.9 Care should be taken that general housing conditions are appropriate to the known or anticipated condition of the tumour-bearing animal, for example, in terms of appropriate bedding, cage structure, and accessibility of food and water.

3.10 Humane end points and other procedures should be refined in the light of experience.

4. Examination of Animals

4.1 The frequency with which animals must be inspected for signs of pain or distress and the extent of each examination will be dictated by:
(i) the known biology of the tumour and/or effects of the inducing agent;
(ii) the effects of any associated techniques;
(iii) the changing clinical status of the animal.

4.2 Rapidly growing or invasive tumours will require more frequent attention, and greater care will be required as the tumour burden increases.

4.3 As a minimum, every tumour-bearing animal should be inspected daily and additional more-detailed examinations undertaken as appropriate. The frequency of the latter should be increased during critical periods where the potential for animal suffering may be anticipated. The experimental design should ensure that these do not occur when staff are absent. Particular attention should be given to animals in poor health.

4.4 Appropriate assessment techniques will include: evaluation of overall clinical condition, including appearance, posture, body temperature, behaviour, and physiological responses; assessment of food and water intake; weighing to determine changes in body weight (both positive and negative changes compared to controls can be associated with increasing tumour burden); caliper measurements to determine tumour volume or mass; and inspection and palpation to locate the sites of tumour growth, as well as to assess distension, ulceration, and compromised mobility.

4.5 Other special examination techniques will be more valuable for specific sites, e.g. breathing rate for lung deposits, neurological disturbances for brain neoplasms, and blood cell counts for leukaemias. Laparotomy or endoscopy may be appropriate in some instances. Estimation of circulating tumour marker substances may also be of value. Autopsy of animals may expose adverse effects undetected by external examinations.

5. Documentation and Publication

5.1 Researchers are strongly urged, for each tumour model in use in their laboratory, to document the expected behaviour of the tumour and host animal under various experimental conditions, including therapy. They should also document humane end points to limit severity with regard to acute and delayed toxicity and maximal tumour burden, and indicate any particular problems which may be encountered in the use of each model. The appropriate response to such problems should be described and the chain of consultation and responsibility clearly defined. Consideration should be given to the inclusion of a numerical scoring system to facilitate decision-making, e.g. when to contact senior staff or to kill an animal. The guidelines for specific tumour models should be readily available to and agreed [on] between all research and animal husbandry staff involved with that model. Particular care should be taken that all procedures are understood by junior and occasional staff. Researchers are also encouraged to share this information with other groups using the same system, for example when providing a tumour cell line to another laboratory.

5.2 Researchers are encourage to publish improvements in humane end points in appropriate journals, so as to ensure wide dissemination of the information.

5.3 Encouragement is given to incorporate animal welfare statements into experimental protocols, and in addition to report compliance with these and other appropriate guidelines (including any local ones) when publishing results. Certain journals require this (e.g. British Journal of Radiology, Cancer Research, and Journal of the National Cancer Institute).

Summary and Concluding Remarks

Researchers have a legal and ethical responsibility to consider the welfare of experimental animals in their care. They must decide whether the use of animals is necessary to answer a particular question, and if so minimize the pain and distress involved. Studies in experimental neoplasia present particular problems. Workers should possess adequate knowledge of the animals and tumour systems to be used. Where unfamiliar procedures are to be employed, information and guidance should be obtained through consultation with experienced colleagues and from the scientific literature. Workers should receive appropriate training and supervision. Pilot experiments should be carried out with small numbers of animals, and numbers should always be restricted to the minimum consistent with the design and purpose of the experiment. Tumour end points should be chosen and refined so as to minimize the adverse effects on the host animals. Survival end points are discouraged. In most instances animals should be killed before they become moribund. Repeated draining of ascitic tumours is discouraged, and a general anaesthetic should be used. Ali staff should understand their individual responsibilities, and a clear chain of consultation should be established so that prompt action can be taken to deal with any problems that arise. Finally, researchers are encouraged to refine end points in experimental neoplasia and to publish such improvements, to incorporate welfare statements in experimental protocols, and to report compliance with appropriate guidelines in publications.

Guidelines issued April 22, 1988. Reprinted from Laboratory Animals(22:195-201, 1988) with permission of the publisher (Royal Society of Medicine) and the UKCCCR Secretariat.

Note

Correspondence and recommendations for future editions of the Guidelines should be addressed to: The Secretariat, UKCCCR, The Medical Research Council, 20 Park Crescent. London W1N 4AL, UK.

References

Animals (Scientific Procedures) Act (1986). HM Stationery Office.

British Council Guidelines on the Use of Living Animals in Scientific Investigations (1984).

Fowler ME (1978). Restraint and Handling of Wild and Domestic Animals. Ames: Iowa State University Press.

Gay WI, ed. (1965). Methods of Animal Experimentation, vol. 1. New York: Academic Press.

Institute of Animal Technology (in press). The Principles of Animal Techniques, vol. I.

Royal Society/Universities Federation for Animal Welfare (UFAW) (1987). Guidelines on the Care of Laboratory Animals and their Use for Scientific Purposes Part l--Housing and Care, Royal Society and UFAW.

Tuffery AA, ed. (1987). Laboratory Animals: An Introduction for New Experimenters. Wiley: Chichester.

U KCCCR [UK Coordinating Committee on Cancer Research] (1980). Guidelines for the Xenografting of Human Tumours. [London: UKCCCR.]

Additional References on Animal Tumour Models and End Points

Denekamp J, ed. (1980). Quantitation of tumour response: a critical appraisal. British Journal of Cancer 41, Suppl. IV, 1-331.

Kallman RF, ed. (1987). Rodent Tumor Models. New York: Pergamon Press.

Kallman RF, Denekamp J, Hill RP, & Kummermehr J (1985). The use of rodent tumors in experimental cancer therapy. Cancer Research 45, 6541-5545.

Martin DS, Balls ME, Fisher B, Frei E, Freireich E J, Heppner GH, Holland JF, Houghton JA, Houghton PJ, Johnson RK, Mittleman A, Rusturn Y, Sawyer RC, Schmid FA, Stolfi RL, & Young CW (1986). Role of murine tumour models in cancer treatment research. Cancer Research 46, 2189-2192.

APPENDIX

The following examples of tumour systems are given for illustration.

1. RIF-1 mouse sarcoma. This is a transplantable sarcoma of C3H/Km mice which is widely used in radiation and chemotherapy studies (Twentyman PR, et al. Journal of the National Cancer Institute 64, 595, 1980). It can be maintained in cell culture and is grown in vivo as a solid tumour by implantation intradermally in the skin of the flank or intramuscularly in the hind leg. The end points used to determine therapeutic effects on the solid tumour are clonogenic survival, regrowth delay, and tumour cure. It is common practice to terminate regrowth delay experiments with leg tumours when the limb diameter reaches approximately 16 mm. At this point the tumour mass is about 3 g or about 10 percent of the body weight and the host animals are in otherwise normal condition. Growth delay is determined from the time to reach four times the treatment size. Metastases occur late and rarely.

2. DMBA-induced rat mammary tumour (Huggins C, et al. Journal of Experimental Medicine 109, 25, 1959). Setting a humane end point with this tumour is more difficult. There is heterogeneity in both the number of tumours which develop and in their relative growth rates, so that individual animals may have widely differing tumour burdens. Close daily monitoring is essential and an overall judgment must be made, based on the aggregate tumour mass, the size and condition of larger tumours, and the general health of the animal. While animals may tolerate an aggregate tumour burden of >10% of body weight if there are many small tumours, a single large tumour can lead to rapid deterioration necessitating humane killing of the animal.

3. MAC 16 mouse colonic adenocarcinoma. This is a transplantable tumour of NMRI mice which is normally grown subcutaneously in the flank (Bibby MC, et al. Journal of the National Cancer Institute 78, 539, 1987). It is of particular interest because it causes progressive cachexia and loss of body weight, beginning at a tumour weight of about 100 mg in a 30-g mouse and increasing over the subsequent 7-14 days. The host mice continue to eat normally over this period. The main difficulty in working with the MAC 16 tumour is the heterogeneity of cachecfic response between animals with similar tumour burdens. Because of this, individual animals are weighed at the time of transplantation and then daily thereafter. Mice are killed when the weight loss is between 20 and 40% (maximum). This careful monitoring procedure prevents the occurrence of death due to cachexia.

4. L1210 mouse leukaemia. This is normally grown as an ascites tumour and used for the evaluation of anticancer agents (Geran RI, et al. Cancer Chemotherapy Reports 3, 1, 1972). The difficulties associated with this model are also shared by other ascites tumour models for which the survival end point has been widely used in the past. Cells (routinely 10⁵-10⁶) are injected into the peritoneal cavity of C57BL × DBA/2 F₁ (BD2F₁) mice. A direct relationship normally exists between the number of viable L1210 cells injected, or remaining after drug treatment, and the subsequent survival of the animal. Implantation of 10⁵ 1210 cells, with a doubling time of approximately 12 h in exponential growth, has been shown to produce life-threatening symptoms by the 8th day after implantation. These symptoms are manifested as a marked abdominal distension produced by peritoneal ascitic fluid, dyspnoea, a hunched posture and poor coat quality, particularly a ruffling of the fur, and mild catatonia. As animals approach this phase of tumour growth, twice daily inspection of tumour-bearing animals is necessary to assess morbidity. The therapeutic substance under investigation is normally administered 24 h after the implantation of the tumour, and may be given at subsequent times. However, the protocol may be modified so as to avoid possible temporal overlap of the toxicity of the substance and the symptoms of morbidity induced by the tumour.

5. Rodent tumour metastasis models. Metastasis may be seeded either "artificially" by intravenous injection of tumour cells, or spontaneously after growth of a solid deposit which can be removed surgically when appropriate. Such models include the B16 and other melanomas and UV-induced fibrosarcomas in mice (Kripke ML, et al. Cancer Research 38, 2962, 1978). It may not be necessary to wait until mice develop symptoms of impending morbidity, and the required information may be obtained after humane killing at an earlier stage (see Kripke et al. cited above). Special attention should be directed to detecting signs associated with clinically significant disease in sites particularly susceptible to metastasis, e.g. dyspnoea due to lung deposits.

6. Chemically induced colonic tumours in rats. Tumours of certain internal organs are difficult to detect by external examination. As an example of the use of special diagnostic techniques, colonic tumours in rats can be identified by endoscopic examination (Merz, et al. Hepato-Gastroenterology 28, 53, 1981; Hermanek PJ & Giedl J Pathology, Research and Practice 178, 548, 1984).

7. Neoplasia in transgenic animals. Problems may be encountered when oncogenes are inserted or activated, or indeed other genetic alterations are introduced into recipient transgenic animals. In particular it may be difficult to predict the consequences of such genetic changes, which may occur other than in the particular organ of interest. An example of this occurred in transgenic mice carrying a hybrid gene comprising the murine aA-crystallin promoter fused to the coding sequence of the oncogene SV40 T antigen. Not only did the expected lens tumours develop, but in addition several animals developed nonlenticular tumours at various sites throughout the body (Mahon KA, et al. Science 235, 1622, 1987). Thus special care must be taken to ensure that such associated sequelae are identified and appropriate measures taken.

¹Member organizations of UKCCCR: Cancer Research Campaign, Imperial Cancer Research Fund, Institute of Cancer Research, Leukaemia Research Fund, Marie Curie Foundation, Medical Research Council, Tenovus Cancer Fund. Observers: Department of Health and Social Security, Ludwig Institute for Cancer Research, Scottish Home and Health Department.

Commentary
Stephen P. Tomasovic and Kenneth N. Gray

Dr. Tomasovic is associate professor and section chief in the Department of Tumor Biology, University of Texas M. D. Anderson Cancer Center, Houston. He served as chair of that institution's institutional animal care and use committee (1ACUC) for three years. Dr. Gray is associate professor and deputy chairman of the Department of Veterinary Medicine and Surgery at the same institution and a current member of its IACUC

The humane care and use of animals in experimental studies of cancer is an issue of special concern to the scientists and animal care and use committees of institutions where cancer treatment and research are the major focus. These guidelines for the welfare of animals in experimental neoplasia prepared for the United Kingdom Coordinating Committee on Cancer Research (UKCCCR) are a welcome addition to the limited literature on this aspect of our field of research and complement their previously issued guidelines for the xenografting of human tumors (1980).

In selecting an animal tumor model and in using that model to answer questions related to cancer, researchers make a series of choices that affect animal welfare. This UKCCCR contribution and several other recent publications now assist beginning and experienced researchers in making decisions that can significantly enhance humane care and use of animals bearing tumors while still allowing efficient and accurate cancer research to be conducted.

This enhancement of humane animal care can occur by replacement of animals with alternative experiments, reduction of numbers of animals used, and refinement of our methods of animal use, including steps to minimize any pain, distress, or suffering (Russell and Burch, 1959). The UKCCCR's guidelines primarily focus on aspects of refinement by emphasizing prediction and recognition of adverse effects, and the use of humane endpoints. We agree with the UKCCCR panel that attempts to develop nonanimal use alternatives are to be continued, but that for now many aspects of cancer biology and treatment cannot be adequately addressed without study of animal tumors in vivo. Reduction in numbers is certainly possible in many instances, and we have addressed this point in our own attempts to provide guidelines for researchers using animal tumor models at the M. D. Anderson Cancer Center and elsewhere (Tomasovic et al., 1988). With regard to other aspects of the refinement of methodology relating to the selection of tumor models and their proper use in cancer research, we highly recommend the recent book edited by Robert Kallman (1987), the recommendations from an international workshop on tumors in experimental cancer therapy (Kallman et al., 1985), and the earlier proceedings of the Ninth L. H. Gray Conference on critical appraisal of the quantitation of tumor response to therapy (Denekamp, 1980). All of these publications address experimental tumors of rodents because they are used in the majority of cancer research studies; however, as pointed out in the UKCCCR guidelines, the general principles are relevant to other animals too.

The decision-making process outlined in the general recommendations of the UKCCCR guidelines and the interaction between the personal licensee and the Home Office are reflective of a similar process that should occur in institutions in the United States between the individual investigator and his/her institutional animal care and use committee (IACUC). We would particularly endorse the two general recommendations on pilot experiments and training of all involved staff. If substandard humane animal care and use is to occur in experimental cancer research, the most likely reasons will be insufficient experience and training in the specific experiments and animal tumor models. A strong IACUC can substantially enhance animal welfare by careful review of the proposals for animal use and establishment of adequate training programs.

The specific recommendations described in the UKCCCR's guidelines address assessment of severity, the biology of tumors, establishing humane endpoints, monitoring of animals, and documentation. Successful application of these recommendations is again heavily dependent on the investigator's experience and training with the specific animals and tumor models and the use of small pilot studies.

The use of humane endpoints is of particular concern in at least two major types of cancer research experiments. These experiments determine the extent and time course of toxicity of therapeutic modalities in vivo and whether these modalities have any inhibitory effect on tumor growth and spread in vivo. The usual nature of these experiments often results in significant but unavoidable distress or discomfort, and animals can die from the treatment or tumor effects. The UKCCCR recommendations on humane endpoints emphasize animals should be killed before predictable death occurs, before they exhibit signs of morbidity, or before tumor burden gets too large.

The UKCCCR's guidelines on this topic complement our own efforts to establish guidelines for euthanasia as an alternative to death as an endpoint in rodents (Tomasovic et al., 1988). We have established some objective signs for judging rodent morbidity (disease/ illness) and the moribund condition (state of dying) in rodents, and these are used in our institution to aid investigators, veterinarians, and the IACUC in determining when to perform euthanasia. Although our guidelines refer to rodents, they can be applied to other animal species with species-specific modifications to the signs of morbidity and the moribund condition. As noted in the UKCCCR's guidelines, care must be taken to maintain flexibility, and procedures should be subject to continuous refinement to reflect experience and the specifics of particular animal tumor models.

Several other UKCCCR recommendations on experimental neoplasia deserve emphasis. The guidance on solid tumor burdens not exceeding 10 percent of normal body weight, ascitic volume not exceeding 20 percent of normal body weight, and weight loss not exceeding 20 percent of initial body weight is reasonable and should be used in negotiations for experimental protocol approval between investigators and their IACUC. Investigators should have specific and compelling justifications for exceeding such guidelines. As noted in the recommendations, animals approaching these preset endpoints should be frequently monitored. Problems can arise during weekend and holiday periods, and these problems can be avoided by establishing clear responsibility for inspection during these times and by small pilot experiments to establish protocols that ensure staff will be present as animals approach endpoints.

Monoclonal antibody production is very common in studies of experimental neoplasia, and the UKCCCR recommendation to use retired breeders for production is good; however, there are a number of additional issues related to the refinement of these procedures or the use of alternatives like in vitro production (although the latter can raise issues of humane animal care too). Animal welfare can be significantly affected by the method of inducing the antibody response (e.g., complete Freund's adjuvant versus nitrocellulose strips versus the Ribi adjuvant system) and other methods, including the frequency of drainage and its duration, the needle size used in drainage, and the interval between drainages. The topic is too complex to discuss fully here, but this is a subject IACUCs and investigators should consider carefully in experimental designs. We recommend a recent article on the refinement of monoclonal antibody production published in this forum (McGuill and Rowan, 1989).

The documentation and publication aspect of the UKCCCR's specific recommendations makes several very useful points. The section on the documentation of tumor model use under various experimental conditions is particularly important because, in our view, implementation of that recommendation could be of great assistance to individual investigators and the IACUC. We noted above that experience and training in the specific experiments and animal tumor model under consideration are critical to high standards of humane animal care and use in cancer research. If the IACUCs could develop a file of short descriptions of the biology of animal tumor models in use at their institution and criteria for assessment of severity and humane endpoints similar to those appended to these UKCCCR guidelines, then this data bank could be used by the IACUC in evaluating future animal research proposals and in training of investigators who propose to use those models. We feel this data bank could be readily established in most cases with the assistance of senior researchers experienced with the animals and tumor models under consideration.

Cancer researchers are involved in the development and implementation of state-of-the-art methods of patient care, and this goal still requires the use of animals in their studies. Concomitant legal, regulatory, and moral responsibilities require that we conduct these experiments with minimum pain, distress, or suffering to the animals. The UKCCCR guidelines on experimental neoplasia should be carefully considered by individual researchers and the IACUCs in the United States as these guidelines provide straightforward and useful recommendations that should help establish and continually evolve high standards of animal welfare in these types of experiments.

References

Denekamp, J., ed. 1980. Quantitation of tumour response: A critical appraisal. Brit. J. Cancer 41, Suppl. IV:1-331.

Kallman, R. F., ed. 1987. Rodent Tumor Models in Experimental Cancer Therapy. New York: Pergamon.

Kallman, R. F., J. M. Brown, J. Denekamp, R. P. Hill, and J. Kummermehr. 1985. The use of rodent tumors in experimental cancer therapy. Conclusions and recommendations from an international workshop. Cancer Res. 45:65416545.

McGuill, M. W., and A. N. Rowan. 1989. Refinement of monoclonal antibody production and animal well-being. ILAR News 31(1): 7-11.

Russell, W. M. S., and R. L. Burch. 1959. The Principles of Humane Experimental Technique. London: Methuen.

Tomasovic, S. P., L. G. Coghlan, K. N. Gray, A. J. Mastromarino, and E. L. Travis. 1988. IACUC evaluation of experiments requiring death as end point: A cancer center's recommendations. Lab. Animal 17:31-34.

UK Coordinating Committee on Cancer Research (UKCCCR). 1980. Guidelines for the Xenografting of Human Tumours. London: UKCCCR.