alternatives to animal experimentation in basic · pdf file7.5 education received 9 october...

ALTEX 21, Suppl. 1/04 3

Content1 Introduction2 Basic Research, what does the term mean?2.1 Definitions and dimensions2.2 Delineation from the applied sciences2.3 Alternative methods need to be funded3 Obstacles and opportunities for alternative

methods in basic research3.1 Definitions3.2 Publication of alternatives in scientific journals3.3 What drives and what hampers the progress

of alternatives?3.3.1 Relevance3.3.2 Availability3.3.3 Visibility3.3.4 Awareness3.4 Advantages and disadvantages of in vitro

compared to in vivo models3.4.1 Advantages of in vivo models3.4.2 Disadvantages of in vivo models3.4.3 Advantages of in vitro models3.4.4 Disadvantages of in vitro models4 Examples of published alternative methods in

basic research4.1 In vivo alternatives4.1.1 The incubated chicken egg as an alternative4.1.2 Transgenic animal models4.1.3 Refinement during procedures, e.g. anaesthetics

and analgesics4.1.4 Refinement during procedures, e.g. humane endpoints4.1.5 Refinement during procedures, e.g. computer-

tomographic models4.1.6 Enrichment in animal keeping, considering the

needs of behavioural biology

Alternatives to Animal Experimentationin Basic Research*Franz P. Gruber1 and Thomas Hartung2

1FFVFF, CH-Zurich; 2ECVAM, EU JRC, I-Ispra and University of D-Konstanz*

4.2 Isolated organs as models in basic research 4.2.1 Animal tissues and organs in research4.2.2 Human tissues and organs in research4.3 Cell culture models4.3.1 The possibilities of hepatocyte cultures4.3.2 Cell cultures in cancer research4.3.3 Stem cell methods4.3.4 Genetically modified cells4.3.5 Artificial membranes4.4 3R principles in antibody production4.4.1 A success story: The in vitro production of

monoclonal antibodies4.4.2 Alternatives to come: The recombinant antibody

families4.4.3 In the meantime: Refinement – the IgY antibody4.4.4 Alternative adjuvants4.5 In vitro techniques in parasitology4.6 In silico alternatives5 Further developments and resources for

information systems on alternatives6 Discussion6.1 Does freedom of science hamper 3R methods?6.2 Animal welfare in the constitutional law6.3 Importance of representative statistics on

animal experimentation7 Solutions, medium and long-term development7.1 Harmonisation7.2 Increasing the quality of in vitro methods7.3 Promoting the use of animal-free methods7.4 The pipeline7.5 Education

Received 9 October 2002; received in final form and accepted for publication 28 October 2004

* The texts and suggestions represent the privat opinions of the authors

SummaryIn contrast to animal testing required by law to guarantee mini-mum safety standards for the licensing of drugs and chemicals,there are no regulations in basic research forcing scientists to per-form animal tests. By (usually) free choice, questions are posedand hypotheses are examined which, in many cases, can only beanswered by means of animal tests. Just as easily, different ques-tions could be asked or different hypotheses could be examinedwhich do not require animal tests. The only criterion for the choice

of a topic is its relevance which cannot necessarily be judged inthe short-term. Thus, it is up to the individual scientist to judgewhat is worth studying and therefore worth animal consumption.The educated mind will consider ethical aspects of this choice.However, on the other hand, this decision is largely influenced by questions of efficacy or (in a negative sense) by the obstaclesposed to an animal consuming approach. Here, peer review andgeneral attitude will strongly influence the methodology chosen.Availability and awareness of adequate in vitro techniques rep-resent the prerequisites for the use of alternative methods. The least one can do in basic research is to avoid tests whichcause severe suffering to animals, as is required in Switzerland

GRUBER AND HARTUNG

ALTEX 21, Suppl. 1/044

and other European countries by binding ethical principles andguidelines. The increasing standard of approval and controlprocedures has improved the situation over the years.There are many examples of successful alternative methods inbasic research. But, the application of such methods is in mostcases limited to the laboratories in which they were developed,calling for technology transfer. Exceptions are procedures thatare used worldwide, like the production of monoclonal anti-bodies, which instead of using the ascites mouse can also beperformed in vitro with some good will. In these cases, com-mercialisation of the techniques has aided their spread withinthe scientific community. Sadly, many methods, even if pub-lished in the scientific literature, are little standardised and re-producible. The suggestion is put forward that publicly accessi-ble databases should make available more detailed descriptionsof methodologies. Due to limitations in space, many scientificjournals cannot publish detailed methodological descriptions.However, nowadays a supplementary central deposit of methodscould easily be linked to the respective article. In numerous cases though, there is simply a lack of will to changeprocedures to methods without animal tests or to pose questionsdifferently in order to avoid the use of animals or to reduce theirnumber or, at least, to reduce stress. In other cases, researchersare simply not aware of the limitations of the animal experimentas such. A thorough review of the validity of critical animal experiments should be carried out and made available publicly.For example, many animal experiments are dramatically “under-powered”, i.e. carried out with groups that are too small to allowconclusions to be drawn from the outcome. This stands in markedcontrast to in vitro experiments where replicate experiments usually represent no major problem.Since in vitro models are generally more prone to artefacts due tothe numerous variables, e.g. of cell culture, the key requirementfor their application is their validation and quality control. Guided by the experience from validation studies for alternativemethods in toxicology, concepts of a Good Cell Culture Practice(GCCP) are currently being developed which aim to define min-imum quality standards for in vitro techniques. This initiativeaiming to increase quality must be complemented by a concept tosystematically assess the relevance of the tests in order to finallyachieve an evidence-based biomedical research.A change in this direction is only possible if those public funds,which were previously assigned predominantly to alternatives tothe animal tests required by law, are now channelled increasinglyinto developing those for basic research. A financial incentive isnecessary to change procedures in basic research to animal free procedures. Ethical considerations alone will bring little movement or change. It is unacceptable that, while numbers ofanimal tests decrease in development and notification of drugsand chemicals, they are increasing in basic research. Due to thecentral role of publishing scientific results, the key options forcontrol are the respective rules of journals for the acceptance ofarticles. By demanding certain standards in the instructions forauthors, e.g. of quality (GCCP), relevance and in case of animalexperiments proof that no alternative is available, pressure couldbe dramatically increased. It is suggested to hold a consensusconference of journals in the life sciences on this topic.

Zusammenfassung: Alternativen zu Tierexperimenten in derGrundlagenforschungAnders als bei gesetzlich vorgeschriebenen Tierversuchen, dieeinen Mindeststandard an Sicherheit bei der Zulassung vonArzneimitteln und Chemikalien sichern sollen, gibt es in derGrundlagenforschung keinerlei Vorschriften, die von Wissen-schaftlern den Einsatz von Tierversuchen verlangen würden.Es werden bestimmte Fragen gestellt, Hypothesen geprüft, diesich vielfach nur mit Tierversuchen beantworten lassen. Mankönnte aber genauso gut andere Fragen stellen und andere Hypothesen prüfen, zu deren Beantwortung keine Tierversuchenötig sind. Das einzige Kriterium für die Wahl ist die Relevanz,die nicht unbedingt kurzfristig vorausgesagt werden kann. Wasman untersuchen sollte und ob dieses Ziel den Einsatz vonTieren rechtfertigt, wird also letztlich individuell vom einzelnenForscher entschieden. Dass dabei ethische Überlegungen eineRolle spielen sollten wird beim Bildungsstand der Wissen-schaftler vorausgesetzt. Die Entscheidung wird aber auch vonNützlichkeitserwägungen beeinflusst und – negativ – von denHindernissen, die beim geplanten Einsatz von Versuchstierenüberwunden werden müssen. Die Begutachtungsverfahren vonWissenschaftsjournalen und der allgemeine Konsens beein-flussen dabei die Wahl der Methode erheblich. Vorhandensein,Zugriff und Kenntnisse von ihrer Existenz sind Grundvoraus-setzungen für den Gebrauch von Alternativmethoden. Man kannin der Grundlagenforschung zumindest auf solche Versucheverzichten, die den Tieren schwerste Leiden zufügen, so wie esdie in der Schweiz und in anderen europäischen Ländernverbindlichen ethischen Grundsätze und Richtlinien verlangen.Zudem verbesserte sich die Situation im Verlauf der Jahre durchdie ständige Erhöhung der Standards bei der Bewilligung undKontrolle von Tierversuchen.Es gibt viele Beispiele für erfolgreiche Alternativmethodenauch in der Grundlagenforschung. Doch deren Anwendungbeschränkt sich meist gerade auf das Labor, in dem sie ent-wickelt wurden. Ausnahmen sind Verfahren, die weltweit üblichwaren, wie die Herstellung monoklonaler Antikörper in derAszitesmaus, ein Verfahren, das sich mit einigem guten Willenauch in vitro durchführen lässt. In diesen Fällen hat die Kom-merzialisierung der Methodik stark zu ihrer Verbreitung in derWissenschaftsgemeinschaft beigetragen. Dagegen sind viele Alternativmethoden, obwohl in wissenschaftlichen Zeitschriftenpubliziert, noch immer kaum standardisiert und reproduzierbar.Es wird vorgeschlagen, dass öffentlich zugängliche Daten-banken detailliertere Beschreibungen der Methoden enthalten.Aus Platzmangel können viele Wissenschaftsjournale dieseLeistung nicht bieten. Heutzutage sollte es jedoch möglich sein,aus einer zentralen Datenbank heraus jeden Artikel mit derentsprechenden Methode zu verlinken.Bei nicht wenigen Versuchen fehlt es aber schlicht am Willen,sie durch tierversuchsfreie Methoden abzulösen oder die Frageanders zu stellen, um die Verwendung von Tieren zu vermeiden,stark zu reduzieren oder wenigstens in ihrer Belastung zu ver-mindern. In anderen Fällen sind sich die Wissenschaftler derbegrenzten Aussagekraft der Tierexperimente als solche einfachnicht bewusst. Es müsste eine sorgfältig durchgeführte Studiezur Validität von kritischen tierexperimentellen Studien

GRUBER AND HARTUNG


1 Introduction

Despite the worldwide advancement ofthe 3R principle, the largest fundingbody in German-speaking countries, i.e.the Deutsche Forschungsgemeinschaft(DFG, i.e. the German Research Coun-cil) has not made use of this term up until the year 2001. In a number ofmemorandums, statements and commu-nications about the situation of animalexperimentation, the term 3R cannot befound (DFG 1971, 1981, 1991, 1993,1996). In the meantime the ESF (Euro-pean Science Foundation), the council ofEuropean Scientific Societies in basicresearch, to which the DFG belongs,published a new codex on the 9th ofSeptember 2000, in which the nationalsocieties are called on explicitly to ac-cept and promote the 3R principle. Usu-ally, it takes some years until such rec-ommendations are put into practice. TheDFG is called to adopt these principlesas soon as possible and include them inthe Rules for funding.

Over the last three years the number ofanimals consumed in basic research has

been increasing in Germany, Switzerlandand most probably also in other industri-alised countries. Therefore, it is very urgent to increase the consideration of3R aspects in basic research.

This study aims to illustrate the lack-ing flow of information, which hampersthe establishment of the 3R principle in basic research. Such a review cannotbe comprehensive or even representa-tive, but wants to give examples andhighlight key aspects. The authors wantto stress the promise and merits of the 3R principle of Russel and Burch(1959) as a practical, feasible compro-mise also for basic research. The defini-tion of the 3R concept is still the bestway to practice humane science and the definition of alternative methods isstill the same as in 1959: “All proce-dures which can completely replace theneed for animal experiments, reducethe numbers of animals required, or diminish the amount of pain and dis-tress suffered by animals in meeting theessential needs of man and other ani-mals.” (Balls et al., 1995).

2 Basic Research, what does the term mean?

2.1 Definitions and dimensionsThe Deutsche Forschungsgemeinschaft(DFG, 1993) defines basic research, i.e.science not aiming for immediate appli-cation of results, as follows (literal trans-lation): “This research springs from thehuman thirst for knowledge, it satisfiesour desire to understand the world inwhich we live and is a characteristic traitof the species Homo sapiens. The resultsof this scientific search for knowledgemay be used – or misused – to very different ends. It will not be suppressed.”

The absolute numbers of consumedanimals are increasing in basic science.Table 1 and Figure 1 show this trend forSwitzerland. The proportion of animalsused for basic research has increased between 1994 to 2000 from 14% to 29%.In total, however, this increase has not(yet) tipped the overall trend to decreas-ing animal numbers in Switzerland,though in the canton Zurich, which isstrong in the field of research, this is already the case. In Germany, the pro-

durchgeführt und der Öffentlichkeit zugänglich gemacht werden. Manche Versuche sind schlicht zu klein angelegt, d.h.sie wurden mit so wenigen Tieren durchgeführt, dass bestimmteAussagen einfach nicht gemacht werden können. In vitroergeben sich solche Probleme so gut wie nie, da Versuchs-wiederholungen üblicherweise keine größeren Probleme bereiten.In vitro Methoden begünstigen wegen der hohen Zahl an Variablen häufig Artefakte, vor allem bei Zellkulturen. Um sieanwenden zu können sind Validierung und Qualitätskontrolleunabdingbar. Angeregt durch die Erfahrungen, die bei toxiko-logischen Validierungsstudien gemacht wurden, werden z.Z.Konzepte für eine Gute Zellkulturpraxis entwickelt (GCCP), umwenigstens einen minimalen Standard bei in vitro Technikengewährleisten zu können. Diese qualitätsorientierte Initiativemuss durch ein Konzept ergänzt werden, das systematisch dieRelevanz von Versuchen erfasst, um letztlich zu einer Evidenz-basierten biomedizinischen Forschung zu gelangen.Eine Veränderung dieser Situation kann es nur geben, wenn öffentliche Fördermittel, mit denen bisher hauptsächlich

Versuche zum Ersatz von gesetzlich vorgeschriebenen Tierver-suchen finanziert wurden, vermehrt für solche in der Grund-lagenforschung fliessen. Es muss auch einen finanziellen Anreizgeben, Verfahren in der Grundlagenforschung in tierver-suchsfreie umzustellen; einzig aus ethischen Überlegungen heraus wird sich nur langsam etwas bewegen. Es ist aber nicht hinzunehmen, dass bei einer ständigen Verminderung von Tierversuchen zur Zulassung von Arzneimitteln undChemikalien gleichzeitig die Tierzahlen in der Grundlagen-forschung ansteigen.Da die Publikation wissenschaftlicher Ergebnisse eine zentraleRolle spielt, steht und fällt der Zugang zu jeder Kontrolle mitden Richtlinien, die sich Journale für die Publikation vonErgebnissen geben. Mit der Forderung, in den Richtlinien fürAutorinnen und Autoren gewisse Standards zu verankern, z.Bder Qualität (GCCP), der Relevanz und im Fall von Tier-versuchen der Nachweis, dass keine Alternative verfügbar ist,könnte der Qualitätsdruck dramatisch erhöht werden. Es wirdvorgeschlagen, zu diesem Zweck eine Konsensus-Konferenz derJournale in den Biowissenschaften abzuhalten.

Keywords: animal testing alternatives, 3R methods, basic research, availability, awareness, relevance, publishing rules

GRUBER AND HARTUNG


portion of animals used in basic researchincreased from 14% in 1991 to 28% in1999 (Tab. 2 and Fig. 2). This increasealready resulted in an increase in the to-tal animal use, which will continue ac-cording to experts (BMBF symposiumMunich-Neuherberg, May 2002). TheEuropean statistics offer detailed num-bers for 1999. According to these num-bers, the proportion of animal use in basic research in Europe is 30% (Fig. 3).

The underlying reason for increasinganimal use can be deduced from theSwiss statistics in the years 1992-2003,i.e. transgenic animals, mainly mice,which added on more than 90,000 animals in Switzerland in 2000 (Tab. 1).This trend appears to flatten now.

In principle, animal experiments in basic research can be divided into thosethat deal with the health and well-beingof humans (medical basic science) andthose which serve the general under-standing of nature (biological basic sci-ence). However, these fields cannot becompletely separated, as biological sci-ence also deals with experiments aimedat understanding the healthy or sick hu-man body and behaviour. A scale couldbe devised demonstrating the immediacyresults from scientific research mighthave for human benefit.

Bateson’s cube (Fig. 4) is a typical ex-ample for such a scale of immediacy(Bateson, 1986). It has the advantage thatit reflects the problem of felling a deci-

sion as a problem of weighing out com-peting interests. It uses three criteria tocome to a decision: (1) scientific qualityof the research project, (2) the probabili-ty that the research project will lead toimportant biomedical results and (3) thedegree of stress and suffering of the ani-mals used. There are a number of severi-ty catalogues by means of which thestress experienced by the animals can begauged before and after the experiment.The most commonly used such cataloguein the German-speaking countries(prospective and retrospective) was com-piled by the Swiss BVET (1994).

In principle, even the most stressfulprocedures may be carried out in basicresearch, as the German animal protec-tion law states, “Experiments on verte-brates, which cause significant long last-ing or repeated pain or suffering, mayonly be carried out if the results whichare aimed at allow the assumption thatthey will be of particular importance forconsiderable needs of humans or animalsincluding the solving of scientific prob-lems.” The solving of scientific prob-lems, even if no application that wouldbenefit humans can be recognised, isgenerally imbued with particular impor-tance. Therefore, considerable burdensmust be born, because this is consideredamong the essential needs of humans andanimals. The EU directive 86/609/EECalso allows animal experiments for scien-tific research without requiring a bio-medical background. Until now, limita-tions in basic research have onlyemerged for the use of primates. In 2002,the Scientific Steering Committee of theEU (s. ALTEX 19, 89) for research re-duced the areas in which primates maystill be used to the development of medi-cations and vaccines for the preventionand healing of diseases such as AIDS,BSE, malaria and influenza, i.e. all areasin which direct medical advances are im-plemented. Pure basic research is notrepresented in this list. However, ofcourse, this list only has an advisory, nota compulsory character.

In Switzerland, funding of projects bythe Nationalfonds (the Swiss ResearchCouncil similar to Deutsche Forschungs-gemeinschaft) requires adherence of thescientists to the ethical principles andguidelines first established in 1994 and

Tab. 1: Development of animal numbers in Switzerland

1996 1997 1998 1999 2000 2001 2002 2003total number 509,755 492,186 452,535 445,682 423,127 446,654 485,969 475,455of animalsIn basic 129,016 143,310 117,426 124,118 123,102 115,717 136,952 147,422research% animals 25,3 29,1 25,9 27,8 29,1 25,9 28,2 31,0in basic res.transgenic n.d. 42,750 52,500 56,250 59,250 60,000 67,945 63500animals*% animals n.d. 8.7 11.6 12.6 14.0 13.4 14.0 13.4transgenic

* estimated from diagramm 3.3 in www.bvet.admin.ch/tierschutz/TV-Statistik/Jasta-03/D/33/Grafik.html

Fig. 1: Development of animal numbers in Switzerland

GRUBER AND HARTUNG


amendments published in ALTEX andfurther journals (SAMW and SANW,1996). Paragraph 4.6 of these guidelinesstates that in case of unbearable sufferingof animals the experiments must be declined. In addition, animal experi-ments, which do not concur with theseguidelines may not be exported to othercountries.

While SOPs (standard operation pro-cedures) are mandatory for animal ex-periments required by laws and regula-tions, the variability of the methodologyused in basic research is great. It has noteven been possible to set up uniform protocols for the production of anti-bodies by immunisation of animals. Thespectrum of alternative methods must beappropriately manifold to establish the3R principles in basic research.

2.2 Delineation from theapplied sciencesIn the applied sciences the regulatory ac-ceptance of alternative methods usuallyundergoes a stringent series of tests, fromwithin-laboratory to between-laboratoryto international comparisons and finallyto the drafting of monographs in pharma-copoeias (drugs) or OECD guidelines(chemicals). Such a series of validationtests is not the norm in basic research.Every scientist can believe the results ofhis colleagues or not. He can try out the

Tab. 2: Development of animal numbers in Germany

1996 1997 1998 1999 2000 2001 2002 2003

total number 1,509,619 1,495,741 1,532,572 1,591,394 1,825,215 2,126,5612,212,376 2,112,341of animals

In basic 308,569 314,782 385,993 438,017 679,026 926,294 826,729 850,710research

% animals 20.4 21.0 25.2 27.5 29.1 25.9 37.4 40.3in basic res.

transgenic 156,876 204,386 222,212 254,155animals*

% animals 8.6 9.4 10.0 12.0transgenic

Fig. 2: Development of animal numbers in Germany

Fig. 3: Distribution of animal numbers in the EU in 1999 Fig. 4: Graph according to Bateson (1986). Quality of research,certainty of benefit and animal suffering should be in abalanced relation. When a research proposal falls into theopaque part of the cube, the experimental work should not be done

GRUBER AND HARTUNG


alternative method or not. If the alterna-tive does not work first time round, nobody forces him to try again. The authorities who must give permission forthe animal experiments in basic researchusually have no overview of the existingalternatives. They must believe the statement of the scientist who often denies the existence of alternatives evenif some do exist.

It is alarming that animal experimentsin medical research apparently often donot lead to major results: Lindl et al.(2001) report a review of 51 studiesbased on animal experiments that wereproposed in the years 1991-1993 at aGerman university. First, the declaredaims of the proposals were evaluatedagainst the applicants’ own statementswith regard to the pain suffered by theanimals. Secondly, it was evaluatedwhether these proposals led to relevantpublications in the scientific literatureand whether these results led to improve-ments in the respective fields of research.A ranking was established from zero (noresults were published within 8 years) tothree (all the proposed goals were effec-tively reached and published). Only onethird of the proposals reached gradethree, and for nearly 40% of the proposalseither no relevant publications werefound or these proposals were never setinto practice. Distinction of the researchin applied and basic research revealedthat basic research per se results in amuch higher percentage of grade three(over 56%). The reason for this might bethat basic research always has something

to report, satisfying a scientific journalsince applicability need not be proven.Noteworthy, as a side aspect it was foundthat in 60% of the studies the scientistsunderestimated the actual level of suffer-ing of the animals.

Table 3 attempts to demonstrate the differences between applied, medical andbiological basic research.

2.3 Alternative methods needto be fundedBasic science strongly depends on publicsupport since it has no immediate appli-cation in medicine. No company wouldeasily invest in such projects. Thus, basicresearch very strongly depends on publicconsensus, since in the long term no so-ciety will finance projects, which are notfully in line with law or morale. A promi-nent example is the current discussion onstem cell research. Similar discussionsare needed when techniques, which raiseanimal consumption, are again increas-ingly applied in basic research.

The German Ministry of Research isprobably the largest national supporter ofalternative methods worldwide. From1980-2000 82 million € were spent forthis purpose. In comparison, the GermanResearch Council (Deutsche Forschungs-gemeinschaft, DFB) alone spent 456.8million € in 2002 on biomedical re-search (Bundesbericht Forschung, 2004)

A study evaluating projects on alterna-tives covering a total of 37 million € offunding was reported some years ago(Gruber et al., 1996): According to theirresearch subjects, the projects were clas-

sified by the following categories: Re-duction of distress or of the number oflaboratory animals used in an experiment(10.6%); reduction of the number of ani-mals by performing an animal-freemethod before the animal test (43.1%),replacement of animal test by an animal-free method (18.6%), basic research(18.1%), promotion of its acceptance byoptimisation of an animal-free method(6.2%) and validation of an animal-freemethod (3.3%). It was remarkable thatonly 18.1 % of the budget was given tobasic research. In 2000 a workshop,which was organised by the ministry inBad Godesberg, aimed to develop newguidelines for funding of research on alternatives with the aim to intensify sup-port for basic research. The follow-upworkshop in 2002 in Munich especiallyaddressed approaches to promote the 3Rprinciple in research with transgenic ani-mals. This technology is responsible formost of the increases in animal numbersin basic research. The same ministry hasthus sponsored reduction of animal useand techniques leading to increased con-sumption at the same time.

But exactly those disciplines, whichcontribute most to increasing animalnumbers, discuss replacement least. Ap-parently, the scientific results promisedby genetically engineered animals aretoo tempting. Disappointing for animalwelfare activists, it must be expected thatanimal numbers will – not dramatically,but steadily – increase over the next fewyears. Hopefully, this will not lead tonew clashes between science and society.

Tab. 3: 3R relevant differences between applied regulatory, medical and biological basic research

regulatory testing medical applied research basic research

initiation/reason of science legal aspects orientated on sickness, proof of scientific hypotheseseffects and side effectsof drugs to human beings and animals

type of science standard operating procedures efficiency of e.g. agent exploratorydiscovery

quality control prevalidation, validation evaluation peer review in journals

dispersion of results report to authorities limited since proprietary publicationproducts

consequences permission or non-permission alteration of national and acceptance or rejection of by authorities international pharmacopoeias new hypotheses

GRUBER AND HARTUNG


The portion of funding basic researchreceives from diverse institutions in Ger-man-speaking countries is displayed inTable 4.

The proportion of basic researchfunded by these institutions variesgreatly. This insight should lead to therealisation that increased funding fromnon-governmental funding organisa-tions should flow into 3R relevant basicresearch. Minimising statutory animalexperiments with 3R methods is theobligation of the institutions of state,which require these experiments. Basicresearch does harbour a higher risk ofinvesting money in unsuccessful pro-jects, but that is a risk one must carry:basic research is often high risk and suc-cess can never be guaranteed.

3 Obstacles and opportunitiesfor alternative methods in basicresearch

3.1 Definitions of test systems used as alternatives First it is necessary to define what wemean by alternative methods or in vitromethods and related terms:● In vitro methods comprise every typeof experiment in the life sciences, whichmakes use of living material but does notinvolve animals or humans except asdonors of tissue or cells. The term “ex

vivo” is usually reserved for the pre-treatment of the donor before tissue sampling. Much confusion is caused bythe fact that molecular biologists oftenuse “in vivo” when talking about cells incontrast to the cell-free systems they usu-ally work with. It would be very helpfulto coin a distinction that is generally accepted.● In vitro tests deliver a defined result,i.e. they measure or classify. In contrast,a method could also comprise the pro-duction of materials (e.g. monoclonal antibodies) or descriptive approaches(e.g. histology). The test result may bequalitative (e.g. mutagenic or not), semi-quantitative (e.g. no, weak, intermediateor strong irritant) or quantitative (e.g. ascore, a percentage or any number sum-marising the result). ● In vitro technologies refer to method-ologies, which enable the establishmentof in vitro methods. For example per-fusion culture, immortalisation of cellsor precision-cut tissue slices representtechniques which enable setting up in vitro methods or tests. ● The term animal-free would implythat no animals are consumed, not evenas donors of cells or tissue.● A filter test is a test which functions asa preliminary test to sort substances, e.g.substances to be tested in vivo, becausethe result is foreseeable (e.g. a simple pH test before applying a substance into

the rabbit eye), thereby reducing thenumber of animal experiments.● An alternative is a method or test,which replaces, reduces or refines an an-imal experiment. Therefore, the respec-tive in vivo experiment must be clearlydefined. An alternative can even be an-other in vivo experiment, in which lessanimals are consumed, there is less harmto the animals or e.g. primate use is re-placed by use of “lower” animals.● A validated alternative is a test which has undergone the procedure ofvalidation, which is best defined by theECVAM criteria (Balls and Karcher,1995). This will only be required rarelyin basic research. Noteworthy, only testsnot methods can be validated, since onlythese allow biometrical analysis. The aimof validation is to establish the repro-ducibility and the relevance of a test.This usually requires a collaborativestudy to assess interlaboratory variabilityand a comparison with (historic) in vivodata. The structured process of validationincludes the standardisation of the testand its biometrical analysis. Finally, aprediction model has to be establishedwhich converts the result of the alterna-tive (e.g. 40% of the cells killed) to theestimated in vivo result (e.g. score of theDraize rabbit eye test). This step hasproven to be most difficult because it canbe based on correlation only, which isstrongly biased by the selection of exam-ples the correlation is based on: If the alternative is established for one class of substances, another prediction modelmight be required for the next class ofsubstances.● An approved or accepted alternativerefers to a method, which is accepted byan authority formerly demanding (bylaw, guidelines or simply practice) an animal experiment as an in vitro substi-tute.

3.2 Publication of alternativesin scientific journalsThere is a lot of confusion on the use ofthe different terms. Authors of paperspublished in the relevant journals for al-ternatives (e.g. ALTEX, ATLA, In vitroToxicology, Toxicology in vitro, etc.) of-ten claim to have developed an alterna-tive method, when they have actually established an in vitro model. If the term

Tab. 4: Portion of funding alternative methods in basic research from diverse institutions in German speaking countries (actual status given as number of projects in 2000)

basic applied projects to percentage of research research spread projectsin basic

3R knowledge research

D-BMB1 23 19 0 55

Ö-BMB2 12 1 0 92

B.-W.MW3 6 1 0 86

ZEBET4 14 4 0 78

SET5 8 3 6 47

CH-3R6 13 6 0 68

FFVFF7 1 0 4 201 German Ministry of Research, 2 Austrian Ministry of Research, 3 Baden-Württemberg Ministry of Research, 4 German Center for Documentation and Evaluation of Alternative Methods to Animal

Experiments, 5 Foundation for the Replacement of Animal Experimentation, Mainz, 6 Swiss Foundation Research 3R,7 Swiss Foundation for Animal Free Research

GRUBER AND HARTUNG


alternative referred to anything you cando without using animals, it would in-clude research in history of art. Review-ers and editors should not accept suchphrasings. The term “alternative” shouldbe reserved for in vitro or in silico mod-els, which aim to replace or reduce aclearly defined in vivo experiment. Simi-larly, following the evaluation of theirmodels, authors tend to attribute validityto their method or even term it validated.It is very important to reserve the term“validated alternative” for those meth-ods, which have undergone a thoroughvalidation study, usually in form of a collaborative study, with a positive out-come.

If these distinctions are not appliedcarefully, confusion and too high ex-pectations of users and the public are theresult, which in the long-term turn intodisappointment and harm the field of alternatives. Too often expectations areraised that animals are spared by a newmethod, without considering the pro-longed and demanding procedure of val-idation and acceptance. The unjustifiedclaims are understandable: they are madeto satisfy the demands of funding bodies,journals or award applications. It is thereviewers’ duty to sort out what has real-ly been achieved.

It should also be mentioned that evenpurely methodological basic research,e.g. to increase the reliability of cell cultures, can also represent a contribu-tion to 3R research. It can increase theacceptance greatly, when such methodsare employed for the replacement of ananimal experiment.

3.3 What drives and what hampers the progress ofalternatives?In principle basic research is free, free to choose topics and free to choose ade-quate methods to answer its questions. Itis therefore important to analyse what isdriving and what is hampering the use ofin vitro techniques instead of in vivomodels.

3.3.1 RelevanceJudgement of the relevance of a givenmodel by the individual scientist andhis/her peers is most crucial: The aim is not to perform a certain technique but

to solve a certain problem. Only occa-sionally the availability of a certainmethod, i.e. already established ex-pertise, will prompt continuation and de-termine the subject to be studied. To interrupt such cycles, i.e. to avoid that ascientist with expertise in vivo simplycontinues to work in vivo because this is his/her speciality, it is necessary thatjustifications for in vivo work mustnowadays be given to ethical reviewboards on a regular basis. This makes the scientist argue (and hopefully alsoquestion) his/her approach and the needfor the animal experiment.

Animal experiments are often overesti-mated with regard to their relevance.Firstly, species-dependent differences tothe human situation are often overlookedor neglected. Secondly, animal use hasdeclined considerably due to the fact thatinbred strains reduce variation (as if onlyidentical twins were used in a clinicalstudy), but the resulting limitation in extrapolation of results is usually ne-glected. Thirdly, most animal experi-ments were never properly validated.Fourth, there is considerable underre-porting of pitfalls and limitations of agiven animal experiment (no one dares towrite that the model he/she is workingwith fails, no journal is interested in neg-ative or, even worse, ambiguous results).Fifth, there is a dangerous trend towardsreporting only the best of repetitive ex-periments or even to leaving out thosethat “did not work”. Sixth, most animalexperiments are strongly underpowered,i.e. the number of animals per group isnot adequate for the inherent variability;proper biometrical analysis shows thatconclusions can often not be drawn andsub-selection of results and inappropriatestatistics are run (non-justified assump-tion of Gaussian distribution, comparisonof two groups only where several groupswere run and ANOVA is required, em-ployment of prospective tests for retro-spective analysis, non-justified assump-tion of one-sided test situations or nottaking into account multiple testing); thesame limitations hold true for many invitro tests, though here replications areusually not limiting.

The obvious consequences of this lineof reasoning are threefold: (1) the limita-tions of in vivo models must be made

clear e.g. by re-assessment or direct com-parison with in vitro tests and clinical data; (2) information on the shortcom-ings of in vivo models must be madepublicly available, e.g. as databases; (3)the advantage of easier validation andquality assurance in vitro should beleveraged to ensure a superiority of invitro approaches. Here, the importance ofinitiatives for Good Cell Culture Practice(GCCP) becomes evident (see below).

3.3.2 AvailabilityIn contrast to most animal experiments,most in vitro techniques require relative-ly sophisticated equipment. While manyuniversities offer facilities to carry outanimal experiments, a common infra-structure for in vitro work is rare.

3.3.3 VisibilityThe various databases developed over thelast decade have made available informa-tion on alternatives (Janusch et al., 1997;Janusch-Roi et al., 2000; Grune et al.,2000). However, mere citations and therespective descriptions in the publishedarticles are usually too brief to allow theadoption of a method. Furthermore, it isoften difficult to trace the series ofamendments of a method over time whenonly the original description is found. A database of methods including Stan-dard Operation Protocols (SOP) such as ECVAM’s Scientific Information Service(SIS) is very helpful to enable the suc-cessful transfer of methods. Such data-bases should involve the laboratory thatdeveloped the method to ensure com-pleteness and permanent updating. Bottrill (2002) gives an overview ofavailable databases on the homepage ofFRAME and Grune et al. (2001) listedsearch strategies for 3R-methods on the internet at the Linz 2001 meeting onalternatives.

3.3.4 AwarenessThe individual scientist must be madeaware of the opportunities and advan-tages of in vitro techniques. Therefore,the body of methods and concepts mustbecome an integral part of any educationin the life sciences. It is timely to imple-ment the subject formally in educationprogrammes, install chairs on alternativemethods and establish courses and pro-

GRUBER AND HARTUNG


grammes (including internet-based of-fers). The educated mind is the best mes-senger to disseminate alternatives. Fol-lowing academic education, studentstake over responsibilities in academia,industry and authorities, leading to anamplification of the taught ideas.

3.4 Advantages anddisadvantages of in vitrocompared to in vivo models

3.4.1 Advantages of in vivo models● Homeostasis – the living organismmaintains the composition of body fluidswith efficiency hardly achievable in vitro● Completeness – the in vivo modelcomprises all aspects of adsorption, dis-tribution, metabolism and excretion ofagents and toxins● Inclusion of systemic responses – con-tributions of other organs and especiallyof the immune system are difficult tomodel in vitro● Low-tech – most animal experimentsfor regulatory purposes and many experi-mental set-ups in basic research do not re-quire very expensive equipment. But dueto the creation of transgenic and clonedanimals, the tendency is moving towards atechnically more complicated phase.

3.4.2 Disadvantages of in vivo models● Ethics – consumption and suffering ofanimals require justification● Species differences – considerable dif-ferences are found between species oreven strains of species, especially incomparison to humans● Costs – both the production and thekeeping of animals under quality-con-trolled conditions is laborious and expen-sive. The demanding care and lowthroughput make animal experimentscostly. ● Statistically underpowered – as dis-cussed above.

3.4.3 Advantages of in vitro models● Costs – in vitro models usually requireinvestment, e.g. to install cell culture lab-oratories, but are in general less expen-sive than a comparable animal experi-ment ● Throughput – since in vitro systemscan be automated (e.g. pipetting robots)they allow standardised performance,e.g. required for high-throughput phar-maceutical screening ● Replicates – the number of replicateexperiments is less limited than in vivo● Simplicity – compared to the complex-

ity of a whole organism, cell modelscomprising one or few cell types aremuch simpler, easing interpretation anddeduction of mechanisms● Accessibility – in most in vitro modelsthe experiment can be stopped instantlyor monitored continuously by bringingmeasurement devices into close contact● Reduction of variance – a rank order ofdecreasing variation leads from a naturalpopulation of animals to standardised ani-mals (age, weight, gender) to out-bredstrains to in-bred strains to primary cells tocell strains to monoclonal cell lines; eachstep on this row represents a reduction invariance which in turn makes significantdifferences in experiments more likely (i.e.lower numbers of replicates required)● Use of human materials – a major advantage with regard to extrapolation to the human situation. However, dedif-ferentiation and in vitro artefacts have always to be excluded.

3.4.4 Disadvantages of in vitro models● Artefacts – the multitude of variablesbrought into a given in vitro system (e.g.ingredients of media, devices, intervalsfor culture steps) influence the status ofthe cells, their responses and thus experi-mental results

Tab. 5: Some alternative websites and databases concerning basic research (not exhaustive)

ALTEX (Alternatives to Animal Experimentation, CH) – www.altex.chAltweb (Alternatives to Animal Testing on the Web, USA) – http://altweb.jhsph.eduAVAR (Association of Veterinarians for Animal Rights, USA) – www.avar.orgAWIC (Animal Welfare Information Center, USA) – http://netvet.wustl.edu/awic.htmCAAT (Center for Alternatives to Animal Testing, USA) – http://caat.jhsph.edu/ECVAM (ECVAM Scientific Information Service, EU) – http://ecvam–sis.jrc.it/FBL3R (Foundation Biographics Laboratory 3R, CH) – http://www.biograf.ch/FFVFF (Foundation for Animal Free Research, CH) – http://www.ffvff.chFR3R (Foundation Research 3R, CH) – http://www.forschung3r.ch/FRAME (Fund for the Replacement of Animals in Medical Experiments, UK) – http://www.frame.org.uk/index.htmHSUS (Humane Society of the United States, USA) – http://www.hsus.orgICCVAM (Interagency Co–ordinating Committee for the Validation of Alternative Methods, USA) – http://iccvam.niehs.nih.gov/NCA (The Netherlands Center Alternatives to Animal Use) – http://www.nca–nl.org/

Norwegian Reference CenterNORINA (Norwegian Inventory of Alternatives) – http://oslovet.veths.no/norina/SFRWAE (Swedish Fund for Research without Animal Experiments) – http://www.algonet.se/SET (Foundation for the Replacement of Animal Experimentation, D) – http://www.tierversuche–ersatz.de/UCCAA (University of California Center for Animal Alternatives, USA) – http://www.vetmed.ucdavis.edu/Animal_Alternatives/main.htmZEBET (Center for Documentation and Evaluation of Alternative Methods to Animal Experiments, D) –

http://www.bgvv.de/cms/detail.php?template=internet_de_index_jsZET (Center for the Replacement of Animal Experimentation, A) – http://www.zet.or.at/

GRUBER AND HARTUNG


● Sub-selection – even primary cells re-present a sub-selection of the naturalpopulation, since more robust cells willbe favoured; the situation becomes moreextreme in case of strains or cell lineswhich are usually oligo- or monoclonaland hardly reflect the natural variety; thereduction in variance (see above) endan-gers the relevance of the model● Dedifferentiation – cells taken out oftheir natural environment with variousfeedback signals and metabolic demandstend to loose differentiation, i.e. the spe-cific set of genes expressed to fulfil organfunction are no longer expressed; thisloss of function is further promoted bythe non-homeostatic culture conditions,e.g. the change of medium which sud-denly takes away the conditioned en-vironment, removes all waste productsand replenishes nutrients. Such suddenchanges favour flexible responses andthus less differentiated pluripotent cells ● Long-term exposure and maintenance– so far no routine solutions for long-term experiments, e.g. exposure to lowlevel toxins, have been devised; perfu-sion culture systems are emerging butstill lack standardisation and validation● Cell density – cell cultures are consid-ered dense at about 1-2 million cells perml (in a typical suspension culture); intissue, at least 100fold higher densities

are achieved and cell cultures thus hard-ly reflect the cell-to-cell interaction ofthe organism● Oxygen supply – the lack of access tooxygen represents a major limitation ofcell culture: the oxygen dissolved infresh culture medium is consumed during the first few hours of culture anddiffusion from air to the cells is limited,resulting in anaerobic metabolism andaccumulation of lactate, shifting the pH.

This list is far from complete. It showsthat although much has been achieved toapproximate organ-like cell culture, manytechnological problems have not yet beensolved. This situation calls for basic re-search in cell and tissue culture to improvethe relevance of in vitro models. The clos-er an in vivo-like behaviour is approxi-mated, the more likely the in vivo situationis reflected in relevant test results.

4 Examples of publishedalternative methods in basicresearch

The following methods are presented toshow examples of the variety of alterna-tives published in the field of basic re-search. The examples chosen originatepredominantly from Switzerland, Austriaand Germany and the manuscripts have

been published in ALTEX. However, despite successful publica-

tion of alternatives, the number of animals used in basic research is still increasing. One reason appears to be the enormous specialisation in basic research, employing very individual set-ups and experimental models. Very oftenthe alternative method is only used in theproject it was developed for, but not inother groups working in similar fields.

Table 6 gives an overview of somepublished methods for alternatives in basic research and an estimation of theirchance to be applied world-wide. In contrast to regulatory tests, which haveto be replaced as exactly as possible tominimise exposing consumers or workersto risk, in vitro techniques which onlypartially mirror the in vivo situation alsohave merits in basic research. While aclear-cut prediction model is required for safety tests, i.e. how to predict theoutcome of a clearly defined animal experiment by the in vitro data, only the relevance of data counts in basic re-search. Often the respective animal experiment cannot even be precisely defined, or a whole class of in vivo ex-periments or different species might beused to study similar phenomena. Mostoften, the modern in vitro techniques offer opportunities far beyond what is

Tab. 6: Methods in basic research described in this article and perspectives in the sense of the 3R

Methods description authors state of development

In vivo

Incubated chicken egg Inoculation of tumour cells on the CAM Kunzi-Rapp (1997) (2)

Incubated chicken egg Production of monoclonal antibodies Hlinak et al. (1997) (0)

Transgenic animal models Genetically modified animals (knock out, first described by (3)knock in, humanised animals) Palmiter et al. (1982)

Better anaesthetics Better pain management reduces stress Flecknell (1994, 1996) (3) (9)and analgesics experienced by the animals

Humane endpoints Typical refinement, death is to avoid as an endpoint Hendriksen and Morton (1999) (3) (9)

Computertomographic Non-invasive observation of brain activity e.g.: Lee et al. (1998) (3)methods

Enrichment in Enrichment of the monotonous cage, Weerd et al. (2002) (3) (6)animal housing satisfaction of elementary behavioral needs

Animal tissue

Brain slice technique 30 preparations out of one rat brain Olpe and Haas (1985) (3) (4)

Rabbit aorta organ 10 aorta preparations can be Finking et al. (2000) (3) (5)culture model made from one animal

Isolated porcine organs Normothermic hemoperfused Nogueira et al. (1999) (3)from the abattoir porcine leg

Normothermic hemoperfused porcine heart Ast et al. (2001) (3)

GRUBER AND HARTUNG


Methods description authors state of development

Human tissue

Human vascular cells Transfilter Co-culture system is used for Siegel-Axel (1999) (3) (4)research on atherosclerosis and restenosis

Human cell-based Cell-vessel wall interactions Reininger (2000) (1)cell function analyser (adhesion, emigration)

Human umbilical Detecting anti-endomysium antibodies Izzi (1998) (3) (5)cord sections

Cell culture models

Hepatocyte bioreactors Comparison of the efficiency of four different Nussler et al. (2001) (3) bioreactors

Co-cultivation of hepatocytes with parenchymal cells Zeilinger et al. (2000) (3) (4)

Cancer research Overview Tritthart (1996) (3)

Automatic analysis system to detect the Gutsche and Stelzner (1997) (3)potential of cancer chemotherapy

Rodent fibroblasts and embryo cells are used to Combes et al. (1999) (3) determine the carcinogenic potential of inorganic chemicals

Stem cell methods Embryonic stem cells (ES) are used to determine Overview by Wobus (3) (9)embryotoxic endpoints and neurotoxicity given in Linz 2001

(s. ALTEX 3/2001)(Seiler et al. 2001)(Arnhold et al. 1998)

Genetically modified cells Insertion of “human” attributes into immortalised cells Doehmer (2001) (3) (4)

Artificial membranes Monolayers or sophisticated co-culture models Overview by (3) (5)allowthe simulation of transport mechanisms across Gindorf et al. (2001)biological barriers

Antibody production

In vitro production of Different cell culture methods are available to Overview by (3) (9)monoclonal antibodies produce very different amounts of antibodies Marx et al. (1997)

Recombinant antibodies Immunoglobulin chains are cloned and recombined Overview by (3) (5)in a bacteriophage vector Stadler et al. (2002)

Use of chicken IgY Specific antibodies can be obtained from the Overview in (3) (6)antibodies (together egg yolk of immunised chickens ALTEX Suppl. (1996)with altern. adj. andappropriate housing)

Use of alternative Replacement of the “hard” Freunds adjuvans by Overview in Leenars (3) (5)adjuvants and moderate “soft” alternative adjuvants is mostly sufficient et al. (1999)immunising methods

Parasitology

In vitro techniques in In vitro cultivation of arthropods or other Falcone et al. (1995) (3) (6)parasitology parasites is often possible Kuhnert (1996)

Matthes and Hiepe (1996)Kuhnert et al. (1998)

In silico alternatives

CADD – computer The Ehrlich’s key and key-hole principle has Vedani (1994) (3) (4)assisted drug design consequences for QSAR methods (quantitative Vedani and Dobler (2001)

structure action relationship)

(0) method has shown no further consequences; (1) method is restricted to the developing laboratory; (2) some other laboratories have adopted the method; (3) method is of great international interest; (4-8) method has already reached different steps of evaluation/validation; (9) method is evaluated/validated; (10) no more animals are in use for this purpose

GRUBER AND HARTUNG


feasible in vivo. Considering them as amere replacement is underestimatingtheir potential. This potential is the majordriving force for change – a scientist willbe reluctant to change his method if thisdoes not entail further advantages. Thus,modern in vitro techniques can be morethan alternatives to animal experiments,but they profit considerably from estab-lishing relevance and reproducibilitysimilar to a validation procedure. Itmight be necessary to coin a new termi-nology, e.g. speaking of valid in vitrotests instead of validated alternatives.

4.1 In vivo alternatives

4.1.1 The incubated chickenegg as an alternative methodThe incubated chicken egg can be used tostudy ocular and mucosal toxicity, angio-genesis, aspects of tumour biology or the effects of heavy metals, functional pa-rameters of the cardiovascular system andeffects of drugs on this system. The em-bryo is insensitive in the early stages of in-cubation; sensitivity develops stepwise on-ly starting around day 7, thus this model isan alternative method in accordance withthe 3Rs if employed within the first weekof incubation (Rosenbruch, 1994, 1997).

The model has been employed to char-acterise tumour growth (Kunzi-Rapp,1997), replacing the use of nude mice.The extra-embryonal vascular system of

the chorioallantois membrane (CAM) of-fers an excellent substrate. Vascularisedorganotypical tissues develop from cellsuspensions applied to the membranewithin a few days. Solid tissue sections,e.g. human biopsy material, are attachedto and supplied by the vascular system ofthe chicken embryo within 48 to 72hours. Results from this model correlatewell with cell culture observations aswell as with in vivo animal experiments.Shoin et al. (1991) used the HET CAMas chemosensitivity test for malignantglioma tumors. To predict the efficacy ofanticancer drugs they graftet human tumors on the chorioallantois membraneof chick embryos (see Fig. 5). In 78% instances the assay response correspond-ed to a clinical partial response, true neg-ative resonse occurred in 100% of allcases.

The CAM method is accepted world-wide for investigations on angiogenesis,however its acceptance and use for theother applications appears only slowly tobe growing. It can be used as an interme-diate assay between cell culture and ani-mal experiment, e.g. for dose-finding ortighter screening of selected substances,reducing the number of animals requiredfor subsequent experiments.

Hlinak et al. (1994) employed the incubated chicken egg in order to pro-duce monoclonal antibodies to avoid theascites mouse system. This approach,

though pragmatic at the time, was quick-ly eclipsed by in vitro methods (see below).

4.1.2 Transgenic animal modelsDelpire and Balls (1998) report the results of an ECVAM workshop aboutthe use of transgenic animals in the European Union. It was obvious that possible advantages as some aspects ofrefinement are threatened by transgenicprocedures which may promote greateranimal use, more-varied applications,and the greater likelihood of animal suf-fering. The generation of transgenicmouse models is the main reason for thecurrent increase in the number of experi-mental animals. However, in certain cases, transgenic mice are considered bysome to be promising tools for refine-ment. The use of transgenic mice insteadof primates could involve less stress tothe animals; transgenic animals with increased susceptibility could shortentoxicity studies. There are some positiveexamples of transgenic animals allowingthe use of fewer animals in total to answer a specific research question (Gruber, 1998). On the other hand, largenumbers of animals are required to “pro-duce” transgenic strains. Also, it is notclear whether these animals suffer morefrom diseases than their wild-type coun-terparts. Thon et al. (2002) showed that36% of the transgenic strains which werereported experienced discomfort (21%minor and 15% severe discomfort). Inaddition 30% of the strains were reportedto have increases in mortality, disease in-cidence and susceptibility to disease. Themost frequently mentioned conditionsbeing increased mortality, decreased fer-tility and diabetes.

However, it should be noted that mostof these animals are required for breed-ing of the transgenic animals and not foranimal experiments. The often disablingeffects of gene knock-outs and knock-insmust not be neglected, but the sufferingcaused is considerably less than in mostother animal experiments. Nevertheless,a discussion has to be initiated urgentlyon how animal numbers can be made todecline again and, as a minimum imme-diate first step, how suffering of these animals can be minimised. Especially,the precision of the microinjection tech-

Fig. 5: Schematic drawing a chemosensitivity test system using the chorioallantoismembrane (HET CAM) of fertilised chicken eggs (Shoin et al., 1991)

GRUBER AND HARTUNG


nique requires drastic improvement to reduce the animal numbers.

A comprehensive form for the stan-dardised documentation of transgenic rodents (genotype, phenotype, welfareassessment, recommendations for refine-ment) was developed by Mertens andRülicke, 2000. They argue that duringthe “creation” of transgenic mousestrains it is impossible to predict theirphenotype with certainty, particularlywith respect to welfare. The form is subdivided into basic and detailed infor-mation and can be updated continuouslyin the form of a computerised database,incorporating growing knowledge andexperience with the strain. The completeform can be downloaded from www.altex.ch/news_e.htm#topics.

With regard to the 3Rs, activities areunderway to define criteria for the dis-continuation of experiments (refinement)and to target gene insertions more effi-ciently (reduction) in order to limit increases in these animal numbers.

Anyway, it must be evaluated on an individual basis to what extent transgenicanimals can contribute to the 3Rs.

4.1.3 Refinement duringprocedures, e.g. anaestheticsand analgesicsBetter anaesthetics and analgesics are theclassical refinement methods. Continuousimprovement of these procedures leads toa significant reduction of stress in the ani-mal experiment (Flecknell, 1994, 1996).Required by law in almost all countriespractical application is often still inade-quate. Authorities and animal welfare offi-cers are responsible to demand and en-force the newest insights and progress intheir institutes. Sadly, there are often greatdiscrepancies between state of the art anddaily practice. Overview in Flecknell andWaterman-Pearson (2000).

4.1.4 Refinement duringprocedures, e.g. humaneendpoints Hendriksen and Morton (1999) editedthe “Proceedings of the internationalconference on humane endpoints in ani-mal experiments for biomedical re-search”, which contains much informa-tion on refinement measures intoxicology and points out how changes

can be made towards more humane end-points in basic research. The Swiss “Eth-ical principles and guidelines” of theAcademies of Natural and Medical Sci-ence (SAMW and SANM, 1996) statethat a procedure, which leads to severesuffering, cannot be weighed up by gainof knowledge and should therefore be ab-stained from. Similarly, Morton (1999)wrote, “... when the level of suffering isso high that it is simply wrong to causethat degree of harm to an animal...”. Aninteresting phrase was coined by Balls(1999). He does not speak of “humaneendpoints” but prefers the wording “lessin-humane animal procedures”.

The estimation of pain and sufferingstill plays a central role in the humaneendpoints. Therefore, it is important toknow how experienced authors see thepossibility of objective evaluation, e.g.Wolfgang Scharmann (1999), “Physio-logical and ethological aspects of the as-sessment of pain, distress and suffering”;Jones, Oates and Trussell (1999), “Anapplied approach to the assessment ofseverity” and Lloyd and Wolfensohn(1999), “Practical use of distress scoringsystems in the application of humaneendpoints”. Wallace (1999) considers hu-mane endpoints in cancer research.

The animal experiment or ethics com-missions or the animal care committeesplay a pivotal role in the definition of cri-

teria for the cessation of the experiment,which should be a central aspect of allapproval procedures (Mench, 1999).

To clarify the situation of defining humane endpoints, Figure 6 shows thechronological events which should trigger the premature termination of anexperiment. Clear termination criteria already defined in the approval of an animal experiment, regular health checksand the continuous presence of a compe-tent and decisive person in charge of theexperiment are important to ensure atimely and animal protection relevant termination of the procedure.

4.1.5 Refinement duringprocedures, e.g. magneticresonance imaging A major challenge for basic research willbe the development of methods to re-place implanted sensors in brains of ani-mals, especially primates, by imagingmethodologies. While in the EU a ban ofexperiments on primates is discussed(Balls, 1995), and some philosophers donot accept invasive procedures on pri-mates at all (Flury, 1999) some researchinstitutions (for example the Universityof Bremen) are expanding their respec-tive primate facilities in order to accom-modate neurophysiologists. The problemof replacing the recording of single cellpotentials from defined brain areas by

Fig. 6: Humane endpoints. The point to end an experiment has to be choosen as near as possible to point (A) and as far as possible to point (B)

GRUBER AND HARTUNG


imaging techniques is still the insuffi-cient resolution of current methods.Therefore, novel approaches to gain neu-rophysiological data by non-invasivemethods require our specific attentionand support.

Lee et al. (1998) studied the changes inthe intrinsic magnetic resonance (MR)relaxation parameters of the marmosethead, imaged before and after death.Knowing the absolute values of the MRparameters made it possible to choose animaging protocol for optimal structuraldifferentiation. The changes between theante-mortem and post-mortem MR pa-rameters provide an insight into thechanging biophysical microenvironmentof the post-mortem brain, and allowsome of the changes that occur in patho-logical conditions to be predicted. Diffu-sion-weighted MR imaging (MRI) wasused to map quantitative apparent diffu-sion coefficient values, and to investigatediffusional anisotropy along the fibre

tracts in pre-mortem and post-mortembrain tissue. A three-dimensional set ofdata of the entire marmoset brain demon-strates the ability of three-dimensionalMRI to differentiate internal brain struc-tures. MRI is a non-invasive techniquethat, in principle, permits the same ani-mal to be re-imaged serially and has thepotential to probe in vivo brain structuraland biophysical changes over an ex-tended period of time. Serial imaging, inwhich each animal acts as its own con-trol, reduces the number of animals re-quired to detect a significant change byminimising inter-subject variance. MRItherefore provides important scientificand ethical benefits.

As important as the introduction ofthese methods is, it cannot substitute forsingle cell measurements required bysome scientists. A further disadvantage isthe expense of this method. However, ifone takes the German Animal ProtectionLaw seriously, the replacement of an

animal experiment may not be put off be-cause the alternative is more expensive,requires more time or more personnel. (§ 9,2, Ziff. 3, TierSchG, 25.5.1998).However, it has many advantages overthe invasive method too. FunctionalMagnetic Resonance Imaging (fMRI)like it is described by Motzkus et al.(2004) even can be used for objective butstressfree pain research.

4.1.6 Enrichment in animalkeeping, considering the needsof behavioural biologyBreeding, transport and keeping of ani-mals have also become targets for mea-sures of refinement. The EU directive86/609/EEC sets minimum standards forexperimental animals in the memberstates. However, already at implementa-tion in 1986, the directive did not reflectthe state of the art standards of be-havioural science but represented a har-monisation at the lowest level. Discus-

Tab. 7: Comparison of the answers of some ethologists and laboratory animal scientists to six specific questions about better housing of rodents

answering Question 1: Q2: Is it allowed Q3: Is there Q4: Is there Q5: What are Q6: Do there scientists Do there exist to compare evidence evidence for the reasons exist debatable

definitively behaviour of for impaired bad scientific not to alter recommendations identified etho- laboratory housing results caused housing for better housing?pathies? animals with conditions? by wrong conditions?

the wildtype? animal housing?

A yes makes no sense yes yes no reasons yes

B difficult yes, but carefully rats and mice no adaptation over yeshave a good 100 generations,adaptation hygiene

C not yet evident shall not be not yet evident, enrichment is only it has to be discussedusedas recom- maybe Hurst recommendable mendation for if real improvementlaboratory animals is to await

D no no no sure validity, noevidence disposability,

practicability, lower costs, missing regulations

E no no yes yes

F some hints yes yes mice may be lower costs, yesmore active, practicability, results therefore but problematic more reliable standardisation

G no (rodents) it is the only no yes, in neuro- hygiene, yespossibility physiology oeconomics

H not evident on a maybe in quality difficult international yesscientific base regulations

I yes yes yes yes yes

K sometimes has to be principally yes no, but mistakes all over approved noconsidered but possible but need for not for recom- better evaluationmendations

GRUBER AND HARTUNG


sions on updating the guideline are ongo-ing, but this is made difficult by widelydiffering opinions of scientists (Tab. 7).

This controversy prompted a workshopinitiated by the Swiss Foundation 3R andthe German Foundation “set” whichreached the following results (s. Tab. 8).

Animals can perform more natural,species-specific types of behaviour in“enriched” environments. Enrichmentprovides the animal with a more stimu-lating and responsive environment, suchas cage mates, nesting material, hidingplaces. It is postulated that animals fromenriched environments cope better withnovel situations, and thus might offermore relevant experimental results withless biological variation (Van der Weerdet al., 2002), despite some studies claiming the contrary though based onless relevant endpoints (Gärtner, 1999).Hopefully, the EU directive will be updated as promised, as this might alsoaffect animal keeping outside of Europe.

It appears to be feasible to achieve amajor advance in appropriate conditionsfor animals by refining standards forhousing laboratory animals. In Figure 7 itis shown that previous quantum jumps inrefining laboratory animal husbandry tothe currently existing regulatory guide-lines largely have been based upon labo-ratory animal science, whereas consider-

ation of ethological knowledge recentlyhas raised the awareness for the need forfurther improvements (Gruber, 1996).

4.2 Isolated organs as modelsin basic research

4.2.1 Animal tissue and organs in researchThe use of isolated organs in experimen-tal procedures is one of the most effective

methods to reduce the number of animalsand the severity of stress. Animal organscan be obtained from laboratory animalsor from slaughterhouses.

One of the first ex vivo models de-scribed in ALTEX was the brain slicetechnique (Olpe and Haas, 1985). Up to30 preparations can be made from thebrain of one rat. These can still be usedfor experiments hours later. Finking et al.(2000) established an organ culture

Tab. 8: Methods and results of a workshop about better housing of laboratory animalsorganised by the Center for Laboratory Animal Husbandry and Welfare (KLab) at the Institute of Animal Sciences, ETH Zürich. Participantswere scientists, representatives of the two funding bodies, and legal authorities from the Swiss Federal Veterinary Office. Participantsagreed on the following priorities for future research:

♦ Behavioural ecology of laboratory animal species, mechanisms underlying behaviour, cognition, interindividual differences, non-invasive indicators of impaired welfare, and husbandry effects on experimental results.

♦ Multidisciplinary and interdisciplinary approaches should be pursued to use present knowledge most effectively.

♦ However, it was also agreed that a framework was needed to facilitate research into how animals’ behaviour in their natural habitatpredicts their environmental needs, and how these needs are expressed as both suffering and pathologic changes when they are notmet by the environment.

♦ As a result of a follow-up meeting with leading ethologists in 1998 (again organised and chaired by the KLab), a research strategy iscurrently being developed with the goal of identifying reliable behavioural indicators of impaired welfare.

Conclusions and relevance for 3R♦ The relationships between suffering and pathology need to be understood from an evolutionary point of view.

♦ It is proposed that natural selection favours the conscious emotional experience of threats to survival and/or reproduction (e.g.pathological changes), if, and only if, such experiences are instrumental in avoiding fitness costs by means of behavioural changes.

♦ This view implies on the one hand that conditions leading to pathological changes do not necessarily induce suffering.

♦ On the other hand, suffering may occur under conditions that do not induce pathological changes.

♦ Thus, whereas some pathological changes may be irrelevant with respect to animal welfare, the mere absence of pathologicalchanges does not guarantee good welfare.

♦ In consequence, behavioural changes associated with threats to fitness (e.g. avoidance behaviour) and behavioural signs that theanimals’ attempts to adjust their behaviour have been stymied (e.g. stereotypies) represent the most direct indicators of sufferinginduced by inappropriate husbandry conditions.

This workshop was sponsered by the German Foundation “set” and the Swiss Foundation Research 3R

Fig. 7: “Quantum Jumps” in the refinement of laboratory husbandry by breedingmethods, laboratory animal science und ethology (Gruber, 1996)

←

←

→

GRUBER AND HARTUNG


model (rabbit aorta) to investigate post-injury estrogen effects in the vessel wall.As 10 aortic segments could be takenfrom one aortic vessel, the number of animals necessary for an in vivo experi-ment could be reduced 10-fold and thestress was reduced to zero. The isolatednormothermic hemoperfused porcineleg, a model to investigate transdermalresorption and vasoactive substances,was first described by Nogueira et al.(1999). Legs and blood can both be obtained from a slaughterhouse. Fromthe same group (Ast et al., 2001) camethe normothermic perfused porcineheart, developed for the experimental de-termination of QT-prolongation in ECG,which can occur as a side effect of different drugs. This could replace invivo experiments on 1500 dogs per yearworldwide in addition to many more rodents (Bruno Christ, personal commu-nication). Pittermann et al. (2000) pub-lished “Proceedings of a workshop onisolated perfused organs”, which washeld 1998 in Hamburg during the 36thScientific Meeting of GV-SOLAS. It wasapparently that with abattoir material theuse of laboratory animals can be reducedand that kinetic parameters of drugscould be measured much more precisely.

Sadri (2002) wrote, “Isolated organperfusion technology offers many advan-tages over other experimental models,such as using live laboratory animals,conducting human clinical trials or test-ing cell cultures. For example, if a re-searcher is studying the beneficial or toxic effect of a particular drug, then perfusing and testing an isolated viableorgan is far more likely to generate un-ambiguous results than using more traditional research methods. Similarly,dosage studies conducted on a perfusedorgan are far more likely to produce reli-able data.”

4.2.2 Human tissue and organs in researchHuman tissue can be used in basic re-search for the prediction of efficacy (e.g.receptor binding assays and enzyme in-hibition), but also for the prediction ofsafety as well as for metabolism andpharmacokinetic studies (e.g. with hu-man microsomes). However, human tissue should only be handled by ade-

quately protected and trained personnel.In some European countries (UK,Netherlands, Germany) the patient mustagree explicitly to the use of the tissuefor scientific purposes, in other countriesthe patient must explicitly refuse(France, Belgium, Poland).

In an ECVAM workshop (Anderson et al., 2001) the creation of an ENRTB(European Network of Research TissueBanks) was suggested, which should ensure high quality standards to ensurereproducibility of results with respect toviability and characterisation. IndividualRTB’s should promote the validation ofalternative methods involving in vitro human tissue models.

Three examples of alternative methodsare presented here.

Siegel-Axel et al. reported in 1999 onan in vitro method of transfilter co-cul-tures of human vascular cells with bloodcells, imitating the morphology of the arterial vessel wall to investigate antipro-liferative and anti-migratory therapeuticstrategies for atherogenesis and resteno-sis. The transfer of the results to the hu-man situation on the basis of these in vit-ro studies seems to be greater than withmost animal models. Other institutes haveadopted this method successfully.

Reininger (2000) described a humancell function analyser (CFA) – a physio-logical in vitro vascular model withwhich cell-vessel wall interactions (ad-hesion, emigration) and cell-cell cohe-sion (aggregation) can be assessed. Thethree components of Virchow’s triad arepresent in a highly standardised and variable form. Specimen fixation and his-tomorphological analysis after the exper-iment is possible. The efficacy of themethod for testing of platelet functionhas been verified by numerous clinicalstudies. The wide variability of test parameters suggests CFA to be suitablefor in vitro analysis of other cell – vesselwall mediated processes, such as inflam-mation, wound healing and tumourmetastasis. However, the CFA providesno direct in vitro alternative to animal ex-periments and several efforts to establishthe method at other institutes were un-successful.

Use of human umbilical cord sectionsto detect anti-endomysium antibodiesfor celiac disease screening was de-scribed by Izzi (1998). The original as-say was limited by the use of endangeredspecies tissue (monkey distal oesopha-gus) as well as by the high cost of commercial kits. The new test is basedon immuno-fluorescence and an ELISA.However, despite the good results of the new test, both kits are still available,as the killing of animals for experimen-tal use is not considered an animal ex-periment in many countries and so theestablished alternative method is notmandatory.

4.3 Cell culture modelsCell culture models as typical in vitromodels represent the backbone of the alternative methods. Here especially theobservation holds true that successespublished in the relevant journals are notnecessarily coupled with a real reductionof the number of experimental animalsused in basic research. The reason isprobably that technically we are still verymuch at the beginning of the culturing ofcells and organs and very often a purelymethodological advancement wakes thegreatest hopes that this method may atsome stage be able to replace animal ex-periments. It is very difficult to separatelegitimate hope from mere assertions,

GRUBER AND HARTUNG


which are made to ensure further financ-ing of the project. Especially privatefoundations cannot be admonishedenough, not to invest funds into suchpurely methodological projects withoutproper expert evaluation.

4.3.1 The possibilities ofhepatocyte culturesThe suitability of hepatocyte culturemodels to study various aspects of drugmetabolism was the topic of an ECVAMworkshop (Blaauboer et al., 1994) andwas reassessed by Nussler et al. (2001).Isolated rat hepatocytes have been in-creasingly used as a model to identifypharmacological and toxicological re-sponses to drugs. However, they retainmost of their functions only for a shortperiod of time. Therefore, numerousmodels and techniques have been devel-oped to study and improve the metabol-ic capacity of hepatocytes in vitro overan extended time period. Nussler et al.compared four different cell culturemodels and found that hepatocytes cul-tured within a collagen sandwich or in a3D-membrane bioreactor retained drugmetabolism over a longer time period.The development of these reactors hasnot yet advanced far enough to allowjudgement of their 3R relevance. It hasbeen shown that primary human livermay also be cultivated successfully, i.e.including maintenance of cell-specificfunctions, for a number of weeks (Wanget al., 2002). Methods to isolate, cultureand differentiate liver stem cells formetabolism studies are being devel-oped. The controlled proliferation anddifferentiation of human liver stem cellsoffers the possibility of sufficientlyavailable and standardisable culturesand could eventually lead to studiescontributing to the replacement of ex-perimental animals in pharmacologicalmetabolic studies. It is unknown howmany animals are used worldwide forsuch studies. A similarly structured butsignificantly simplified model has nowalso been developed, which can be operated in standard incubators and iscurrently in the test phase.

A liver cell co-culture of hepatocytesand liver macrophages, the Kupffercells, was established in order to modelinflammatory processes (Hartung and

Wendel, 1993). Mediators released byKupffer cells when stimulated by bacte-rial endotoxin induced hepatocyte celldeath. The inflammatory cascade wasfurther approximated by introducingneutrophilic granulocytes into the co-culture (Sauer et al., 1996). The modelwas pre-validated with more than 100pharmacological agents in collaborationwith industrial partners. The overall correlation of in vitro results with the an-imal experiment were remarkable, bothwith regard to effects of substances andmechanisms. This project shows thateven complex inflammatory reactionscan be modelled in vitro.

4.3.2 Cell cultures in cancerresearchAn increasing variety of in vitro test sys-tems is used in cancer research to detectmutagenic or carcinogenic factors, to improve diagnosis and therapy and tostudy cancer cell functions and meta-stasis (Tritthart, 1996). These in vitrotechniques are not only superior to someanimal experiments but often are theleading techniques. Improved 3D cellculture techniques, studies of membraneand organelle functions as well as molec-ular biology and genetic techniques are

discussed in light of cancer researchwithin strategic recommendations of theEuropean Community (Europe AgainstCancer). The multistep cascade of eventsduring metastasis can be studied elegant-ly by in vitro techniques, making manypainful animal experiments avoidable.Metabolically competent human cells aresuperior tools to uncover carcinogeniccompounds and mutagenic factors.

Gutsche and Stelzner (1997) devel-oped an automatic analysis system ofsuspended and adherent cells to detectthe potential of cancer chemotherapy.Without the use of animal experiments,they were able to exclude dangerous substances and to find new pharmacolog-ically active agents in cell cultures. They also provided information on the cyto-static activities of the agents. Their end-points were the statistical distributions ofthe cell diameters of K-562 and L-929cells using an electronic cell analyser(CASY1).

An ECVAM Workshop on “Cell Trans-formation Assays as Predictors of Hu-man Carcinogenicity” was held in 1999(Combes et al., 1999). A new, mechanis-tically based, in vitro strategy involvingBalb/c 3T3 clone A31-1-1 mouse em-bryo fibroblasts was proposed at thisworkshop for the determination of thecarcinogenic potential of inorganicchemicals, in order to establish priorityof metal compounds to be tested and,whenever possible, to compare the in vitro results with the corresponding in vivo data. After harmonisation and stand-ardization of the Balb/3T3 protocol, cells were exposed for 72 hours to a fixeddose (100 µM) of 58 different metal compounds. The cytotoxicity induced by some metal compounds was found tobe related to their chemical structure.The results of the systematic study on themetal-induced cytotoxic effects in theBalb/3T3 cell line could be arbitrarilyclassified into three groups according tothe degree of cytotoxicity (Mazzotti etal., 2001). In the ECVAM workshop pro-tocol, the potential of mouse fibroblastsand hamster embryo cells to be used togauge the carcinogenic potential of met-al compounds is described. The possibil-ity of automation and the costs are dis-cussed. But regrettable the 3R potentialis not discussed or estimated.

GRUBER AND HARTUNG


In 2004, an ECVAM expert meeting onthe cell transformation assay was con-vened, which reviewed the status of de-velopment and recommended a pre-vali-dation of both the Syrian hamsterembryo cell (SHE) test and the Balb/c3T3 assay, which is currently organised.In parallel, an OECD test guideline waselaborated, and therefore these novel as-says promise to complement soon the invitro battery for the assessment of car-cinogenic health effects.

4.3.3 Stem cell methodsGreat successes have already beenachieved with animal stem cells, whichcan develop into all 3 germ layers. AnnaM. Wobus received the Felix-Wankel-animal protection-research prize in 2001for her contributions to this field. HumanES would offer an even greater researchpotential, but the ethics of their use arestill under public debate.

The developmentally controlled ex-pression of tissue-specific genes, pro-teins, ion channels and receptors duringstem cell differentiation is the basis ofseveral in vitro approaches: 1) “loss of function” assays with ES cells

containing homozygous mutations ofspecific genes,

2) “gain of function” assays with EScells overexpressing exogenous genes,

3) developmental analysis of terato-genic/embryotoxic compounds in vitro,

4) pharmacological assays and theestablishment of model systems forpathological cell functions, and

5) the application of differentiation andgrowth factors for induction of se-lectively differentiated cells which, inthe future, may be used as a source of tissue grafts.

Strübing et al. (1995) established an invitro model system for the differentiationof synaptically coupled neurons frommouse embryonic stem cells (BLC-6),which is more efficient than using prima-ry embryonic cells from animals or tumour cell lines. The electrophysiologi-cal characterisation of the BLC-6 derivedneurons shows that these cells carry thecomplex electrical properties of post-mitotic neurons and are coupled by in-hibitory and excitatory synapses. Arn-hold et al. (1998) also used BLC-6 cells

to study neurodifferentiation and neuro-toxicity in vitro. Characterisation of neu-ronal cells revealed characteristic pheno-types equipped with neurotransmitters,e.g. glutamate. The effect of glutamateintoxication by the receptor agonist NMDA during early neuronal develop-ment could be studied by videomicro-scopic time lapse in these cells.

Guan et al. (1999) also used embryon-ic stem cells to establish standardisedprotocols for cardiogenesis, myogenesis,neurogenesis and vascular smooth mus-cle cell differentiation.

The use of stem cells to establishmolecular endpoints for the screening ofembryotoxic chemicals was reported bySeiler et al. (2001), and Buesen et al.(2004). This test was developed at ZEBET (Berlin) and was successfullyvalidated by an ECVAM study (Gen-schow et al., 2002, 2004). Contractingheart cells differentiated from ES wereused to assess the embryotoxic potentialof test compounds and compared to cytotoxic effects in differentiated 3R3mouse fibroblast cells.

4.3.4 Genetically modified cellsThe usual immortalised cell lines havemuch reduced metabolic capacities.Therefore, either primary cells are used,which require the killing of animals, orleftovers from operations in the clinicsare employed, or another way is opened:Cellular gene technology, improvementof the metabolic capacity by humangenes. The possibilities for combina-tions are endless. Doehmer received the FISEA Prize for his geneticallymodified V79 hamster cell cultures in1990. The methodology is now usedcommercially (www.genpharmtox.de)and promises a further reduction of ani-mal experiments in drug screening butalso in basic research.

Since 1986, V79 Chinese hamstercells have been genetically engineeredby introducing rat and human enzymesfor studies on metabolism-dependent ef-fects in toxicology and pharmacologyunder defined conditions (Doehmer andJacob, 1994). The metabolic activationof potentially harmful chemicals is of in-terest in toxicology, because metabolitesmay have their own toxic potency. An-abolic and catabolic pathways of drugs

are of interest in pharmacology, becausemetabolism has a decisive impact on theefficacy of drugs. As an example for theusefulness of V79 cells genetically engi-neered for metabolic competence, andtheir advantage over animal experimen-tation, metabolic activation of polycyclicaromatic hydrocarbons has been studied.Comparative studies using V79 cells expressing rat and human enzymes re-vealed striking differences in themetabolite profile, which made it evi-dent that human beings are more at riskthan rats, when exposed to polycyclicaromatic hydrocarbons. This suggests atthe same time that results obtained fromanimal experimentation must be inter-preted with caution if they form the basis of risk assessment. Doehmer(2001) suggests that all V79 constructstogether form the V79 Cell Battery®

which is capable of checking drugmetabolism early in preclinical drug de-velopment with immediate relevance forthe clinical testing of drugs. It would beimpossible to obtain all these data fromanimal testing. In this way, the intrinsicproblem of extrapolating data generatedfrom animal testing to humans has beencircumvented in this instance.

4.3.5 Artificial membranesFundamental biopharmaceutical parame-ters like transport across different biolog-ical barriers are essential to understandthe availability of drugs in an early stageof development, as any pharmaceuticallyactive agent must be able to overcomethe body’s natural protective mecha-nisms. Far-reaching replacement of allanimal experiments with in vitro meth-ods is to be expected, already because ofthe great efficiency of the new methods(high throughput screening).

Gindorf et al. (2001) carried out an ex-cellent survey on the possibilities to sim-ulate transport mechanisms across bio-logical barriers in vitro. They comparedcell culture models for the gastrointesti-nal barrier, the blood-brain barrier andthe alveolar epithelium of the lung. In order to respond to the flood of new ac-tive ingredients currently being generat-ed by combinatorial chemistry or molec-ular biological synthesis, selectionprocedures must be able to filter out drugcandidates with the highest development

GRUBER AND HARTUNG


potential rapidly and economically. Abroad variety of biological barriers canbe simulated in the laboratory by cellmonolayer models. Apart from ethicalaspects, the advantage of these in vitrotest systems is that permeability studiescan be performed at high throughputrates under controlled and reproducibleconditions. The validity of such modelsis ultimately reflected in their ability toaccurately predict the behaviour of activeingredients at the corresponding in vivobarrier.

Especially the blood-brain barrier hasgained more importance in the last yearssince neuropathological symptoms areon the advance, due to more elderly people in our population. There is agreater need to check the brain availabil-ity of substances, which was done untilnow in very detailed animal experiments.Bartmann et al. (2001) reported a verysophisticated artificial blood-brain-barri-er model consisting of co-cultures ofcapillary endothelium cells and astro-cytes, which gave good correlations within vivo data.

4.4 3R principles in antibodyproductionThere is hardly a discipline in basicbiomedical research that can do withoutthe use of specific antibodies. As the production of antibodies in animals wasoften not considered an animal experi-ment, it went almost unnoticed by the

regulatory authorities and the practice ofcruel forms of immunisation, e.g. in theball of the foot or using Freund’s adju-vant (FA), are still practiced in somecountries. Often, the regulatory authori-ties do not know about available alterna-tives and believe the scientist’s statementthat the method is necessary without requiring proof that one of the many established refinement methods was at-tempted.

It should also be considered that ex-tremely high antibody titres are often notrequired and it is also often overlookedthat the antibody titre is not the true con-trol for the success of the immunisation,because the real goal is to have immu-nised spleen cells for a fusion to becomehybridoma cells. Ferber et al. (1999)showed that the alternative adjuvantpApU (polyadenylic-polyuridylic acid)can have a better fusion efficiency thanFA (see Fig. 8) but has a much lowerhistopathology score.

4.4.1 A success story: The in vitro production ofmonoclonal antibodiesAntibodies are very useful tools to detectdifferent structures in basic research,clinical diagnostics and also therapy.Monoclonal antibodies (mAb) were firstdescribed by Köhler and Milstein (1975)and were originally produced in ascitesmice. Ascites induction causes very se-vere pain and distress to the mice used.

In most cases, non-animal or in vitro al-ternatives can be employed to reduce oreliminate the use of animals for mAbproduction. After a number of years ofsingle national initiatives to ban the pro-duction of monoclonal antibodies usingthe extremely painful in vivo method (ascites mouse), an ECVAM workshop in1997 brought the break-through (Marx etal., 1997). It could be proven that thereare enough in vitro laboratory methods tocater for all requirements regardingquantity and quality. Meanwhile lists ofmanufacturers who produce monoclonalantibodies solely by in vitro methods can be found at www.frame.org.uk/ardfmab.htm. The painful method of pro-duction using the ascites mouse is nowonly performed in backward countrieswithout any legal animal protection. Thenumber of animals spared is enormous,millions of mice per year.

Hendriksen (1998) therefore discussesa ban of the ascites method in the wholeof Europe, as is the case de facto in Germany, Switzerland and the Nether-lands. He writes that “prohibition of theuse of animals in the production of mAbsis recommended, except when the re-placement in vitro methods prove to beinsufficient, and in a limited number ofother well documented cases, such as an exceptional need for an emergencytherapeutic application”. He emphasisesthe establishment of core facilities for in vitro mAb production.

Even in the USA, where mice are notprotected under animal welfare legisla-tion, the production of monoclonal anti-bodies in ascites mice is no longer “stateof the art” and governmental agenciessince 2000 do not give money to researchprojects in which ascites mice are used, sothat one can speak of a de facto ban.

4.4.2 Alternatives to come: Therecombinant antibody familiesAt the Linz 1995 congress and at the 3rd

World Congress in Bologna, Stadler(1995) and Stadler et al. (2000) describedthe replacement of monoclonal anti-bodies by recombinant antibodies. Thistechnique has developed into a useful approach to get antibodies of nearly allspecificities. It is based on the cloning oflight and heavy immunoglobulin chains,which are subsequently recombined in a

Fig. 8: Fusion efficiency quantified as number of growing clones/fusion (see Ferber et al., 1999)

GRUBER AND HARTUNG


bacteriophage vector. The source of the immunoglobulin-variable genes canbe the genome of an animal or a human,immunised or after normal exposure ordisease.

Furthermore, hypervariable regions ofthe immunoglobulins can also be artifi-cially inserted, so the generation of syn-thetic antibody libraries is possible. Theadvantage of the methodology lies in thefact that these antibody libraries have acomplexity of 108 in the case of immunelibraries or 1012 in the case of syntheticlibraries, a much greater range than canbe isolated by hybridoma technology.Stadler writes that the new technologystill needs a higher degree of laboratoryskills than the production of monoclon-als. He also considers that recombinantantibodies cannot end the use of animalsfor the production of antibodies com-pletely. Antibody production that assiststhe understanding of pathophysiologicalprocesses in animals could not be feasi-bly performed in humans. And animalswould still be needed for reagents thatare not immunogenic in humans or thatwould be unethical to use in humans. Butthese are minor exceptions. The future ofantibody production in basic researchpossibly brings a nearly complete renun-ciation of animal use in the next 10-15years – if scientists support these devel-opments. This may bring a reduction inthe numbers of laboratory animals in di-mensions not known before. Estimationstally a number of several million animalsper year worldwide.

4.4.3 In the meantime:Refinement – the IgY antibodyImmunisation of chickens and extractionof antibodies from egg yolk belongs tothe alternative methods since the animalsuffering is reduced by non-invasive anti-body-sampling (Gassmann et al., 1990;Schade and Hlinak, 1996). One year lat-er Schade et al. (1997) wrote “It has beenknown for over a century that specific an-tibodies can be extracted from the eggsof immunised chickens. However, it wasonly when animal welfare became a sub-ject of public debate that the chicken wasconsidered as an alternative source of antibodies due to the possibility of non-invasive antibody sampling. Unfortu-nately, the welfare of animals alone is not

sufficient to attract the interest of scien-tists, it is therefore important to demon-strate to potential users that avian anti-bodies can be used successfully in avariety of scientific investigations”. Aconsiderable quantity of antibodies canbe obtained from one chicken. Thismethod is quick and simple to perform.The animal welfare advantage is that thechicken must not be bled as the antibod-ies can be isolated from the eggs. But asthe chicken still has to be immunised, itis not an animal-free method. ALTEXdedicated an entire supplement to theproduction of avian antibodies (ALTEX,1996). But it should be reiterated thatthere is no refinement in the productionof avian antibodies if the chickens are notheld in an absolutely chicken-friendlyenvironment, which allows them to liveout all their behavioural specialities(Scharmann, 1996). And, of course, it is not a refinement method if adjuvantsare used which cause adverse effects in the chicken or which are given by aninhumane route (i.e. Freund’s adjuvant given intramuscularly or intraperitoneal-ly).

4.4.4 Alternative adjuvantsCurrent immunisation procedures are often based on habit and tradition. Veryoften the “hard” Freunds adjuvant is used but “soft” alternatives would alsodo (a list of “soft” alternatives is given inLeenars et al., 1999, Appendix 2). Theprocedures need to be harmonised toguarantee a maximum of refinementworldwide. Great consideration shouldbe given to the appropriate housing ofimmunised rabbits. There is no scientificreason to lock these animals (usually rab-bits) in cages. Keeping them on theground in small groups increases theirquality of life significantly.

4.5 In vitro techniques inparasitologyAlmost the most forgotten animals in science are those not used to produce scientific results but “only” to multiplycertain organisms in the field of para-sitology. Animals used for the growth ofparasites suffer much more than general-ly believed. Great efforts are required to promote the development of in vitrocultures. These are often not undertaken

for reasons of convenience. The “re-placement” potential is very large.

Falcone et al. (1995) described the invitro cultivation of Brugia malayi, a parasitic nematode that causes humanlymphatic filariasis. Kuhnert (1996) described results on in vitro feeding ofthe hard ticks Ixodes ricinus, Boophilusmicroplus, Amblyomma variegatum andAmblyomma hebraeum with bovineblood and presented an in vitro test forsystemic acaricides. Larvae of B. micro-plus were bred to the unfed adult stage.Females gained half of the natural bodymass and laid eggs. The lifecycle of A. hebraeum was completed in vitro. Thereproductive capacity of most adult tickspecies was strongly impaired, possiblyas a consequence of the low quality of the blood. This method offers new invitro possibilities for tick rearing as wellas for tests with systemic acaricides, anti-feedants and repellents and allowsexperiments on the transmission ofpathogens.

In later experiments Kuhnert et al.(1998) showed the successful feeding ofmales and females of A. hebraeum in apartly automated, thus essentially main-tenance-free, in vitro system. The feed-ing data did not differ significantly from

GRUBER AND HARTUNG


that achieved with bovine hosts.Matthes and Hiepe (1996) reported on

the laboratory breeding of lice, fleas andother bloodsucking arthropods. To date,it is not possible to replace rabbits as natural blood sources for breeding lice colonies (Pediculus spp.). A so-called “artificial dog” has been devel-oped for in vitro breeding of cat fleas(Ctenocephalides felis). This may re-place numerous animals used as hosts of laboratory flea colonies in pharma-ceutical companies and other researchlaboratories.

4.6 In silico alternativesLike key and keyhole, biologically activesubstances coming into contact with receptors on cell surfaces either fit andhave an effect or do not. Vedani (1994) is one of the pioneers in this field whodeveloped a model, which allows the calculation of effects even if one of thereacting partners, the receptor, is notknown. This technique is called pseu-doreceptor modelling, and was a veryoriginal concept within the field of Com-puter-Aided Drug Design. It allows thereconstruction of the three-dimensionalstructure of an unknown bioregulatorbased on the structures of its ligands(known bioactive compounds). It com-bines present techniques but significantly

extends their possibilities by the genera-tion of an explicit receptor model. Thismodel may subsequently be used for the qualitative prediction of the bindingstrength of novel drug molecules. Therelevance of pseudoreceptor modellingfor reducing and replacing animal mod-els is given by the fact that the techniquecan be applied under circumstanceswhere little or no information is availableon the true biological receptor, a situa-tion where to date only in vivo tech-niques were promising. As a first tech-nique, pseudoreceptor modelling allowsthe screening of potential drug moleculesbinding to an unknown bioregulator.

The newest project of this group is thecall for a co-operation to establish an internet laboratory for predicting harmfuleffects triggered by drugs, chemicals andtheir metabolites (Vedani and Dobler,2001). Presently, the database of Vedaniand Dobler includes models for only fivebiological targets. It will be extendedcontinuously to include surrogates forany bioregulator known or presumed tomediate harmful effects. Free access tothis virtual laboratory shall allow any interested party to estimate the harmfulpotential of a given substance prior to itssynthesis. See also www.biograf.ch.

QSAR methods have become very important in pharmacological basic re-

search. They have contributed signifi-cantly to reducing the animal numbers in the pharmaceutical industry. Thesemethods are faster, cheaper and more efficient. Possible metabolic products ofthe test substances must also be consid-ered in the QSAR methods in future. It isto be hoped that the side effects of sub-stances will eventually be predictable.The potential for reduction of experi-mental animal use by QSAR methodslies in the region of millions of animalsper year. According to an EU press re-lease (16 October 2003), the introductionof QSAR methods could substitute foranimal experiments costing 950 millionEuro in the EU programme REACHalone. According to a German chemicalcompany, the introduction of QSARmethods could lead to a reduction of the experimental animal numbers by afurther 30% within a few years. Alreadythe use of isolated organs had previouslyalso reduced the animal numbers by 30%(see Fig. 9).

5 Further developments and resources for informationsystems on alternatives

The importance of databases to find pos-sible alternatives easily and fast canhardly be overestimated. Only with suchdatabases can the regulatory authoritiesbe given the right instruments to makecompetent decisions about the approvalof animal experiments. However, simplelists of alternatives for a specific proce-dure are not enough. Comments on thefeasibility and relevance of the alterna-tive methods must be added. While theaccepted methods for mandatory drugand chemical testing are eventually setdown in pharmacopoeias or OECDguidelines, these databases represent theonly possibility to promote the use of 3Rmethods in basic research.

In 2000, ZEBET (German Centre forthe Documentation and Validation of Alternative Methods) at the Federal Institute for Consumer Health Protectionand Veterinary Medicine (BgVV) put theZEBET-database on alternative methodsto animal experiments on the internet viaDIMDI, the German Institute for Medi-cal Documentation and InformationFig. 9: Development of laboratory animal numbers at a German chemistry plant

Jahr

%

GRUBER AND HARTUNG


(www.dimdi.de) (Grune et al., 2000). Ac-cess to the database is free, moreoverDIMDI’s complete service is available to visitors of the ZEBET-database. The ZEBET-database contains documents on alternatives to testing in animals,which have been carefully evaluated byZEBET’s staff according to the 3R con-cept. In addition, the ZEBET-databaseprovides information on the current stageof development and validation of amethod and on its acceptance for eitherscientific or regulatory purposes. Eachdocument is characterised by the follow-ing criteria: the title of a method, key-words, assessment, summary and biblio-graphic references. To searchDIMDI’s-database and host system“grips” soft- ware must be used. Current-ly about 125 alternative methods meetthe criteria of the ZEBET-database. In-ternational fellow organisations, e.g.FRAME (Fund for the Replacement ofAnimals in Medical Experiments) in theUK and CAAT (Johns Hopkins Centerfor Alternatives to Animal Testing) in theUSA, have established links on their websites to pro-vide visitors free access to the ZEBET-database.

Similarly, the European Centre for the Validation of Alternative Methods(ECVAM) has established a database onalternatives, which was recently released(http://ecvam-sis.jrc.it/).

A number of amendments with respectto databases on alternatives might be en-visaged for the future:● The different databases should be in-terlinked.● They should include Standard Opera-tion Protocols in collaboration with thedeveloping laboratory in order to easetaking over the method.● The databases should be extended topromising approaches which are not yetevaluated.● A search in the databases should be-come mandatory when asking for ap-proval of an animal experiment.● Guidance for searching these databas-es should be developed.

Furthermore, it is suggested to estab-lish in parallel databases, which listdrawbacks and limitations of in vivomodels. Similar to gene banks, free entries of comments and additional infor-

mation from scientists should be possi-ble, since limitations and shortcomingsare usually underreported.

A new study which should analysehow the results of a research for publi-cations on alternatives indicated that thecurrent indexing systems do not providethe required information. Not all the relevant information is indexed under“alternative methods”. So the partici-pants of a workshop held at ZEBET inBerlin 2004 developed recommendationsto develop suitable search strategies onalternative methods for scientists (Gruneet al., 2004).

6 Discussion

6.1 Does freedom of sciencehamper 3R methods?Contrary to the animal experiments re-quired officially, which are listed and described clearly in the pharmacopoeias,DIN / ISO instructions and OECD guide-lines, in basic research we hardly everfind any laboratory that carries out theidentical animal experiment as an anoth-er laboratory, even if working on thesame subject. In the regulatory animalexperiments, the solution to overcomeanimal consumption is clearly defined,i.e. the acceptance of a valid alternativemethod, but in basic research, scientistsmostly do not even know what the termvalidation means. This was one of the results of an assessment of researchfunding for replacement of animal exper-iments in basic research (Gruber etal.,1996).

Only a few techniques like the in vitroproduction of monoclonal antibodies(see 4.4.1) have really succeeded. Most-ly, the application of the alternative tech-nique only takes place in the institutionin which it was developed; more oftenthe application of the alternative methodin this institution is given up after chang-ing employees. This development mustbe paid increasing attention especially bythe allocation of research means. Noth-ing and nobody can force a scientist inbasic research to develop an alternativemethod or to apply a method developedelsewhere. Smallest differences can al-ways be found in the researched materialor in the collected data, which from a le-

gal point of view allow the continuationof an animal experiment.

Where are the difficulties in basic research, why are animal tests not dimin-ishing like in the field of applied sci-ences? How shall validation work in basic research? Hartung and Spielmann(1995) wrote about the experience of re-cent attempts to validate alternativemethods to animal experiments, pointingto the differences in the field of safetytesting (toxicology) and identification ofputative drugs (pharmacology, basic re-search) and stating that these require dif-ferent validation strategies. In the end, itis most important in basic research thatthe alternative method is relevant andmore efficient than the respective animalexperiment. It does not have to be the exact mirror of the animal experiment,which is usually required for safety tests,where the fear is that a change in themethod will increase the risk for people.Thus, beside availability and visibility ofthe method, which is best achieved whenmarketed by a company, the limitationsof the animal experiment and the advan-tages of the replacement must be madeclear. Therefore, on the one hand system-atic reviews of the methods are required,ideally made available both in paperform and in a database, and the animalexperiment itself should be included andreassessed in every validation study.There have already been examples inwhich the gold standard, i.e. the animalexperiment to be replaced, turned out tobe less golden than expected. If possible,this should be done without carrying outadditional animal experiments, by em-ploying either historical data or takingadvantage of animal experiments whichare currently carried out anyway and op-timally running the in vitro test in paral-lel on the same material. If neither can beachieved, this might indicate that the re-spective animal experiment is seldomdone or that there is little willingness toreplace it by the user – either should bean indicator that there are more pressingareas for the development of alternatives.

Freedom of research and teaching isembodied in the German Constitution(the European Constitution is expected tocontain at least a paragraph on the free-dom of research). Nevertheless, freedomof science in most cases is only written

GRUBER AND HARTUNG


on the paper, even at universities. The ar-guments for the freedom of science arebased on assumptions, which largely nolonger correspond with the social realityof the research. The respective article inthe constitution, once promulgated in or-der to protect science from governmentalarbitrariness, still implies a powerfulstate and a delicate plant called science.Meanwhile, rather powerless states areentirely helpless when faced with the in-novations of an international science.Bayertz (2000) writes (literal transla-tion): “The delicate plant called sciencehas grown into a massive tree whichthrows its shadows on the entire commu-nity. Today, the community is influencedand changed more by scientific-technicalinnovations than by political decisions.Similar to the economy (and in a paralleldevelopment) it has become a democrat-ically uncontrollable social-politicalpower. The key problem and challengelies in the control of science as a subsys-tem of modern society. Not only, but to agreat extent, finding a solution for thisproblem will determine whether the so-cial evolution grasps humanity like a nat-ural force and drags it along or whetherthis evolution is steered by sensible andhumane criteria. Currently, there are fewindications that this problem can besolved, either theoretically or practical-ly.”

In case of animal experiments the 3Rprinciple has offered a compromise be-tween science and society with regard toanimal use for many years. Unfortunate-ly, this principle is not even mentioned inthe memorandums of large German sci-entific societies. A change in attitude isrequired following the recommendationsof the European Science Foundation ofthe year 2000 in order to promote the application of the 3R principle.

But there are exceptions in which sci-ence decided to forego some of the basicrights of its freedom. In Zurich the academic teachers decided to waive ex-periments involving excessive suffering ofanimals, even when useful results can beexpected. This is a new attitude in basicresearch. Details can be found in the “Listof animal tests which are no longer al-lowed at institutes of higher learning inZurich” (Task force for animal welfarequestions at Zurich University and ETH,

Dietz et al., 1997). Following a series ofintensive discussions, the members of the task force and representatives of SwissAnimal Welfare Organisations agreed on aso-called “negative list”. Operations andtreatments listed therein are not to be per-formed any longer, even if a high gain ofscientific knowledge is to be expected.This is in accordance with paragraph 4.6of the “Ethical principles and guidelinesfor scientific animal testing” of the SwissAcademy of Medical Sciences and theSwiss Academy of Natural Scienceswhich states that animal testing must notbe performed if such testing leads to severe suffering of animals. This list is up-dated at regular intervals.

6.2 Animal welfare in theconstitutional lawAfter several frustrating attempts animalwelfare has now been given constitution-al rank in Germany. However, this doesnot alter the fact that there is a primacy of the freedom of basic research over animal welfare. According to the opinionof the authors it would now be muchmore effective to authorise freedom ofresearch and teaching with a legal provi-sion. A mention of animal welfare in Basic Law in the medium-term does notyet mean basic rights for animals. Al-though legal conflicts on controversialanimal experiments (e.g. the Berlinjudgement) could turn out differently,more and more decisions on individualcases would have to be enforced. A pro-vision in that point is necessary whichgives the animal welfare law the meaningthat it should have according to the opin-ion of the legislator. In any case, a legalprovision should be inserted in article5/III of the constitution of the GermanFederal Republic supplemental to themention of animal welfare.

In Switzerland animal welfare has hadconstitutional rank for many years. In legal practice, this has not resulted in obvious improvements. In Switzerland –different to Austria and Germany – theanimal was an object until 2002. In theSwiss animal law the vertebrate animal isprotected from pains, sorrows, injuriesand fears, but its life is not protected. Thekilling even of a primate is not againstthe law as long as it is done properly.

1998 Caspar published an article “An-

thropocentrism versus Pathocentrism –On the Integration of Animal Protectioninto the System of Safeguarding BasicDemocratic Rights”. Modern animal pro-tection laws must be firmly rooted in adecisive pathocentric (similar treatmentfor similar suffering capacity) fundamen-tal declaration. In accordance with this, itis not a case of protecting human inter-ests concerning a particular way of treat-ing animals but rather of protecting ani-mals from the exploitation interests ofhumans. In order to attain a balance be-tween animal protection and the lawspertaining to animal use, the regulationsgoverning pathocentric animal protectionmust rest on a legal foundation backed byconstitutional law. In the final instance, astatutory framework either in the form ofan objective legal obligation or – betterstill – in the form of subjective-legal ani-mal protection must be considered. Thelatter solution has the advantage that thelaw enforcement deficits of the authoritiesresponsible for animal protection may besupplemented by the fiduciary safeguard-ing of animal rights by a third party.

In non-European countries animalrights are frequently discussed very in-tensively but are often tied up with cer-tain cognitive abilities of the animals(primates), a debate that rather appears inthe light of an extended speciesism dis-cussion. In those countries terms like the“dignity of the creature” are not under-stood. It is astonishing that some US societies of science now as before refuseto give rodents and birds the ”status of ananimal” at all, fearing that these labora-tory animals would then have to becounted in statistical surveys. Superfi-cially it is argued that this is a financialproblem (Goldberg, 2002).

6.3 Importance ofrepresentative statistics onanimal experimentationRevised tables for the collection of statis-tical information on the number of ani-mals used in experiments and for otherscientific purposes in the EuropeanUnion are finally available (Casati andHartung, 2003). The new tables to beused for future statistics on animal exper-iments in the European Union presentedby the European Commission have beensignificantly extended and improved in

GRUBER AND HARTUNG


comparison with the old ones. Severalcategories of little relevance and unclearexpressions have been revised. However,some important data (inter-alia cate-gories for pain and distress as well as forseveral purposes of use; the origin ofsome animal species; types of institu-tions) are still not required, even thoughthey are highly relevant with respect toanimal welfare aspects and must be regarded as indispensable for a targetedapplication of the statistics to set prioritiesconcerning the 3Rs.

7 Solutions, medium and long-term development

7.1 HarmonisationStandards of animal welfare are very dif-ferent in various countries: While scienceis a very international business, the stand-ards for reasonable and humane controlof animal use in basic research have beenlittle harmonised yet. The situation thatSwitzerland calls for “the dignity of (all)animals” and that in the US birds, miceand rats are not even considered to be animals is unsatisfactory. The majority ofEuropean countries with a more humaneanimal welfare legislation, however,should not be brought down to a lowerharmonised international level. In orderto implement fully humane practices inscience, it is necessary to join the pro-gressive forces of all countries includingthose lagging behind. Self-restriction ofscience could be a driving force of theseunifying efforts.

In the next years, it should be possibleto develop a standardised practice for hu-mane treatment of laboratory animalswithin the EU. In addition, the new in-structions of the European Science Foun-dation of the 9th of September 2000 haveraised the hope that basic research will atleast accept the 3R principle. This mayhowever only be understood as a transi-tional phase. In the long-term, only thesuperiority of in vitro tests over animalexperiments will lead to their application(see below). The shortcomings and fail-ures of specific animal experiments inbasic research have to be stressed in or-der to provoke rethinking. It will takegenerations of scientists to change the

general attitude as to the value of animalexperiments.

It is of critical importance that researchis not exported to countries with lessstrict regulations with regard to animalwelfare. Where this cannot be achievedby harmonisation, binding rules set byauthorities and funding bodies must en-sure this. It is not acceptable that stand-ards are circumvented by “the help offriends abroad”.

7.2 Increasing the quality of in vitro methods The most convincing argument for a sci-entist to employ a given method is its rel-evance and the likelihood of convincingresults. Quality assurance is therefore ofutmost importance. Based on this idea,the initiative was taken to establish aguideline of Good Cell Culture Practice(GCCP). In contrast to Good LaboratoryPractice (GLP), which is broadly estab-lished in industry and is hardly feasiblein academic research, only the minimumrequirements for quality control are de-fined. In 1999, on occasion of the 3rd

World Conference on Animal Use and itsAlternatives in the Life Sciences, a work-shop and subsequent plenary session endorsed a declaration towards the estab-lishment of GCCP. A taskforce was in-stalled which worked on the drafting ofsuch guidelines. The respective taskforcereport is available (Hartung et al., 2002).The idea shall be promoted by an ECVAMworkshop to develop the guidelines further. In a broader sense, this idea ofquality control leverages the experiencesmade in validating alternative methodsfor the whole field of biomedicine. Never before have similar efforts beenundertaken to validate in vitro test sys-tems and establish both reproducibilityand relevance. This concept parallels thedevelopment towards evidence-basedmedicine in the clinics: Only by provingthe validity of results – whatever effort ittakes – meaningful science can beachieved.

7.3 Promoting the use of animal-free methodsIn many areas the field calls for in vitromethods; an earlier analysis of the areaof pharmacology by Hartung and Wendel

(1998) came to a positive overall assess-ment: Asking the question “is replace-ment of animal experimentation in phar-macology a goal or a social constraint,utopia or reality?”, the authors come tothe conclusion that “the judgementcomes clearly in favour of the alternativemethods, needless to say in a field wheremost animal experimentation occurs.”They further “conclude that the disci-pline profits considerably from switchingto in vitro models, even though the com-plete replacement of animal experimen-tation would be unrealistic”.

Science follows the mechanisms of themarket, since it aims to achieve the re-sults with least investment with regard toresources (personnel, consumables, time,equipment). The price for animal experi-ments might thus be raised considerablyto reduce animal experimentation. Thismight include animal welfare require-ments, required training of personnel, de-lays in approval of animal experimentsand bureaucracy of approval.

The availability and visibility of alter-natives must be increased by promotingtechnology transfer. So far, the term ispredominantly employed for transfer ofmethods from academia to industrial use.The Steinbeis technology transfer centreInPuT (In vitro Pharmacology and Toxi-cology) in Konstanz, Germany, hasshown over the years that this can beachieved efficiently and profitably. Incase of basic research, similar technolo-gy transfer must be enabled, but herepublic sponsoring is required in contrastto industrial financing, which is morethan willing to pay for a relevant method.Databases with indispensable cross-checks for the availability of an alterna-tive when asking for approval to carryout an in vivo study must be implement-ed. Here guidance how to search thedatabases properly should be developed.Setting the respective alternative intopractice should be aided by providinglaboratory infrastructure as well as de-tailed information on the respectivemethod, e.g. by making available de-tailed SOPs or training courses.

Journals play a critical role in directingscience and funding bodies since theycan urge the scientist to apply or not ap-ply certain standards. A binding declara-

GRUBER AND HARTUNG


tion similar to the Declaration of Helsin-ki for clinical trials on patients and vol-unteers, which defines the standards foroptimised and minimal animal use isnecessary. Such a consensus declarationwould go far beyond a GCCP guideline;it might be based on the Bologna decla-ration on the 3R principle and could becreated on occasion of a future confer-ence of this series.

7.4 The pipelineThe pipeline is a term used in pharma-ceutical companies for the products un-der development and considered close tothe market. For alternative methods insome areas such as chronic toxicity, thepipeline is mostly empty. Still, the devel-opment of alternatives from scratch in areas of need (e.g. cosmetics, chemicals,chronic toxicity, neurotoxicity, immuno-toxicity) should be encouraged by offer-ing funding and awards, or the gap be-tween development and validation ofmethods already available has to beclosed. It is little tempting for a re-searcher to standardise, optimise andcritically review his/her experimentalmodel. Too often this will not lead topublishable data but to a questioning of

the results gained so far. In science,where “publish or perish” is a commonsaying, too little attention is paid to theproper evaluation (not even asking forvalidation) of the methods employed. A biometrical analysis to establish theopportunities and limitations of a test ishardly ever carried out for any methoddescribed in a scientific paper. In theory,this is a prerequisite for the interpretationof results, in practice, nobody cares,which might be the reason why so manyresults reported cannot be reproducedand vanish over time. Since these steps ofevaluation and optimisation, which arerequired for an increase in the quality ofthe performed science, are exactly thesteps to be carried out between develop-ment and validation of a test, both aimswork hand in hand. Seen the other wayaround, this is what the field of alterna-tive methods has to offer basic science: aproper concept of how to establish the reproducibility and the relevance of amethod. Applying this in a broader formwill automatically fill the pipeline, mak-ing the gap towards validation smaller onthe one hand and on the other hand, im-proving the reliability and thus relevanceof the method and making its applicationmore likely.

7.5 EducationConcepts are best furthered from teacherto pupil. At a stage when collectingideas, which will last a lifetime, i.e. whenthe scientific mind is formed during theearly years of university, the concepts ofalternatives to animal experiments, re-levance and quality control must be im-planted. This purpose is served best byestablishing the subject in the study cur-ricula of the life sciences. A chair on alternatives might be a first step towardsdeveloping and implementing the con-cepts and to optimising the respective educational offers (Wendel, 2002). Theeducated mind of future academics is thebest messenger to spread these ideas -when these young people take over responsibility in their profession and further develop the body of these ideas.See also the following chapter by Gruberand Dewhurst.

ReferencesALTEX Supplement (1996). Dottersack-

Antikörper, Egg Yolk Antibodies, IgY-Technology. ALTEX 13 Suppl., 1-85.

Anderson, R., Balls, M., Burke, M. D. etal. (2001). The establishment of hu-man research tissue banking in the UKand several western European coun-tries. The report and recommendationsof ECVAM workshop 44. ATLA 29,125-134.

Arnhold, S., Andressen, C., Hescheler, J.und Addicks, K. (1998). Mikro-kinematographische Neurodifferen-zierungs- und Neurotoxizitätsstudienan einem in vitro Modell unter Ver-wendung embryonaler Stammzellender Maus (Microcinematographicstudies on neurodifferentiation andneurotoxicity in vitro using mouse em-bryonic stem cells). ALTEX 15, 59-66.

Ast, I., Heydeck, D., Mothes, E. andChrist, B. (2001). Standardized in vit-ro electrophysiologic measurementsusing isolated perfused porcine hearts– assessment of QT interval alter-ations. ALTEX 19, 3-8.

Balls, M. (1995). The use of non-humanprimates as laboratory animals in Eu-rope: moving toward the zero option.ATLA 23, 284-286.

Balls, M. and Karcher, W. (1995). Thevalidation of alternative methods. ATLA 23, 884-886.

Balls, M., Goldberg, A. M., Fentem, J. etal. (1995). The Three Rs: the way for-ward. The report and recommenda-tions of ECVAM workshop 11. ATLA23, 838-866.

Balls, M. (1999). The biomedical sci-ences and the need for less in-humaneprocedures. In C. F. M. Hendriksenand D. Morton (eds.), Humane end-points in animal experiments forbiomedical research (1-4). London:The Royal Society of Medicine Pressltd.

Bartmann, A., Sladek, M., Schimpf, T. etal. (2001). Die Blut-Hirn-Schranke invitro und ihre Anwendung: Standard-isierung, Validierung und in vitro-/invivo-Korrelation (The blood-brain bar-rier in vitro and its applications: stan-dardisation, validation and in vitro-/ invivo correlation). ALTEX 18, 181 (ab-stract).

GRUBER AND HARTUNG


Bateson, P. (1986). When to experimenton animals. New Scientist, 20 Febru-ary, 30-32.

Bayertz, K. (2000). Drei Argumente fürdie Freiheit der Wissenschaft (Threearguments for scientific freedom). ALTEX 17, 59-65.

Blaauboer, B. J., Boobis, A. R., Castell,J. V. et al. (1994). The practical appli-cability of hepatocyte cultures in rou-tine testing. ATLA 22, 231-241.

Bottrill, K. (2002). A Guide to searchingfor alternatives to the use of laboratoryanimals. http://www.frame.org.uk/Guide/index.htm

Buesen, R., Visan, A., Genschow, E. etal. (2004). Trends in improving theembryonic stem cell test (EST): anoverview. ALTEX 21, 15-22.

Bundesbericht Forschung (2004). Bun-desministerium für Bildung undForschung, www.bmbf.de/pub/bufo2004.pdf

BVET (1994). Einteilung von Versuchennach Schweregraden vor Versuchs-beginn (Belastungskategorien). Infor-mation Tierschutz 1.04. http://www.bvet. admin.ch/0_navigation-d/0_in-dex-intern.html

Casati, S. und Hartung, T. (2003). DritterReport der EU zu Versuchstierzahlenliegt vor (Third report of the EU on experimental animal numbers). ALTEX 20, 93.

Caspar, J. (1998). Anthropozentrismusversus Pathozentrismus – zur Stellungdes Tierschutzes im System des grun-drechtlichen Freiheitsschutzes (An-thropocentrism versus pathocentrism –on the integration of animal protectioninto the system of safeguarding basicdemocratic rights). ALTEX 15, 205-208.

Combes, R., Balls, M., Curren, R. et al.(1999). Cell transformation assays aspredictors of human carcinogenicity.ATLA 27, 745-767.

Delpire, V. and Balls, M. (1998). EC-VAM Workshop über die Verwendungtransgener Tiere in der EU. EinÜberblick (Use of transgenic animalsin the European Union). ALTEX 15,23-26.

DFG (1971). Mitteilung I der Kommis-sion für Versuchstierforschung. Grund-informationen über Versuchstierfra-gen. Bonn: DFG.

DFG (1981). Mitteilung III der Kommis-sion für Versuchstierforschung. Tierex-perimentelle Forschung und Tier-schutz. Boppard: Harald Boldt Verlag.

DFG (1991). Mitteilung IV der Senats-kommission für Versuchstier-forschung. Novellierung des Tier-schutzgesetzes 1986. Informationenfür den Forscher. Weinheim: VCH.

DFG (1993). Denkschrift der Senats-kommission für tierexperimentelleForschung. Tierversuche in derForschung. Weinheim: VCH.

DFG (1996). Deutsche Forschungsge-meinschaft. Forschungsfreiheit. EinPlädoyer für bessere Rahmenbedin-gungen der Forschung in Deutschland.Weinheim: VCH.

Dietz, V., Eppenberger, H. M., Glaser, D.et al. (1997). Liste nicht mehr zuläs-siger Tierversuche an den ZürcherHochschulen (List of animal testswhich are no longer allowed at insti-tutes of higher learning in Zurich). AL-TEX 14, 61-62.

Doehmer J. und Jacob J. (1994). Gen-technologisch konstruierte V79 Zell-linien in Kombination mit chemisch-analytischen Verfahren als Ersatz undErgänzung zu Tierversuchen (Geneti-cally engineered V79 cell lines in com-bination with analytical-chemical pro-cedures for replacing and refininganimal experimentations). ALTEX 11,141-147.

Doehmer, J. (2001). Moderne Arznei-mittelentwicklung mit molekular- undzellbiologischen Methoden (Moderndrug development by molecular andcellbiological methods). ALTEX 18, 9-12.

Falcone, F. H., Schlaak, M. und Haas, H.(1995). In vitro Kultur von Brugiamalayi, einem Erreger der lymphatis-chen Filariose beim Menschen (In vit-ro cultivation of Brugia malayi, a para-sitic nematode that causes humanlymphatic filariasis). ALTEX 12, 179-187.

Ferber, P. C., Ossent, P., Homberger, F.R. and Fischer, R. W. (1999). The gen-eration of monoclonal antibodies inmice: influence of adjuvants on the im-mune response, fusion efficiency anddistress. Laboratory animals 33, 334-350.

Finking, G., Wolkenhauer, M., Lenz, C.

and Hanke, H. (2000). Post-injury exvivo model to investigate effects andtoxicity of pharmacological treatmentin rings of rabbit aortic vessels. ALTEX17, 67-74.

Flecknell, P. A. (1994). Refinement ofanimal use – assessment and allevia-tion of pain and distress. LaboratoryAnimals 28, 222-231.

Flecknell, P. A. (1996). Laboratory Ani-mal Anaesthesia: An Introduction forResearch Workers and Technicians. 2nd

ed. London: Academic Press.Flecknell, P. and Waterman-Pearson,

A.(2000). Pain Management in Ani-mals. London: Harcourt Health Sci-ences.

Flury, A. (1999). Sind operative Eingriffein Gehirne lebender Primaten in derGrundlagenforschung moralisch ver-tretbar? (Is brain surgery on primatesin basic research morally acceptable?).ALTEX 16, 267-270.

Gärtner, K. (1999). Cage enrichment occasionally increases deviation ofquantitative traits. Proceedings of theinternational joint meeting, twelfthICLAS general assembly & confer-ence, seventh FELASA symposium,26-28 May 1999, Palma de Mallorca,Spain. London: Laboratory AnimalsLtd.

Gassmann, M., Weiser, T., Tömmes, P.und Hübscher, U. (1990). Das Hühner-ei als Lieferant polyklonaler Antikörp-er (The egg yolk as resource for poly-clonal antibodies). Schweiz. Arch.Tierheilk. 132, 289-294.

Genschow, E., Spielmann, H., Scholz, G.et al. (2002). The ECVAM interna-tional validation study on in vitro embryotoxicity tests. Results of thedefinitive phase and evaluation of pre-diction models. ATLA 30, 151-176.

Genschow, E., Spielmann, H., Scholz, G.et al. (2004). Validation of the embry-onic stem cell test (EST) in the ECVAM international validation studyon in vitro embryotoxicity. ATLA 32,in press.

Gindorf, C., Steimer, A., Lehr et al.(2001). Markertransport über biolo-gische Barrieren in vitro: Vergleichvon Zellkulturmodellen für die Dünn-darmschleimwand, die Blut-Hirn-Schranke und das Alveolarepithel derLunge (Marker transport across bio-

GRUBER AND HARTUNG


logical barries in vitro: comparison ofcell culture models for the gastroin-testinal barrier, the blood-brain barrierand the alveolar epithelium of thelung). ALTEX 18, 155-164.

Goldberg, A. M. (2002). Use of animalsin research: a science – society contro-versy? American perspectives: animalwelfare issues. ALTEX 19, 137-139and in F. P. Gruber and K. Brune(eds.), ALTEX-Buch 2002. ALTEX19, Suppl. 2, 41-45.

Gruber, F. P. (1996). Alternativmethodenim Sinne eines Refinements (Alterna-tive methods for Refinement). In F. P.Gruber and H. Spielmann (Eds.), Al-ternatives to animal experiments (273-300). Berlin, Heidelberg, Oxford:Spektrum Akademischer Verlag.

Gruber, F. P. (1998). Transgene Tiere im3R-Konzept (Transgenic animals andthe 3R). ALTEX 15, 26-28.

Gruber, F. P., Günzel, P., Rusche, B. undSchwabenbauer, K. (1996). Studieüber die Forschungsförderung desBMBF zur Entwicklung von Er-satzmethoden zu Tierversuchen (Studyabout the grant programme of the Ger-man BMBF for the promotion of alter-natives to animal experiments). ALTEX13, 55-67.

Grune, B., Herrmann, S., Dörendahl, A.et al. (2000).The ZEBET-Database onAlternative Methods to Animal Exper-iments in the Internet – a Contributionto the Protection of Experimental Ani-mals. ALTEX 17, 127-133.

Grune, B., Dörendahl, A., Skolik, S. undSpielmann (2001). Suchstrategiennach 3R-Methoden im Internet (Searchstrategies for 3R-methods in the inter-net). ALTEX 18, 182.

Grune, B., Fallon, M., Howard, C. et al.(2004). Report and recommendationsof the international workshop „Re-trieval approaches for information onalternative methods to animal experi-ments”. ALTEX 21, 115-127.

Guan, K., Schmidt, M. M., Ding, Q. et al.(1999). Embryonic stem cells in vitro– prospects for cell and developmentalbiology, embryotoxicology and celltherapy. ALTEX 16, 135-141.

Gutsche, W. and Stelzner A. (1997). Pre-selection of potential cancerostatics byautomatic analysis of suspended andadherent cells incubated in mi-

croplates. ATLA 25, 45-54.Hartung, T. und Wendel, A. (1993). Ent-

wicklung eines Zellkulturmodells fürdas Organversagen im septischenSchock (Development of a cell-culturemodel multiorgan failure in septicshock). ALTEX 10, Nr. 18, 16-24.

Hartung, T. und Spielmann, H. (1995).Der lange Weg zur validierten Er-satzmethode (The sophisticated pro-cess of validation). ALTEX 12, 98-103.

Hartung, T. und Wendel, A. (1998). Ersatz von Tierversuchen in der Phar-makologie – Ziel oder Zwang, Utopieoder Realität? (Is replacement of ani-mal experimentation in pharmacologya goal or a social constraint, utopia orreality?). ALTEX 15, 213-215.

Hartung, T., Gstraunthaler, G., Coecke,S. et al. (2001). Good cell culture prac-tice (GCCP) – eine Initiative zur Stan-dardisierung und Qualitätssicherungvon in vitro Arbeiten. Die Etablierungeiner ECVAM Task Force on GCCP(Good Cell Culture Practice (GCCP) –an Initiative for Standardisation andQuality Assurance of in vitro Studies.The Establishment of an ECVAM TaskForce on GCCP). ALTEX 18, 75-78.

Hartung T., Balls, M., Bardouille, C. etal. (2002). Good cell culture practice.ECVAM good cell culture practise taskforce report 1. ALTLA 30, 407-414.

Hendriksen, C. F. M. (1998). A call for aEuropean prohibition of monoclonalantibody production by the ascites pro-cedure in laboratory animals. ATLA26, 523-540.

Hendriksen, C. F. M. and Morton, D.(eds.) (1999). Humane endpoints inanimal experiments for biomedical re-search. London: The Royal Society ofMedicine Press ltd.

Hlinak, A., Marx, U. und Jäger, V.(1994). Experimente zur Herstellungvon monoklonalen Antikörpern überdas Hühnerei (Production of mono-clonal antibodies in chicken eggs). AL-TEX 11, 85-91.

Izzi, L. (1998). Anti-endimysium anti-bodies detection in the celiac diseasescreening: Indirect immunofluores-cence pattern using umbilical cord sec-tions as substrate. ALTEX 15, 141-143.

Janusch, A., van der Kamp, M. D. O.,Bottrill, K. et al. (1997). Current statusand future developments of databases

on alternative methods. The report andrecommendations of ECVAM work-shop 25. ATLA 25, 411-422.

Janusch-Roi, A., Libowitz, L., Grune, B.and Kreger, M. (2000). Alternativemethods databases – specialised infor-mation sources on alternatives to sup-port scientists and authorities responsi-ble for granting project licences. In M.Balls, A.-M. van Zeller and M. Halder(eds.), Progress in the Reduction, Refinement and Replacement of Animal Experimentation (1731-1736).Amsterdam: Elsevier.

Jones, H. R. P., Oates, J. and Trussell, B.A. (1999). An applied approach to theassessment of severity. In C. F. M.Hendriksen and D. Morton (eds.), Hu-mane endpoints in animal experimentsfor biomedical research (40-47). Lon-don: The Royal Society of MedicinePress ltd.

Köhler, G. and Milstein, C. (1975). Con-tinuous cultures of fused cells secret-ing antibody of predefined specificity.Nature 256, 495-497.

Kuhnert, F. (1996). Feeding of hard ticksin vitro: new perspectives for rearingand for the identification of systemicacaricides. ALTEX 13, 76-87.

Kuhnert, F., Issmer, A. E. andGrunewald, J. (1998). Teilauto-matisierte in vitro Fütterung adulterSchildzecken (Amblyomma hebraeum)(Partly automated in vitro feeding ofadult Amblyomma hebraeum (Acari:Ixodidae)). ALTEX 15, 67-72.

Kunzi-Rapp, K., Westphal-Frösch, C.,Akgün, N. et al. (1997). Die Chorioal-lantoismembran des befruchteten Hühnereis als in vivo-Ersatzsystem fürdie Photodynamische Therapie (Thechorioallantoic membrane of the fertil-ized hen’s egg as an in vivo system forphotodynamic therapy). In H. Schöffl,H. Spielmann, H. A. Tritthart et al.(Hrsg.), Forschung ohne Tierversuche1996 (154-158). Wien, New York:Springer Verlag.

Lee, V.-M., Burdett, N. G., Carpenter, T.A. et al. (1998). Magnetic ResonanceImaging of the Common MarmosetHead. ATLA 26, 343-356.

Leenars, P. P. A. M., Hendriksen, C. F.M., Leeuw, W. A. de et al. (1999). Theproduction of polyclonal antibodies inlaboratory animals. The report and rec-

GRUBER AND HARTUNG


ommendations of ECVAM workshop35. ATLA 27, 79-102.

Lindl, T., Weichenmeier, I., Labahn, D.et al. (2001). Evaluation vongenehmigten tierexperimentellen Ver-suchsvorhaben in Bezug auf dasForschungsziel, den wissenschaft-lichen Nutzen und die medizinischeRelevanz (Evaluation of authorised ex-periments on laboratory animals withregard to the following frame of refer-ences: the aim of the research to becarried out, its scientific usefulnessand its medical relevance). ALTEX 18,171-178.

Lloyd, M. H. and Wolfensohn, S. E.(1999). Practical use of distress scor-ing systems in the application of hu-mane endpoints. In C. F. M. Hendrik-sen and D. Morton (eds.), Humaneendpoints in animal experiments forbiomedical research (48-53). London:The Royal Society of Medicine Pressltd.

Marx, U., Embleton, M. J., Fischer, R. etal. (1997). Monoclonal antibody pro-duction. The report and recommenda-tions of ECVAM workshop 23. ATLA25, 121-137.

Matthes, H.-F. und Hiepe, T. (1996).Zum Stand der in vitro Züchtung vonLäusen und Flöhen (The state of in vit-ro breeding of lice and fleas). ALTEX13, 130-135.

Mazzotti, F., Sabbioni, E., Ghiani, M. etal. (2001). In vitro assessment of cyto-toxicity and carcinogenic potential ofchemicals: evaluation of the cytotoxic-ity induced by 58 metal compounds inthe Balb/3T3 cell line. ATLA 29, 601-611.

Mench, J. (1999). Defining endpoints:The role of the animal care committee.In C. F. M. Hendriksen and D. Morton(eds.), Humane endpoints in animalexperiments for biomedical research(133-138). London: The Royal Societyof Medicine Press ltd.

Mertens, C. und Rülicke, T. (2000). Um-fassendes Formular zur strukturiertenCharakterisierung gentechnisch verän-derter Tiere (A comprehensive formfor the standardised documentation oftransgenic rodents). ALTEX 17, 15-21.

Morton, D. (1999). Humane endpoints inanimal experiments for biomedical re-

search: ethical, legal and practical is-sues. In C. F. M. Hendriksen and D.Morton (eds.), Humane endpoints inanimal experiments for biomedical research (5-12). London: The RoyalSociety of Medicine Press ltd.

Motzkus, N., Budinsky, L., Sergejeva,M. et al. (2004). New animal model forstressfree and objective pain research.ALTEX 21, 168-169 (abstract).

Nogueira, A. C., Wagner, S., Riebeling, I.und Klug, S. (1999). Die isolierte normotherme hämoperfundierteSchweineextremität als Modell fürpharmakologische und toxikologischeUntersuchungen (The isolated nor-mothermic hemoperfused porcine legas model for pharmacological and tox-icological investigations). ALTEX 16,90-94.

Nussler, A. K., Wang, A., Neuhaus, P. etal. (2001). The suitability of hepato-cyte culture models to study variousaspects of drug metabolism. ALTEX18, 91-101.

Olpe, H.-R. und Haas, H. (1985). DerHippocampus in vitro im Dienste derEpilepsieforschung (The Hippocam-pus in vitro in the service of epilepsyresearch). ALTEX 2, No. 2, 5-14.

Palmiter, R. D., Brinster, R. L., Hammer,R. E., Trumbauer, M. E., Rosenfeld,M. G., Birnberg, N. C., and Evans, R.M. (1982). Dramatic growth of micethat develop from eggs microinjectedwith metallothionein-growth hormonefusion genes. Nature 300, 611-615.

Pittermann, W., Kietzmann, M. andGrosse-Siestrup, C. (eds.) (2000). Pro-ceedings of a Workshop on IsolatedPerfused Organs (Hamburg 1998, 36thScientific Meeting of GV-SOLAS).London: Laboratory Animal Ltd.

Reininger, C. B. (2000). The cell func-tion analyser (CFA) – a physiologicalin vitro vascular model and potentialalternative to animal experiments. AL-TEX 17, 115-125.

Rosenbruch, M. (1994). Frühe Entwick-lungsstadien des bebrüteten Hühner-eies als Modell in der experimentellenBiologie und Medizin (Early stages ofthe incubated chicken egg as a modelin experimental biology andmedicine). ALTEX 11, 199-206.

Rosenbruch, M. (1997). Zur Sensitivität

des Embryos im bebrüteten Hühnerei(The sensitivity of chicken embryos inincubated eggs). ALTEX 14, 111-113.

Russel, W. M. S. and Burch, R. L.(1959). The principles of humane experimental technique. London:Methuen & Co LTD. (1992). Specialedition by UFAW.

Sadri, F. (2002). Letter to the editors ofthe proceedings of a workshop on iso-lated perfused organs. ALTEX 19, 110-111.

SAMW und SANW, (1996). EthischeGrundsätze und Richtlinien. Schwei-zerische Akademie der MedizinischenWissenschaften, Basel und Schwei-zerische Akademie der Naturwis-senschaften, Bern. ALTEX 13, 3-6.

Sauer, A., Hartung, T., Aigner, J., Wen-del, A. (1996). Endotoxin-induciblegranulocyte-mediated hepatocytotoxi-city requires adhesion and serine pro-tease release. J. Leukoc. Biol. 60, 633-643.

Schade, R. and Hlinak, A. (1996). Eggyolk-antibodies, state of the art and future prospects. ALTEX 13, 5-9.

Schade, R., Hlinak, A., Marburger, A. etal. (1997). Advantages of using eggyolk antibodies in the life sciences:The results of five studies. ATLA 25,555-586.

Scharmann, W. (1996). Tiergerechte Haltung von Legehennen unter Labor-bedingungen (Accomodation of layinghens in the laboratory in accordancewith animal welfare requirements).ALTEX 13, 136-139.

Scharmann, W. (1999). Physiologicaland ethological aspects of the assess-ment of pain distress and suffering. InC. F. M. Hendriksen and D. Morton(eds.), Humane endpoints in animalexperiments for biomedical research(33-39). London: The Royal Society ofMedicine Press ltd.

Schweizer Akademie der MedizinischenWissenschaften und SchweizerAkademie der Naturwissenschaften(1996). Ethische Grundsätze undRichtlinien für wissenschaftlicheTierversuche (Ethical principles andguidelines for scientific experimentson animals). ALTEX 13, 3-6.

Seiler, A., Visan, A., Pohl, I. et al. (2001).Etablierung molekularer Endpunkte

GRUBER AND HARTUNG


zur Weiterentwicklung des Embry-onalen Stammzelltests (EST) mit em-bryonalen Stammzellen der Maus(Zelllinie D3) (Improving the embry-onic stem cell test (EST) by establish-ing molecular endpoints of tissue spe-cific development using murineembryonic stem cells (D3 cells)). ALTEX 18 (Suppl. Linz 2001), 55-63.

Shoin, K., Yamashita, J., Enkaku, F. et al.(1991). Chick embryo assay aschemosensitive test for malignantglioma. Jpn. J. Cancer Res. 82, 1165-1170.

Siegel-Axel, D. I., Viebahn, R., Betz, E.L. and Barsch, R. (1999). Progress inthe research of atherosclerosis andrestenosis using an in vitro method:transfilter cocultures with human vas-cular cells. ALTEX 1999, 117-122.

Stadler, B. (1995). Herstellung von humanen monoklonalen Antikörpernin vitro durch Repertoire-Klonierung(Production of human monoclonal antibodies in vitro by repertoirecloning). In H. Schöffl, H. Spielmann,H. A. Tritthart (Eds.), Research with-out animal experiments 1995 (301-306). Vienna, New York: Springer.

Stadler, B. M., Vogel, M. and Miescher,S. (2000). Recombinant antibodies by

phage display: the end of animal use?In M. Balls, A.-M. van Zeller and M.Halder (eds.), Progress in the reduc-tion, refinement and replacement ofanimal experimentation (873-878).Amsterdam: Elsevier.

Strübing, C., Wobus, A. M. und Heschel-er, J. (1995). Etablierung eines in vitroModellsystems zur Differenzierungsynaptisch verknüpfter Nervenzellenaus pluripotenten embryonalen Stam-mzellen der Maus (Establishment of anin vitro model system for the differen-tiation of synaptically coupled neuronsfrom mouse embryonic stem cells).ALTEX 12, 129-137.

Thon, R., Lassen, J., Kornerup Hansen,A. et al. (2002). Welfare evaluation ofgenetically modified mice - An inven-tory study of reports to the Danish An-imal Experiments Inspectorate. Scand.J. Lab. Anim. Sci. 29, 45-53.

Tritthart, H. A. (1996). In vitro Modellein der Krebsforschung (In vitro testsystems in cancer research). ALTEX13, 118-123.

Van de Weerd, H. A., Aarsen, E. L., Mul-der, A. et al. (2002). Effects of envi-ronmental enrichment for mice: varia-tion in experimental results. Journal ofApplied Animal welfare Science 5, 87-109.

Vedani, A. (1994). Das Konzept desPseudorezeptors für das pharmako-logische Screening (Pseudoreceptormodeling – a tool in the pharmacolog-ical screening process). ALTEX 11, 11-21.

Vedani, A. and Dobler, M. (2001). Inter-net laboratory for predicting harmfuleffects triggered by drugs and chemi-cals. Concept and call for co-opera-tion. ALTEX 18, 110-114.

Wallace, J. (1999). Humane endpoints in

cancer research. In C. F. M. Hendrik-sen and D. Morton (eds.), Humaneendpoints in animal experiments for bio-medical research (79-84). London: TheRoyal Society of Medicine Press ltd.

Wang, T., Nüssler, A. K., Grebe, A. et al.(2002). Three-dimensional co-cultureof primary human liver cells in biore-actors for in vitro studies: effects of theinitial cell quality on the long-termmaintenance of hepatocyte-specificfunctions. ATLA 30, 525-538.

Weerd, Van de, H. A., Aarsen, E. L., Mulder, A., Kruitwagen, C. L. J. J.,Hendriksen, C. F. M. and Baumans, V.(2002). Effects of environmental en-richment for mice on variation in ex-perimental results. Journal of AppliedAnimal Welfare Science 5, 87-108.

Wendel, A. (2002). Do we need a “chairof alternative methods”, and where?ALTEX 19, 64-68.

Wobus, A. M. (2001). Potential of em-bryonic stem cells. Molecular aspectsof medicine 22, 149-164.

Zeilinger, K., Auth, S., Unger, J., Grebe,A., Mao, L., Petrik, M., Holland, G.,Appel, K., Nüssler, A., Neuhaus, P.und Gerlach, J. (2000). Leberzellkulturin Bioreaktoren für in vitro Studienzum Arzneimittelmetabolismus als Alternative zum Tierversuch (Livercell culture in bioreactors for in vitrodrug studies as an alternative to animaltesting). ALTEX 17, 3-10.

Correspondence toPD Dr. Franz P. GruberFFVFFHegarstr. 9 P.O. Box 1766CH-8032 ZürichSwitzerlandE-mail: [email protected]

alternatives to animal experimentation in basic · pdf file7.5 education received 9 october...

Documents