The workshop on validation of alternative methods for the potency testing of vaccines was organised by ECVAM (European Centre for the Validation of Alternative Methods) and AGAATI (Advisory Group on Alternatives to Animal Testing in Immunobiologicals), and it was held in Angera, Italy, on 14-16 November 1997 under the co-chairmanship of Coenraad Hendriksen (RIVM, Bilthoven, The Netherlands) and Jean-Marc Spieser (EDQM, Strasbourg, France). The 17 participants, all experts in vaccine quality control and/or validation procedures, came from international regulatory bodies and organisations, national control laboratories and vaccine manufacturers. The aim of the workshop was to discuss, in an informal atmosphere, the complex issue of guidelines for the validation of alternative methods to potency testing of vaccines in laboratory animals, in particular: a) to review existing guidelines for validation; b) to discuss specific guidelines for the validation of alternative methods to the potency testing of vaccines and to agree on strategies for the preparation of guidelines; and c) to discuss and make recommendations for the implementation of validated alternative methods for regulatory purposes. The outcome of the discussions and the recommendations will be published in the ECVAM workshop series in ATLA in late 1998.

The monographs of the European Pharmacopoeia (Ph.-Eur.) are legally binding. Namely in immunobiologicals there is still need to perform animal trials in testing for safety and potency. Only an alternative method which is introduced into the Ph.-Eur. is able to substantially reduce the number of animals. To introduce a new method into the Ph.-Eur., mainly to substitute the assay but also as part of the test- or in process-section of the monographs, the requirements of a formal follow up have to be met. To understand these mechanisms, some of the underlying principles and the way a monograph is established are explained. To meet the requirements of a validated alternative method the steps from the first idea to the request for revision are recapitulated. The development of such a method should be performed in close co-operation with the respective group of experts and the biological standardisation group of the Ph.-Eur. Some ideas for acceleration of the implementation are given.

When cells of the immune system, especially blood monocytes and macrophages, come into contact with pyrogenic (fever-inducing) contaminations, they secrete messenger molecules which initiate an hyperthermic reaction in the organism. Of this group of endogenous pyrogens, most is known about interleukin-1 (IL-1). A new pyrogen test makes use of this reaction as a system for detection: The substances which are to be screened are incubated with a small volume of blood from a healthy donor. Any pyrogens present induce the production of IL-1 which can be detected by ELISA.
This test has a higher sensitivity and is more economical than the conventional pyrogen test in rabbits and furthermore reflects the reaction of the relevant species. In contrast to the customary alternative method, the Limulus amoebocyte lysate test (LAL), this test is not restricted to endotoxins from Gram-negative bacteria and is also not hindered by substances which bind endotoxins, such as blood proteins, to the same extent. Consequently, more than 50 non-endotoxin pyrogens have already been traced by this test. The whole blood test is even superior to the LAL in regard to the detection of endotoxins: in a comparison of about 60 endotoxins, there was a correlation of the potency of the individual endotoxins between the whole blood test and the pyrogen test in rabbits, but neither test correlated with the LAL test. In some cases, endotoxins with equal effects in the LAL test differed in potency in the human blood model by a factor of 10 000.
A method has been developed by which cryopreserved blood can be put to use in the test. In this way, blood donations from a donor can be pre-tested so that uniform material may be employed in the test.
This test opens up entirely new perspectives on pyrogen testing for Gram-positive or fungal pyrogens as well as in medicinal products. In addition, it could fill the dangerous security gap which might result from the limitations of testing medications and blood products with LAL only.
The project was supported by ZEBET, D-Berlin, BMBF, D-Bonn, and set, D-Mainz.

When establishing the whole blood assay for pyrogens in our laboratory we did several investigations on the standard pyrogen to be used. Because of the hydrophobicity of the lipopolysaccharid (LPS) it is difficult to get a homogeneous solution. Therefore we used ultrasonic and biphenyl hydroxyl amine chloride for a better dispergation.
Surprisingly the reactivity of the LPS in the whole blood assay for pyrogens dropped the more the LPS was dispergated. That's right inverse to the situation in the LAL.
There are first experiences in the whole blood assay for pyrogens with biological pharmaceuticals. This includes immunoglobulins, vaccines and coagulation factors.
Except of the vaccines and albumin the rabbit assay is the only way to test these pharmaceuticals for pyrogens.
The albumins show a little amplification of the reactivity of LPS in the whole blood assay for pyrogens. The testing of albumin in the whole blood assay for pyrogens showed highly reproproducible results and a detection limit under 50 pg/ml LPS.
The immunoglobulins show an amplifying effect as well. Immunoglobulins spiked with LPS were detected as pyrogenic in all cases.
Some vaccines show a slight pyrogenity in the whole blood assay which gets visible when diluting the sample. Undiluted vaccines may cover a LPS contamination in the whole blood assay for pyrogens.

The human whole blood assay utilises the natural fever response to detect pyrogens by determination of the release of IL-1b.
In order to replace the official method, the rabbit pyrogen test, a validation of the whole blood assay is necessary. A comparison of the results obtained from many blood samples has revealed the following:
1) Blood not stimulated by LPS does not produce IL-1b.
2) Stimulation by LPS induces a concentration-dependent release of IL-1b beginning at a concentration of between 2-5 pg/mL LPS.
3) The amount of IL-1b released varies greatly between samples obtained from different individuals.
4) Storing blood samples results in a right shifted LPS/IL-1b curve with a steeper gradient and higher maximum value of IL-1b.
In this paper we suggest an experimental method for the determination of pyrogens based on the established semi-quantitative LAL gelation method as detailed in the European Pharmacopeia.
Using this methodology, we were able to show that the amount of endotoxin in a number of different infusion solutions was below the LAL-endotoxin limit concentration. LPS was quantitatively determined from spiked samples.

The testing of drugs for pyrogen contamination e.g. the exclusion of fever inducing impurities, was standardised in the Forties in the form of animal pyrogen assay. We recently proposed an alternative method using human whole blood from healthy donors. The substances to be tested are incubated in the presence of blood and the generation of endogenous pyrogens (Interleukin 1, IL-1) are measured reflecting the primary fever reaction in man. With this method it is possible to detect the reaction of the relevant species. However, pronounced differences were reported for the potency of several bacterial pyrogens in different species (e.g. up to factor 10.000 for an individual endotoxin in different mammals). As a part of the validation of the new method discrepancies between the whole blood system and the rabbit test might occur. In these cases it is not possible to distinguish between an in-vitro-artefact and species differences.
Therefore, the goal of this project is the development of an in-vitro-test based on rabbit blood. For this purpose rabbit IL-1 was expressed. At present chickens and mice are immunized and possibly crossreacting anti-human-IL-1-antibodies are tested. In this way, an ELISA for the rabbit cytokin will be established. Using a temporarily commercially available rabbit TNF-ELISA, the feasibility of the test-system could already be shown: Pyrogens induced the generation of TNF in rabbit blood. In the human model, the endpoints IL-1 and TNF were very similar. The sensitivity of human blood and rabbit blood appeared to be very similar. However, individual pyrogens (only different endotoxins were tested up to now) showed considerable deviations. However, these results require the verification in the final test.
The rabbit test in preparation will be implemented immediately into the ongoing evaluations by the Paul-Ehrlich-Institute (BMBF) and at the University of Heidelberg (European Pharmakopoea). Furthermore, such a test also represents an interesting alternative in the veterinarian medical field using the correct target species for pyrogenicity testing.
This project war supported by the Foundation Research 3R, CH-Muensingen.

Each oral poliovirus vaccine lot contains attenuated and small amounts of wildtype viruses. Vaccines containing more than a certain limit of wildtype viruses may cause a vaccine associated poliomyelitis. To provide safe vaccines for humans, each newly manufactured vaccine lot is tested in the monkey neurovirulence test. A certain point mutation on the poliovirus genome has been shown to be responsible for the attenuated phenotype of the vaccine virus.
We developed a quantitative PCR method to determine the wildtype proportion at position 472 of poliovirus type 3 genome.
This method possibly can be used as an alternative for the monkey neurovirulence test.

For safety testing and release for attenuated oral poliomyelitis virus (OPV) vaccine lots, national and international authorities require the in vivo neurovirulence test (NVT). This test is extremely expensive, time consuming and is carried out in socially sensitive monkeys. In this test the test vaccine is injected into the spinal chord of monkeys and the mean lesion score of neurovirulence is evaluated and compared to that of a reference vaccine. If the test vaccine is not significantly different from the reference, the vaccine passes the test.
Data from several investigations suggest that the accumulation of distinct reversions in a vaccine lot correlate with an increased mean lesion score of neurovirulence in the monkey test, and thus are the causa sine qua non of reversion of attenuation to wild type. The relevant reversions are for serotype 1, 480 G to A and 525 U to C, for serotype 2, 481 A to G, and for serotype 3, 472 U to C. To quantify these reversions on a relative basis in vitro, the MAPREC test (Mutation Analysis by PCR and Restriction Enzyme Cleavage), developed by Chumakov et al. (FDA), seems to be the only suitable test. For serotype 3 it could be demonstrated, that vaccine lots failing the NVT comprise a reversion rate ³ 0.9%. For serotypes 1 and 2 such a limit could not yet clearly be defined. MAPREC quantifies only single mutations, which have been shown to be important for neurovirulence. There is a possibility, however, that also other mutations/reversions, not yet identified, could principally contribute to neurovirulence, but would not be detected by MAPREC. Therefore, at date, there is no scientific basis to replace the monkey NVT by MAPREC. Howeser, MAPREC can be used to check the consistence of reversion rates in vaccine lots under current production conditions, and the impact of changed production conditions on the reversion rates, and thus can be used to get more and more experience.
It further ongoing studies and analyses of manufacturers and national and international control laboratories will show that the predictions of the MAPREC results on the outcome of the monkey NVT are sufficiently safe, this in vitro test would be an excellent alternative for the monkey test.

The neurovirulence test established by Albert B. Sabin is required according to different pharmaceutical monographies (WHO Technical Report Series, US-Pharmacopoeia, European Pharmacopoeia, etc) and is used for evaluation of sufficient and consistent attenuation of life monovalent poliomyelitis virus bulk lots (Sabin) which are used for the manufacture of oral poliomyelitis vaccines (OPV). Since this animal experiment is based on the use of relatively high numbers of primates (at least 22 animals per test vaccine for the virus serotypes 1 and 2 and at least 36 animals for serotype 3) it is searched for alternatives since several years. Besides a molecular approach (MAPREC) a test based on the use of transgenic mice seems to be a promising alternative for the primate model.
The isolation and introduction of the human poliomyelitis virus receptor gene into the mouse genome made small laboratory animals susceptible for poliomyelitis for the first time. After infection with attenuated virus, the transgenic animals develop the same clinical symptomatic and the respective lesions in the central nerve system (CNS) as observed in monkeys. In a WHO collaborative study it could be demonstrated that the transgenic mouse neurovirulence test for OPV serotype 3 is at least as sensitive as the monkey neurovirulence test. Furthermore, it was shown that there is a strong correlation between the strength of the clinical defects and the spread of the specific histopathological lesions in the CNS. Therefore, the observation and description of the clinical symptoms seems to be sufficient for the evaluation of the experiment. Thus the time-consuming histopathological investigation can be avoided, since it does not provide additional information to the results of the clinical investigations. In a first evaluation, similar results were shown for serotype 2, however there are not yet sufficient data on serotype 1.
In 1997 an extended validation study under the participation of 5 laboratories from 3 continents has been initiated by the WHO with the objective to test a sufficient number of OPV serotype 3 charges from different manufacturers and to gain further informations on the consistency and validity of the mouse neurovirulence test. Results obtained from recent experiments suggest that the neurovirulence test of serotype 3 in transgenic mice may be included into the monographies as an alternative method for the monkey NVT.

The selection of possible candidate immunosuppressive antibodies to prevent graft rejection is performed in vitro. Additionally, due to the species specificity of these monoclonal antibodies (MABs), pre-clinical studies in non-human primates are necessary. If a positive correlation between the in vitro and in vivo findings would exist, these tests can act as a pre-screening before new reagents are tested in vivo. The correlation of the in vitro and in vivo efficacy of an anti-T-lymphocyte globulin (ATG) and an anti-CD80 MAB is evaluated in a rhesus monkey skin transplant model. The results show that lymphocytotoxic titers (NIH-test) do not predict the outcome of in vivo effectiveness of ATG in rhesus monkeys. Additionally, no evidence of tolerance to a skin allograft could be shown to correlate with inhibition of a secondary mixed lymphocyte culture (MLC) by anti-CD80 and cyclosporin A (CsA). Thus, these in vitro assay used can not predict the in vivo efficacy of new immunosuppressive antibodies.

Antilymphocyte globulins (ALG) are immunosuppressive agents of animal origin currently used in clinical transplantation medicine and for the treatment of severe aplastic anemia. The potency of each batch is tested in vivo using primates as hosts for allogeneic skin transplantation. The test is done with a maximum of three animals, one as a control and two after the treatment with ALG. The two in vitro methods in use are a cytotoxic assay and the rosette inhibition assay. These methods are evaluated with the microscope. Besides wellfare aspects these methods require a lot of experience, are subjective, difficult to validate and the information about the biological potency of the sera is questionable.
The aim of our study is a better biological characterisation as a prerequisite to subsequently define an in vitro alternative for the potency test in monkeys.
Using a competition assay with monoclonal antibodies we can identify several specificities directed against functional molecules on T cells (e.g., CD2, CD3, CD5, CD28), B Cells (CD19), macrophages and natural killer cells (CD16) and nonlineage specificities such as CD18, CD25, CD29, CD95. This method could describe a part of the biological potency and control homogeneity of batches. The cytotoxic capacity of ALG either with or without complement as well as DNA-fragmentation characteristic for apoptosis can be analysed by flowcytometry using propidiumiodide- (PI) incorporation. Immunoprecipitation of cell-lysate with ALG´s and subsequent incubation with radioactive ATP (kinase-assay) shows specific bands which seem to be identical between different batches of one product.

The quality control of human tetanus immunglobulin requires animal experiments according to European Pharmacopoeia monograph 398. The potency estimation has to be done in a toxin neutralisation test in mice (MNT) or guinea pigs. Immunoassays could also be used if they show a suitable sensitivity and specificity. The first results of our study verify that an indirect enzyme linked immunosorbent assay (ELISA), a rocket immunelectrophoresis (RIE) and a toxin binding inhibition test (ToBI) could be used as serological alternativ methods to the MNT. Studies on the reproducibility of the in vitro methods resulted inter-assay coefficients of variation between 2 and 27%. The ELISA is more sensitive (limit of detectability: 0,005 IE/ml) than the ToBI (0,04 IE/ml) and the RIE (5 IE/ml). The transferability of the ELISA to other labs is proofed. The transferability of the RIE and the ToBI will be tested in the near future.

We have developed the Pertussis Serological Potency Test (PSPT) as an alternative to potency testing of pertussis Whole Cell Vaccines (WCV) in the intracerebral Mouse Protection Test (MPT). The PSPT is based on the humoral antibody response against the whole range of B. pertussis surface-antigens per vaccine dose and correlates well with the MPT. WCV-induced-antibody responses against "protective" antigens; or the biological activitv of pertussis antibodies poorly correlated or did not correlate at all with mouse protection. Compared to the MPT the PSPT is more reproducible, reduces the animal distress and the number of animals in use. Moreover, the number of animals in use could be reduced even more by simplifying the multiple dose design to a single dose model and by combining in vitro assays for potency testing of tetanus, diphtheria and pertussis components in one animal model.

Toxoplasma gondii is an intracellular protozoan parasite of worldwide distribution. Parasites that habour a complete antigenic profile, that is necessary for the serological diagnosis of human Toxoplasma infections, are provided by in vivo culture methods only. It seems that the host immune pressure is responsible for the expression of a total antigen pattern. Thus, in most laboratories the asexual proliferative stage of the parasite, the tachyzoite, is maintained by successive intraperitoneal passages in highly susceptible animals such as mice.
We would like to develop an in vitro method to provide sufficient amounts of high quality parasite antigen suitable for diagnosis of Toxoplasmosis. Using the RAPD-PCR (random amplified polymorphic DNA-PCR) technique, we were able to show that during different culture conditions (in vivo and in vitro culture) the parasites undergo no clonal selection. That means the entire parasite population changes the protein expression pattern due to the in vitro culture conditions. Based on the result described previously the work could be continued as follows. One strategy could be to simulate the host immune pressure during the in vitro cultivation of the tachyzoites. A more convinced approach may be the production of recombinant parasite antigens useful for diagnosis of a Toxoplasma infection in human adults and newborns.

Vaccines are successfully used to prevent rabies infections in humans and animals. The quality requirements for these products are high. After licensing it is still necessary for batch release to perform animal challenge tests to demonstrate the potency of the vaccines. In these experiments it is registered how many animals in the immunized groups die or show symptoms of rabies within a defined time frame.
An ECVAM working group has examined whether clinical signs can be used in challenge tests for vaccine control to replace lethality as criterion.
The severe distress and the typical signs of the disease made the rabies infection a promising example for this study. Clinical signs, body temperature and body weight were the criteria used to define humane endpoints.
It could be shown, that the reduction of body weight and the appearonce of typical clinical signs are suitable to define humane endpoints. Pain and distress could be considerably reduced by the implementation of these criteria in to the legal requirements for potency tests of rabies vaccines.

In the study in vitro alternatives to a nonvalidated and harmful animal test for the absence of extraneous virus in live vaccines were investigated.
For evaluation of a suitable in vitro method the porcine herpesvirus (Aujeszkyvirus, Pseudorabiesvirus) was used as a model virus. In artificially contaminated live vaccines the aujeszkyvirus could be detected by moleculargenetical and cellular methods. Regarding the threshold values of virus detection in vitro tests showed to be more efficacious than animal testing.
Meanwhile the European Pharmacopoeia Commission deleted the animal test for extraneous virus from two monographs. The discussion, if respective animal testing can be cancelled for the other live vaccines as well, is still ongoing.
The study was supported by the German Ministry of Education, Science, Research and Technology.

At present, the complete inactivation of rabies virus in rabies vaccines ad us. vet. is proven by an animal experiment which causes severe suffering, the intracerebral injection of mice. This animal experiment yet is not validated. We have quantified the sensitivity of the mouse test and examined whether the animal experiment may be replaced by the immunofluorescence assay (IFT) as an in vitro method. Detection limits of both assays were determined depending on the examined product, i.e. prior to and after the addition of adjuvans and preservative, respectively. Furthermore, symptoms of the rabies desease were recorded and their severity was classified on a range of 1-5. Symptoms of rabies-infected mice were clear and highly specific. Symptoms classified as greater or equal as 2 in context with a loss of greater or equal as 15% of the initial weight were defined as humane endpoints of the desease. The quantitative detection of active virus was not inhibited in the presence of even high concentrations of inactivated virus. The detection limit of the mouse test was 10 viruses ml-1 independent of the examined product. The detection limit of the IFT prior to the addition of adjuvans and preservative was 10 viruses ml-1 as well. After the addition of these substances the detection limit rose to 10³ viruses ml-1. Advantages and disadvantages of the mouse test and IFT are discussed.

The testing of live avian enzephalomyelitis (AE) virus vaccines according to the prescriptions of the European Pharmacopœia (EP) requires a number of animal trials. The purpose of these trials presented is the replacement respectively of two of the animal trials. The first test to be replaced is the virus titration in eggs which includes the observation of the hatching and survival capacity of infected embryos. Therefore a non-competitive ELISA was developed which allows the quantitative determination of AE-virus preparations.
The second test to be refined is the efficacy testing of AE-vaccines. The challenge described in the European Pharmacopœia shall be replaced by the quantification of the serological response of the chickens and their progeny after vaccination. Therefore the minimum amount of protecting humoral antibodies has to be defined. First trials compared the pathogenicity of two challenge strains and evaluated the kinetic of maternal antibodies in chickens derived from vaccinated hens.

The aim of our study was to investigate if a cell culture assay can replace the toxin neutralisation test in mice for the potency testing of alpha toxoid containing clostridial vaccines for veterinary use.
The basis for the development of our cell culture assay was the detection of the cytotoxic/cytopathic action of alpha toxin on cells in culture. Nine permanent cell lines were examined for their reaction to the alpha toxin. The action of the toxin was determined after three days by microscopic examination and MTT assay. The alpha toxin exhibited the strongest effect on the ESH-L cells. We were able to show that the cytopathic effect was neutralised by the international standard for gas gangrene antitoxin (C.novyi) but never by heterologous antisera. Our results showed that the ESH-L cell line was a suitable indicator for the detection of the cytotoxic effect of alpha toxin.

Clostridium (C.) septicum, a spore-forming bacterium of the soil, is the classical causative agent of the malignant oedema, a fatal infection. As immunogenic compound, vaccines contain the toxoidized form of the soluble a-toxin, a lethal exoprotein.
A cytotoxin inhibition test based upon cell culture for the detection of toxin neutralizing antibodies was developed as an alternative to the neutralisation test in mice, which has to be done as measure of quality control according to DAB 10.
Sera derived from cattle that had been vaccinated with a-toxoid-vaccine, and sera from rabbits, from the official quality control of six different clostridial vaccines, were tested.
The in vitro method was able to detect antibodies in the sera of the cows as well as in the sera of the rabbits. The results were reproducible.

The requirements for the quality control of C. perfringens vaccines for veterinary use are described in the monograph 363 of the European Pharmacopoeia (Ph. Eur.). In the current used potency test neutralising antibodies against C. perfringens beta- and epsilontoxin are determined in a mouse neutralisation test (MNT). Two ELISA methods were developed for the replacement of the MNT. Both methods use monoclonal antibodies to determine the quantity of specific antibodies against beta toxin (Capture-ELISA) and epsilon toxin (Competitive-ELISA) in vitro. In parallel to the routine batch potency test in mice, the beta- and epsilonantitoxin levels in 523 samples were estimated in the ELISA procedures. A high specificity and a good reproducibility are evident for both test systems. An interlaboratory prevalidation study was carried out to evaluate the relevance and the transferability of the ELISA procedures. It is concluded that both ELISA systems seems to be suitable alternative methods for assessing the potency of beta- and epsilontoxoid in batches of vaccines for veterinary use.

Botulism is a severe, often lethal intoxication, which affects man and animal. It is caused by the neurotoxins (BoNT) of Clostridium botulinum, an anaerobic spore-former. At present the toxin types A-G have been identified. The extreme toxicity of these substances still make the animal testing inevitable. No alternative method could reach the sensitivity of the mouse bioassay. Based on the investigations presented, a serological test shall be developed, which concentrates the sample and detects the toxin in one step. Preliminary results of an in vitro test detecting BoNT/D are discussed.

Endotoxin (lipopolysaccharide, LPS) is a constituent of the cell walls of gram-negative bacteria and is found in many vaccines produced from these bacteria. High levels of endotoxin can give rise to a range of pathophysiological reactions, and adverse reactions tend to be seen in animals following vaccination.
In this study, pigs of various ages and weights were vaccinated with licensed porcine vaccines and the endotoxin content of the vaccines was determined with the LAL, conducted according to DAB 10. The experiments followed the DAB guidelines relating to the safety testing of veterinary vaccines. The animals were monitored for 48 h after vaccination, their body temperatures were measured, and blood samples were taken for analysis and for the determination of plasma endotoxin levels.
There was a clear relationship between vaccine endotoxin content and changes in blood cell counts and in the clinical picture. Elevated plasma endotoxin levels correlated with the occurrence of initial leucopenia followed by leucocytosis as well as with clinical symptoms ranging from refusal of food and depression to shock-like symptoms. After 24 h, normal physiological values were regained. Young animals weighing between 10 and 40 kg were found to be very sensitive to elevated endotoxin content in vaccines. The differences in individual reactions could be due not only to differences in vaccine endotoxin content, but also to differences in the reactivity of the organism, and in the type of bacteria used or in the composition of the vaccine.

Endotoxins from gram negative bacteria are known to be potent inducers for the synthesis and the release of cytokines such as tumour necrosis factor (TNF) and interleukin 6 (IL-6). The amount of these proinflammatory mediators in plasma from animals and human patients suffering of an acute infection or sepsis, however, is well correlated with the outcome and the prognosis of such diseases (Hack et al., 1989; Ostermann et al., 1997; Rigato, 1996). In connection with regular testing of vaccine lots we determine the release and the kinetic of TNF and IL-6 in piglets after immunisation with different vaccines containing endotoxin. The current results suggest that the amounts of both cytokines increased with elevated endotoxin concentration given with the doses. TNF peaked in plasma after one hour, IL-6 peaked between two and four hours p.appl.. We did not find any influence of the gender of the animals. In contrast, the body weight seems to affect the cytokine release in different ways. Determination of cytokine changes in plasma is a sensitive tool for the evaluation of systemic reactions and supports data about the clinical and haematological signs.

The legal requirements for veterinary vaccines require the testing of each new vaccine batch for safety. In the European Pharmacopoeia a target animal safety test (TAST) is regularly required as a final product test for almost all vaccine batches. This test is also required in the "General requirements for the production and control of live mammalian viral and bacterial vaccines (GRLMV)" and the "General requirements for the production and control of inactivated mammalian viral and bacterial vaccines (GRIMV)" by the European Union.
In several publications the relevance of this TAST has been questioned. Particularly, it is questionable whether this animal test is able to give a significant contribution to the safety of immunologicals considering today's standards of the quality control of animal vaccines, e.g. Good manufacturing practice and Good laboratory practice. However, due to a lack of sufficient amount of data the relevance of this animal test is discussed contradictorily.
1934 batch release protocols were sent to the Paul-Ehrlich-Institut between January of 1994 and December of 1997 for bacterial vaccines including combinations with viral antigens, vaccines against parasitic or fungal diseases and sera. During that period the manufacturers needed 5250 animals to perform the TAST.
Obviously, the requirement for the TAST has a great impact on the animal numbers needed for the quality control of veterinary vaccines. The deletion of the TAST as a routine test could reduce the number of animals considerably. Preliminary results of a survey on this problem are presented and discussed.

Lethal-challenge experiments are provided by German law till now for testing fish vaccines. Finding an in vitro-assay which could replace this experiments was subject of our work.
Two questions must have been answered to develop the lymphocyte blastogenesis assay: Do lymphocytes of immunised fish react the same way as lymphocytes of naive fish, being incubated with antigen and is there a coherence between the lymphocyte's in-vitro-response and the immune status of challenged fish.
Methods to take blood samples from anaesthetised fish and to obtain lymphocytes by centrifugation on gradients were worked out and the lymphocyte's demands on culture requirements were evaluated thoroughly because the ability to proliferate depends on many exogenous parameters. Proliferation was represented and measured by a tetrazolium based colorimetric assay (MTT).

As an alternative method for challenge experiments in fish a sandwich-ELISA has been established for the detection of potentially protective antibody populations. Assay specificity depends on monoclonal antibodies (mabs) directed against different antigens of Aeromonas salmonicida, the causative agent of furunculosis. Comparable levels of salmon antibodies against LPS and the A-layer protein were recorded by ELISA and antigen binding assay. These studies, together with the first analysis of trout sera refer to the specificity and suitability of the ELISA-system for monitoring of individual antibody populations. A study of different mabs against the A-layer protein points to the possibility for detecting at least partially different antibody populations against this virulence factor.

The antibody-stimulating activities and side effects of commercially available adjuvants were compared in BALB/c mice immunised with the immunosuppressive (ISU) peptide of HIV I. Clinical and pathological-histological parameters as well as behavioural changes, were used to assess the distress and pain caused to the animals. Complete Freund´s Adjuvant (FAk) was used as the positive control and PBS as the negative control. The oil-based adjuvants Montanide ISA 51® (M 51), Specol®, and Hunter´s TiterMax Gold® (HTMG) were used with the ISU-peptide conjugated to KLH. The water-soluble Gerbu Adjuvant® was administered together with the antigen KLH conjugate and also with the ISU-peptide conjugated to cholera toxin B subunit (ChTxB). The Dutch "Code of Practice" was used as a guideline for all immunisations. No changes in animal activity or behaviour was observed in any of the groups. All the oil-based adjuvants gave rise to swellings and encapsulations, which were most pronounced in the HTMG-group. Although body weight increased throughout the study, no increase was seen in any adjuvant group for a short period after each booster immunisation, nor after the first immunisation in the HTMG group. FAk induced a light fever after all immunisations. FAk, Specol and M 51 as well as Gerbu-ChTxB induced antibody titres which were detectable in the ELISA, but no detectable antibody the water-soluble adjuvants or following administration of HTMG applied with the ISU-peptide-KLH conjugate.

Immunisation experiments are performed on a large scale in laboratory animals e.g. for the production of antibodies. Depending on the immunisation protocol, severe discomfort may be induced. In several countries, guidelines for the immunisation of laboratory animals were drafted to refine immunisation protocols. One of the aspects that effects the immune response as well as the side effects is the route of injection. Careful selection of the route of injection is therefore crucial to refine immunisation protocols. In several papers the influence of the route of injection on the immune response is studied but generally the induced discomfort in the animals is not evaluated.
In several comparative experiments we studied the influence of the route of injection on efficacy and side effects. The most widely used routes of injection were compared in rabbits (subcutaneous and intramuscular injection) and in mice (subcutaneous and intraperitoneal injection). After the injection of several adjuvant- (Freund's adjuvant, TiterMax, RIBI, Specol, or Montanide ISA50) and antigen- (synthetic peptide, glycolipid, or particulate antigen) combinations by different routes, we studied antibody production to evaluate efficacy of the injection and clinical-, physiological- and behavioural parameters and pathology (macroscopy/microscopy) to compare the induced side effects.
In both rabbits and mice minimal differences in antibody titers were found between the two routes under study. However, differences in the severity of the induced side effects did occur. In rabbits, no significant differences were found in clinical, physiological and behavioural parameters but severe pathological changes were found depending on the route of injection. In mice, the intraperitoneal injection induces significant discomfort as compared to subcutaneous injection as shown by the severity of the pathological changes in the peritoneum and by the decreased activity of the animals. Important disadvantages of intramuscular and intraperitoneal injections are that these injections are difficult to perform properly and that the injection site is hard to monitor. Since minimal differences in antibody production were found between the two routes of injection, we concluded that the subcutaneous route of injection is preferred for the induction of polyclonal antibodies in rabbits and mice when an adjuvant is applied.