Bidogicals

(1991)

19, m-159

MEETING REPORT International 8-9 November

Symposium on Vi.rological A Safety of Biological Products

1990 at The Zoological

Society

Session I. Safety of biologlcals: Regulatory approach The first session was dedicated to the historical and regulatory background of the safety of biologicals. The problem of the safety of biologicals originates from Louis Pasteur’s work on vaccines, since he was the first to demonstrate that a well-defined human microbial pathogen can be used as starting material to prepare drugs capable of having a specific prophylactic action. However, historically the viral vaccines were the first biologicals that raised the complex problem of safety, since they were at the origin of: (11 accidents caused by the incomplete inactivation of the virus used as immunogen; (2) accidents caused by the cell component substrate in which the virus is replicated; and (3) accidents caused by viral contamination of cell substrates or reagents used in vaccine preparation. These adverse effects are usually encountered when a new generation of biological products becomes available. In many instances, the changes in substrate and technology have been the cause of unforeseen hazardous events. The development of vaccines and other biologicals uses technologies today that are increasingly complex. Consequently, it is necessary in every case to analyse the biological in order to identify and to evaluate its particular risk factors. By this approach, it will be possible to diminish the undesirable effects of new biologicals. J. Petricciani discussed the problems raised by the use of substrates, such as embryonated eggs and primary cells, for the production of viral vaccines. Avian leukemia virus (ALV) has been identified as a contaminating agent in two products obtained from chick embryos. A more dramatic example of contamination of a vaccine with an adventitious virus was the presence of SV 40 in polio and other viral vaccines. Unlike ALV, which is not considered to be a human pathogen, SV 40 can induce tumors in newborn hamsters. The follow-up of this incident did not KM-1056/91/020151+09

$03.00/O

of London,

Regent’s

s of the Park, London

U.K.

implicate SV 40 as a source of clinical complication. A turning-point in the cell substrate concept was, in the W&s, the development of human diploid cell lines. The advantage of this system was that well characterized cell banks could be established. The possibility of producing interferon in human lymphoblastoid cell lines raised the question whether cell lines having an indefinite life span could be accepted for the manufacture of virological products. In the 19808, this category of cell lines was exploited in rDNA technology as ‘factories’ of novel products. Hybridomas and CHO cells are examples presently used in the biotechnology industry. The advantage of this system is the feasibility of identifying, by thorough studies, the potential risk factors and of purifying the product in such a way that potential viral and other contaminants can be removed or inactivated. In J. Petricciani’s view, the discussion in the past focused inadequately on the theoretical risk posed by contaminating residual cellular DNA, while the risk of adventitious viruses is real as outlined above. The regulatory agencies should have a flexible attitude in analysing each new biological product case by case. According to Dr Magrath, continuous cell lines (CCL) are attractive alternatives to primary cells for producing large quantities of viral vaccines. It has to be taken into consideration that CCL have an oncogenie potential and that their genome can contain viral DNA sequences. However, their products can be subjected to a purification process which has to be carefully validated for being capable of inactivating and/or removing viruses and other contaminants. The presence of transforming viral and genomic sequences is not considered a major risk factor because cellular DNA cannot induce tumors in sensitive newborn or immunosuppressed animals. In connection with that, the problem of the acceptable quantity of residual genomic DNA was discussed. Actually, a WHO document1 evoked a quantity of 100 pg per human dose of residual DNA as an amount @ 1991 The

International

Association

of Biological

Standardization

152

Meeting

that was not considered to be a risk. However, WHO and other regulatory institutions never put a limit to residual cell DNA. The question is what can be achieved and what can be justified for each product. Dr Brunko introduced the complex interplay between the Commission of the European Community (EC), the Council of Ministers, the European Parliament and other bodies of the EC, which are all involved in the process of legislation. Beginning in 1965, directives have been put into force, the aim of which is the harmonization of national laws regarding the conditions for licensing medicinal products. These directives have now been extended to cover all industrially produced medicinal products, including vaccines and serums, blood products and radiopharmaceuticals. The technical directive describing the tests which have to be performed with these products is in the stage of finalization. The authorization of biotechnological products was regulated in 1987. These products are subjected to a concertation procedure prior to any national decision. However, at present, the Ad Hoc Biotechnology/ Pharmacy Working Group of the EC can only release non-binding opinions. Therefore, a future EC system is being developed which will ensure a uniform authorization in all Member States. According to the type of product, the licensing procedures may be centralized or decentralized. They will be supervised by a European Agency. If necessary, there will be a binding decision by the appropriate EC body. A question raised during discussion concerned the amount of residual cellular DNA considered to be acceptable by the WHO Study Group in products produced on continuous cell lines. The WHO document, explained Dr Petricciani, states that the risk is negligible if the amount of DNA does not exceed 100 pg. But there could be a milligram of DNA and the product could still be safe, under appropriate conditions. Dr Brunko reiterated that quality, safety and efficacy are the factors considered in existing pharmaceutical legislation; social utility may be used as an additional criterion in the agricultural sector sometime in the future. Finally, Dr Melnick raised the issue of whether there should not be a common regulatory agency for the US, Canada, and Western Europe, where the same kinds of problems exist, in order to avoid duplication of regulatory functions. Session

II. Virus

inactivation

In order to assure the safety of biological products, it is essential that actual and potential viral contaminants are either removed or rendered non-infectious.

report

The sessionon virus inactivation was introduced by Dr Melnick who reminded the meeting of the dangers in assuming that the kinetics of viral inactivation followed a simple straight-line curve. The failure to correctly interpret laboratory data led to the release of a few lots of formalin inactivated polio vaccine containing small amounts of infectious virus with the consequent induction of paralytic polio in some of the recipients. Furthermore, the expectation that residual live virus would be detected by the inspection of samples into the CNS of monkeys was also found to be mistaken, since viral replication could occur without induction of lesions in the CNS, and detection in cell cultures was also rendered more difficult by the formalin treatment. Dr Brown reported similar problems with footand-mouth disease (FMD) vaccine where a number of outbreaks of FMD could be traced, by means of oligonucleotide mapping procedures, to the use of incompletely inactivated vaccines. It is interesting that the European Community reacted to these observations by abolishing vaccination, accepting that natural outbreaks were preferable to those induced by vaccine, despite evidence that reliable inactivation was possible using imines. The mechanisms of inactivation were reviewed by Drs Adamowicz and Horowitz. Viral inactivation may be achieved by procedures which either interfere with the attachment of the virus to the cell receptor, by altering the coat protein to make it unrecognizable to the target cell, or by destroying the biological activity of the viral nucleic acid, thus preventing replication. It should be noted that methods resulting in the removal or alteration of the coat protein leave the genome intact and retain the infectious potential of the virus, although in practice it may not be infectious. Thermal inactivation has been successfully used for many years to treat a number of biological products and acts by denaturing double stranded nucleic acid as well as by disturbing both surface and internal proteins. However, the kinetics of inactivation are often not easy to establish, furthermore, viruses exhibit a wide range of sensitivity and the exposures necessary may be too rigorous for some products. Both gamma and UV radiation have been employed for a number of years for the inactivation of viruses and seem to involve the whole virus structure. Although resistance varies, in general members of the same class exhibit similar levels of sensitivity. The most commonly used chemical inactivating agents have been formaldehyde and beta propiolac-

M-m report tone but imines and tri-N-butyl phosphate are also used. The latter is particularly effective in the disruption of enveloped viruses in blood products. In summary, successful inactivation of viral contaminants in biological products requires that the method of inactivation is appropriate to the product and the anticipated type and load of contaminating viruses. It is also necessary to make a careful evaluation of the kinetics of inactivation and to establish that all infedious particles are equally exposed and sensitive to the selected agent. The first point raised in discussion was that public pressure can sometimes influence the decision as to when to put a vaccine on the market. What occurred in the Cutter incident might be repeated with another vaccine, if the scientific community and the public health authorities were swayed to authorize the vaccine before it had been adequately tested. Dr Melnick confirmed that poliovirus in today’s vaccine is effectively inactivated. He mentioned, that in the past, active virus could sometimes be recovered from the blood of monkeys used for testing vaccine, even though their spinal cords were free of virus. Dr Horaud pointed out that the RNA of the poliovirus genome is not completely inactivated by formaldehyde. The RNA in a vaccine can produce infectious particles, but does so slowly. In Sweden, a long period (up to 1 month) at a temperature lower than 37°C has been used to inactivate poliovirus. The vaccine obtained by this method was highly immunogenic, as evidenced by the persistence for many years of antibody in those vaccinated. Sessions III and IV. BSE and related agents A group of diseases of animals and man share a similar pathology involving spongiform degeneration of the brain. The diseases include Kuru, CreutzfeldtJakob disease and Gerstmann-Straussler syndrome of humans, scrapie of sheep, transmissible mink encephalopathy and bovine spongiform encephalopathy of cows, as well as others. The nature of the causative agents was reviewed by Dr S. Prusiner. They are characterized by a high resistance to treatments expected to destroy nucleic acids and sensitivity to treatment with proteases, while being generally extraordinarily hardy. No readily applicable in vitro culture method or assay is available, so that studies of the agents depend on in uiuo assays of infectivity in mice or hamsters. For distinguishing strains these agents have been termed prions, and the protein which copurifles with infectivity has been termed the prion protein. A celhrlar form of the prion protein has been identified which is indistinguishable from that

153

found with the agent except that it is non-infectious and the form associated with the agent is partially resistant to protease treatment. By preparation of transgenic mice carrying the celhrlar prion gene from hamsters it was possible to obtain animals which would give rise to hamster-like strains if inoculated with mouse adapted scrapie. Human individuals carrying the gene for Cerstmann-Straussler syndrome develop CreutzfeldtrJakob-like symptoms at a relatively early age and at high frequency, and it has been shown to be a genetically determined dominant characteristic. Transgenic mice were prepared carrying a gene from a Gerstmann-Straussler carrier, and it was found that they developed scrapie without infection. The data collectively suggest that the prion protein forms at least a component of the infectious agent. A number of distinct strains of the agent may exist, even when they all possess the same prion protein, however. Three models have been proposed. The agent may be a filamentous virus, although its resistance makes this unlikely. Secondly, it may be a protein capable of nucleating an abnormal ‘crystallization’ of a normal cellular protein. If this is the case the cellular protein must be able to ‘cry&all&e to a number of similarly stable forms to account for the phenomenon of strain variation. Finally, it may consist of a small piece of nucleic acid coated with a normal cellular protein, although no such nucleic acid has ever been demonstrated, despite repeated efforts. The nature of the agent is therefore currently not resolved. Bovine spongiform encephalopathy is a disease restricted to the U.K. or U.K.-linked animals and is caused by a prion like agent. The current position was reviewed by Dr R. Kimberlin. The first case was recognized in 1986, although earlier cases have been identified retrospectively. The epidemiology is consistent with a common source epidemic with a first exposure in 1981, and is attributed to changes in feeding practices, in which the treatment of protein supplements derived from sheep was altered by the omission of solvent extractions and steam treatments. The age of the affected animals has fallen progressively since 1986, consistent with adaptation of the agent to cows, as infected cows were recycled as animal feed. The factors affecting the ease of transmission of prions include the dose and the route of administration. In addition, prions originating in one species may be difficult to transmit to another, a phenomenon termed the species barrier. The tissues which are expected to have the highest titre of infectivity in BSE-infected animals are, by analogy with scrapie, brain and tissues of the lymphoreticular

154

Meeting report

system while serum is low risk, with a titre expected to be at least 6 loglo lower than that in the brain. While the situation is filled with uncertainties, experience with scrapie suggests a number of ways of minimizing the risk of transmission, including appropriate selection of source animals and tissues. Serum, in particular, would be thought a low-risk issue. Based on transmission of scrapie, and likely exposure to bovine materials, it is generally considered that the probability of transmitting bovine spongiform encephalopathy to humans is very low. However, iatrogenic transmission of CreutzfeldMakob disease is well documented, and was reviewed by Dr R. Will. Human growth hormone extracted from pituitary glands from cadavers was first used in 1960, and about 25 000 patients were treated with it before it was withdrawn in 1985. A total of 14 cases of CJD in recipients of this material have now been reported; several different batches and sources have been implicated, and the final total may be higher. The disease followed an abnormal course including the clinical symptoms shown. Iatrogenic transmission by other means has also been reported, including dura mater transplants and the use of surgical instruments which had not been adequately sterilized before use in neurosurgery. The difficulty with cadaveric human growth hormone was overcome by the use of recombinant DNAderived products. However, validation of production processes for the removal of prions remains potentially important for other products. Dr M. Pocchiari reviewed studies on the validation of a modified production process of cadaver derived human growth hormone to assess its ability to remove scrapie agent, and thus presumably also the agent of CJD. It was shown that column purification and urea treatment would remove at least 9 logio of infectivity, and it is possible to assess the likely reduction in cases which would be expected as a result. Dr D. Taylor described studies of the inactivation of scrapie-like agents, with a view to validation of the ability of processes to inactivate or remove the agent. Extremes of pH, such as 1 M NaOH or formic acid, were fairly effective, whilst irradiation with ultraviolet or gamma irradiation left infectivity even under extreme conditions. Formaldehyde in particular was ineffective, and might even stabilize the agent to subsequent heat treatment. Dry-heat sterilization was not satisfactory, while autoclaving under the appropriate conditions was effective. However, strains differed significantly in their sensitivity to the different treatments. The difficulties associated with contamination with scrapie-like agents were therefore considerable,

although certain treatments gave some degree of confidence. To the question of whether calf sera used in cell cultivation can transmit the BSE agent, Dr Kimberlin made the following comments. Infectivity has not been detected in the blood of naturally infected sheep and goats, even at the clinical stage of the disease. Infectivity has been detected irregularly in the blood of some laboratory animals that were experimentally infected with scrapie, showing that, even in these circumstances, titres are very low; less then 10 i.c. LDso units/ml of blood (the detection limits of bio-assays). The recovery rate can be increased using concentration methods to increase the sensitivity of detection. One of the simplest of these is to prepare buffy coat cells, which shows that such infectivity as is present is cell associated. This is supported by the very rare recovery of infectivity in serum of experimentally infected animals. The expectation, from studies of natural scrapie, is that no infectivity would be detectable in any tissue in foetal or neonatal animals. The only likely source of blood contamination would be from foetal membranes, which are a known source of the scrapie agent in sheep. However, this would only apply if BSE is maternally transmitted (like scrapie). In any case, contamination could be avoided at the time of collection and reduced during the preparation of serum. Endemic BSE has only occurred in the British Isles. Therefore the risks from animals in other countries must be regarded as very low, even if some are infected. In many situations, the risks from a few infected animals outside the British Isles would be subject to enormous dilution factors. Additional margins of safety would arise from: (a) the likely removal of contamination with such a ‘sticky’ agent by procedures, like centrifugation of precipitates, membrane filtration, and column chromatography; (b) the small dose of product in relation to any initial contamination; and (cl the use of non-neural routes of injection, which are less efficient than the intracerebral route. Session

V. Retroviruses

The human immunodeficiency virus (HIV) infection is spreading rapidly almost world-wide and has reached endemic proportions in Central Africa, in parts of South America and in certain populations in industrialized countries with risk behaviour for contracting venereal or blood-born infections. The mechanisms of HIV-induced immune suppression are

-ing vfi still not entirely clear, as direct T-lymphocyte destruction after viral infection cannot account for the almost complete loss of CD4 cells in the final stages of disease. Various indirect mechanisms of immune cell destruction were discussed. Two types of endogenous retroviral-like sequences related to type C or Intracisternal A Particle (IAP) genes were detected in the Chinese hamster (CH) genome. Their structure and expression were analysed by virological and molecular approaches in different CH cell lines. When clones isolated of type C or lAP from Chinese hamster ovary (CHO) cells were analysed, retrovirallike sequences were revealed. However, only a few retroviral-like type C particles and no IA particles were detectable by electron microscopy. The inactivation of type C viral production by halogenated pyrimidines or other classical viral activation agents was not highly efficient. No replication of this virus could be obtained in rodent, canine, simian and human cell lines. Data on CHO retrovirus-like particles were reported also by another group. Two families of sequences related to IAP of mice and Syrian hamsters were identified in cytoplasmic RNA from CHO cells. None of the four clones which were sequenced exhibited intact gag, pol, or env open reading frames. Only family II IAP sequences were present in purified extracellular particles of CHO cells. Several cDNA sequences related to mammalian C-type retrovirus genomes were isolated and cloned from gradient-purified, extracellular particles from recombinant CHO cells. All were homologous to the conserved endonuclease domain of murine leukemia virus. Nucleotide sequence analysis of the largest cDNA revealed multiple interruptions of the endonuclease-encoding reading frame, providing one possible explanation for the non-infectious nature of the particles observed in CHO cells. Both types of retrovirus-like sequences identified in purified extracellular particles of CHO cells (CHO IAP family II and C-type) were present as conserved, moderately repetitive sequences in the DNA of alI CHO cell lines examined, as well as in DNA from a Chinese hamster liver. It is therefore likely that the extracellular retrovirus-like particles of CHO cells are the products of endogenous provirus elements present in the germline of Chinese hamsters. Validation of virus removal during product manufacture is one of several techniques which can be used to help establish product safety. The usefulness of the validation technique will be illustrated for the manufacture of immunoglobulins by the Cohn-Oncley

155

ethanol fractionation procedure in which the partitioning and inactivation of HIV was studied, and for the manufacture of Factor VIII. In the case of Factor VIII, which was contaminated with HIV and transmitted infection to product recipients, several manufacturers explored means of eliminating infectious virus by using a validation procedure involving a scaled-down manufacturing process and the spiking of individual steps to determine the extent of removal from product or inactivation of virus during manufacture. Assays for virus remaining after each step were performed within tissue culture. Some products were also tested in chimpanzees. Results of tissue culture assays could be correlated with safety testing in animals, and also with subsequent retrospective analysis of actual product use in patients. Correlation was good, suggesting that process validation is useful. Current use of virus validation to assess virus removal from products made from continuous cell lines, such as human hybridomas which could potentially contain viruses infectious for humans, as well as some limitations of these validation techniques, was also discussed. Retroviruses and herpesviruses pose particular problems in safety testing for two principal reasons. First, these groups of viruses are associated with the aetiology of neoplasia, immunosuppression and other serious conditions. Second, both groups have the capacity to remain as latent infections within cells, particularly lymphoid cells. Moreover, these viruses are not usually contaminants. Contamination by these virus groups has to be considered whenever human cell lines are used and in some instances when DNA-constructs of human origin have been employed. The major sources of concern are human-human hybridomas and murine-human hybridomas, particularly where Epstein-Barr virus (EBV) has been used in their generation. EBV remains as an unintegrated episome in B-lymphocytes where it expresses a limited number of transcripts including those for the nuclear antigen EBNA-1. Detection of EBV in cells is most accurately determined by DNA hybridization or by polymerase chain reaction (PCR) amplification. Until recently, cytomegalovirus was the other major herpes virus of concern. However, in the last few years, two new lymphotropic herpesviruses, HHV-6 and HHV-7, have been isolated. The discovery of these viruses emphaaises the need to include infectivity assays in evaluating the safety of a cell line rather than relying solely on the detection of viral genomes.

156

Meeting report

Although endogenous sequences related to human retroviruses are present in the human genome, the known infectious retroviruses are exogenous, i.e. acquired by infection. They include the human spumavirus, the human leukaemogenic retroviruses HTLV-I and HTLV-II and the human lentiviruses HIV-l and 2. A general scheme of testing can be proposed in which Master Cell Banks of Manufacturing Working Cell Banks are screened, both by virus isolation and by DNA hybridization. The extended cell bank is re-analysed by hybridization in case a low-level contaminant has spread in the culture. PCR is particularly valuable for the analysis of small quantities of final product. The discovery of new human herpesviruses and retroviruses emphasises the importance of validating the clearance of these viruses in the downstream purification process. Although HIV is more resistant to a number of chemical and physical agents than retroviruses like MuLV, we do not recommend the use of HIV in validations. A safe and reliable model of HIV is available in the form of the ovine lentivirus maedi-visna, which is available in high titre and poses no risk to the operators. In addition to a retrovirus, a herpesvirus like HSV should be utilized, and finally a resistant small virus-like poliovirus should be included to act as a severe test of clearance and/or inactivation. Since HIV may occur in human plasma and other retroviruses in established cell lines used for the production of monoclonal antibodies or recombinant proteins, respectively, it is necessary to exclude these infectious viruses from the source material or, at least, by the manufacturing procedure from the final product. Screening of human plasma donations and detailed analysis of master seed lots of mammalian cells considerably reduces the risk of contamination of the final biological products with retroviruses. Data on the inactivation of HIV-l and HIV-2 in aqueous protein solutions by heat treatment, by ethanol precipitation at low temperature and by nonionic detergents, respectively were reported. Dr Kurth explained that a person with HIV or a monkey with SIV can become antibody-negative. Virus cannot always be detected in the lymph nodes of such individuals, even with PCR. Because it has been observed that animals that are antibodynegative may become positive again, he concluded that blood from a once-infected animal should never be used in the manufacture of biological products. The major point of discussion of Dr Anderson’s paper was whether proviruses present in cells fused

with CHO cells could recombine so as to produce infectious particles. Dr Anderson explained that, although some CHO cells produce extracellular retroviral-like particles, these particles do not infect other cell lines. Dr Petricciani pointed out the importance of not having such particles in the final product, regardless of whether they are present in the cell line used as substrate. Two points were raised in discussion following the communication of Dr Marcus-Sekura. The first was the difficulty of evaluating whether two methods of virus removal actually have an additive effect if each alone removes nearly all the virus. It was suggested that in order to test for the additivity of the two methods used to remove or inactivate virus, each method could be tailored so that it would remove only half the virus present in a sample. Dr Marcus-Sekura thought, however, that this would not correspond to the real conditions of manufacturing. The second point, mentioned by Dr Melnick, was that the first 99% of virus present in a sample is usually easily inactivated, but there is always a remaining, resistant fraction. He suggested using the resistant fraction to spike the succeeding step. Finally, in response to Dr Hilfenhaus, Dr Horaud stated the desirability of using more than a single type of virus in validation procedures and of co-ordinating purification and validation procedures. Session

VI. Viral contamination

and detection

(1)

This session dealt primarily with problems of viral contamination introduced to cell culture systems or laboratory personnel via sera or laboratory animals. The prevalence of bovine viral diarrhoea virus (BVD) contamination in commercial lots of fetal bovine serum was emphasized in the first two presentations by Drs Erickson and Levings. Indeed, unless careful preventive measures are taken in selecting reagents used from origination through establishment and expansion of cell lines, one may expect this virus to be present in all lines of susceptible species (bovine, porcine, feline, simian and others). One report of infection of human infants by an unclassified pestivirus’ was referenced to point out potential for serious concern to the vaccine manufacturing industry. Suggestions for pre-treatment of sera and other tissue culture reagents to eliminate BVD included heat inactivation or, more effectively, gamma irradiation with 26-34 M rads at -40°C. More sensitive assay procedures have also been described recently.3 In the third presentation, Dr Mahy reviewed new incidences of infection of animal caretakers with lym-

wing f-Port phocytic choriomeningitis (LCM) virus and Ebolalike f’lloviruses. The former study was the first to implicate nude mice as a potential source of transmission for human LCM infection. The speaker emphasized the extreme severity of public health hazard associated with filovirus infection in imported animals, documenting 53-100% primate fatality depending upon the virus strain involved. Fortunately, no human fatalities occurred with the 1989-90 outbreaks recorded and discussed. During the discussion of his talk, Dr Erickson emphasized that to assure the quality of fetal bovine serum it is desirable both to check that the immunological response to the BVD vaccine is high and to screen donor animals for persistent BVD infection. He also mentioned the usefulness of a method for differential diagnosis, to distinguish vaccine and field strains of BVD virus. For example, if a strain with a glycoprotein 1 deletion is used for vaccination, then animals with antibody to this glycoprotein could be considered infected. Controls are required after gamma irradiation in order to ascertain the efficiency of BVD virus inactivation. In addition, the effects of irradiation on the protein components of fetal bovine serum might lead to some modifications of these components associated with different immunogenic properties. The infectivity of bovine semen and its role in BVD transmission were discussed. Preliminary approaches to BVD vaccine using glycoprotein-depleted virus were mentioned. Lastly, the question of cross-reactivity between BVD and human hepatitis C virus, which is assumed to be a pestivirus, was raised but no answer could be provided yet. Dr Mahy proposed that a more reliable and easier to perform assay be developed for testing for filovirus. In the existing serological tests he described, there may be viruses, as yet uncharacterized, which crossreact but which are silent as far as pathogenicity is concerned. He reported that a good correlation was observed between the immunofluorescence assay and the neutralization test, when the immunofluorescence assay was carried out with dilutions higher than 256. However, in some animals much higher titres of antibody were observed, although the animals showed no signs of disease and the virus could not be isolated. This is thought to represent a transient seroconversion, which remains unexplained. Session VII. Viral contamination detection (2) In the session on viral contamination

and and detection,

157

Dr R. Hay described the problems which can arise when cell cultures are contaminated by the introduction of other animal cells or by microbial agents. Contamination with other animal cells can be monitored by several methods including fluorescent antibody staining, isoenxyme analysis and DNA fingerprinting using molecular probes. Detection of microbial contamination is more straightforward, although infedion with mycoplasms is a significant and chronic problem. It can best be controlled by using pre-screened reagents without added antibiotics. Contamination with viruses is also a major problem, especially when they do not lead to a cytopathic effect. Such contamination can also be a hazard for the staff handling the cultures, as shown by their infection with lymphocytic choriomeningitis virus from transplantable tumour cell lines. Dr Robertson described the polymerase chain reaction and its impact on diagnostic methodology. The power of the technique, i.e. its capacity to amplify specific nucleic acid sequences from extremely low levels, makes it attractive, not only because of its sensitivity but because it can be applied in a complex background of mammalian cell genomes. There can be little doubt that the technique will have widespread application within the next few years. Dr Agut drew attention to the difference between virus infection and pathogenic&y and its importance to the persistence of the viral genome in tissues. He stressed that the growth of a virus in a cell culture and the associated cytopathic effect may not reflect its pathogenicity for man. Consequently, animal models are likely to provide more relevant information. However, in many cases, markers of pathogenicity are often lacking, whereas the combination of serology and molecular methods is now powerful enough to diagnose most viral contaminants. Two suggestions were made concerning the identification of cell lines. The first was that-IWR should be used across the variable regions in Jeffrey’s probes to pick out strain variations. The second was that an easy and rapid technique to identify particular lymphoid cell lines is to look for antigen receptor gene rearrangements. In spite of the acknowledged sensitivity of PCR, several limitations of this technique came to light. So far, PCR is only applicable when the genetic sequence (for example, of a virus) is known, although techniques are presently being developed with which unknown sequences could be detected. Furthermore, results obtained with PCR are positive, regardless of whether virus is live or inactivated : a negative result is obtained only in the absence of viral nucleic acid.

A. Chauhan

158

PCR is therefore not useful for determining whether virus has been inactivated by a given procedure. Currently, PCR is being used in epidemiological studies to distinguish wild-type from vaccine strains of poliovirus. Dr Schild suggested that in order to make the most effective use of the new techniques, such as PCR, they be targeted towards defined goals. Scientific workshops could be held to discuss which viruses ought to be tested for in various manufacturing procedures, and when PCR could be usefully applied. One application of PCR, for example, would be the detection of contaminating mycoplasm in cell lines. Session VIII. Round of virus safety

table on regulatory

aspects

The Sth, and last, session of the Symposium dealt with regulatory aspects of virus safety of high technology biological products. The first two presentations discussed the concept of assuring the freedom of viral contamination of products. The argument was well made that, except for testing unacceptably large volumes of the product itself, safety can only be assured by demonstration that the manufacturing process was designed in a way which still remove or inactivate possible viral contaminants to insignificant levels. For instance, if a guideline was to be applied requiring demonstration of a theoretical contamination level of not more than one infectious particle in one liter, then a product with an initial contamination level of 3-O log&ml would have to give an overall clearance factor of at least 9.0 loglo to provide an adequate margin of safety. It might be desirable to require even greater reduction factors to account for possible high resistance fractions in the viral progeny. Clearance factors of this magnitude can be demonstrated in validation studies in which sizeable amounts of virus are added deliberately to each manufacturing step and a determination is made on how efficiently each step will remove or inactivate the virus. The manufacturing processes may be validated using relevant viruses, i.e. viruses likely to contaminate the product. Examples are HIV, HTLV-I and II, herpes viruses or hepatitis viruses in products derived from human blood. Because it is difficult to grow and titer efficiently many of these viruses, preference is often given to ‘model viruses’ which grow well, can be accurately assayed and are safe for handling. Examples are pseudorabies virus as a model for the human herpes virus group, attenuated poliovirus for hepatitis viruses and murine leukemia

et al.

virus for HIV. Validation with one of the relevant viruses may still be necessary. Notes for guidance on ‘Validation of virus removal and inactivation procedures’ were drafted by the Ad Hoc Working Party on Biotechnology/Pharmacy of the European Community in February 1990 and are to be approved shortly. The technology allowing for the production of human monoclonal antibodies for therapeutic or diagnostic purposes has been perfected and a number of preparations have been licensed. They have a clear advantage over antibodies of murine origin as medicinal products for human use. Nevertheless, one of the fusion partners is frequently a murine myeloma line because of the lack of a suitable human myeloma cell line. Experience has shown that fusion will lead to activation of infectious murine leukemia viruses. Consideration also has to be given to contamination by ovine agents such as Visna-maedi or scrapie due to rosetting of human B cell with sheep RBC. EBV may be activated if used in the immortalization of the human fusion partner. Other viruses known to infect and persist in human lymphocytes include hepatitis B virus, HIV, HTLV-I and II and viruses of the herpes virus group. Determination of the health status and geographical origin of the human donor may help to exclude more exotic types of viruses, e.g. Ebola or Marburg. More than with any other product, materials derived from human lymphoctes must be assessed individually depending on their quality and intended for use. A note for guidance for ‘Production and Quality Control of Human Monoclonal Antibodies’ has been recently approved by the appropriate authority of the European Community. The same group which developed the note for guidance for human monoclonal antibodies finalized the draft of the note for guidance for ‘Medicinal Products Derived from Human Blood and Plasma.’ The document was presented at the symposium by Prof. Florian Horaud. Many of the validation issues discussed by previous speakers bear directly on products derived from human blood, including immune serum globulin, coagulation factors, plasminogen and other complex serum derivatives. Moreover, in view of the multitude of source materials entering a production batch, screening of donors and testing for transmissible agents is an important and formidable task. Therefore guidelines for the make-up of the physical facilities, the collection and documentation process are integral parts of the guidelines. Validation of virus removal or inactivation may be complex in view of the diverse steps used in the manufacture of human blood-derived products which include alcohol

Maetlng rqmt

precipitation, treatment with detergents, column purification and pasteurization. The use of HIV in the validation process is obligatory. The WHO policy in assuring freedom of biological products from viral contamination was presented by Dr Victor Grachev. It is based mainly on two documents: ‘General Bequirements for Manufacturing Establishments and Control Laboratories’; and ‘Gen-

159

eral Bequirements for Sterility of Biological Substances.’ These requirements regulate not only general questions of production of biological products but also the safety, including viral contamination. 1. WHO. Technical Report Series 1987; 747. 2. Potts BJ et al. Lancet 1987; 972. 3. Potts FhJet al. J Virol Meth 1989; 26: 119-124.