Introduction
The history of epidemics and their control has highlighted the numerous oppositions generated by vaccines, against the social control they impose, in defiance of the public authorities or because of potential secondary side effects.Footnote 1 The recent controversies about the efficacy of vaccines against SARS-CoV-2 illustrate well what Jeremy Ward calls conditional critiques Footnote 2 that the vaccine was developed ‘too quickly’; that there was insufficient consideration of potential side effects; that hygiene and other sanitary measures (e.g. masks, social distance, ventilation) would be more appropriate to prevent contagion. Said differently, are these vaccines actually, that is, in everyday life, safe and efficacious? What range of evidence enables assessment of their properties and their stability? It seems that all these questions require an ‘objective’ answer that should come from scientists to put an end to public fears or resistance. But since Ludwick Fleck’s work on the social construction of medical facts in the 1930sFootnote 3 and its rediscovery in the late 1980s,Footnote 4 numerous studies in the history and sociology of science and medicine have analysed the social dimension of illness, disease, and medical knowledge. In particular, they have shown how illnesses have both biomedical and experiential dimensions, or what qualifies as biomedical evidence is socially negotiated and interpreted.Footnote 5 Many comparative histories of human vaccination in Western countries have been written within that frame.Footnote 6 Perhaps the most emblematic work is Linda Bryder’s comparative history of the anti-tuberculosis vaccine BCG (Bacillus Calmette–Guérin).Footnote 7 She shows how the data gathered by Albert Calmette, the co-founder of BCG, were bitterly criticised by British medical statisticians. According to Bryder, the opposing positions of the scientists were underpinned by ideological differences (a liberal ideology based on individual freedom of choice to be vaccinated or not) and by different approaches to disease prevention (favouring an efficient public health system with sanatoriums). Conversely, Calmette’s data were not problematic in Scandinavia, where vaccines were accepted. This ideological analysis seems to revivify Erwin Ackerknecht’s thesis, according to which liberal countries such as Britain favoured sanitary measures instead of control over bodies.Footnote 8 However, as Peter Baldwin has pointed out, a careful comparative analysis of different diseases in various countries shows that states’ ways of controlling diseases vary widely, both within and amongst countries, between diseases and over time, according to many issues – social, economic, political, medical, or scientific.Footnote 9 Historians of human and veterinary medicine have since confirmed his claim in fascinating accounts that analyse diverse contexts and situations.Footnote 10 However, all of these case studies have a common ground: they all depict strong administrative structures and the state’s commitment to impose their policies, whether sanitary (mainly based on diagnosis/isolation/disinfection) or medical (based on vaccination or serum therapy). Medical policies imply a state oversight of vaccines’ safety and efficacy, no matter whether the producers of these biologicals are public institutions or private pharmaceutical companies.
Focusing on the trajectories of vaccines against foot-and-mouth disease (FMD) in France and the UK between the 1920s and the 1960s, this paper is an attempt to complete this social history of vaccinations, showing that vaccination policies are not always a top-down process emanating from the states but can also emerge as the result of private and local initiatives that eventually are endorsed by the whole country. These local initiatives are rooted in various experiences of FMD (those of farmers, veterinary practitioners, and virologists) that nourish imaginaries (i.e. the intuitions, perspectives, discourses, and utopias that accompany the design of a new technology and its uses amongst populations),Footnote 11 which play an important role in the assessment of the vaccines’ efficacy.
FMD primarily affects cattle but also pigs. Its transmission to humans is exceptional.Footnote 12 The disease, usually not lethal, is characterised by ulcerations on the hooves, jaws, and udders of contaminated animals that prevent them from eating, standing, and being milked. FMD thus impacts the whole agricultural economy of a country, especially since there are seven serotypes of the FMD virus, including many variants that not only provoke symptoms of diverse intensity,Footnote 13 but also render natural immunity impossible. For all these reasons, up to the nineteenth century, the virus was rampant in Europe, and populations were used to dealing with the disease. Abigail Woods has written a complete history of FMD control in the UK. According to her, it is from the last quarter of the nineteenth century that the UK considered FMD a contagious disease, against which it was possible to establish sanitary measures. Indeed, the UK had large agricultural exploitations and was a leader in the pedigree livestock international trade. In order to maintain this lucrative trade towards countries free from FMD (such as the United States or Canada), the British government, backed by pedigree livestock farmers – often members of the House of Lords – adopted a ‘stamping out’ strategy: a sanitary policy based on the isolation and culling of contaminated animals and their contact, in exchange for compensations.Footnote 14 Woods clearly demonstrates why the British Ministry of Agriculture, Fisheries and Food (MAFF) remained convinced that slaughter was the best and most economical solution to control the disease. She mentions serums and vaccines as continental options that the MAFF could not accept and justifies that position by ‘national differences in geography, commerce and scientific tradition’ that influenced the way researchers in the UK and the Continent judged vaccines, their risks, and their benefits.Footnote 15 However, research for efficient vaccines against FMD in Britain engulfed massive public funds since the 1920s and led to a commercial partnership in the 1960s with the Wellcome pharmaceutical industry to produce and sell the vaccines abroad.Footnote 16 While Abigail Woods explains this with reference to military reasons, she does not say much about how British virologists apprehended this research, settled their protocols, and assessed the vaccine’s efficacy. This untold side of the story is, however, worth analysing in comparison with another vaccinating country such as France. Indeed, in the late nineteenth century, France faced different issues. FMD was rampant, jeopardising milk production and the animal workforce.Footnote 17 But the French government did not seem to really consider FMD a national issue: some measures of isolation and disinfection of affected premises had been enacted but without sanctions if not applied. As a consequence, many folk remedies were used to alleviate the symptoms as much as possible.Footnote 18 Some pharmaceutical industries chose to invest in vaccine production as early as the late 1930s.Footnote 19
The purpose of this paper is to go beyond the economic and national cultural explanations of the rejection of vaccines, to analyse the knowledge and practices developed to assess the vaccines’ efficacies in each country. There were indeed many similarities: 1) French elites – politicians, veterinarians, and scientists – shared the same conception as the MAFF about the way animal diseases should be managed; 2) both countries launched research on FMD immunity and vaccines; 3) the vaccine preparation process was similar,Footnote 20 and French and British virologists shared the same frame of thought regarding the regime of evidence – statistical – that had to be put in place in order to assess their efficacy. However, despite these similarities, the scientists in charge of testing the vaccines did not agree on their degree of efficacy, nor on the way they had to be used as a control tool. As demonstrated by Christoph Gradmann and Jonathan Simon, sera and vaccines are biological substances, difficult to standardise, for which the preparation through the attenuation of their pathological properties needs a constant clinical evaluation to ascertain their innocuousness and efficacy.Footnote 21 Standardising these substances requires various systems of control – scientific, industrial, and clinical – which mobilise a large range of evidence (e.g. statistical and physiological) to ascertain the homogenisation of the effects in practice. These processes have always been difficult to establish and are the result of alignment between the work done in the laboratory and the clinical observations. Moreover, in The Invisible Industrialist, Jean-Paul Gaudillière and Ilana Löwy analyse how standardisation possesses a double signification – a ‘restricted’ meaning that ‘denotes a formal agreement’ on the techniques, substances, or instruments to use ‘to enforce the replicability of measurements and operations’Footnote 22 – and a ‘metaphorical meaning’, which refers to the ‘moral and social values’ that a technology posesses, which is ‘a certain understanding of the quality, the excellence or the value that gives the basis to assess and reach consensus’.Footnote 23 This paper follows Gaudillière and Löwy to show how vaccine efficacy also possesses two meanings. The first one, technical – or experimental – refers to test protocols and regimes of evidence that are settled to assess the vaccines’ capacity to protect from infection. The second, practical – or experiential – refers to the experience various actors have of diseases and their direct or indirect consequences on economic and social exchanges, as well as on representations and imaginaries they share about vaccines and vaccination.Footnote 24 These two meanings, experimental and experiential, historically and culturally situated, contribute to the populations’ perceptions of risks due to diseases or vaccinations. As a result, the very same vaccine could be perceived as completely efficacious by certain actors but rejected by others, both between and within nations.
This paper will analyse the issues and imaginaries underlying these scientific, political, and economic choices in France and Britain, and show how the two meanings of efficacy are deeply intertwined, explaining why France used vaccines despite their rejection by political and veterinary elites, and also why British produced vaccines that were not used inland. In its first two parts, it will focus on how the two meanings of efficacy – experimental and experiential – were characterised for each group of actors involved in the control of foot-and-mouth disease. A third part will show how, in the French case, these two meanings were aligned in favour of the absence of animal symptoms, contributing to the redefinition of the standards of efficacy, and to the generalisation of vaccines. On the contrary, in Britain, the standardisation of safety and efficacy tests discarded vaccines from their inland use, whereas they were exported to Britain’s commercial partners as control tools of the disease, redefining the conditions for trade in meat and animals on international markets.
This paper is based on the exploration of public sources, i.e. the French and British National Archives (Centre des Archives Contemporaines/CAC and Public Record Office [PRO]), private sources (from the Mérieux Institute),Footnote 25 as well as on professional agricultural press and veterinary journals.
In Britain: biological warfare and medical research
In Britain, research into means of immunisation against FMD emerged at the crossroads of various interests. The Treasury was looking for a less expensive solution than slaughter, and the Medical Research Council (MRC) was looking forward to developing new knowledge of viruses that could be relevant to human diseases. Thus, despite the hostility of the Ministry of Agriculture, a research station dedicated to the study of the FMD virus was established in the tuberculin production centre of Pirbright, Surrey.Footnote 26 The study of viral properties, modes of transmission and resistance to external agents, as well as the development of typological tests, occupied most of the station’s researchers. This new knowledge was used to trace the origins of virus importations and to strengthen border controls. As it was still impossible to cultivate the virus in vitro, its culture in vivo on susceptible bovines made the implementation of large-scale or even medium-scale experiments within the station difficult. As a result, in the 1930s, only a very small number of experiments, conducted mostly with the consent of farmers on surrounding farms, sought to establish the value of serum therapy in protecting pedigree cattle.Footnote 27 The Treasury planned to use this process to avoid expensive compensations in case of an FMD epizootic, but it was in no way a question of abandoning slaughter, the only method considered by the Ministry of Agriculture, by breeders of pedigree cattle and by veterinarians, as being able to eradicate the virus and keep the territory disease free.Footnote 28 Indeed, as Susan D. Jones has shown, the British mapping of epidemics associated the concept of health with the absence of germs – viruses or bacteria – in a given area.Footnote 29 Hygiene and sanitary policies were therefore the main weapons against diseases. Moreover, according to Abigail Woods, the Ministry of Agriculture and the pedigree cattle breeders shared ‘deep-seated, … not even consciously recognized’ convictions that slaughter represented a ‘virile symbol’, a ‘moralizing’ and ‘educational force’, making Britain a strong nation facing the virus enemy (often imported from abroad) while nations allowing germs to spread on their soil (like France) were seen as ‘disorganized, inefficient, ignorant and immoral’.Footnote 30 There, the preservation of pedigree cattle through these sacrifices also referred to imaginaries related to the preservation of British aristocratic lineages and the prestige associated with them.Footnote 31
Things changed when the hostilities of the Second World War broke out. A few months before, in 1938, a German research team led by Otto Waldmann had managed to develop an effective vaccine against FMD that could be produced on a large scale. Within the war context, the British Ministry of Defence and Her Majesty’s Cabinet feared that Nazi Germany, protected by the vaccine, would use FMD viruses as biological weapons.Footnote 32 Brian Balmer has demonstrated how livestock diseases (such as anthrax or rinderpest) were seriously considered as effective military weapons, as they could cripple the economy of the enemy and cause food shortages.Footnote 33 In that context, mastering the production of a vaccine against FMD seemed strategically essential. The Cabinet and the Ministry of War secretly decided to induce research in that direction, allocating massive grants to the Pirbright station. First, new buildings were set up to enable virus culture in large farm animals; then a permanent team was hired, with Ian Galloway, a veterinarian trained at the National Veterinary School of Alfort (ENVA) and at the Pasteur Institute in France, as director. The aim of the team was not only to study the value of existing vaccines but also to achieve the production of safe, reliable vaccines in Britain. Indeed, the Waldmann vaccine was a live attenuated vaccine. If the mitigation process were not carried out correctly, the vaccine could give rise to new outbreaks, especially since British bovines were particularly susceptible to FMD, as they had not been previously exposed. Similarly, the use of a vaccine with limited effectiveness could have allowed the virus to circulate in the territory, i.e. within animal bodies, and both members of the government and pedigree cattle breeders would not accept it. The experiential efficacy of a vaccine would thus have been a vaccine able to block the circulation of germs within the animal body, as imaginaries about the preservation of pedigree cattle met imaginaries about germ intruders within bodies.
In order to produce usable vaccines, the research team chose to commit to the development of standardised protocols for evaluating their innocuousness and efficacy. As Jean-Paul Gaudillière pointed out with regard to the preparation of hormonal products at Bayer and Schering , this standardisation was not only aimed at evaluating the effectiveness of the products, but it was also intended to monitor and control all the stages of the production process,Footnote 34 which required a huge effort to stabilise the experimental conditions by studying all the parameters. Experiments were therefore carried out to determine which were the most optimal experimental parameters to approximate the natural conditions of infection: the length of exposure; the conditions of exposure (e.g. size of boxes, number of infected bovines, symptoms of contaminated animals); and the atmospheric and ventilation conditions of the premises.Footnote 35 At the same time, the veterinarian William Macgregor Henderson developed an intradermal tongue test for titrating the virus on living bovines.Footnote 36 From there, the team established a standard method for quantitatively estimating the infectiousness of viruses.Footnote 37 The team also worked on the typification and culture of new FMD virus strains (named SAT 1, SAT 2, SAT 3, and Asia 1) and developed the concept of subtype to explain why bovines were only partially immunised against the viruses. All these results contributed to rewarding the Station with the status of the world centre of expertise on FMD viruses. On that basis, the research team sought to determine the best vaccination protocols; all the various parameters were examined in eighty experiments mobilising more than five thousand bovines.Footnote 38 With the Cold War, the Ministry of Defence became even more concerned with achieving the production of effective vaccines in Britain. The United States and Canada shared this concern, while the Soviet Union replaced Nazi Germany as the chief threat.Footnote 39 In a three-way alliance similar to the early Manhattan Project or to the production of antibiotics in time of war, Britain intensified its vaccine research on behalf of the three countries that did not have a laboratory offering satisfactory conditions of isolation, and invested more money into research for FMD vaccines. As Abigail Woods explains, the Department of Defence also saw this engagement as an opportunity to improve its relations with the United States.Footnote 40 The objective was not only to protect oneself from a potential enemy attack but also to produce biological weapons in case of a hot conflict.Footnote 41 This research had, of course, to remain confidential, and given the importance of the military stakes, it was placed under the supervision of a military officer who had to report regularly to the Ministry of Defence, whilst notifying the Ministry of Agriculture as little as possible.Footnote 42 In addition, work to isolate the station was undertaken. Cumulatively, the work on the buildings and tensions due to military supervision led to significant delays in research, so that in the early 1950s, it was still impossible to produce vaccines on a medium scale,Footnote 43 although many advances were made in in vitro culture and in local production.Footnote 44 But as the military goal remained secret, so were the delays and difficulties.
During the winter of 1951–1952, a very serious FMD outbreak occurred. As it affected a large number of farms, strong protest movements emerged against the massive slaughter of herds. Through a press campaign, some veterinarians, small breeders, and journalists called for a more ‘modern, scientific and humane’ process: the application of vaccination, as it was carried out on the continent.Footnote 45 However, the Ministry of Agriculture could not comply with this request for several reasons. First, Pirbright was not able to supply the vaccines, and the Ministry could not admit it because of the colossal sums invested. Second, vaccines would have had to be imported, and the Treasury was struggling to limit imports as the country was going through a financial crisis. Third, it would have been necessary to give up, at least temporarily, the pedigree cattle trade that, in addition to making the country proud, brought in many currencies, essential to the status of the country:
We must balance our external accounts. If we do not, the £ will collapse. People in this country don’t realise what this means because for centuries we have had a stable currency. (…) The collapse of the £ would mean widespread and hopeless unemployment because we should not be able to afford to buy the raw materials to keep our factories going. It would mean that people’s savings would melt away; and, finally, it would mean the disappearance of Great Britain as a power in the world.Footnote 46
The preservation of the pedigree cattle trade seemed deeply linked with the power and prestige of the country, values that could not be negotiated. In response to the press campaign, the Ministry of Agriculture, supported by the National Farmers’ Union – on behalf of pedigree cattle breeders – insisted on the many practical limits of vaccines:Footnote 47 1) the duration of the immunity conferred by the vaccines was short (six months); 2) it was impossible to ensure that a vaccinated animal would not become a ‘healthy carrier’: an asymptomatic animal capable of contaminating unvaccinated cattle; and 3) the vaccines were ineffective on pigs. If the virus were to circulate in the territory, pigs would then become major propagators. The population had to trust its FMD experts; vaccines were too risky for the country.Footnote 48 Despite this communication campaign, many small farmers kept asking for vaccines.Footnote 49 To answer these public pressures, an expert committee, the Gowers’ Committee – named after its chairman, Sir Ernest Gowers – was set up to compare the value of the sanitary and medical policies against FMD. After hearing a large number of experts (British and continental) and representatives of agricultural and veterinary associations, the Committee concluded that the vaccines were not suitable for controlling the disease in Britain since they were not fully efficacious. This assertion was not a pretext to justify the maintenance of stamping out at all costs, as the Committee considered vaccination possible in the event of an epidemic so severe that slaughter would be impossible.Footnote 50 It was rooted in the experience of veterinarians and large-scale farmers who adhered to this cartographic representation of epidemics and who considered virus eradication possible through a drastic control of imports, thanks to British force and will. But it was also based on the expertise developed at Pirbright, where drastic standards of safety and efficacy tests had been settled, contrary, according to Ian Galloway, to all other producing centres.Footnote 51 Only the tests carried out at Pirbright offered, according to him, the maximum guarantee within the known experimental limits: the safety tests for a batch of vaccine were based on Henderson’s method, on eight bovines. The first six were inoculated with four dilutions of the vaccine placed at five different points on the tongue, so that twenty points could be observed for each animal, therefore amounting to 120 points in total. The dilutions were calculated to embrace the viral infecting dose for fifty per cent of the animals (i.e. the appearance of ulcerations on half of the inoculated points). If the virus had been properly attenuated, no ulcerations should be visible. To ensure that the bovines used in this safety test were susceptible to FMD, they were then tested with two other control animals with a virus of a different type than the one used to manufacture the vaccine. All should develop FMD symptoms. According to Galloway, however drastic, these tests could not guarantee that the vaccines were perfectly safe:
‘It must be remembered that while it is true that the greater the number of samples tested may be extended, a negative, in this case the absence of virus, can never be proved. Even 100 observations in our intradermal tongue test may not detect an infective level of 3% of samples positive. This may be taken as a reasonable calculated risk to take for all practical purposes for there must be a practical limit.’Footnote 52
The conditions of clinical experimentations, which involved a limited number of animals, made the statistical value of one hundred per cent impossible to reach: as a consequence, some vaccines may therefore not have been properly attenuated but have passed safety tests. Galloway considered that continental vaccines, designed to protect animals against two or three types of FMD virus and whose safety tests did not meet Pirbright conditions, were two or three times more dangerous than monovalent vaccines. It was therefore impossible to import these vaccines without taking the risk of creating new outbreaks. Galloway was even more negative with efficacy tests. To meet British standards, three groups of eight bovines had to be inoculated with one batch of vaccine; the first group would receive the pure vaccine, the second the vaccine diluted to a quarter, the third group the vaccine diluted to a sixteenth (i.e. inoculated with 1/16th of the recommended protective dose). Once the immunity was acquired, i.e. three weeks later, the bovines were tested by an injection of virus, at a dose 10,000 times higher than that contaminating fifty per cent of the animals (a fifty per cent infective dose). A week later, all animals had to be free of any lesions on the feet, even those that had received the diluted vaccine. The severity of these tests was justified by the risk of introducing a control method that could potentially worsen the epizootic. But as a consequence, it eliminated all existing vaccines, not only those produced on the continent, but also those from Pirbright!
However hard these conclusions, research continued at Pirbright, not only to maintain the knowledge and know-how related to virus culture, titration, and vaccine production, but also because British vaccines against FMD were soon intended for foreign countries, which wanted to export cattle to the UK.Footnote 53 This remote control was perfectly in line with the British frame of thought, consisting in protecting ‘the island’ from contagion while preserving its trade by establishing quarantines and remote inspections of its borders, as it had been the case, for example, in Gibraltar or Malta in the nineteenth century, with cholera.Footnote 54 Dangerous importations of meat or animals carrying FMD would be controlled within their own countries, thanks to British vaccines whose efficacy was guaranteed by Pirbright standards:
Anything done at Pirbright to develop vaccines and to facilitate their use abroad cannot fail to help us in the long run. Our nearness to the continent is one very good reason for our interest in the African and Asian types of virus. They could readily spread to us if introduced to the Continent because the vaccines currently in use there afford no protection against them.Footnote 55
Used ‘abroad’, the vaccines were thus reliable tools, making it possible to keep British herds free of germs, whether attenuated (in vaccines) or not.
In France: industrial safety and efficacy tests ‘adapted to reality’Footnote 56
In France, the Ministry of Agriculture, like the veterinary elites such as Gaston Ramon or Emmanuel Leclainche who were involved in the management of epizootics on an international scale within the Office International des Epizooties (OIE) – formerly the World Organization for Animal Health – and in close contact with British experts, shared their opinion on the merits of slaughter against FMD.Footnote 57
Emmanuel Leclainche studied veterinary medicine at the National Veterinary School of Alfort (ENVA) and was trained as a bacteriologist at the Pasteur Institute of Paris. Professor at the ENVA and then at the National Veterinary School of Toulouse, he co-founded the Toulouse Institute of Serum Therapy (IST), a pharmaceutical laboratory specialised in the production of vaccines and serums for physicians and veterinarians. He was also a member of the Advisory Committee on Epizootics (Comité Consultatif des Epizooties, CCEFootnote 58) and the initiator and first director of the OIE. Gaston Ramon studied veterinary medicine at the ENVA and became an eminent member of the Pasteur Institute. He discovered the anatoxins (against diphtheria and tetanus), flocculation reactions, and how to boost immunity as a reaction to vaccines through the use of adjuvants. He was a member of the French Academy of Medicine and the French Academy of Science. He was also a member of the OIE and its second director. Both of them were renowned microbiologists and possessed a certain expertise on FMD: the IST sold vaccines against the disease, and Ramon had investigated FMD vaccination at the Pasteur Institute (with no much support from the Direction nor much success)Footnote 59. But despite their experience with vaccines, both veterinarians were convinced that slaughter was the best way to eliminate the disease, and they made numerous communications at the CCE, in the Senate, or at the Assembly in that direction. Despite the authority of these two veterinarians, the French government refused to implement the sanitary policy not only because of the amount of compensation required, but also because most farm work relied on bovines (mechanisation was still limited and horses were expensive and scarceFootnote 60). A mass culling of farm animals would have jeopardised French agriculture in general. In addition, slaughter had no support from veterinarians. According to their professional union, veterinarians were too attached to their liberal status to mimic British veterinarians and get involved with a sanitary policy that would provide them with a status close to that of civil servants.Footnote 61 As a result, the isolation and disinfection policy was poorly implemented, so that FMD viruses circulated freely in the country. Animal owners were used to having contaminated animals recovering from FMD, and they generally considered this disease a mild condition (apart from a few exceptional episodes). They treated their sick animals with various curative procedures inherited from their families or purchased from healers, peddlers or pharmacies. These products were most often made from antiseptics such as vinegar, thyme, or wild thyme, to avoid the most problematic complications (e.g. myocarditis, metritis, abortions, mastitis, enteritis, neonatal mortality). Farmers could also speed up the infection process to get rid of the disease more quickly, by moistening the gums of their animals with contaminated saliva, a practice called aphtisation. Finally, the sanitary veterinarians (working for French departments) and certain veterinary pharmaceutical manufacturers had developed a process of passive immunisation, called serum therapy, which consisted of inoculating the animals with the serum of convalescent animals. Despite its short-term protection, this technique was very popular amongst farmers and liberal veterinarians in the 1930s.Footnote 62 The immunisation processes – whether active or passive – referred to the bacteriological culture characteristic of a form of French expertise, supported by the Pasteur Institute and which was shared by veterinarians, either due to their trainingFootnote 63 or to their association with Pastorians in various public and private enterprises.Footnote 64 But it also fitted very well with a pre-bacteriological style of thought Footnote 65 still present in France, a sort of holistic representation of illness that integrated germs (viruses and bacteria) together with humans and animals within their environment.Footnote 66 Getting rid of infections could only be done, therefore, according to the Pastorians, by replacing pathogenic germs with harmless germs (killed or attenuated) within individuals, through vaccines.Footnote 67 This replacement ideology has to be understood in the same manner that the French nineteenth-century botanists Augustin and Alphonse De Candolle spoke of ‘nature’s war’ or the ‘plant war’ to describe plants’ competition to occupy the same environment. In the case of France, this ecological representation of the germs’ relations within an organism met the possibility to experiment with alternative means of control of animal diseases, and the farmers’ aspirations to preserve animal bodies as much as possible. Various bacteriological laboratories which marketed serums and vaccines, such as the Pasteur Institutes in Lille and Paris, the Roger Bellon Laboratories, the Laboratory of serum therapy of L’Aigle, or the Bacteriological Institute of Tours, to name but a few, therefore encountered many successes with farmers. Amongst these laboratories, the Mérieux Institute (IM) in Lyon, sold so much serum against FMD that it decided after the war to join forces with the Toulouse Institute of Serum Therapy (IST) to produce on a large scale a vaccine against FMD. The process developed by Otto Waldmann supposed an in vivo culture of the virus in bovines. Once the characteristic ulcers reached a peak, the animals were killed, their lingual epithelium removed and ground, and then the virus was mitigated. The IM and the IST, therefore, founded the French Institute for Foot-and-Mouth Disease (IFFA) in Lyon, a vaccine production centre large enough to serve as a slaughterhouse. The IFFA sold trivalent vaccines (to the State, the IM, and the IST), made by the successive preparation of monovalent vaccines – each active against one type of virus – that were combined at the time of injection. The culture of each viral strain had to be done one after the other to avoid any contamination.Footnote 68 The yields were very low, as the slaughter of a 500-kg bovine led to the production of thirty to forty grams of virus, enough to produce fifty doses of vaccine.Footnote 69 This forced the IFFA to slaughter an ever-increasing number of animals to achieve industrial production, which was not without problems, in particular, concerning the sale of meat.Footnote 70 Once produced, vaccine batches underwent three safety tests:Footnote 71 a physico-chemical test (the pH of the vaccine had to be close to nine); a bacteriological test to detect potential bacterial impurities; and a biological test on three bovines. After being inoculated with the vaccine, these animals were placed under observation for fifteen days, in isolation. A herdsman – not a veterinarian – was responsible for observing any visible symptoms that would lead to the rejection of the batch of vaccine. Two weeks after these tests, the efficacy test was performed on the three animals plus a fourth, as a control (all of them were infected with a virulent FMD virus). The control animal had to show symptoms of generalised FMD. Without the possibility of titrating the virus or the vaccine, the only method of analysing the efficacy of the vaccine was therefore the clinical comparison of the biological modifications caused by the inoculation of the virus in the animals. To this end, a veterinarian performed a clinical examination: if no reaction was observed or, failing that, if only a local ulcer appeared, the vaccine was considered ‘good’. In all other cases – multiple mouth ulcers, ‘thermal reaction’, generalised foot lesions – the vaccine was rejected.Footnote 72 The results of these tests mobilised a register of clinical evidence that depended considerably on the susceptibility of bovines to the FMD virus. According to Maurice Fedida and his collaborators, employees at the Laboratory of Animal Virology – where vaccines manufactured by the IFFA were controlled in the 1970s – only fifty per cent of the vaccines designed in the 1950s were considered ‘good’ at the end of these efficacy tests.Footnote 73 Even if this testimony probably aimed to reinforce, by contrast, the quality of current vaccines, it nevertheless allows a glimpse of what the conditions of production and evaluation of vaccines were like in the early 1950s.
Between 1951 and 1953, a very strong epizootic broke out, and isolation and the suspension of fairs and markets, as well as vaccination around outbreaks (‘ring vaccination’ operations), were insufficient to control contagion.Footnote 74 As a result, individual vaccine demands skyrocketed as farmers attempted to save animals that had not yet been affected. Vaccines were sent to veterinarians in an emergency, often too late.Footnote 75 Worse, new variants appeared, rendering obsolete the vaccines that had first succeeded in protecting some animals.Footnote 76 All agricultural work seemed paralysed, and food shortages increased. Faced with the urgency of the situation, the Minister of Agriculture opposed the recommendations of the French Academy of ScienceFootnote 77 and restored the circulation of bovines on the territory to avoid the financial crisis caused by the cessation of all commercial activities.Footnote 78 The Ministry of Agriculture mobilised the army to help bring in the crops, as too many animals were unable to work,Footnote 79 and ordered them to carry out trials of any possible curative treatments. This order reveals a significant cleavage: on the one hand, liberal veterinarians and breeders, depending on animal traction that looked for individual protection of bovines through any means of preservation. On the other hand, the veterinary scientific elites members of prestigious academies demanded the culling of herds as the only possible means of control.Footnote 80
From the farmers’ view, slaughter was even more unpopular than isolation and the suspension of fairs and markets, because of their experience of the disease as a mild disease. A large majority of them used bovines to work the fields. The recovery of their animals was important as bovines were not easily replaced. Training them to work under the yoke took time and presupposed that the farmer already had an animal accustomed to working in this way, which ‘guided’ the young cattle in learning farm work. The slaughter of two animals was often simply impossible for these farms. For dairy farmers, this was more or less the same, as compensation would not cover the loss of their animals. Faced with such a reality where the culling of animals could jeopardise many farms, farmers were in favour of vaccines because they deemed any animal spared or suffering from mild symptoms of the disease a prophylactic success.Footnote 81 Liberal veterinarians shared the same point of view, as their motto, ‘first, do no harm’, involved caring for animals – or at least sparing them as much as possible.Footnote 82 Their representatives in Parliament asked the government to promote vaccine production to meet national needs.Footnote 83 But for the veterinary elites, as for the Ministry of Agriculture, which doubted the real capacity of vaccines to control an epizootic, it was unthinkable to endorse this medical policy as long as vaccines lacked a real assessment of their efficacy, carried out by an independent laboratory.Footnote 84 Unfortunately, the only public veterinary laboratory (at the ENVA) lacked the means to fulfil this role, and the Ministry of Agriculture did not plan to modify this situation, showing that FMD vaccines were not a national priority.
Aligning experimental and experiential efficacies, generalising vaccination
The IM and the IST sought to respond to these objections while forging new commercial partnerships with liberal veterinarians. They reinvested the profits from the sales of vaccines from the 1952 epizootic in new in vitro culture systemsFootnote 85 and funded the creation of an ‘independent’ laboratory to control all the foot-and-mouth disease vaccines produced in France, including the IFFA’s. This laboratory – named the Veterinary Research Laboratory (VRL)Footnote 86 – operated thanks to the IFFA’s payment of two per cent of its annual profits.Footnote 87 In the 1950s, the international reputation of Pirbright researchers was very important, and their standards legitimising the rejection of vaccines in the UK were, of course, known to the director of the VRL, the veterinarian Félix Lucam, and to his team. However, these standards were incompatible with the industrial production developed by the IFFA and the other laboratories manufacturing FMD vaccines because their safety and efficacy tests required the use of too many animals. The VRL then adopted other standards, ‘rapid, easy to learn and cost-effective’, thus compatible with industrial production.Footnote 88 Nevertheless, to legitimate this choice, the French researchers referred to the statistical tables published by Henderson, which showed how even the most drastic safety controls left a chance of missing a badly attenuated batch. But if this assertion based on statistical evidence justified the vaccine’s rejection in Britain, in France, on the contrary, it made acceptable the existence of a margin of uncertainty for safety and efficacy tests. Two types of tests were therefore performed. The more rigorous of the two involved a larger number of animals but was applied less often, ‘once in 10 or 15 vaccine preparations ’:Footnote 89 each valence was tested on eight bovines (plus four as controls that should develop generalised FMD). The vaccinated animals were clinically examined and the results converted into statistical values, in an operation masking the clinical variabilities.Footnote 90 The vaccines were then classified according to the lesions observed, from ‘authorised’ – if no lesion was observed – or ‘acceptable’ – if only one animal was clinically affected – to ‘rejected’ – if two or more animals were affected. ‘In the worst case’, Félix Lucam and his team estimated that the vaccines were at least eighty per cent effective, which they considered ‘quite good’, consequently accepting in practice a twenty per cent failure rate.Footnote 91 Despite its approximate nature, this first type of control was still considered too costly in terms of bovines to control the industrial production of IFFA. Most of the time, a second type of control was performed on two to four bovines (plus one or two as controls). If all of them, or three out of four, or two out of three animals resisted the virulent test, the vaccine was accepted. If one out of two, or two out of three or four bovines did not resist, the batch of vaccine was not destroyed, but new tests were performed. It was only in this second case, when ‘the same results or worse’ were obtained, that the vaccine was rejected.Footnote 92 Otherwise, the vaccine was considered sufficiently efficacious and marketed. Of course, the researchers at the VRL knew that these tests involved too few animals to be statistically significant. Nevertheless, they justified their practices by adding up all the tests carried out as if they had tested only one batch, in defiance of all statistical rules:
‘Results interpretation. Safety control: out of 271 batches of vaccine checked, none proved to be virulent. We noted above the fairly large degree of uncertainty that surrounds each control. But now consider all of those 271 vaccine lots. Since these are similar materials, we can very well imagine them gathered in a single batch. For this single batch, we used 542 bovines which each received 20 samples. We have therefore, in total, tested 542 x 20 = 10,840 samples from this single batch. Since none of these samples proved to be virulent, we can apply the formula (1 – p)n = P, which made it possible to establish Table I and in which n is the number of samples tested and which all were shown to be avirulent, p is the maximum possible percentage of those which, nevertheless, in the entire batch of vaccine could be virulent, and P the probability of missing the latter.Footnote 93 We find, under these conditions, that in our supposed batch, there could only be at most 0.003% virulent samples with a one in twenty chance of not having been able to detect them. […] It follows that if, at the time of each test, the interpretation of the result must be subject to reserve, the accumulation of tests makes it possible, a posteriori, to get a fairly exact idea of the average value of an important production.’Footnote 94
According to these researchers, this series of approximations was justified by the industrial nature of the production and the controls to be carried out. Nevertheless, it helped to disseminate a numerical value that made the uncertainties surrounding the controls invisible, leading them to demonstrate that the vaccines produced in France were completely (99.997%) safe, and their efficacy foolproof. It was not a question of a lack of knowledge of the statistics (as the reference to Henderson’s statistical analysis is explicit). The logic underlying the implementation of these tests has to do with what Nathalie Jas describes as a principle of ‘adaptation to reality’ when she analyses the implementation of the acceptable daily intake tool by the WHO/FAO Committee on Food Additives in the 1950s–1960s.Footnote 95 She shows how, knowing well that there are no threshold doses below which carcinogens would not endanger public health, the members of this committee legitimised the addition of carcinogens to food because their use was considered essential for the community as a whole. Similarly, the VRL knew perfectly well that the standards used in France were not statistically capable of guaranteeing one hundred per cent safety and efficacy. But their results approached the experiential efficacy of many farmers and veterinarians, who needed to see their animals survive FMD and avoid the slaughter requested by the French government. In this social and cultural context, a one hundred per cent effective vaccine was a vaccine capable of justifying the non-application of culling measures, even if this meant that some animals were poorly protected.
However, this was not true in all regions of France. In Finistère, the breeders who traded with Britain had decided to slaughter as soon as the first outbreaks appeared, to isolate contaminated farms and to control all the movements of the population within the department.Footnote 96 This experience of eradication relied entirely on farmers, with the help of the local authorities, who compensated for the losses endured. Everyone agreed that this experience had been a frank success. For the Ministry of Agriculture, it was the confirmation that the sanitary policy was the solution. However, it was impossible to make it mandatory as the government was financially involved in very costly operations of the eradication of bovine tuberculosis.Footnote 97 To protect the FMD vaccines market, the Director of the Mérieux Institute, Charles Mérieux, contacted liberal veterinarians. They indeed had no interest in seeing the sanitary policy against FMD generalised to the whole country. First, their involvement was not required either for the diagnostic or the application of the sanitary measures (as it was shown by the Finistère experiment). Second, they would lose a great number of their clients, as the vaccines, sera, or treatments against the disease would become useless or forbidden. Charles Mérieux was a close relative of Marcel Quentin, the President of the National Veterinary Union. He knew that liberal practitioners were discussing with farmers, breeders’ associations, and unions (such as the Confédération Nationale de l’Elevage, the Fédération Nationale des Groupements de Défense Sanitaire, or the Union Nationale des Coopératives Laitières) about the possibility of setting up vaccinations on a large scale.Footnote 98 Charles Mérieux developed an alliance with the National Veterinary Union to establish partnerships with breeders opposed to slaughter. This led to vaccination contracts linking the industries producing vaccines to veterinarians and farmers for a period of five years. The settlement of such a partnership relied entirely on the relations liberal veterinarians had with their clients or with agricultural cooperatives. The cost of the prophylaxis would rest on farmers, but with negotiated prices for vaccines and veterinary inoculation, it would be cheaper than the consequences of eradication. Besides, the liberal sensibility of veterinarians (their ability to set the price of their services) was spared: their honoraria varied according to their local conditions of exercise (the terrain, the animal dispersion, and the number of animals on the farm).Footnote 99 Vaccination operations and their annual boosters guaranteed the presence of veterinarians in the herds, the ‘syringe’ appearing as a powerful form of technical mediationFootnote 100 between vets and farmers. These ‘vaccination contracts’ linking industrialists, farmers, and liberal veterinarians were not just a simple commercial operation for industrialists. They also contributed to aligning experimental efficacy with experiential efficacy, with the continuous improvement of vaccines. At the end, the government found itself with the setup of a medical policy with no civil servants or public money involved. Farmers interested in these measures saved themselves the losses caused by the disease, and veterinarians gained a regular income and the loyalty of their customers. Manufacturers, supported by regular purchases of vaccines and having access to large-scale safety and efficacy tests carried out through the use of vaccines in practice, increased and improved their production capacities. Eventually, the administrative arrangements were aligned with the experience of farmers who had freely chosen to subscribe to vaccination contracts: in 1961, vaccination became mandatory in France, slaughter being kept as a secondary tool, in case of an outbreak of FMD.Footnote 101
Conclusion
As Joan Fujimura pointed out, ‘technology alone cannot make problems doable’. She argues that ‘doability is conceptualised as the alignment of several levels of work organization’ – experiments, the laboratory, and the social world.Footnote 102 The articulation of tasks between these three levels of work organisation makes problems doable. Following Fujimura, this paper has shown the tinkering of the VRL scientists in order to align these three levels and produce vaccines that are one hundred per cent efficient. Indeed, although falling under the same methods of preparation of FMD vaccines, and therefore possessing an experimental efficacy that can be considered similar in France and Britain, the experts on their efficacy in the two countries reached opposite conclusions. In Britain, techniques for the standardisation of preparation, inoculation, and evaluation of vaccines, such as the register of statistical evidence mobilised, made it possible to reject all vaccines, including those produced in the country, because none of them could statistically achieve one hundred per cent efficacy. The degree of uncertainty, even small statistically, reduced the possibility of using the vaccines in practice, as it objectivised the risk of immune escape and made possible an estimation of the economic losses and of the losses in power and prestige that Britain would endure. But it was also impossible for vaccines to meet the experiential efficacy required by British members of the Government and pedigree cattle breeders, with no germs inside the animal body. Indeed, vaccinated cattle appeared as healthy carriers, potentially dangerous for unvaccinated animals. Nathalie Jas showed how the construction of risk regulation tools – here, the statistical evaluation of the efficacy of vaccines according to the Henderson protocol – goes through a series of accommodations which ultimately make public health a relative issue, dependent on the need to adapt to reality.Footnote 103 The British reality referred not only to an economic-political logic, aimed at preserving very lucrative export markets that benefitted the reputation of the country, but also to the imaginaries associating hygiene, body control, culture, and power of a nation, so that to be effective, a vaccine would have had to eradicate the germ responsible for the disease, which the FMD vaccines could not do. As a consequence of that experiential effectiveness, the experimental effectiveness was statistically impossible to reach, legitimating the practical reluctance to use vaccines. On the other hand, in France, the experimental efficacy of vaccines was articulated with the experiential efficacy of breeders who were looking for a way not to suffer from the disease’s effects. Perfectly adapted to the reality of breeders in the 1950s and 1960s, French vaccines imposed themselves even though they only imperfectly protected the entire national herd. As Loconto and Demortain argued, standards do not necessarily come from the top. They also emerge from diversity, through the appropriation of the possible applications of these standards, so that uses determine effectiveness.Footnote 104 French vaccines, therefore, being more and more used, have gradually improved their efficacy, both through the improvement of technical processes and through the coverage effect of a greater number of vaccinated animals. This double improvement contributed to the success of vaccination campaigns up to the modification of animal disease control standards in France, through the establishment of a compulsory medical policy against FMD in 1961. Thus, the example of FMD vaccination in France clearly demonstrates that public policies can also emerge from local and private initiatives, without any intervention by the state.
Open access
