Therapeutic Vaccines and Medical Innovation

To grasp the innovative character of therapeutic vaccines, we must keep in mind that the ways they were researched, produced, tested, and handled by health administrators were novel in their own day. Looking at French and German history, which is mostly the focus of this issue, the diphtheria antitoxin was part and parcel of important changes in the regulation, research, and production of medicines.

In relation to testing and regulation, medicines were traditionally listed in a register controlled by health officials, the pharmacopeia, and had to be tested for proper preparation and non-toxicity by pharmacists (see Hess in this issue for more detail), while the evaluation of their efficacy was left to the clinical experience of the applying physician (Tröhler Reference Tröhler and Köbberling1993). With the diphtheria antitoxin it became common practice to carry out quantitative and experimental testing of therapeutic efficacy in an animal model, where a vaccine's potency could be measured against so-called test-poisons. As a result the serum's activity was measured quantitatively and ultimately standardized. This measurement could be seen as a forerunner of what came to be called the therapeutic index in the twentieth century since it was based on the establishment of minimal and lethal doses in animal models – the therapeutic index being essentially a proportion of these two. Measurement was also basic for standardizing the essential parts of the process such as the test-poisons or indeed the vaccines themselves. Beyond its immediate historical context, this evaluation or Wertbestimmung should be seen as a standardization of a medicine of biological origin which – as will be argued in the articles collected in this issue – became a model not only for the invention of more vaccines in subsequent years, but also for a variety of other biologicals in the twentieth century.

Changes in the development and production of medicines related to the introduction of therapeutic vaccines were no less radical: Traditionally, drugs had been produced in a workshop setting surrounded by an apothecary. Such a process had been controlled by recipes written down in the pharmacopeia from which the single pharmacist would prepare medicine (Huhle-Kreutzer Reference Huhle-Kreutzer1989). Vaccines as such, of course, were not a late nineteenth-century innovation; in fact the smallpox vaccine had already been in use for almost a century. Yet this medicine differed from therapeutic vaccines in that it was not an industrially produced, standardized medicine. The practical application of the many stages of development of the smallpox vaccine had been stabilized before the experimentalization of medicine and the industrialization of pharmaceutical production.Footnote ¹ This industrialization – starting in the mid-nineteenth century and not limited to therapeutic vaccines – fundamentally changed the traditional set-up (Fleischer Reference Fleischer1984; Seckelmann Reference Seckelmann2006; Wimmer Reference Wimmer1994). What became visible to the wider public in the second half of the nineteenth century with the arrival of new industrial producers like the Hoechst company and novel classes of products such as the tablets were structural changes in the ways in which medicines were invented. Research on new drugs was moved to the laboratory; production of pharmaceuticals became a matter of industrial technology and marketing; property rights were secured by patents rather than by keeping the composition of a medicine secret; apothecaries became retailers rather than producers of medicine; finally in relation to defining individual medicines, the role of the recipe was gradually taken over by chemical analysis (Müller-Jahnke and Friedrich Reference Müller-Jahnke and Friedrich1996; Sneader Reference Sneader2005) and in the case of biological medicines by an evaluation of their efficacy. What evolved was a system of industrially divided labor, and standards played a crucial role in the system's integration. For instance the use of statistics in evaluating medicines required first of all a standardized product in order to compare its performance in a series of applications. At the same time this standardization was, of course, an essential feature of large-scale industrial production itself.

Standards such as Wertbestimmung served as an interface of communication between those who engaged in carrying out research, testing, producing, or applying therapeutic vaccines. This has to be seen as a local variation of a trend of standardization in societies and industries of the age – ranging from the standard measure of the meter, to railway-timetables that enforced the standardization of space and time, all the way to technical norms of industrial production (Cahan [1977] Reference Cahan1992; Schievelbusch [1977] Reference Schievelbusch1989; Wise Reference Wise1995). In relation to pharmacology and medicine this meant that testing a novel medicine could no longer be arranged on a personal basis between apothecary, physician, and patient. Complex biological medicines like the diphtheria antitoxin were developed in laboratories and produced in factories. They did not easily fit into the existing pharmacopoeias as regular drugs, nor could they be safely handled as so-called specialities of unclear composition like Galenic medicines. Instead, testing took a different path. What developed starting with the diphtheria antitoxin was a quantifying, standardizing measurement of efficacy which facilitated, in addition to its chemical analysis, its safe use.

Finally, the invention of therapeutic vaccines was bound up with changes in medicine itself, changes that are commonly summarized under the heading of the laboratory revolution in medicine (Cunningham and Williams Reference Cunningham and Williams1992). While this process had started with experimental physiology in the mid-nineteenth century, in the eyes of contemporaries, the medical sciences acquired clinical relevance with the arrival of medical bacteriology. As therapeutic vaccines intended to neutralize a pathogen's deadly course, the toxins could be assessed as an immunological consequence following the novel face that medical bacteriology had applied to infectious diseases in the 1880s. A familiar epidemic disease now had a modern, plausible, and concrete face as a bacterial infection. As a consequence, the cure of infectious diseases via the shortcut of controlling what had been identified as their necessary causes appeared promising. As Robert Koch put it in 1882 in relation to tuberculosis, mankind no longer had to “deal with an undefined something, but with a concrete parasite, whose living conditions are uncovered at large” (Koch [1882] Reference Koch and Schwalbe1912, 444; cf. also Faber Reference Faber1930, 110–111). As another author observed about a century later: “Find the cause, then find the cure: this is the mission of biomedicine in a nutshell” (Epstein Reference Epstein1996, 31). This mantra of modern biomedicine was phrased in the final third of the nineteenth century (Cunningham Reference Cunningham, Cunningham and Williams1992) and the therapeutic promises following from the identification of pathogenic microbes are an essential part of it (Schlich Reference Schlich, Gradmann and Schlich1999).

However, while it is certainly appropriate to assess the businesses of therapeutic vaccines and their evaluation as an integrated, dynamic, and innovative system, one should not overlook two important features of the historical situation that fundamentally limited this integration. First, we have to keep in mind that this history took place in a field of tension and of uneven temporality with contemporary medicine and pharmacology. The novel ways in which physiology or bacteriology envisioned the human body and its pathologies did not easily translate into practice. Innovations such as the thermometer to measure fever or the bacteriological diagnosis of infectious diseases were competing with established diagnostic techniques and were moreover reliant on a context in which their use was considered plausible by contemporaries (Bynum Reference Bynum1994; Hess Reference Hess, Jorland, Opinel and Weisz2006). The new medical sciences were not fully compatible with clinical medicine. It can be argued that clinical medicine from the late nineteenth century on developed simultanueously along two different and sometimes conflicting lines: While developments towards a “scientific” clinical medicine were visible in the creation of internal medicine as a field (Marks Reference Marks1997; see also Warner Reference Warner1995), tensions between the basic sciences and clinical practice still became a distinctive feature of medicine (Amsterdamska Reference Amsterdamska1987; Gradmann Reference Gradmann2005, 171–251).

Secondly, those who were interested in the development, production, or evaluation of therapeutic vaccines by no means constituted a homogenous group. Such medicines were the product of a variety of medical, industrial, and administrative actors, each pursuing interconnected, yet differing interests. Some of these actors, such as health administrators or industrialists, it should be added, have found too little attention in historiography (Gaudillière and Löwy Reference Gaudilliere and Löwy1998). Thus, though the essays in this issue argue that with therapeutic vaccines the testing of medicine underwent fundamental changes and that thereby a model for twentieth-century medicine was created, the very term “model” does not, however, imply any monolithic structure in terms of what followed and, in fact, not even with respect to the “original” frameworks. Handling the diphtheria antitoxin involved the participation of a variety of actors and – as will be explained in some more detail below – was heavily influenced by local cultures from the outset.

If we employ the picture outlined above for a more systematic argument, therapeutic vaccines in connection with the practices of their evaluation are best understood as a dynamic communicative tool for those who were dealing with therapeutive vaccines. In this sense, the procedure can reasonably be accessed as a boundary object (Star and Griesemer Reference Star and Griesemer1989). A medical researcher trying to define a lethal dose in a guinea pig, a clinician attempting to cure a child suffering from diphtheria, and an industrialist trying to produce and sell a vaccine, even though they were all pursuing rather different interests, could integrate their efforts via a procedure of measuring and quantifying a vaccine's curative efficacy. This resulted in its standardization as an industrial product. Developing and applying tests was part and parcel of communicating and mediating interests between medical researchers, clinicians, industrial producers, and health administrators. Boundary objects in this sense are loosely defined and achieve stability by means of active interchange among the various groups of actors who handle them. They can also serve to fuel and integrate dynamic historical processes, a function that is particularly valid in the case of the consequences of the evaluation procedure for the diphtheria antitoxin.

Approaching vaccines and sera as boundary objects will allow us to emphasize several interconnected features of “scientific” medicine in relation to standardization starting from about 1880. Thus, such medicines were at the same time novel objects of knowledge, industrially produced medicines, and the subject for innovative forms of control and regulation. Indeed, while the issue of standardization is implicit in all these aspects, the standardization of therapeutic vaccines in its peculiar form was dependent upon their hybrid character. On the one hand, the therapeutic vaccines were biological substances of an inherently heterogeneous nature. On the other hand, they were the product of a sophisticated process of research, development, and industrial production that could hardly be integrated into the traditional repertoire of the local pharmacist.

Paris and Berlin

One of the peculiarities of the history of the diphtheria antitoxin is that it came into being as a parallel invention in two distinct and sometimes even opposing cultures that had formed around two research institutes: the Institut Pasteur in Paris and the Institut für Infektionskrankheiten in Berlin.Footnote ² A brief glance at those two notoriously rival schools of medical microbiology can help us understand the different paths in the development and handling of the antitoxin. It also explains why and how this rivalry, which had shaped the 1880s controversies of Koch and Pasteur, took the more civilized form of a productive competition when the two groups researched the diphtheria antitoxin in subsequent years.

In France, Louis Pasteur and his school had focused intensively on investigating and producing medicines for some time and had invented some vaccines along the way. As is well known, Pasteur's ideas about the relation of man and disease went beyond explaining infections. They included ideas on the common natural history of men and microbes that was manifested in the growing and waning of epidemics brought about by an increasing and decreasing virulence of microbes (Geison Reference Geison1995; Mendelsohn Reference Mendelsohn1996). The controlled reproduction of this naturally occurring phenomenon as attenuation and increase of virulence of microbial cultures in the laboratory paved the way to successful vaccines for anthrax and rabies in the 1880s. The success of the rabies vaccine, and more precisely its revenues, literally laid the financial basis of the Institut Pasteur which had been designed to research and produce medicine from the start (Cassier Reference Cassier2005; Löwy Reference Löwy1994). In three important respects – identity of research and production, self-control of the producer, and monopoly of production – the Institut Pasteur held control of the research and production of earlier medicines, which pre-shaped the handling of the diphtheria antitoxin in France (Cassier Reference Cassier2005). However, some important features were lacking in comparison to handling therapeutic vaccines. As substances produced from microbial cultures, the first generation of Pasteurian vaccines differed from the diphtheria antitoxin, as the latter was produced from blood serum and was thus a far more complex biological substance. While the vaccines for anthrax and rabies were essentially preventive medicines, which could be perceived in the tradition of older practices such as the smallpox-vaccination, the diphtheria antitoxin as a curative medicine was immunological in its scientific basis. Its history in Paris began with Roux's and Yersin's work on toxins in 1888 (Roux and Yersin Reference Roux and Yersin1888; for an in-depth analysis, see Ulrike Klöppel's essay in the present issue). It was not only part and parcel of an elaborate experimental and industrial system; it also led the Pasteurians beyond their methods of attenuating microbial cultures into investigating the interaction of microbes and their hosts, that is into immunology (Bulloch [1938] Reference Bulloch1960, 255–283; Silverstein Reference Silverstein1989).

In that respect, comparing Pasteur's earlier vaccine with the diphtheria antitoxin is illuminating. In his contribution to this issue, Maurice Cassier reaches the conclusion that although there were attempts to standardize the anthrax vaccine it was not yet an industrial product in the sense outlined above for the diphtheria antitoxin. There was no clear distinction between the labor of research and the labor of production; both activities were held together under Louis Pasteur's personal tutelage. Property rights for the vaccines would essentially be protected by the traditional means of keeping the methods of preparation secret rather than by acquiring patents. Unsurprisingly, the diphtheria antitoxin involved a more advanced level of systematic clinical testing as compared to the rabies vaccine, where testing was very limited. Testing for the rabies vaccine had been performed in the theatrical fashion of the public treatment of a supposedly infected boy, Joseph Meister (Geison Reference Geison1995, 206–233). Finally, it can be argued that the transition from Louis Pasteur to his successor Émile Roux, who directed the research on the diphtheria antitoxin and the Institut Pasteur from the 1890s, brought with it a shift in focus. While Pasteur focused on the natural history of diseases and epidemiology, Roux focused more on the medical-clinical aspect producing medicines for individual patients (ibid., 234–256).

In a comparable way to the French story outlined above, Emil Behring's and Karl Wernicke's work on the antitoxins indicates the generational and focal shift of the German-Prussian school of medical bacteriology led by Koch.Footnote ³ Before the late 1880s this group had become famous for what has been labeled “microbe hunting” (De Kruif [1926] Reference De Kruif1927). Starting in 1881 as a bacteriological team in the Imperial Health Office, Koch and his collaborators and colleagues, including Georg Gaffky and Friedrich Löffler, identified the bacterial pathogens of important infectious diseases such as tuberculosis, cholera, and diphtheria (see Löffler Reference Löffler1884 for the identification of diphtheria). The approach of the bacteriologists in Germany differed markedly from the approach of the microbiologists in France: While the Parisians focused on the variability of microbial species and the natural history of diseases, Koch and his group examined the stability of these bacterial species and their properties (see Mendelsohn Reference Mendelsohn1996).Footnote ⁴ Pathogenicity, the ability to produce disease in an organism, was such a property, hence infection was understood as a bacterial invasion of a passive human body that somehow resembled a nutritive medium (Gradmann Reference Gradmann2005, 124–34; Mendelsohn Reference Mendelsohn1996, 73–166). Measures for control, as they were developed by Koch's group at that time, reflected this approach. Disinfection had initially been shown to be effective against bacteria outside the human body and became a model for therapy from the mid-1880s (Simon Reference Simon2007). While disease was understood as analogous to a cultivation process, therapy could be understood as a form of inner disinfecting which would – by way of injecting suitable substances – turn the formerly nutritive medium of the body into an infertile soil. As Georg Cornet, upon reviewing the state of the art of such therapy, noted in 1889, the intention was “to soak the tissues with antiseptics and to turn them into an unfavourable culture medium – in a way to sterilise them” (Cornet Reference Cornet1889, 100–101). Jonathan Simon recently argued that Behring's and Wernicke's work on immunizing substances in the blood was less a departure from this tradition than a creative prolongation of it. It was based on the assumption that blood itself possessed disinfecting properties. Such properties moreover had the advantage of being specific in relation to single diseases. The proposed carriers of these traits, the anti-toxins, could be extracted via blood serum and used to treat infections (Simon Reference Simon2007). Behring's and Kitasato's famous paper on the diphtheria and tetanus-antitoxins of 1890 is a direct outcome of this approach (Behring and Kitasato Reference Behring and Kitasato1890).

Behring's research was clearly a product of the late nineteenth-century culture of experimental medicine in Berlin. His understanding of the antitoxic properties of blood serum was closely related to the basic assumptions about infections purported by the Berlin group and the application of these assumptions to immunology. At the same time, Behring's work was tied to a division of labor between health administrators, chemical industrialists, and medical researchers – each field having its own independent actors. This division of labor was actively promoted by Prussian government officials. While in France the Institut Pasteur would perform all three functions, in Germany we find a division of labor among independent actors who shared the objects they worked with (such as aniline dyes). This division of labor predated the arrival of therapeutic vaccines in the 1890s and was expanded further with the employment of the vaccines and the rapid growth of pharmaceutical industries.Footnote ⁵

In the late 1870s, Carl Weigert, soon to be followed by his nephew Paul Ehrlich, started to use aniline dyes and other chemical substances for histological staining and – inspired by the specific affinity of such molecules to bacterial cells – later on tested some of these, such as methylene blue, as possible therapeutics against infectious diseases (Parascandola Reference Parascandola1981; Prüll Reference Prüll2003). The same period saw a trend towards the production of raw materials or intermediate products by chemical industrialists such as the Hoechst Company along with the invention of industrially produced, ready made drugs (Huhle-Kreutzer Reference Huhle-Kreutzer1989; Wimmer Reference Wimmer1994). With the rise of tabloids and the popularity of patent medicines since the late 1880sFootnote ⁶ – whereby industrially produced medicine in relation to property rights finally departed from the so-called specialities or secret remedies – medical bacteriology became the intellectual and technological resource for innovations in the drug market. Major companies like Hoechst, Schering, Merck, and Bayer opened bacteriological laboratories shortly after 1890.Footnote ⁷ The diphtheria-antitoxin, produced based on a division of labor between the researchers in Berlin and the industrialists in Frankfurt (Hoechst), was part and parcel of this process of expansion of the pharmaceutical market (Throm Reference Throm1995). This particular German division of labor entails yet another interesting comparison with the French situation: In France – as in the case of the rabies vaccine earlier – the control of the diphtheria antitoxin remained exclusively in the hands of the Institut Pasteur and its local branches; in Germany, competition between producers – initially Hoechst and Schering, but later more companies joined in – made state mediation or control seem to be a requirement.

Finally in the 1890s, as Volker Hess and Ulrike Klöppel show in their contributions to the present issue, researchers in Paris and Berlin were brought closer together by their common work on antitoxin therapy. During the decade earlier, the 1880s, their research on bacteriology differed markedly in approach and methodology. Nationalistic rhetoric had been employed in disputes between the leaders Pasteur and Koch and there was even a certain amount of secrecy and – resulting from it – intelligence work. For example, in 1885 a member of Pasteur's team was sent to Berlin incognito to participate in a bacteriological training course held at Koch's institute (Mendelsohn Reference Mendelsohn1996, 265–86; Métraux Reference Métraux and Maines1991). This, however, should be seen as one rather short period, and the close proximity of their work on antitoxin therapy brought the two schools into tune. Despite the fact that the years around 1890 were overshadowed by yet another debate between Paris and Berlin in relation to cellular or humoral theories of immunity (Silverstein Reference Silverstein1989, 38–58), the days of mutual separation were over. Medical microbiology in the 1890s, as Ulrike Klöppel's essay makes clear, could suitably be called a more open, but still heterogeneous field. The factors that brought this about were manifold: the generational changes mentioned above, increasing availability of handbooks in the field, and related work on antitoxins and their medical applications (Simon Reference Simon2007). After all, Roux's work on the antitoxin of diphtheria in 1892 began by repeating the experiments done by Behring and Kitasato. In notable contrast to the silence and sometimes open hostility between Pasteur and Koch in the early 1880s, Roux and Behring were careful to acknowledge each other's contributions.

Control

There was yet another feature that was visibly different in the serological cultures in Berlin and Paris. Faced with the task of regulating drugs such as sera, which were biological products from laboratories that did not fit into the frame of existent pharmacopoeias, health administrators in France developed regimes of regulation that were rather different from those of health administrators in Germany. In France, since the Institut Pasteur was a respected producer of vaccines, the quality and safety of therapeutic vaccines such as the diphtheria antitoxin was essentially left under the control of the Institut or its branches. As Volker Hess shows in his contribution to this issue, the Institut Pasteur, although not the sole producer of the diphtheria antitoxin, secured its position as a gatekeeper to a market which it dominated. Control by outsiders such as medical professionals was symbolic and they served to safeguard the institute's pivotal position instead of imposing limits. Christian Bonah's contribution shows that this was still the case when Albert Calmette marketed the BCG vaccine in the 1920s. In Germany, however, an elaborate division of labor developed and testing by a state commissioned institute was made obligatory for various producers. The Controlstation created for that purpose started as a branch of the Institute for Infectious Diseases in Berlin in 1894, quickly acquired independence, and became the nucleus of today's Paul Ehrlich Institute. What started with the diphtheria antitoxin had, by the outbreak of World War I, been extended to regulate antitoxic sera as such (Otto Reference Otto1906). Its history and further development beyond 1918 are described in Axel Hüntelmann's article in this issue.

In this introduction, the following questions remain: To what historical situation did the implementation of such a strict system of control respond and why it was exclusively designed for what was called “bacteriological medicine” at the time? In fact, the notion of such medicines had been profoundly influenced by events in 1890 and 1891 that were widely considered scandalous. The need for tight control may well have seemed plausible to public health officials, researchers, and producers. In Germany, Robert Koch's remedy for tuberculosis, which became known as tuberculin, had been one of the first attempts to develop medicine on the basis of what the bacteriological laboratory had elucidated about infectious diseases (Gradmann Reference Gradmann2001; Gradmann Reference Gradmann2005, 105–170). Intended as a cure for what was arguably the most widespread bacterial infection of the age, tuberculin took the medical community and the wider public from euphoria to deep disappointment in just a few months. The marketing of tuberculin, which started in November 1890, marks a crucial point in the history of antibacterial therapies. Until that time such medicines had been more a matter of imagination and expectation, nursed by the microbe hunting of the 1880s to which Koch himself had contributed by identifying the tubercle bacillus in 1882 (Schlich Reference Schlich and Schott1996). With tuberculin, the reality did not live up to the medical community's expectations. Initially Koch's cure seemed to be effective; it triggered an unprecedented euphoria that seemed to match the frenzy that had surrounded Pasteur's rabies vaccine in 1885. At that time few people questioned the fact that the inventor, to safeguard his financial and intellectual property, had decided to present tuberculin as a secret remedy – this after all had been the regular though somewhat anachronistic means to protect private interests in the days before patents on drugs. Doctors noted the violent reaction that the medicine could trigger in the human organism, but saw little ground for concern at the time. Initially, this particularly strong reaction even seemed to testify to its curative potential. Ernst von Leyden, a Berlin clinician, noted: “it is immediately obvious to the observer that this is a highly peculiar kind of liquid. The smallest dose of 1–2 milligrams upon injection produces a large and violent reaction. The temporary but tremendous transformation of the human organism is such that, in fact, we are surprised to the highest degree” (Die Wirksamkeit des Koch'schen Heilmittels 1891, 1145).

Nobody objected to the hands-on theatrical fashion in which the new medicine was tested a mere two months before its release: A small group of trustworthy clinicians had been supplied with the remedy, which was then tested on an estimated 50 patients. In a notable parallel to events surrounding the introduction of the rabies vaccine in 1885 and the famous treatment of Joseph Meister by Louis Pasteur, the testing of Koch's remedy in 1890 culminated in publicly staged applications that demonstrated the capacities of the new medicine (Gradmann Reference Gradmann2005, 189–197). However, towards the end of that year it became known that numerous relapses and a number of deaths had occurred in immediate connection with the organism's reaction to tuberculin. Moreover, the severity of the reaction did not appear to be in a stable, plausible relation to the dosage and thus the effects of tuberculin came to be regarded as unpredictable. Towards the spring of 1891, many doctors such as Friedrich Schulze from Bonn refrained from further application “since there are no criteria whatsoever on how the patient will fare if this unknown substance is applied” (Deutsche Medizinische Wochenschrift 1890–1891, 51–2). There is a striking contrast between this hasty launching of tuberculin and the careful launching of the diphtheria antitoxin three years later. When Behring tested the diphtheria antitoxin, he undertook lengthy clinical trials before marketing the medicine and even after this the serum was not hurried into medical use, but was introduced slowly over a long period and accompanied by a fair amount of critical discussion (Hüntelmann Reference Hüntelmann2005; Weindling Reference Weindling and Pickstone1992a).

In 1894 the head of the Imperial Health Office, Rudolf Köhler, ordered an expert review on the value of the diphtheria antitoxin for public health. The author of the report – pointing to the tuberculin episode – made a strong argument for state control of the medicine while maintaining that the diphtheria antitoxin had already been tested with a lot more scrutiny than Koch's infamous cure (Hardy Reference Hardy2006, 63–64). Just a few years after the event, the introduction of tuberculin was touted as a lesson on how not to proceed. In 1893 in the second edition of his textbook on the side effects of drugs, Leo Lewin devoted an entire chapter to Koch's wonder cure and characterized its uncontrolled application as “an episode raving with unscientific bustle” (Lewin [1881] Reference Lewin1899, 391).

Thus, the tuberculin scandal ended in a grand disillusionment. In Germany, this bacteriological medicine was known to have a strong effect. Yet it was of unknown composition and was unpredictable and dangerous in application. The most important lesson to be learned was that the therapeutic promise of laboratory medicine carried with it certain specific novel forms of serious, inherent dangers. Dangers of this nature were almost inconceivable prior to 1890 but they became a distinctive feature of twentieth-century medicine. It can be argued that this experience came to assume particular importance in the twentieth century, for it triggered a number of changes in the evaluation of therapies based on experimental medicine that shaped the way therapeutic vaccines were handled in Germany.

Thus, in addition to the proverbial Prussian-German passion for regulation, some lessons drawn from the tuberculin fiasco are visible in the requirements for handling therapeutic vaccines initiated just three years later: compulsory in-depth testing by a third party who was neither the inventor nor the producer of such medicine; the intention of this evaluation to establish – in the animal model – a stable and predictable relation of dosage and the organism's reaction; coherent documentation about the method of preparation and composition; and liability of producers to the control of public health officials. What can be described as the essential features of the evaluation of the diphtheria antitoxin and subsequent therapeutic vaccines had been all but missing in the case of tuberculin. It thereby becomes clear that the testing of the diphtheria-antitoxin had features of a technology of trust, designed to restore faith in medical drugs invented in the laboratory. In this sense it was certainly welcome to producers and researchers, since it secured the public standing of the object they dealt with. Serum therapy, after all, was under fire, for example in the British medical world, precisely because it reminded people of tuberculin (Hardy Reference Hardy2006, 63). Furthermore, the testing of diphtheria-antitoxin was expanded to other comparable medicines and became generalized as Wertbestimmung, which was made compulsory for such medicines. This move effectively blocked the marketing of antibacterial therapies of biological origin in other ways. Even tuberculin itself was in 1897 incorporated into that system, when a protocol for its testing was developed by Wilhelm Dönitz, a collaborator of Paul Ehrlich (Gradmann Reference Gradmann2005, 166).

Finally, in a rather different way, viewed from the perspective of the future producers of therapeutic vaccines, tuberculin can actually be said to have had beneficial effects. As mentioned above, links between Koch's Berlin group and the Hoechst Company had existed prior to the 1890s, when pharmaceutical companies generally aimed at expanding their medical bacteriological branches. In fact, precisely because it was not a financial success, tuberculin offered the Farbwerke Hoechst a chance to enter into pharmaceutical research without great financial investment (Bäumler Reference Bäumler1989, 67; Wetzel Reference Wetzel1991, 189–190). August Laubenheimer, a chemist at Hoechst, described the company's venture into pharmaceutical research by means of tuberculin:

In May 1892, an agreement was signed which led to the transfer of research on tuberculin to Hoechst. Koch's collaborator in producing the medicine, Arnold Libbertz, was employed by Hoechst for the purpose of supervising the production of tuberculin and became the head of the bacteriological laboratory (Wimmer Reference Wimmer1994, 157–58).

Thus, we can close this paper by noting this one final connection between tuberculin and the history of therapeutic vaccines. By attracting attention in medicine, public health, and industry, tuberculin created a space into which the application of diphtheria antitoxin would evolve, from 1895, when the marketing of the diphtheria antitoxin started and the Controlstation was created as a branch of the institute that was headed by the inventor of tuberculin.