- Split View
-
Views
-
Cite
Cite
Carol O. Tacket, Myron M. Levine, CVD 908, CVD 908-htrA, and CVD 909 Live Oral Typhoid Vaccines: A Logical Progression, Clinical Infectious Diseases, Volume 45, Issue Supplement_1, July 2007, Pages S20–S23, https://doi.org/10.1086/518135
- Share Icon Share
Abstract
Typhoid fever remains an important public health problem in many parts of the world. Despite the availability of oral Ty21a (Vivotif; Berna Biotech) and parenteral Vi polysaccharide vaccine (Typhim Vi; Aventis Pasteur), improved typhoid fever vaccines have been sought. These include a series of vaccine candidates developed at the Center for Vaccine Development, University of Maryland, based on attenuation of Salmonella enterica serovar Typhi by deletions in the aroC, aroD, and htrA genes. These vaccine candidates, designated “CVD 908,” “CVD 908-htrA,” and “CVD 909,” have been developed and tested in volunteers with variable success. This review summarizes the clinical data that directed the logical progression of this vaccine development strategy.
The tradition of excellence in infectious diseases and international medicine at the University of Maryland began with the keen clinical observations and investigations of Theodore E. Woodward. In the 1940s, Dr. Woodward found “virtually by accident” that typhoid fever responded to chloramphenicol, the first specific effective treatment for this disease. Dr. Woodward conducted seminal studies that elucidated the pathogenesis of typhoid fever and the efficacy of oral vaccines, both killed and live, in healthy volunteers challenged with Salmonella enterica serovar Typhi (S. Typhi). Dr. Woodward was a strong supporter of the University of Maryland's Center for Vaccine Development (CVD) in its early formative days in the mid-1970s, and he is the academic father of the hundreds of faculty, fellows, residents, students, and staff who have studied and worked at the CVD over the years.
Typhoid fever was a common disease in Europe, America, and elsewhere before the sanitation reforms of the late nineteenth and early twentieth centuries. Vaccination was pursued as a means of typhoid control soon after the discovery of the etiologic agent of typhoid fever near the end of the nineteenth century. The first killed whole-cell parenteral vaccine against S. Typhi was effective, but the local and systemic adverse effects were frequent and sometimes severe. An oral vaccine, attenuated by chemical mutagenesis and designed to mimic natural mucosal infection, was developed in the 1970s [1]. This vaccine, Ty21a, currently available as Vivotif (Berna Biotech), is well tolerated and efficacious but has a number of drawbacks. These drawbacks include lack of definition of the attenuating gene mutations and a requirement for multiple doses for optimal immunogenicity.
In the 1980s, with the dawn of the era of molecular biology, new attenuated S. Typhi strains with defined genetic mutations and heartier immunogenicity than Ty21a were sought. A number of such strains were developed, and several have been tested in humans, including Ty800, which is mutated in phoP/phoQ [2]; χ3927, which is mutated in cya and crp [3]; and MO1ZH09, which is mutated in aroC and ssaV [4]. During this time, a series of strains based on mutations in aroC and aroD was also developed and tested in humans over a 12-year period beginning in 1990 at the CVD. In the present article, we describe the history and lineage of these 3 vaccine strains. The story of these strains illustrates the tribulations of phase 1 vaccine testing and describes a failed vaccine that was too reactogenic, one that was subimmunogenic, and one that may strike the requisite balance between immunogenicity and reactogenicity.
CVD 908
One approach to attenuating S. Typhi has been to develop nutritional auxotrophs by interrupting the pathway for biosynthesis of aromatic metabolites, which renders Salmonella organisms nutritionally dependent on p-aminobenzoic acid and 2,3-dihydroxybenzoate, substrates not available to bacteria in mammalian tissues [5]. Mutants with deletions in the genes of this pathway are unable to synthesize chorismic acid (precursor of the aromatic compounds p-aminobenzoic acid and 2,3-dihydroxybenzoate), and no other pathways exist that can overcome this deficiency in Salmonella organisms. As a result, the aro-deleted bacteria cannot proliferate within mammalian cells, but the organisms grow intracellularly long enough to stimulate immune responses. Inactivation of either aroC or aroD independently results in attenuation, but deletions in both provide a high level of safety against restoration of pathogenicity by recombination.
Hone and Levine at the CVD developed an aroC/aroD-deleted derivative of S. Typhi strain Ty2 (the parent of Ty21a) and designated the strain “CVD 908” [6]. Two phase 1, dose-escalating studies of CVD 908 were conducted in inpatient volunteers at the CVD [3, 7]. After a single oral dose of 5 × 104 or 5 × 105 cfu, the vaccine was well tolerated, and vaccine bacteremia was not detected in multiple serial blood cultures. Six of the 12 volunteers who received CVD 908 at these doses developed serum IgG anti–lipopolysaccharide (LPS) antibodies, and IgA anti-LPS antibody-secreting cells (ASCs) were detected in 9 of 11 volunteers [3]. However, in subsequent clinical studies, clinically silent, self-limited vaccine bacteremia occurred in 50% of volunteers after a single dose of 5 × 107 organisms and in 100% of volunteers after a single dose of 5 × 108 organisms; this bacteremia occurred 4–8 days after vaccination [7,8–9]. The explanation did not involve genetic reversion to virulence, because the recovered blood and stool isolates retained their auxotrophic phenotype when grown on minimal medium.
The significance of the asymptomatic vaccine bacteremias was much discussed. On the one hand, transient vaccine viremia was well known among individuals who received live rubella and type 2 poliovirus vaccines, and vaccine viremia likely contributes to the robust immunogenicity of these successful vaccines [10,11–12]. Silent bacteremia, with seeding of organs such as the liver and spleen as well as of bone marrow, is one of the steps in the course of S. Typhi infection and is likely to stimulate vigorous protective immune responses. On the other hand, vaccine bacteremia was thought perhaps to portend unacceptable reactogenicity in future studies involving larger numbers of volunteers, children, elderly persons, and immunocompromised individuals. Even without overt clinical reactogenicity, the occurrence of vaccinemias would render the path to licensure more difficult. There was room for improvement.
CVD 908-HTRA
Additional mutations were sought to enhance the attenuation of CVD 908 (i.e., to eliminate vaccinemia) without compromising its immunogenicity. In a fruitful collaboration with Gordon Dougan of Imperial College of Science, Technology, and Medicine and Stephen Chatfield of Medeva Group Research, a further mutation in the gene htrA was constructed in Salmonella organisms and looked promising in preclinical studies. htrA encodes a heat-shock protein in Salmonella organisms; when the gene is deleted, the resulting mutant is less virulent because of an impaired ability to survive and replicate in host tissues [13]. In vitro, htrA mutants of S. Typhimurium are more susceptible to oxidative killing within macrophages. A further derivative of CVD 908, deleted in htrA and designated “CVD 908-htrA,” was constructed and studied in inpatient volunteers.
CVD 908-htrA retained the immunogenicity of the parent CVD 908, but no vaccine bacteremias were detected among 22 volunteers who received a single dose of 5 × 107, 5 × 108, or 5 × 109 cfu, despite intense surveillance for bacteremias after vaccination [8]. The immune responses measured by seroconversion rates with IgG anti-LPS after vaccination with CVD 908 and CVD 908-htrA were similar. Almost all the volunteers developed ASC responses, regardless of the vaccine strain received. Significant lymphoproliferative responses to flagella and particulate antigen after CVD 908-htrA vaccination were also observed [14].
To this point, the clinical studies of CVD 908 and CVD 908-htrA involved freshly harvested organisms grown from frozen stock cultures. On vaccination days, the vaccine strains were harvested from agar plates, washed, and diluted to the appropriate inoculum size. The crucial next step in the progression was to develop a practical formulation of CVD 908-htrA for future phase 2 and phase 3 studies. In building on the initial success of freshly harvested CVD 908-htrA, an industrial collaborator was identified (Peptide Therapeutics Group), and a practical formulation of CVD 908-htrA was produced by Evans Medical at their manufacturing facility in Speke, United Kingdom. This formulation consisted of live lyophilized organisms in single-dose glass vials to be resuspended in buffer solution. A phase 2 safety and immunogenicity study of CVD 908-htrA, formulated as a lyophilate, was approved for use in outpatient volunteers [15].
In this study, 80 healthy adult outpatient volunteers were randomized to receive, with buffer, a single oral administration of high-dose CVD 908-htrA (4.5 × 108 cfu), lower-dose CVD 908-htrA (5 × 107 cfu), or placebo. On day 28, there was a crossover in which the volunteers who had ingested CVD 908-htrA vaccine on day 0 received placebo and those who had received placebo on day 0 received CVD 908-htrA vaccine in either the high or the lower dose.
The incidence of adverse effects was the same in recipients of placebo, high-dose vaccine, or lower-dose vaccine after vaccination. Diarrhea, a potential concern observed in a small number of vaccine recipients in the uncontrolled, phase 1 study, occurred in 3 (4%) of 76 placebo recipients, 1 (3%) of 39 lower-dose vaccine recipients, and 4 (10%) of 39 high-dose vaccine recipients (a statistically insignificant difference). However, in a secondary analysis of the incidence of symptoms in the first 7 days after vaccination, recipients of the high-dose vaccine were found to be slightly more likely to have had diarrhea (8%) than were recipients of placebo (0%) (P = .04). No recipient of the lower-dose vaccine had diarrhea in the first 7 days after vaccination.
Importantly, all blood culture results were negative in this study. After ingestion of the lower-dose vaccine, 46% of volunteers shed the vaccine strain in their stool for a mean of 1.8 days. Among recipients of high-dose vaccine, 77% shed vaccine for a mean of 2.3 days, with a maximum shedding of 8 × 106 cfu/g.
In this phase 2 study, ASCs producing IgA anti-LPS were detected in 100% of recipients of the high-dose vaccine (geometric mean, 189 ASCs per 106 PBMCs) and in 92% of recipients of the lower-dose vaccine (geometric mean, 62 ASCs per 106 PBMCs). ASCs producing IgA anti-H antigen occurred in 79% and 73% of high- and lower-dose vaccine recipients, respectively. Serum IgG anti-LPS responses occurred in 49% and 46% of high- and lower-dose vaccine recipients, with geometric mean titers of 197 and 207, respectively, after immunization. Notably, none of the volunteers developed an increase in serum IgG Vi polysaccharide capsule antibodies. Lymphoproliferative responses to S. Typhi flagella and to particulate whole-cell S. Typhi by PBMCs obtained 14 days after immunization were detected in approximately half of the vaccine recipients, regardless of the dose received.
At this point, strain CVD 908-htrA appeared to meet many of the criteria for an improved oral typhoid vaccine. Unlike multidose Ty21a, single-dose CVD 908-htrA stimulated vigorous mucosal, humoral, and cellular immune responses that equaled or surpassed those measured after multiple doses of Ty21a [16].
CVD 909
Meanwhile, during this time, the intramuscular Vi polysaccharide vaccine (Typhim Vi; Aventis Pasteur) was developed, found to be efficacious [17, 18], and licensed in 1994. However, oral CVD 908-htrA offered substantial potential advantages over the parenteral Vi polysaccharide vaccine, in that CVD 908-htrA elicited an array of immune responses, including cellular responses, and immunologic memory not elicited by the polysaccharide. None of the new oral vaccine candidates (such as CVD 908, CVD 908-htrA, Ty800, and MO1ZH09), however, was consistent in their ability to stimulate serum Vi antibody.
One hint to explain this observation is that the expression of Vi is highly regulated. Vi expression in vivo occurs when the bacteria exist in extracellular environments, such as blood and bile, where they are protected from complement-mediated O antibody–dependent bacterial killing [19]. Vi expression is turned off when the bacteria reside within macrophages and dendritic cells in the gut-associated lymphoid tissue. At least 2 separate 2-component systems, rcsB-rcsC [20, 21] and ompR-envZ [22], regulate Vi expression. M.M.L. proposed that, by rendering expression of Vi constitutive in CVD 908-htrA, the oral strain might elicit serum and mucosal anti-Vi antibody responses in addition to responses to whole-cell antigens.
M.M.L. and colleagues replaced the highly regulated PtviA in S. Typhi vaccine strain CVD 908-htrA with the strong constitutive promoter Ptac, resulting in a strain designated “CVD 909” [23]. In a clinical study, 24 healthy adult outpatient volunteers received a single oral dose of CVD 909 of 106–109 cfu, and 8 healthy adults received 2 doses of 6 × 109 cfu per dose with buffer 14 days apart [24]. In this case, the vaccine was freshly harvested from a frozen stock culture.
The serologic responses to S. Typhi LPS and H antigen were vigorous, as had been seen after administration of CVD 908-htrA. After 1 or 2 doses of CVD 909, almost all volunteers developed IgA anti-LPS ASCs. This was encouraging, because the presence of the Vi polysaccharide capsule could have interfered with immune responses to other cell surface structures. However, only 1 (4%) of 24 volunteers who ingested a single dose and only 1 (12.5%) of 8 volunteers who ingested 2 doses developed serum IgG anti-Vi [24]. However, IgA anti-Vi ASC responses were detected in 80% of the 20 volunteers who received 108–109 cfu of CVD 909; in all but 1 volunteer, this response occurred after a single dose. These Vi-specific ASC responses provided convincing evidence that the Vi antigen was expressed and immunologically processed, even if serum antibody responses were meager.
What is the significance of ASC Vi responses in the absence of serum IgG antibody to Vi? It is possible that the immune system was primed to this antigen and a protective anamnestic response would follow infection with wild-type S. Typhi. Studies to examine this question and the future of CVD 909 vaccine are ongoing.
Conclusion
The World Health Organization estimates that 16 million cases of typhoid fever occur annually, resulting in ∼600,000 deaths [25]. Despite the availability of Ty21a (Vivotif) and the Vi polysaccharide vaccine (Typhim Vi), the World Health Organization and the wider community recognize that improved typhoid fever vaccines are needed. Both Ty21a and Vi polysaccharide are only moderately effective (50%–70%), and their efficacy in children <3–5 years of age is not acceptable or unknown. In at least some areas of the world, typhoid fever occurs frequently in children <5 years of age. The urgent need for a safe and effective vaccine is even more compelling in areas where antibiotic-resistant S. Typhi is prevalent. The new Vi-protein conjugate vaccines [26], as well as some new oral vaccine strains, including CVD 908-htrA and, possibly, CVD 909, offer reason for optimism. The stuttering but logical process of discovery, illustrated by the development of CVD 908, CVD 908-htrA, and CVD 909, is a progression that Dr. Woodward understood and celebrated.
Acknowledgments
We acknowledge the invaluable assistance of the staff of the Center for Vaccine Development and the University of Maryland General Clinical Research Center (National Institutes of Health/National Center for Research Resources grant MO1 RR016500).
Financial support. National Institute of Allergy and Infectious Diseases, National Institutes of Health (grants NO1 AI25461 to M.M.L. and NO1 AI40014 to C.O.T.).
Supplement sponsorship. This article was published as part of a supplement entitled “Tribute to Ted Woodward,” sponsored by an unrestricted grant from Cubist Pharmaceuticals and a donation from John G. McCormick of McCormick & Company, Hunt Valley, Maryland.
Potential conflicts of interest. M.M.L. is the coinventor of the patent for attenuated Salmonella enterica serovar Typhi strains that constitutively express Vi capsular polysaccharide antigen. However, no company has licensed this technology. C.O.T.: no conflicts.
Comments