Vaccinating against Helicobacter pylori in the developing world

Abstract

Helicobacter pylori infects more than half the world’s population and in developing nations the incidence can be over 90%. The morbidity and mortality associated with H. pylori-associated diseases including ulcers and gastric cancer therefore, disproportionately impact the developing world. Mice have been used extensively to demonstrate the feasibility of developing a vaccine for H. pylori infection, and for testing antigens, routes of immunization, dose, and adjuvants. These successes however, have not translated well in clinical trials. Although there are examples where immune responses have been activated, there are few instances of achieving a reduced bacterial load. In vivo and in vitro analyses in both mice and humans demonstrates that the host responds to H. pylori infection through the activation of immunoregulatory mechanisms designed to suppress the anti-H. pylori response. Improved vaccine efficacy therefore, will require the inclusion of factors that over-ride or re-program these immunoregulatory rersponse mechanisms.

Keywords: :

• Helicobacter pylori
• vaccine
• dyspepsia
• gastric cancer

H. pylori Disease and Epidemiology

Helicobacter pylori infects the stomach of more than half of the humans on Earth and is a primary cause of pathologies including some dyspepsias, peptic ulcer disease, and gastric cancer.^1-5 Prevalence ranges greatly from 10 to 70%, even within Western countries, and can approach close to 100% in developing countries.^6,7 Infections are believed to be primarily acquired in early childhood and predominantly through oral-oral transmission.^6,7 Infection lasts for the life of the host if left untreated, and while most hosts remain asymptomatic, 10–20% develop peptic ulcer disease, 1% will develop gastric adenocarcinoma, and less than 1% will develop mucosa-associated lymphoid tissue lymphoma (MALT).⁸ The World Health Organization has classified H. pylori as a definite class I carcinogen.⁵ Gastric cancer remains the second leading cause of death due to cancer worldwide and in large geographic regions including South America, Eastern Europe, and the Far East these levels range from 20 to 40 per 100 000.⁹

The incidence of infection is strongly influenced by socioeconomic factors.^6,10 One illustrative survey employed the use of seroposivity levels of subjects collected between 1988 and 1991 which demonstrated an overall prevalence in the US of 32.7%.¹⁰ The frequency rose to 48% for those below the poverty level, and to over 50% for those living in overcrowded homes or with little education. Another study has documented the wide variation observed in the U.S, among various ethnic populations between 2004–2008.¹¹ When the prevalence in males was investigated the incidence among Whites was found to be 9.5 per 100 000, while the incidence among Hispanics, American Indians, and Alaska Natives was over 14 per 100 000. African Americans and Asian and/or Pacific Islanders had an incidence of over 17 per 100 000 men.¹¹ The mortality rate among most of these minority populations was twice the rate of Whites for both men and women.

The Economic Benefits of an H. pylori Vaccine

H. pylori-related gastric diseases create a heavy burden on health care systems world-wide. In its most recent analysis, the US. Department of Health and Human Services estimates that in 2004, the direct costs associated with peptic ulcer disease in the US alone were approximately $2.6 billion.¹² Additional indirect costs associated with loss of work productivity were estimated to be over $518 million. The costs due to gastric cancer were estimated to be $487 million with indirect costs reaching $1.4 billion.

Similar analyses on associated costs in developing nations are not available, but where the incidence of infection and disease are significantly higher, the costs associated with monitoring and treating H. pylori-associated disease can be inferred from existing studies performed in developed nations. Several independent studies have been performed to compare the costs associated with treating dyspeptic patients or patients with peptic ulcer disease. Two such studies found an average cost for testing for H. pylori presence and then treating with antimicrobial therapy at over $500 per dyspeptic patient, with an additional $400 added for those diagnosed by endoscopy^13,14 A similar analysis performed in China for patients with ulcers estimated an approximate cost of US $1778 per patient.¹⁵ A model analysis using economic data to determine the cost effectiveness of a population screening and treatment for H. pylori infection in 1 000 000 45 y olds would yield a savings of over 6 000 000 pounds sterling as well as 1300 y of life.¹⁶ In low socioeconomic populations where H. pylori is common, it is easy to see such costs as prohibitive on a large scale.

Antimicrobial therapy for treatment of H. pylori infection therefore can makes up a significant part of the economic burden for a developing nation, even where cost saving strategies have been developed and recommended.⁷ However, even if the costs associated with therapy were nominal, there are several reasons why such therapies still do not represent the most cost effective means of dealing with H. pylori and its associated diseases. Antimicrobial therapy requires multidrug regimens, including several antibiotics plus a proton pump inhibitor.² Additionally, as treatment is taken several times a day for at least seven days, patient compliance can become a significant obstacle to success. Antibiotic resistance to some of the more commonly used antibiotics such as clarithromycin, and metronidazole has become increasingly problematic in some countries.¹⁷ It is also important to consider that even successful eradication with antimicrobial therapy does not provide resistance to subsequent H. pylori infection, and in some countries the rate of reinfection is as high as 15‒30% per annum.^17-19

Perhaps the most important reason the test and treat strategies are suboptimal is that most patients that develop H. pylori associated gastric cancer typically remain asymptomatic until progression of the cancer is in advanced stages. Treating patients with antimicrobial therapy as they present with symptoms would fail to identify infection in time to intervene with carcinogenesis. Vaccination therefore, offers the most cost-effective means for reducing the incidence of H. pylori and its associated diseases. A prophylactic or possibly even a therapeutic vaccine could be administered to populations with high incidence of gastric cancer to circumvent significant morbidity and mortality. It has already been demonstrated that where H. pylori is detected and treated in patients with early gastric cancer, the incidence of secondary cancer is reduced by 65%.²⁰

A theoretical model has been developed to predict the effects of a vaccine administered to children.²¹ The authors assumed a vaccine would be available in 2010 for young infants in the US. Based on the currently declining prevalence of H. pylori infection, they predicted vaccination would decrease the prevalence of H. pylori an additional 6-fold to only 0.7% by 2100. The incidence of gastric cancer was estimated to decrease more than 3-fold compared with a population with no vaccine to only 0.4 per 100 000. Application of this model to a population with a higher incidence of gastric cancer such as Japan could decrease the incidence from 17.6 to 1.0 per 100 000 by 2100. For a country with an incidence exceeding 30.0 per 100 000, continuous vaccination could decrease incidence by 80% to 5.8 per 100 000 by 2100.

Vaccination in developing nations may also impact childhood health in general. One potential consequence of early H. pylori infection is the temporary increase in gastric pH and consequently the loss of the gastric barrier to other pathogens following acquisition of H. pylori.²² Children in the Gambia for instance frequently develop diarrhea, malnutrition, and growth retardation as they begin eating solid food. This is presumably due to contamination of food or hands with enteric pathogens.²³ H. pylori infection has also been associated with iron deficiency anemia in children.²⁴ A vaccination campaign therefore, would benefit the developing and developed world and improve overall health of pediatric and adult populations against diseases associated with high morbidity and mortality. While such a vaccine would ideally be administered as a prophylactic vaccine to young children, as described below, the technology suggests that a vaccine could even be administered therapeutically to adults following a diagnosis of infection. It should be noted, that the need for vaccination in developing nations due to the high incidence of gastric cancer, or high cost of diagnosis and treatment, will require logistical, economic, and practical aid from the World Health Organization as well as national governments of the respective nations involved.

Lessons from Small Animal Models

Murine models of H. pylori and H. felis infection have been used for over 20 y to investigate and refine candidate vaccines. Multiple laboratories have utilized these models to consistently demonstrate that protective immunity can be achieved through vaccination as measured by either sterilizing immunity or significant reductions in bacterial load. A comprehensive review of all the H. pylori vaccination parameters that have been investigated in mice is beyond the scope of this article but has been reviewed recently.²⁵ Several observations however are worth highlighting for their potential importance with regard to developing a human vaccine.

First, although the majority of animal and clinical studies have been performed with a single antigen, an efficacious vaccine for use in humans will most likely require multiple antigens. An early study by Ferrero et al. demonstrated that combining Heat shock protein A with a Urease subunit vaccine protected 100% of mice against challenge with H. felis whereas immunization with either component individually yielded protection in only 80% of mice.²⁶ Additionally, whereas complete protection has been difficult to achieve in larger animal models, a subunit multivalent vaccine incorporating VacA, CagA, and Neutrophil Activating Protein (NAP) was shown to be protective when administered therapeutically in beagle dogs as measured by immunohistochemical detection of H. pylori in histologic sections.²⁷ Since a vaccine might require multiple antigens, it is encouraging that almost any H. pylori protein that has been tested so far seems to work as well as any other, including urease subunits,²⁸ CagA,²⁹ VacA,²⁹ catalase,³⁰ flagellin,³¹ heat shock proteins,²⁶ H. pylori adhesion A,³² NAP,³³ and others. Although whole cell lysate preparations have also been used extensively in mice as described in the initial Helicobacter vaccine reports,^34,35 the development of a human vaccine will require a better characterized, well defined product. The large number of tested proteins indicates that selection of antigens for a multivalent vaccine will not be problematic. Given the antigenic variation between isolates of H. pylori however, a vaccine incorporating antigens with sequences known to be highly conserved and uniformly found in all strains would find the most utility for widespread use.

Another important finding from murine studies is the success achieved when administering an experimental vaccine therapeutically. Several laboratories have demonstrated that vaccination of mice following experimental infection with either H. felis or H. pylori is as effective as prophylactic immunization.^36-38 Extension of this technology to the domestic ferret model of indigenous gastric H. mustelae infection resulted in 30% of the animals achieving protective immunity as measured by the inability to culture organisms from gastric pinch biopsies.³⁹ These studies were the basis for subsequent efforts to test a therapeutic vaccine against H. pylori in nonhuman primates and in clinical studies described below. The ability to eliminate or reduce bacterial load when immunizing therapeutically is an important observation as it suggests that it is not the pre-emptive kinetic quality of the vaccine that is important but rather the induction of some aspect of host immunity not induced in the host during the course of infection.

Mouse models have also been helpful for experimentation with different routes of immunization. Initial studies focused on oral immunization, or immunizations designed to optimize mucosal immunity since H. pylori colonizes the gastric mucosa and is noninvasive. Thus the first successful demonstrations of Helicobacter vaccine feasibility in the mouse model included live H. felis delivered intraperitoneally to mice,³⁴ and H. felis lysate with cholera toxin (CT) adjuvant delivered by oral gavage,^34,35 all of which reduced Helicobacter colonization. Although oral vaccination dominated subsequent vaccine trials in the 1990s it was soon determined that protection could be induced by other routes of mucosal immunization including orogastric, intranasal, and rectal^{28,29,34,35,40-42} The use of intranasal immunization has been particularly interesting as it allows for the use of less H. pylori antigen and in several laboratories has induced complete protection, as opposed to a significant reduction in bacterial load,^43,44 although sterilizing immunity has been reported through routes including rectal immunization in mice,⁴⁵ and oral immunization in gerbils.⁴⁶

Surprisingly, vaccination has also been shown to be as efficacious when delivered by systemic routes such as the subcutaneous or intraperitoneal.^47-49 Although unusual for a mucosal pathogen, this may reflect the nature of Helicobacter challenge resulting in the recruitment of peripheral lymphocytes to the gastric mucosa in response to infection. Evidence for such recruitment comes from models in which infected SCID or transgenic rag knockout mice are reconstituted with splenic T cells and the demonstration of subsequent gastritis and reduction or eradication of H. pylori from the mucosa^50-52 The induction of protective immunity therefore may be possible through more traditional parenteral vaccination strategies.

Finally, the mouse model has been instrumental in testing a wide variety of adjuvants for Helicobacter vaccination. The earliest immunization studies on Helicobacter employed the well established mucosal adjuvants, CT and LT. These bacterial exotoxins consist of non-toxic pentameric B (binding subunits) noncovalently linked to an enzymatically active A subunit which mediates toxicity.⁵³ These adjuvants are used extensively in animal models of enteric vaccination and have been crucial for improving our understanding the protective immune response to H. pylori. Unfortunately, as discussed below, the same doses that can be tolerated by mice remain toxic in humans. Therefore one strategy toward improving the prospects for human use vaccine has been to develop a nontoxic form of LT through genetic manipulation. Such nontoxic mutants have been evaluated against several pathogenic microorganisms, including H. pylori^29,54,55 but their use has not been extended to human clinical trials against H. pylori.

The potential problems associated with exotoxin adjuvants in humans presents a major obstacle to vaccine development. To the extent that mucosal immunization is deemed desirable for vaccinating against H. pylori, a suitable mucosal adjuvant will have to be developed. The body’s immune system is designed to tolerate most proteins that pass through the alimentary tract since food and commensal microbiota are always present. Similarly for H. pylori protein subunits, they will remain weak immunogens when given mucosally unless a suitable adjuvant is present. No such adjuvant exists that is both efficacious and can be delivered safely to humans. A great deal of effort therefore has been expended to explore the use of alternative adjuyants. Some of the technologies that have been explored include the use of CPG olignucleoties, biodegradeable microspheres, saponins, chitosan, muramyl dipeptide (MDP), M cell lectins, and recombinant attenuated Salmonella or viral vectors (reviewed in ref. 25). While almost all of these technologies have induced various levels of protection, few have been further tested in nonhuman primate models or in clinical trials. In this light, the observation that significant reductions in bacterial load can be achieved by parenteral immunization is potentially very important. As mentioned above, aluminum oxide based vaccines induced protection in mice comparable to the use of CT with mucosal immunization.^47-49 Since alum is already approved for use in humans, advances in an injectable vaccine against H. pylori may prove to be the most straightforward way to achieve success. Although, as discussed below, an alum-based vaccine did not protect humans from H. pylori in a clinical trial, the core basis of a subunit vaccine can be made, even if ultimately it may require the addition of additional subunits that promote specific types of immunity.

Lessons from Clinical Trials

There have been four clinical trials in which experimental vaccines have been tested for their ability to either eradicate an existing H. pylori infection or to prevent an experimental challenge, although several additional clinical studies have been performed to test safety and immunogenicity for various vaccine formulations.⁵⁶ The first published report was telling of the difficulties that will be necessary to overcome in an oral vaccine. The authors tested a subunit vaccine consisting of the urease holoenzyme in combination with LT adjuvant delivered by the oral route in a therapeutic vaccine trial.⁵⁷ Study participants received four weekly doses with either 5 or 10 µg LT. Only one study group achieved a significant reduction in bacterial load as measured by CFUs in gastric biopsies. Complications due to the use of LT were apparent as many of the study volunteers experienced various levels of stomach cramping and gastrointestinal distress. The use of 10 µg doses of LT had to be discontinued prior to completion of the study, but even the 5 µg dose induced diarrhea in many of the subjects. Follow up studies in volunteers were performed for the express purpose of identifying an immunogenic yet acceptably safe dose of LT to use in subsequent studies. LT doses ranging from 0.1 μg to 2.5 μg in combination with urease demonstrated that as little as 2.5 μg of LT retained adjuvanticity but 50% of the volunteers still experienced diarrhea.⁵⁸ Rectal immunization was tested as an alternative and safer route of administration but while the LT itself remained immunogenic, the urease antigen failed to induce a humoral or T cell response after several doses.⁵⁹

Shortly thereafter, another therapeutic study was performed by a different laboratory using a formalin killed whole cell antigenic preparation in combination with recombinant LT_R192G, an LT variant in which an amino acid substitution has been made in the trypsin cleavage site to reduce its toxicity.⁶⁰ No study participants eradicated H. pylori as measured with the noninvasive ¹³C urea breath test 7.5 mo post-immunization. It is possible that reductions in bacterial load were achieved but such information cannot be determined with the ¹³C urea breath test. Unfortunately, the LT_R192G, induced diarrhea in 28% of vaccine subjects and 33% of placebo recipients, levels comparable to the use of wild type LT described above. These two clinical studies therefore demonstrate the need for continued development of nontoxic mucosal adjuvants.

Studies using recombinant, avirulent Salmonella strains expressing H. pylori urease have also been performed to test immunogenicity in volunteers.^61,62 Only one formulation was found to induce any urease-specific antibodies but this response varied between subjects. The use of urease expressing Salmonella vaccine strain Ty21a demonstrated only weak responses in H. pylori-negative volunteers with no induction of urease-specific antibodies and only a few subjects developing T cell memory.⁶³ However, subsequent use of the Salmonella Ty21a based recombinant vaccine in a prophylactic clinical challenge study using H. pylori-negative volunteers did provide some mixed results.⁶⁴ Administration of multiple oral doses of either the urease-expressing Salmonella strain, or another recombinant strain expressing the HP0231 H. pylori antigen prior to challenge with a H. pylori strain previously characterized for use in human challenge studies⁶⁵ resulted in the development of immune responses typically associated with H. pylori infection.⁶⁴ The number of organisms in gastric biopsies decreased in eight of the vaccinated subjects, but this decrease also occurred in the nonimmunized control group. These attenuated vaccines were well tolerated however and may be worth continued investigation, perhaps with additional antigens expressed to increase valency.

Most recently, a multivalent subunit vaccine consisting of VacA, CagA, and NAP mixed with Alum have been used to explore the potential for systemic vaccination in clinical studies. Initially, a therapeutic study using intramuscular immunizations to vaccinate beagles was performed in which they demonstrated eradication of H. pylori by testing for urease activity in biopsies and examining histlogic sections.²⁷ In a subsequent study for safety and immunogenicity in human, H. pylori negative volunteers demonstrated strong immunogenicity, inducing strong antibody and cellular recall responses.⁶⁶ Volunteers received three intramuscular immunizations with the multiunit vaccine plus aluminum hydroxide. No untoward side effects were noted and a booster immunization applied almost 2 y later induced a strong memory response. The use of this vaccine in a Phase III clinical challenge model however were disappointing. Whereas approximately 50% of volunteers receiving the subunit vaccine were H. pylori negative 12 weeks post-challenge, similar rates of protection were observed in volunteers receiving the placebo.⁶⁷

The Need to Overcome the H. pylori-Induced Immune Response

The clinical trials described above can be characterized as unsuccessful although there are instances of immune induction and in some cases a reduction in bacterial load. There is still reason to believe however, that vaccine-induced protective immunity can be achieved in humans despite this series of disappointing clinical trials. Development of an improved vaccine may be possible not only due to our improved understanding of the protective immune response in animals models, but also because of our evolving understanding of the host response to H. pylori infection and why it fails to eradicate H. pylori from the gastric mucosa.

Numerous laboratories have helped to define the nature of the protective immune response and these efforts began soon after the mouse models of vaccine induced immunity to H. felis and H. pylori were reported. A series of studies performed in transgenic knockout mice were instrumental in determining the essential role of the T helper cell response. Mice lacking the mucosal immunoglobulin IgA, or indeed any antibody molecules were shown to be equally well protected as wild type mice by vaccination.^38,68,69 Conversely, transfer of a H. felis-specific CD4⁺ T cell line reduced the bacterial burden in naïve recipient mice following challenge with H. felis⁷⁰ and an essential role for CD4⁺ T cells was confirmed in experiments in which vaccinated CD4⁺ T cell deficient MHC class II knock out mice failed to be protected from H. pylori challenge.^69,71Protection in CD8⁺ T cell deficient MHC class I knock out mice was not ameliorated. It has also been demonstrated that adoptive transfer of CD4⁺ T cells into immunodeficient SCID mice followed by challenge with H. pylori results in pronounced inflammation that reduces the bacterial load compared with wild type control mice.^50,51 Thus, the T helper cell response is not only necessary to induce protective immunity, in the absence of any other arm of adaptive immunity, it is sufficient to induce protection.

The nature of the protective T helper cell response has been studied in detail and it is now generally accepted that the induction of increased inflammation via Th1 or Th17 cells is the driver of protective immunity. Although several reports on CD4⁺ T cell mediated protection were supportive of a role for Th2 cells,^48,70,72,73 other studies indicated that Th2 responses were not protective and provided evidence that protection was dependent on Th1 responses^74-77 It should be noted that even where Th2 cells were deemed to be important, the mice did develop robust histologic gastritis upon challenge, consistent with a proinflammatory response. The essential role of proinflammatory T helper cells has been demonstrated by several laboratories through the use of p40 deficient mice which fail to generate protective immunity against H. pylori infection following vaccination.^74,76 The p40 subunit is shared by two heterodimeric cytokines IL-12 (p40 + p35) and IL-23 (p40 + p19) that promote divergent T helper cell responses, and therefore these experiments did not address the relative contributions of Th1 and Th17 cells respectively.

There is now evidence to suggest that either a strong, vaccine-induced Th1 or Th17 response can be protective, and that in the absence of one, the other will compensate. This helps explain why models deficient in Th1 activity either due to lack of IL-12 or IFNγ have generated conflicting data with regard to the importance of Th1 cells.^74,76,78-81 It also explains why models of Th17 deficiency through either lack of IL-23, IL-17, or the IL-17R indicate that Th17 cells are not necessary for protection.^81,82 Ultimately, experiments in which IL-17 was neutralized during the challenge phase in immunized mice demonstrated that Th17 cells are vital for protective immunity.^83,84 It appears that mice deficient in either pathway will compensate following immunization by relying on the alternative T helper cell type. However, if wild type mice are immunized and develop immunity, neutralization of either IFNγ or IL-17 can ameliorate protection because the animal has already developed immunity relying on the T cells that were present during the immunization phase. The effector mechanisms that actually eradicate H. pylori remain ill-defined and it is likely that multiple mechanisms are involved. The depletion of neutrophils however, has been demonstrated to significantly limit bacterial clearance in both vaccine-induced immunity models as well as a model of spontaneous eradication that occurs in IL-10 deficient mice.^79,85,86 These results suggest that vaccines designed to specifically enhance the Th1 or Th17 response to H. pylori may help to improve vaccine efficacy if such technologies can be developed.

Knowing the type of immunity required for eradication of H. pylori will be important for improved vaccine efficacy. It is perhaps equally important to take into account the propensity of the host to downregulate the immune response to H. pylori following infection. A model in which the host response to H. pylori is suppressed may seem to conflict with some of the most basic observations regarding the H. pylori infection. Infection is always accompanied by robust histologic gastritis consisting of a mixed polymorphonuclear cell and lymphocytic infiltration.⁸⁷ And T cells from this inflammation consist predominantly of Th1^50,88-92 and Th17 cells.^93-95 Functional analysis in vitro however, has revealed that the H. pylori-specific recall response of blood or gastric T cells from H. pylori-infected subjects is equivalent to, and in some cases less than for lymphocytes isolated from H. pylori negative donors.^91,96-99 Additionally, depletion of CD25⁺ T cells (Treg cells) from isolated cells prior to stimulation with H. pylori significantly increases the IFNγ response.¹⁰⁰ Histologically, FoxP3⁺ Treg cells are observed in patient gastric biopsies and H. pylori-infected children have increased numbers of these Treg cells and significantly less gastritis than H. pylori-infected adults.¹⁰¹

The mouse model has been instrumental in defining the role of Treg cells in H. pylori infection. Several laboratories have taken advantage of the expression of high levels of CD25 by Treg cells. When isolated lymph node cells depleted of CD25⁺ cells were adoptively transferred into nude mice, the recipient mice generated significantly more severe gastritis and reduced bacterial load compared with mice that received an unmanipulated population of lymph node cells.¹⁰² In a subsequent study, antibody-mediated depletion of CD25⁺ cells in mice prior to challenge also resulted in increased gastritis and significantly reduced bacterial loads.¹⁰³ More recently, since the Foxp3 transcription factor has been identified as a better marker for Treg cells, depletion of Treg cells through the injection of diphtheria into FoxP3-DTR transgenic mice has yielded similar results.¹⁰⁴ Preventing Treg activation through the blocking of cell surface CTLA4 also results in decreased bacterial load and significantly increased gastritis.¹⁰⁵

An active role for Treg cells in suppressing the host response to H. pylori infection is consistent with descriptions of several mouse models in which spontaneous eradication of H. felis or H. pylori have been reported. Immunologic homeostasis at environmental interfaces is meditated by production of IL-10 by Treg cells.¹⁰⁶ IL-10 deficient mice have been demonstrated to respond to H. felis or H. pylori infection with unusually severe gastric inflammation and a significant reduction or eradication of the bacteria from the gastric mucosa.^50,85,86 Challenge of NADPH oxidase deficient mice with either H. felis of H. pylori yields similar results,¹⁰⁷ presumably due to the role that reactive oxygen species play in the development of Treg cells.¹⁰⁸ Finally, transfer of splenic T cells into otherwise immunodeficient SCID mice results in unregulated T cell hyperplasty that can also eliminate H. pylori from the gastric mucosa, an effect that is ameliorated when the transfer is limited to memory T cells already under the influence of Treg cells.^50,52

H. pylori may indeed be a significant pathogen throughout the world, but it bears remembering that greater than 80% of infected individuals do not develop H. pylori associated disease. It is becoming apparent that despite its role in peptic ulcer disease and gastric cancer, the host may actually recognize H. pylori as “nondangerous” bacteria and respond the same way it does to other non-invasive colonizers of the epithelium, by suppressing the host immune response through a network of regulatory mechanisms. It is this predisposition of the host that must be overcome in order to improve the efficacy of an H. pylori vaccine. In this light, contrary to concerns about the increase in histologic gastritis experienced in immunized mice following challenge, a vaccine must be designed to increase this cellular immune response, with the appreciation that such inflammation will dissipate after H. pylori has been eradicated from the mucosa.^43,50 Mouse models of vaccine-induced Helicobacter immunity, and spontaneous clearance through immune dysregulation, and even the clinical trial employing the Salmonella Ty21a based recombinant vaccine described above⁶⁴ describe an overall significant increase in gastric inflammation that coincides with a reduction in bacterial load.²⁵ Such an increase in inflammatory infiltrate is essential since the stomach is largely devoid of organized or diffuse lymphoid tissue. Additionally, observations that even asymptomatic H. pylori-infected individuals develop histologic gastritis provide evidence that an increase in gastric inflammation will not necessarily be detrimental to the host.⁸⁷

There are two examples to support the notion that host immunity can be manipulated in favor of H. pylori eradication and that the efficacy of an H. pylori vaccine could be enhanced. Murine models have been used to test the potential of cytokine administration to increase inflammation during Helicobacter infection and reduce the bacterial burden. In one model, mice infected with H. felis were administered IL-17 on five consecutive days post-infection and analyzed on day 14.⁸³ The bacterial load of IL-17 treated mice was found to be significantly less than vehicle treated mice as determined by the rapid urease test on biopsies. In a separate study, the potential of a vaccine independent Th1 response was evaluated by injection of H. felis infected mice with IL-12.⁸¹ Mice were treated with frequent doses of IL-12 at two week intervals, with two week rests between treatments, and evaluated at three months post infection. The IL-12 treated mice developed significantly more severe gastritis and in most cases eradicated H. felis from the gastric mucosa as determined by direct inspection of silver-stained histologic sections. Such studies suggest that strategies can be adopted to complement vaccine construction and overcome the propensity of the host to suppress the immune response and specifically bias the immune response toward a Th17 or Th1 character.

Potential Detriments to Immune Based Eradication of H. pylori

There are at least two potentially detrimental ramifications to large scale eradication of H. pylori through vaccination. The first involves the risk of developing pathologic sequela under specific circumstances. As mentioned above, lack of IL-10 in mice results in significantly enhanced gastritis and where the bacteria are not eradicated, hyperplastic gastritis developed including changes in proliferative kinetics and tissue architecture.¹⁰⁹ The production of IL-10 is a primary means by which Treg and Tr1 cells downregulate immunity. Therefore, in the absence of complete protection, a vaccination that reduces regulatory T cell activity and increases the Th1 and Th17 response has the risk of predisposing the host to more severe Helicobacter associated pathogenesis. It is unclear whether similar events might occur, at least transiently, where sterilizing immunity is achieved but the host continues to be challenged by repeated exposure to H. pylori. These developments in the gastric mucosa can only be detected through the collection and examination of gastric biopsies which makes monitoring for such events prohibitively impractical and expensive.

The second argument against eradication of H. pylori from large populations is the benefit conferred on the host by infection. Numerous studies have been performed demonstrating an inverse correlation between H. pylori infection, particularly CagA positive strains, and Barrett’s Esophagus and some meta-analysis indicate a similar relationship between CagA positive strains of H. pylori and esophageal cancer. The physiologic factors that relate H. pylori infection to esophageal health are complex and not well understood. However, an inverse correlation between H. pylori infection and early childhood asthma has also been reported,¹¹⁰ and murine models of experimentally induced asthma demonstrate that H. pylori-associated protection against asthma is related to the regulatory T cells induced by H. pylori infection.¹¹¹ These Treg cells are able to disseminate to the lungs and suppress the response to the allergin used to induce allergic asthma. A similar mechanism may explain an inverse correlation between H. pylori infection and inflammatory bowel disease.¹¹² While there may be benefits to H. pylori infection however, a vaccination initiative targeted to populations in which gastric cancer remains a significant cause of morbidity and mortality would represent a worthwhile trade off.

Conclusion

It is the developing world that could benefit the most over the long-term from an efficacious vaccine against H. pylori. The development of such a vaccine is complicated by the propensity of the host to downregulate the immune response to H. pylori upon infection. Thus, it will be necessary to use technologies that re-program the host immunregulatory response in addition to the selection of antigens and adjuvants capable of inducing a robust, proinflammatory, T helper cell response. The application of cytokines that promote either Th17 or Th1 immunity can result in bacterial clearance even in the absence of immunization. Such cytokine adjuvants, when combined in low doses with candidate subunit vaccines may provide a rational strategy to achieve improved efficacy, and ultimately be successful at decreasing the incidence of H. pylori in the developing world.

Acknowledgments

These studies were funded by NIAID, NIH, DHHS grant U19 AI082655 (Cooperative Center for Translational Research in Human Immunology and Biodefense; CCHI) to Czinn SJ and Blanchard TG, and by NIDDK, NIH, DHHS grant R01 DK 46461 to Czinn SJ. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Allergy and Infectious Diseases or the National Institutes of Health.

10.4161/gmic.26943

Disclosure of Potential Conflicts of Interest

No potential conflict of interest was disclosed.