Panel on Caries Vaccine

January 28, 2003

Sponsored by: The National Institute of Dental and Craniofacial Research
National Institutes of Health

I. Background:

The National Institute of Dental and Craniofacial Research (NIDCR) is the primary sponsor of research and research training in oral, dental and craniofacial diseases. It fulfills its mission through both intramural research programs and the support of extramural investigators and mentors in academic institutions and other research organizations. Its research support portfolio includes a strong component in dental caries and oral infectious conditions. Part of this program involves support of research on the development of a caries vaccine. As there are a number of scientific, economic and ethical issues in this activity, the Institute Director asked that an Expert Panel be convened to help the Institute explore these issues and make recommendations on how to address them and delineate the manner in which NIDCR can move research in this area forward.

The Panel was convened on January 28, 2003 and this represents a summary report of its deliberations. The agenda included presentations by groups currently doing research in caries vaccine development and an extended discussion by the Panel of the various issues outlined above.

The Institute Director opened the meeting by thanking the participants for their contributions to the Panel. He indicated that NIDCR has supported research related to the development of a caries vaccine for several years and that it is at a point where decisions have to be made about the support of clinical trials in this area. The Institute will continue to support fundamental research in immunobiology but has also reinvigorated its efforts in the support of clinical research, particularly clinical trials. Caries vaccine trials would be an obvious candidate within this effort but this comes with a series of questions from the scientific, ethical and economic angles. That is why the Panel includes a number of content experts with experience in vaccines from all those different angles. The recommendations from the Panel will be presented to the Institute’s National Advisory Council in June.

SCIENTIFIC PRESENTATIONS

There were five presentations from various research groups throughout the country. Dr. Martin Taubman from the Forsyth Institute indicated that dental caries is a major health problem in the U.S. Eighty percent of Americans have dental disease by 18 years or age and two-thirds of disadvantaged children have untreated decay, so there is a lot of unmet needs. The problem is worldwide and numbers on prevalence are high in countries like China, Japan and Brazil. The disease is caused by a group of organisms called Streptococcus mutans (S. mutans) and occurs in 3 phases: an initial interaction with the tooth surface mediated by adhesins; the accumulation of the bacteria in a biofilm and the production of glucose and glucans by the bacterial enzyme glucosyl transferase; and the formation of lactic acid.

The target for vaccine development in his research group has been the glucosyl transferases (GTF) and the glucan binding proteins (GbP). The basic hypothesis is that mucosal induction of salivary IG antibody to GTF interferes with the accumulation of S. mutans in hard surfaces like glass and teeth. The enzyme is inhibited by this antibody both in vitro and in vivo. Their research has extended to sub unit vaccines, delivery systems, mucosal adjuvants and various routes of application. They have synthesized a variety of peptides from the catalytic domain and from the glucan-binding domain and prepared mono and diepitopic constructs. A peptide from GbpB has been shown to protect against caries.

The group has utilized several delivery methods, including subcutaneous injection, intra gastric intubation, oral capsules and topical application. More recently. they have used intranasal administration in aluminum phosphate or PLGA micro particles. They have also utilized attenuated salmonella typhus expressing the glucapeptide.

The primary vaccine is the GTF from Streptococcus sobrinus, but other materials have also been tested. Two clinical trials were conducted in adolescents 10 years ago, and they now propose a pediatric vaccine in a target population of children approximately one year of age. In the phase I trial they measured safety, antibody activity and the reaccummulation of indigenous mutans species. The data indicated significant reductions in S. mutans and antibody increases after 42 days in individuals in the GTF group.

At birth, there are no bacteria in the oral cavity, but there is subsequent colonization with certain “pioneer” oral bacteria that come essentially from the mother, including S. midus and S. salivarius. Teeth emerge at about 6 months and certain teeth-associated flora begins to be found, such as S. sanguis. S. mutans is not present (in a certain portion of individuals) until between 18 and 36 months of age. IG levels also rise and by 12 months children have adult levels of IG and of antibody to certain antigens. At 5 weeks of age one can find antibody salivary IgA to S. midus antigen in saliva. When infected, children can respond to a variety of antigens associated with microorganisms.

The group is convinced of the safety of the GTF, as exposure to multiple doses of orally and topically applied GTF did not cause any health problems. GTF is the only streptococcal component in vaccine materials. The plan is to use GTF as the antigen, administered with PLGA micro particles by the intranasal route, 2-3 doses between 12 and 24 months of age. This would be a phase I trial. The cost of the trial would be between 2 and 3 million dollars.

In response to questions, Dr. Taubman indicated that there is no precise correlation data, but that a general correlation exists between IgA levels and the reduction in bacterial titers. The dose of GTF used in the previous trials was approximately 10 times the amount being normally swallowed and there were no theoretical basis to anticipate untoward effects. There is no accurate data on the proportion of children under the poverty level who become colonized with S. mutans but in studies done in Alabama and Brazil, the evidence suggests that the dose increases the risk of colonization and that there is a direct correspondence between the level of infection and the caries. There are other acidogenic bacteria that can cause caries, but S. mutans is the most heavily implicated and solving the mutans problem would solve 90 percent of the caries problem. The end point would be safety, plus some type of surrogate outcome, such as colonization and antibody efficacy.

Dr. Noel Childers from the University of Alabama indicated that the pathogenesis of dental caries is complex, but that the idea is that, if you can prevent the initial colonization, this will have an effect on the disease process. Their studies have focused on the two antigens already mentioned. The first is that involved in the initial attachment (antigen I-II) and the second, the one associated with the more tenacious attachment mediated by GTF. Preclinical studies in rodents have shown that oral or nasal immunization with bacterial lysates, purified antigen I-II, GTF, combinations of antigen and GTF, and recombinant GLU can confer protection. Antigen alone was poorly immunogenic and modifications of antigen preparations with adjuvant or using other delivery systems have been necessary to improve immunogenicity. One such system involves the use of biologically safe liposomes of 100 nm diameter. This preparation has been used in phase I trials with small numbers of subjects to determine safety and immunogenicity. Seven studies have been conducted, three involving oral immunization and four intranasal applications. Volunteers ingested enteric-coated capsules containing 500 µg of antigen or placebo and the same protocol was repeated after 28 days. Follow up was for 8 weeks to monitor changes in antibody activity. There was a 50 percent increase in IgA response in the experimental group, occurring between day 28 and day 35. In the intranasal immunization, individuals received 250 µg of the same antigen either as free antigen or in liposomes; twice, seven days apart. The response in the nasal wash secretions occurred in both groups and began about day 14 and peaked around day 21-28. It also appeared to persist longer. The salivary IgA response was no different in the 2 groups, where a 75 percent increase occurred on day 21.

Oral, nasal and tonsilar administration of the liposomal antigen was found to be safe. The nasal spray vaccine induced the best specific mucosal IgA responses and these appeared to be dose-specific. The IgA response may be associated with a delay in recolonization. Their next step is to carry out a phase I trial in pre-adolescents and the ultimate goal is to conduct phase I, II and III studies in an infant population. The idea is to immunize them prior to the eruption of the teeth, as there may be multiple windows of infectivity as specific teeth erupt.

The prospect of industrial support is unclear because of patentability issues and because these would be long-term studies that extend past a patent time frame.

In response to questions from the Panel, Dr. Childers indicated that different immunization protocols may be used in the phase II trial in infants. Issues of the safety and efficacy of the nasal spray system were discussed and a study will be performed in mice to clarify these issues. The vaccine would be administered between 12 and 16 months of age in the infant trial being planned. There is conflicting information on the benefits of breast-feeding in preventing dental caries.

Dr. Michael Russell, SUNY at Buffalo, indicated his work has focused on the antigen I-II. The antigen has been cloned and sequenced and occurs only in S. mutans and not in other cariogenic bacteria, except S. sobrinus. It is a large protein on the surface of the mutans and has a characteristic gram-positive wall anchor and involves the cleavage and insertion of the c-terminal tail into the cell wall. His research has focused on the saliva-binding region where certain residues appear to be important in attachment to the salivary pellicle tooth surface. Antibodies against this part of the molecule can exert an anti adherence function. Antibodies against antigen I-II, are effective anti adherence antibodies. There is no evidence that antigen I-II has heart cross reactivity.

Interest is now focused on developing mucosal vaccines exploiting the immunogenicity of cholera toxin and its B subunit. Potent immune responses have been found in monkeys and rats by coupling the antigen to cholera toxin. The same is true in the rat model when the saliva-binding region is genetically coupled to the non-toxic components of antigen I-II. Mice immunized intragastrically with antigen coupled to the B subunit of cholera toxin maintain memory of the event. Antigen has been chemically conjugated to cholera toxin B and so-called chimeric antigen A1-B5 immunogens have been subsequently developed. A similar approach using segments of GTF has proven more difficult because of the way that E. coli processes these segments. Trials need to be performed to demonstrate that chimeric immunogens are effective in humans. They can be incorporated into liposomes or other micro particles. These chimeric antigens skirt the toxicity issue associated with enterotoxins as adjunvants or delivery vehicles.

The issue of a correlation between antibodies to antigen I-II and a decrease in S. mutans colonization was again brought up. No longitudinal study from infancy onward has been done that would clearly answer this question. Saliva is contaminated with other bacteria that have cross-reactive antigens and it is very difficult to show that kind of correlation. A small-scale human trial to demonstrate that immunogens coupled to CTB induce a potent response needs to be done as a prelude to more extensive trials.

Dr. Debra Trantolo from Cambridge Scientific, Boston, spoke about her work in developing a delivery system for GTF through a phase I-phase II SBIR grant. The issues being faced are similar to those in the development of any vaccine. The delivery system uses the biopolymer polyactide (PLGA), which is a bioabsorable substance used in sutures and in drug delivery. It is used as an adjuvant and as something that provides some release characteristics. A lot of preclinical work has been done with this delivery system. Because they are looking at mucosal delivery, the other piece of the system is the incorporation of a bioadhesive. The three components of the delivery system are compatible for the FDA clearance process and the Company has a patent on the system.

The system is called a matrix system and is a non-encapsulation system where the biological or drug is dispersed throughout the polymer. There are no organic solvents used in the manufacturing process. An aqueous solution of the biological is sucked under vacuum into the polymer foam, which is then lyophilized and compressed to yield a spaghetti-like rod that can be ground into a particulate for suspension. There are several issues in terms of how much of biological and of the bioadhesive to use and also manufacturing considerations like pressure, temperature, etc. SOPs have to be prepared for the combination of the polymer with GTF as part of a technology transfer document for INDs.

In terms of funding, the SOPs and putting the material together would add another million dollars to the estimate given by Forsyth. Another piece is corporate partners to help in the manufacturing of the product.

In response to questions, Dr. Trantolo indicated that there is no major additional cost in scaling the technology up and that there is no definitive information about the long-term stability of the product. There has been no interaction with the pharmaceutical industry so far in terms of a large scaling-up of the process. There have been some clinical trials using polyactide micro particles in a matrix material with disappointing results. This may be related to adhesive issues. A better uptake is seen in the oral and nasal applications in the presence of an adhesive. Another issue may be particle size.

Dr. James Larrick from Planet Biotechnology, California, explained their work in developing secretory IgA antibodies in a product called Cario Rx. The Company has focused on manufacturing monoclonals in plants that can make large amounts of IgA. The antibodies block adherence of microbes and neutralize toxins and viruses. IgA has more stability on the tooth surface and in the GI tract and is used to produce large amounts of antibody in tobacco plants. The extraction process has been worked out and yields products of high purity in a relatively short period of time.

Cario Rx is a nominal therapy to reduce the adherence of S. mutans to teeth and it’s devoid of any adverse effect. The hypothesis is that in an altered biofilm, the antibody blocks the repopulation dynamics of S. mutans. There were animal studies that supported the clinical development of these SA I-II antibodies. There was a reduction in the recolonization of Strep mutans and protection against caries. Elimination of caries will likely be required as the end point by the FDA in this type of passive immuno-therapy. There have been about 100 patients in phase 1-phase 2 trials and the data indicate that passive immunotherapy can eliminate or significantly reduce S. mutans.

The first trial was with antibodies that blocked SA I-II biding. One day after a regimen of carsodyl or gel and mouthwash, the reagent was applied directly to the teeth. The antibody prevented recolonization for up to a year. Other trials have shown that some antibodies are not effective and that multivalency was required to prevent recolonization. Another study has shown that combinations of antibodies can be effective against S. mutans but not against actinomyces (species specificity). A confirmatory trial used a different protocol for the initial antisepsis and the outcome was not uniform, as 35-40 percent of subjects did not clear the S. mutans. Still, there was a large portion of subjects that did not recolonize. The antisepsis will be standardized and optimized in the future to completely eliminate the biofilm.

In response to questions, Dr. Larrick indicated that the purity of the final antibody is 98-99 percent and that the antibody cannot be detected after 2 months. The approach does not work if the strep already there is not eliminated. Theoretically, the antibody can be put into something like a formula. The plan is to start with adolescents and then move down to infants.

GENERAL DISCUSSION

The general discussion following these presentations focused on the following topics:

1. The issue of safety and how to evaluate it. Do small children have different safety issues or different side effects than older children and adults because of the extent of the development of their immune system? There is not a lot of data, but it is known that a detectable mucosal immune response can be seen within 2 weeks of oral colonization. Mucosal antibody to S. mutans develops within 2 months. Immune response in children is less active than in adults and intense untoward reactions would be less likely.

   There is public concern about the number of vaccines that infants/children currently receive. The assurance of safety is paramount. Any vaccine that is targeted for use in children will need to take into account this factor, as well as the impact that such a vaccine will have on other routinely administered vaccines.

2. The lack of colonization in a subset of the infant population. Is the observation that children who do not become colonized with S. mutans the result of the biofilm they develop preventing the colonization or are there other factors (i.e., genetic factors)?
   
   There is no answer to this question. Children, who are heavily colonized, have more than one genotype, so the lack of colonization could be related to many factors. There are clearly different immune responses among siblings and some respond more favorably to GTF. Adopted children do not get the adoptive parents’ S. mutans. Then there is a second window of infectivity when the permanent teeth erupt and everybody is colonized.

3. The burden of tooth decay and its effect on the quality of life.
   
   There is some idea on the cost of treatment, but there is really no accurate information on the real burden (i.e., lost school or work days) or the cost in terms of pain and suffering. A large portion of the economic burden comes from the latter. Only 25 percent of children eligible for Medicaid are able to get care in any given year and probably even less in the pre-school population. The burden will not be fixed by the care delivery system. Medicaid mandates that all children have preventive and corrective dental services, but the reality is that they do not get the necessary care.

4. The lack of longitudinal studies that identify risk factors for colonization and outcomes.
   
   One risk factor (sucrose in the diet) is well known and it is also true there is probably no correlation between antibody levels and infectivity.

5. The issue of economic costs.
   
   The last IOM report on prioritizing vaccine development on the basis of economics does not highlight caries. It will require lots of money to reach licensed products. Children below the poverty line are the big problem and they also have lower access to vaccines. A vaccine or vaccines would need to be compared in economic terms to other approaches, such as giving vouchers to the children to visit dentists. One of the barriers is that reimbursement is not sufficient. The other issue is the parental sense of the importance of visiting the dentist. There is no evidence that a system of annual visits will be very helpful in this regard.

6. The question of the role of fluoride. Is fluoride a solution or just an ancillary approach?
   
   The issue is actually access to fluoride. We don’t know what the true behavior is (as opposed to that described in answers to surveys and questionnaires) in terms of use of fluoride. There is a subset of children who experience high rates of tooth decay even though they live in fluoridated communities. Fluoride cannot be, therefore, the sole solution.

RECOMMENDATIONS

The Director of NIDCR reviewed briefly the major characteristics of dental caries and its links to behavior and to access issues. The clinical research portfolio of NIDCR is mainly in caries and periodontal diseases, but phase III trials have never been done in caries. The treatment of caries has not changed significantly over the years, although new chemotherapeutic agents and efforts to disturb signaling in the plaque are now of mounting interest. There is a lack of longitudinal studies and the need to clearly identify those who are most at risk.

Panel members then brought up several general issues in vaccine development:

1. Elements in successful vaccine development. A good reference is the CDC’s retrospective study on what was involved in creating a successful vaccine.
2. The economic/risk benefit issue. Where is the product going to be utilized? The rationale is that introduction of a vaccine in a community will be cost-saving. If the regimen requires too many contacts with the health care provider, this may make the vaccine unfeasible.
3. Industry partnerships. From the point of view of industry, the market will drive development and ACIP recommendation for routine use in everybody is a plus. An industrial partner is needed beyond phase II, but here may be limited partner potential in terms of pilot lots, clinical trial development and manufacturing. An approach is to develop a solid science foundation, a system for the economic surveillance of the epidemiological data and a complete infrastructure (including animals models and assay development and standardization) and then bring in the industrial partner. Another approach is to target small industry, particularly small biotech companies and to provide support through an SBIR award to do some initial production and carry the research through phase I and II trials.
4. The way a caries vaccine may fit into the existing model of dental care (self pay mode) and the need to come up with a relatively inexpensive, but efficient delivery model for a vaccine.

After extensive additional discussion, the Panel made the following broad recommendations:

1. There is intrinsic value in learning more about the science in terms of the mucosal immune system and NIDCR should continue to support basic research in immunobiology.
2. Real world barriers have to be considered and overcome if starting from the premise that a product will be delivered. Perhaps, NIDCR should frame the goal differently and provide guidance to the community. The approach can be to only reach to proof of principle in phase III trials.
3. There might be some intrinsic advantage to a passive immunity approach, both in terms of cost and of acceptance.
4. There is definitely a need for more longitudinal epidemiology correlates. This can be achieved through a “center’ where expert consultants can work with the core staff in addressing the various problems.
5. Advantage should be taken of natural experiments, especially children who are not colonized despite significant exposure. More research is needed on possible differences in innate (i.e., saliva) factors and on longitudinal follow-ups of how the oral environment changes.