Human Lyme Disease Vaccines: Past and Future Concerns

Dean T. Nardelli; Erik L. Munson; Steven M. Callister; Ronald F. Schell


Future Microbiol. 2009;4(4):457-469. 

In This Article

Concerns With the OspA Vaccine

Immunogenicity & Maintenance of Protection

An OspA vaccine may be effective at preventing the pathologic manifestations of infection with B. burgdorferi in humans. However, concerns about the vaccine's bactericidal immunogenicity and duration of protection were only partially addressed in the Phase III OspA vaccine trials.[28,32] Maintenance of high titers of protective borreliacidal antibody is vital for OspA vaccine efficacy.[36] A second booster of the lipidated, recombinant OspA vaccine, administered 12 months following initial vaccination, yielded significantly higher titers, which remained elevated 8 months later.[28] However, while the presence of these antibodies was established, their biological action against B. burgdorferi was not directly assessed. Indeed, the ability of antibodies to kill or inhibit growth of B. burgdorferi may be the only factor by which to determine the protective potential of a vaccine. The authors acknowledged that further studies were necessary to determine whether these antibodies were protective; moreover, they stated the necessity to further characterize the potential duration of protection.[28]

Similarly, an investigation of sera from OspA vaccine recipients enrolled in Sigal et al.'s study showed that although the immune sera inhibited growth of B. burgdorferi, its ability to do so waned below presumed protective levels prior to the second booster.[37] Following administration of the second booster, the ability of immune sera to inhibit borrelial growth increased significantly; however, the serum activity was assessed 4-6 weeks after this booster, with no further characterization of antibody activity provided.[37] Therefore, the duration of these antibody titers and the requirement for subsequent boosters were not established. Determining these factors is vital for the success of a future Lyme disease vaccine.

Other evidence suggests that the antibody-mediated protection, induced by an OspA vaccine, may be insufficient or too short-lived to be effective. An early recombinant OspA vaccine trial in humans assessed the ability of vaccine-induced antibodies to prevent borrelial growth.[38] Individuals were administered an OspA vaccine, with or without adjuvant, and a corresponding booster 1 month later. Total serum OspA IgG antibodies increased in the weeks following the booster but waned significantly 5 months later. More importantly, the borrelial growth inhibition titers reflected this trend. The ability of immune sera (from recipients of either vaccine preparation) to inhibit growth of B. burgdorferi was indistinguishable from baseline titers 5 months after administration of the booster. In addition, recipients of the nonadjuvanted vaccine were given a second booster 5 months after the first booster. The growth-inhibiting titer of serum drawn 1 month after this second booster was even lower than that observed 2 weeks after the first booster.[38] This study demonstrates that the protection against infection provided by OspA vaccination may be fleeting. Only a narrow window of effectiveness was induced by repeated vaccination with OspA.

The observation of short-lived, vaccine-induced OspA antibodies able to inhibit borrelial growth was supported by Padilla and colleagues.[36] An OspA vaccine comparable to that used by Sigal et al. was administered to humans in order to determine the duration of induced protective borreliacidal antibodies.[32] Participants were administered a single booster 1 month after initial vaccination, and the levels of total OspA antibodies and protective borreliacidal anti-OspA antibodies were assessed.[36] The production of total OspA antibodies peaked approximately 2 months after the booster before declining. However, the borreliacidal OspA antibody titer peaked at low levels only 1 month after the booster and waned rapidly.[36] Similar results were observed in hamsters vaccinated with a higher recombinant OspA vaccine dosage. Total OspA antibodies peaked at a significant titer 2 weeks following the booster, rapidly waned and gradually recovered approximately 6 months later. Similarly, borreliacidal OspA antibodies peaked (at much lower titers) 2 weeks following the booster and rapidly declined.[36] There was no recovery of borreliacidal OspA antibodies, suggesting that the protection afforded by these antibodies was short-lived. Although this study investigated the effects of two vaccine injections, the study by Keller et al. demonstrated that administration of an additional booster still did not augment protection.[38] Obviously, a major concern was the number of vaccinations required to generate, at best, only modest protection. Vaccinated persons would be at risk of infection during the vaccination administration period. Furthermore, completion of the course of vaccination did not ensure that recipients had acquired long-term protection from infection.

Tick Anticomplement Factors

In addition to providing sufficient and sustained protection, a vaccine must induce biological activities in the host that overcome factors exploited by the pathogen to cause disease. The reported effectiveness of the OspA vaccine was attributed primarily to the ability of induced antibodies to enter the tick during a blood meal and eliminate the spirochetes before their passage into the mammalian host.[21] Indeed, infected ticks feeding on OspA-immunized mice were shown to be sterilized as a result of ingesting anti-B. burgdorferi antibodies.[21,39,40] This mechanism requires such antibody-mediated killing to occur independently of complement, as factors in tick saliva have anticomplement properties.[41,42] In addition, Borrelia organisms themselves may be able to directly inactivate the effects of complement.[43,44,45] However, Schmitz et al. established a role for complement-dependent and complement-independent antibodies in the protection against infection with B. burgdorferi.[46] In addition, Munson et al. showed that antibodies directed against OspA can dispose of B. burgdorferi by either mechanism.[47] In support of this hypothesis, Rathinavelu et al. showed that vaccine-induced, OspA-specific antibodies in mice kill B. burgdorferi in feeding ticks regardless of the presence of host complement.[48] A suggested mechanism for complement-independent, antibody-mediated killing of B. burgdorferi is through opsonization and phagocytosis of the spirochetes.[46] However, phagocytosis does not occur in infected ticks. Complement-independent antibody may aggregate the Borrelia organisms, and this prevents passage of the spirochetes to the mammalian host. By contrast, in vitro studies of vaccine-induced, OspA-specific antibodies in serum of rhesus monkeys showed that the presence of an intact complement system greatly enhances spirochete killing.[49,50] In fact, complement-dependent borreliacidal antibodies are the major defense mechanism that eliminates B. burgdorferi from the host and ticks. This raises a major concern about OspA vaccination. A patient history of OspA vaccination would confound diagnosis of B. burgdorferi infection should anticomplement factors cause the vaccine's protective mechanisms to fail, as these OspA-vaccinated individuals would be seropositive for OspA.

Downregulation of Target Antigen

Assuming that protective antibodies, induced upon vaccination with OspA, are able to breach the anticomplement properties of tick saliva, a question remains as to whether the target antigen on B. burgdorferi is available for the antibodies to act upon within the tick midgut. While OspA is highly expressed by B. burgdorferi residing in the midgut of infected Ixodes ticks, it is down-regulated in the midgut of ticks ingesting a blood meal.[51] It is important to note that prior to the Phase III OspA vaccine trials,[28,32] several studies demonstrated that animals develop low OspA antibody responses following infection with B. burgdorferi. Mice injected with B. burgdorferi organisms displayed a low level of serum IgM reactivity and no IgG reactivity to OspA in the week following infection.[52] Golde et al. extended these findings to show that while OspA antibodies were detected in mice following needle challenge, those infected via a tick bite produced low OspA antibody titers up to 60 days after tick exposure.[53] Similar results were reported in animals[54,55] at significantly longer periods following infection. In addition, canines naturally-exposed to infected ticks showed no consistent OspA antibody reaction.[56] This suggested that B. burgdorferi expressed OspA at low levels or for a short duration during tick attachment and feeding on the mammalian host.

Similar findings were shown in humans infected with B. burgdorferi. Although antibodies to OspA, either free or complexed to the spirochete, were observed in patients with erythema migrans,[57] several investigations indicate that OspA is not a significant target of the human immune response following infection.[58,59,60,61] Indeed, the frequency of serum reactivity to OspA even in patients with Lyme arthritis or late-phase neurological abnormalities was typically less than that to other borrelial antigens.[59] Taken together, these findings suggest that a Lyme disease vaccine consisting solely of OspA may be ineffective, and that additional components of B. burgdorferi should be included in a comprehensive vaccine. Most importantly, the downregulation of the vaccine's target antigen, OspA, in the ticks during feeding was a major drawback of the OspA vaccine. In support of this assertion, rabbits immunized with OspA were not protected from infection following transplantation of skin from erythema migrans lesions,[62] highlighting the importance of considering antigens on B. burgdorferi that are induced, rather than downregulated, in the host. These observations alone should have precipitously decreased the value of a solely OspA-based vaccine. Although OspA is gradually downregulated during tick feeding, Battisti et al. showed that OspA can block other surface proteins that can be used as a vaccine to generate protective antibodies.[63] This highlights the importance of retaining OspA as a component for future combination vaccines.

Potential for Adverse Effects

A paramount concern in the creation of a vaccine is that it does not induce harmful side effects, despite its ability to prevent disease. Indeed, ample evidence in animal models and humans demonstrates that vaccination with OspA provides at least short-term antibody-mediated protection against infection with B. burgdorferi. However, studies in animals and humans conducted prior to and after the Phase III OspA human vaccine trials showed that the immune response to OspA may come with the caveat of autoreactivity against the recipient. Recently, Steere and Glickstein hypothesized that structural homology between host peptides and OspA induces an autoimmune mechanism, which mediates antibiotic treatment-resistant Lyme arthritis.[64] In addition, reports of Lyme disease symptoms[65,66] in recipients of OspA vaccines raised questions about the safety of such a vaccine. Concerns about these potential side effects may have played a factor in the withdrawal of the Lyme disease vaccine from the market.

Although the humoral response to OspA appears to be effective in the clearance of B. burgdorferi infection, several studies have linked the development of OspA-specific antibodies to the manifestation of late-stage Lyme disease symptoms - particularly chronic arthritis. Steere et al. demonstrated a genetic predisposition to the development of chronic, antibiotic treatment-resistant Lyme arthritis.[67] A vast majority of patients with chronic arthritis were shown to possess either, or both, human leukocyte antigen (HLA)-DR4 or -DR2 alleles, whereas individuals with short- or moderate-duration arthritis were less likely to possess these alleles.[67] The onset of prolonged bouts of arthritis in HLA-DR4-positive individuals coincided with development of strong IgG responses to OspA months to years after manifestation of initial Lyme disease symptoms.[68] In addition, most patients with chronic arthritis developed a strong IgG response to a C-terminal OspA epitope, and this anti-OspA antibody response reflected the duration of arthritis.[58] These findings were supported by Akin et al., who showed that the production of anti-OspA IgG antibodies correlated with the duration and severity of chronic Lyme arthritis, while the IgG responses to most other B. burgdorferi proteins were indicative of early-stage disease manifestations.[69] These studies indicate that the development of chronic Lyme arthritis is related to immune recognition of OspA months after the onset of initial disease symptoms. The correlation between late-stage OspA reactivity and arthritis development suggests epitopes of OspA may induce an adverse response in humans.

This assertion is supported by the association of OspA-reactive T cells in the development of chronic Lyme arthritis. The T-cell response to B. burgdorferi polypeptides in patients with chronic Lyme arthritis was stronger than that of infected patients with transient arthritis,[70] and T-cell clones derived from patients with Lyme arthritis were shown to bind to OspA epitopes.[71,72,73] In addition, T cells from the synovial fluid or peripheral blood of patients with antibiotic treatment-resistant Lyme arthritis were able to preferentially recognize OspA, whereas those of individuals with treatment-responsive arthritis were not.[74] OspA-reactive T cells localized in the joints of patients with chronic Lyme arthritis were subsequently identified as Th1 cells.[75] However, despite an abundance of circumstantial evidence for OspA-induced T-cell autoreactivity in the development of chronic Lyme arthritis, a host peptide with homology to any borrelial antigen had yet to be indentified.

Various HLA-DRB alleles associated with rheumatoid arthritis[76] are frequently found in individuals with Lyme arthritis several years after receiving antibiotic treatment.[77,78] Using a transgenic mouse model expressing a human major histocompatibility complex class II HLA-DRB allele, an immunodominant OspA epitope (OspA165-173) was identified.[79] The structure of this OspA epitope was shown to be homologous to a sequence of human leukocyte function-associated antigen-1 (hLFA-1),[79] providing the first evidence for possible autoimmune-inducing molecular mimicry between borrelial and human peptides. Subsequently, certain HLA-DRB molecules of the DR4 subtype were shown to bind OspA165-173 and its hLFA-1 homolog,[77] and an association between OspA165-173-binding HLA-DR molecules and antibiotic treatment-resistant Lyme arthritis was postulated.[78] By contrast, LFA-1 was shown to be a weak agonist for T cells reactive against OspA165-173, making it an unlikely source of autoimmune-inducing molecular mimicry.[80] Nonetheless, the hypothesis of arthritis-inducing molecular mimicry to host tissue is intriguing.[64] In addition, reports of adverse effects in humans after OspA vaccination raise a red flag that OspA is involved in the induction of those effects.[65,66] Moreover, nearly 1000 reports of adverse effects following OspA vaccination were documented by the US FDA less than 2 years after the vaccine was placed on the market.[81] Although most of these cases were not reported as published case studies, these events of adverse effects contributed to the withdrawal of the vaccine for use in humans.

In addition, the idea that vaccination with B. burgdorferi or its components may carry the risk of direct or indirect arthritic side effects was known prior to the FDA-approved human OspA vaccine field trials. Lim et al. showed that hamsters vaccinated with whole Borrelia organisms developed a severe, destructive osteoarthropathy following infection with a heterologous borrelial strain.[82] Arthritis in these vaccinated hamsters was mediated by Borrelia-specific T cells.[83] In addition, severe, destructive arthritis developed following heterologous infection of Borrelia-vaccinated mice.[84,85,86,87] Most importantly, Croke et al. showed that vaccination with recombinant OspA also primed animals for arthritis development following heterologous challenge.[88] In addition, OspA vaccination has also been linked to development of arthritis and neurological disease in humans.[65,66] Collectively, these findings represent various lines of evidence of OspA-induced adverse effects in animals and humans. Interestingly, these adverse effects were ignored since they may not have reflected the preliminary safety data collected in humans by Keller et al.[38] These adverse reports highlight the necessity for evaluating all evidence on the safety of prospective Lyme vaccines, particularly those composed of OspA.

New approaches for creating safe and effective OspA vaccines for Lyme disease are in development. A modified OspA-based vaccine showed potential for protection against infection without induction of autoreactive T cells.[89] An epitope on OspA is reportedly able to induce autoreactive T cells owing to its significant structural homology to that of hLFA-1.[79] These investigators mutated the potentially cross-reactive OspA epitope while maintaining the structure of an epitope reported to bind protective antibodies.[89] Mice vaccinated with the modified OspA were protected from needle challenge with B. burgdorferi, while OspA-reactive human T cells did not respond to the mutated OspA protein.[89] Although more work is required to characterize the vaccine-induced immune response and protection afforded against tick-borne infection, removal of this potentially autoreactive epitope may be a key requirement for any future Lyme disease vaccine containing OspA. However, this vaccine still does not overcome the fact that OspA is downregulated in the feeding tick.