In the United States and other countries, leptospiral vaccines are tested in hamsters before being used in domesticated species. Drawing on this experience, Johnson and coworkers [2] did seminal studies in 1986 showing that Syrian hamsters could be successfully immunized against B. burgdorferi using an inactivated whole-cell vaccine without adjuvant. In 1990, a whole-cell, chemically inactivated B. burgdorferi vaccine formulation with a proprietary polymer-based adjuvant (Fort Dodge Laboratories, Fort Dodge, Iowa) was conditionally licensed by the U.S. Department of Agriculture for use in dogs. This vaccine was fully licensed in 1992, but little information on its effectiveness and safety has been published [3].

There has been little enthusiasm for developing an inactivated whole-cell vaccine for humans, primarily because of theoretical concerns that such a vaccine may be too "reactogenic," that is, potentially inducing immunologic responses to cross-reactive human antigens [4]. Subunit vaccines consisting of single recombinant proteins are considered to have the greatest promise for preventing Lyme disease in humans. In animal experimentation, purified recombinant proteins, particularly certain outer-surface proteins (Osp) of B. burgdorferi such as OspA, B, or C are highly protective [5-7].

The most extensively studied of these preparations is OspA. Active immunization studies with OspA immunogens have shown high-level protection in mice [5, 6]. In these studies, mice vaccinated with OspA preparations were protected against clinical disease and tissue damage after either syringe or tick challenge with B. burgdorferi. Humoral immunity was sufficient for protection. Although binding to an antigenic determinant (conformational epitope) located within the carboxyl end of OspA appears to be important for protection [8], the functional activity through which protection is conferred by antibodies has not been established. In other experimental systems, antibody-induced protection has correlated directly with the borreliacidal antibody titer [9, 10].

Considering the well-recognized feature of latency in human B. burgdorferi infections, surprisingly little attention has been paid to the ability of these experimental vaccines to prevent asymptomatic infection (recognized by the development of antibodies to borrelial antigens other than OspA) [7]. Although it is unknown whether a vaccine candidate should be held to a standard that would disallow asymptomatic infection, there is precedent for this degree of efficacy with other human vaccines [11].

On the basis of the observation that B. burgdorferi is cleared from infected ticks after they ingest blood from immunized mice, researchers have hypothesized that an OspA Lyme vaccine has a dual mode of action [6, 12, 13]. Transmission of the spirochete from an attached and feeding tick to the host is delayed by at least 36 to 48 hours [14]. This delay may allow sufficient time for serum antibodies from the blood of the host (possibly in concert with other factors such as complement) to interact with spirochetes while still in the tick; transmission of B. burgdorferi would thereby be reduced or eliminated. This distinctive second mode of action implies that the vaccine-induced immune response might begin to take effect well before the spirochete enters the host.

Other vaccine approaches are being studied [4]. For example, the BCG vaccine has been considered a potential vehicle for delivering antigens for immunization against various diseases, including Lyme disease. In one study, a recombinant BCG vaccine expressing OspA lipoprotein effectively protected mice against intradermal challenge with B. burgdorferi [15].

A potential drawback to OspA vaccine candidates is the heterogeneity of these proteins among isolates of B. burgdorferi in nature. In mice, the OspA immunogens are highly protective when the challenge isolate is either identical or closely related to the one from which the OspA protein was derived. Protection against syringe challenge with other more diverse isolates is minimal or nonexistent [16]. European OspA strains are more diverse [17] than those from the United States. The homogeneity of the U.S. strains has led to optimism that only a monovalent OspA vaccine (that is, a single OspA antigen) would be required in this country. Furthermore, diversity of strains may be less important when the organism is introduced by ticks, possibly because of the antispirochetal actions of blood in the tick or because of a lower inoculum of B. burgdorferi [18].

In the United States, two single-protein recombinant OspA preparations, one manufactured by Connaught Laboratories, Inc., and the other by SmithKline Beecham Biologicals, are undergoing phase III testing for efficacy and safety. Both vaccine preparations are immunogenic and can elicit a borreliacidal or growth-inhibiting antibody response in humans [19, 20]. The most common adverse reactions in phase I and II studies have been pain and tenderness at the injection site.

In natural infection in humans, high titers of IgG OspA antibody are usually detectable only in patients with late Lyme disease; for example, the antibody is present in approximately 70% of patients with Lyme arthritis [21]. Because of this association, researchers have suggested that an OspA immune response (humoral or cellular) may play an etiologic role in the pathogenesis of arthritis. If so, a vigorous response to an OspA vaccine might have the undesirable effect of causing synovial inflammation in the vaccine recipient. Based on the preliminary results of previous studies, this concern has not been proved, even in persons who have previously had Lyme disease or in those who carry the HLA DR4 allele, the genetic marker that has been associated with cases of chronic Lyme arthritis that are refractory to antibiotic agents (19, 22, 23; unpublished data).

Persons who receive a recombinant OspA vaccine preparation may become seropositive when tested for antibodies to whole-cell B. burgdorferi by enzyme-linked immunosorbent assay. Vaccine recipients must therefore rely on immunoblotting (a technique that determines antibody reactivity to borrelial antigens other than OspA) when serologic testing is necessary to assist in establishing the clinical diagnosis of Lyme disease.

In conclusion, a vaccine approach using recombinant OspA preparations has been highly successful in protecting laboratory animals against challenge by strains of B. burgdorferi closely related to the one from which the OspA was derived. Humoral immunity has been shown to be sufficient for protection. The vaccine-induced immune response against natural infection introduced by ticks may begin to take effect in the tick itself, before the spirochete enters the host, and may extend to more diverse strains of B. burgdorferi. Single-protein recombinant OspA vaccine preparations are currently being evaluated in two large efficacy trials in adults in the United States.

Although all this seems promising, the vaccine is in an early stage of development, and the prospects for a safe, clinically effective vaccine for human use are still uncertain. Even if safety and efficacy are established, questions regarding duration of immunity, need for booster doses, and safety and immunogenicity in children must still be answered. Because European strains of B. burgdorferi strains are more diverse, a vaccine containing multiple antigens will probably be necessary in those countries; this is also a consideration in the United States.