Vaccine Efficacy and Controversies

Veterinarians need to think critically about vaccine recommendations and brand choices. Many factors should be weighed, including likelihood of exposure, and severity of each infectious disease, efficacy of each vaccine, potential side effects, and implications for future disease diagnosis. Data on these characteristics should be provided by the manufacturers for all new vaccines. Vaccine selections should be evaluated annually and reviewed with clients to gain maximum protective benefits.

1. Introduction

Veterinarians today are faced with a myriad of choices regarding vaccines to prevent equine infectious diseases. Clearcut data derived from well-designed clinical trials and readily available information regarding adverse reactions to vaccines are not common. Commercially available vaccine combinations are not always logical, because one component may be more frequently indicated than another. Pricing incentives often encourage veterinarians to stock products from a single pharmaceutical company rather than only stocking the most efficacious vaccine for a specific disease.

Adding to the quagmire of choices for veterinarians, increasing numbers of horse owners insist on making illogical vaccination decisions. Some are cognizant of the risks posed by vaccinations and may seek to minimize or eliminate vaccinations for their animals. A few ask for titers on which the practitioner should base their vaccination decisions. Owners may also make foolhardy decisions based on rumors or purchase their vaccines through non-veterinary outlets, which decreases the number of opportunities available for the veterinarian to educate them about the vaccinations. New guidelines for small-animal vaccinations have led clients as well as practitioners to question vaccination protocols advocated for horses [1]. These thought-provoking comparisons have led to similar strategies for developing equine vaccination recommendations.

The objectives of this paper are (1) to consider criteria that should be used to guide vaccine choices and client education, (2) to show the application of these criteria to important diseases for which vaccines are available, (3) to address specific vaccine controversies, (4) to encourage reporting of adverse reactions, and (5) to highlight the need for more prospective clinical trials of vaccine efficacy and safety.

2. Criteria for Vaccine Selection

In the United States, products are available and licensed for immunoprophylaxis of viral, bacterial, and protozoal equine infectious diseases. Viral vaccines are available to protect against Eastern equine encephalomyelitis (EEE), Western equine encephalomyelitis (WEE), Venezuelan equine encephalomyelitis (VEE), West Nile encephalomyelitis, influenza, equine herpes 1 (EHV1), equine herpes 4 (EHV4), rabies, rotavirus, and equine viral arteritis (EVA). Hyperimmune serum and plasma products are also on the market for West Nile virus. For protection from bacterial diseases, vaccines exist against anthrax, botulism, strangles, tetanus (toxoid), Neorickettsia risticii, and endotoxin (Salmonella derived). Antitoxin for tetanus and hyperimmune plasma or serum products for endotoxin, E. coli, and Rhodococcus equi are also available. To date, only a conditionally licensed vaccine is offered for prevention of equine protozoal myeloencephalitis (EPM). Vaccines licensed for other species also have been employed in an extralabel manner for Lyme disease, leptospirosis, and clostridial enteritis [2-4].

Risk Assessment

The first question a veterinarian should address in selecting vaccines for their practice is whether there is a risk of the infectious disease for which a vaccine is available. Tetanus is a good example of an infectious disease to which all horses are exposed and should be vaccinated. In contrast, VEE has not been present in the United States since the early 1970s, and therefore, it is an illogical choice for inclusion at present. Surveillance for VEE viral activity in Mexico should provide sufficient warning should an epidemic again develop and threaten horses in the southern United States. For other infectious diseases, regional risks may vary and should be considered. Strangles is very common in certain locales, such as the upper Midwestern states, but it is quite uncommon in other areas. Veterinarians devising vaccination programs for clients whose horses travel to other regions of the country should keep regional differences in mind. Similarly, increased susceptibility of neonates, weanlings, and geriatric horses with Cushing's-induced immunosuppression should influence vaccination choices for those groups.

The second issue to consider is the likelihood of exposure to an infectious disease for a specific horse or a specific facility. Annually, the farm-specific risks should be reviewed with the owner or manager. Fewer or less frequent vaccines may be indicated for a closed herd of horses as opposed to a farm where horses travel off the premises to venues where exposure to infectious agents may occur, either through direct aerosol transmission or environmental contamination. Facilities like racetracks, where a high risk of exposure exists, may institute specific vaccination requirements. Local prevalence of an infectious disease should also be evaluated, because the proximity of a closed herd to a higher-risk farm may raise concerns that aerosol or vector transmission is possible. A farm history of a specific disease, such as Potomac Horse Fever, equine protozoal myeloencephalitis, or rotavirus, may prompt a decision to vaccinate on just that farm. An individual horse with heaves that is fed haylage should be vaccinated against botulism. The broodmare, headed to Kentucky to be bred after foaling, might need rotavirus and botulism vaccines pre-partum to protect her foal.

The veterinarian is often asked to make blanket recommendations for a facility, which is challenging, particularly when multiple owners want to provide input. Perhaps equine practitioners can follow the lead of small-animal colleagues and use a form for each owner to fill out. This form assesses the risk of exposure before annual vaccinations (Fig. 1). Discussion of the identified risks gives the veterinarian an opportunity to explain why more than yearly vaccines may be warranted as well as to discuss other aspects of preventive medicine.

Figure 1. Form for vaccination risk assessment for horse owners to use before consultation with their veterinarian.

Vaccine Efficacy

The efficacy of vaccines available for specific diseases and the potential impact of the disease, should the horse develop it, should be weighed against cost and risk of adverse effects of the vaccine. For example, EHV4 infections seldom result in death, but an untimely infection in a show horse might preclude an important performance or athletic event. On the other hand, most killed vaccines for EHV4 offer only a reduction in severity of clinical signs. however, the diminution in viral shedding may be important for limiting spread within a herd or stable. This reasoning suggests that EHV4 vaccination is logical for open barns but is an optional vaccine for herds with little likelihood of exposure. In contrast, EEE kills ~90% of unvaccinated horses; thus, in areas at risk for EEE, this should be a core vaccine recommended for all horses and not be considered optional.

A minor point that veterinarians also need to consider is the potential future impact of seroconversion produced by vaccination. Horses that seroconvert to Sarcocystis neurona vaccine will have titers that are indistinguishable from natural exposure [5,6]. Horses, particularly stallions, that are vaccinated against EVA may face future export restrictions [7].

How can true efficacy be determined? Efficacy in the laboratory in an experimental model of a disease, as required for vaccine licensure, may not be comparable to naturally occurring disease. Laboratory trials for licensure of West Nile vaccines required only prevention of viremia. This can occur in the absence of clinical signs and can be produced by as few as 14 infected mosquitoes [8]. Hundreds of mosquitoes have been observed simultaneously feeding on horses that subsequently developed WNV in Minnesota. Surveillance on affected farms has shown high numbers of WNV positive mosquito pools in traps.

Experimental reproduction of clinical disease can provide valuable information on vaccine efficacy as well as vaccine comparisons that can guide practitioner's product recommendations. The influenza experimental challenge model developed by Chambers et al. 9], is an excellent example of a clinically applicable model. Viral exposure through nebulization reliably produced clinical illness in naïve ponies and allowed industry-independent determination of efficacy of a number of influenza vaccines [9,10]. Their data suggested good efficacy for both an intranasal flu vaccine (Flu Avert I.N. [a]) and an IM vaccine (Calvenza EIV [b]). This model has also allowed assessment of cross protection between influenza strains, because antigenic drift leads to new A2 strains.

Multicenter prospective clinical trials, as conducted for assessing vaccine efficacy and safety in humans, would be ideal, but these trials have seldom garnered significant funding from unbiased sources to provide equivalent data. More typically in equine medicine, field trials provide data to interpret vaccine safety and efficacy in specific locales [11]. These are often funded by vaccine manufacturers to generate safety data for vaccine approval, and they do not always reach peer-reviewed, scientific journals. However, trials that compare serologic responses to various vaccines can yield surprising differences that may support selection of specific products. A recent study in mares showed higher titers post-vaccination with specific vaccines from a variety of manufacturers, including significant differences in response to tetanus, influenza, EEE, and EHV1 [12]. Higher interferon gamma concentrations were also noted for one EHV1 product after the third dose, which is another correlate for protective immunity for viral infections.

Protection from clinical disease, as a definition of efficacy, merits scrutiny. Realistically, no vaccine, including tetanus toxoid, can prevent clinical disease in 100% of vaccinated animals, a concept often misunderstood by clients. Vaccinated animals can also serve as carriers for the organism after exposure. Vaccines most often reduce the severity of clinical signs and/or duration of the illness, which translates to lower morbidity and lower mortality. These important points must be stressed to clients, who are quick to blame vaccine manufacturers when a vaccinated animal develops clinical disease or seems to have brought strangles, influenza, or herpes back to the farm after an excursion or competition. When modified live vaccines are used, such as Arvac [c] for EVA or Pinnacle I.N. [c] for strangles, vaccinated animals may also spread the vaccine strain shortly after vaccination, which could have untoward effects [13,14]. If this is a concern, vaccinates should be segregated from non-vaccinates for an appropriate time.

From the herd perspective, one of the most important benefits of vaccination is the reduction in shedding, which may minimize spread to naïve animals in the herd, stable, or racetrack. This has been well documented in both experimental models and in the field for the influenza virus [15-17]. To promote "herd immunity", vaccination requirements should be strongly recommended for all commercial facilities, and if possible, the most efficacious vaccines should be specified. Ideally, new arrivals should have received the requisite vaccines before arrival, and if not, they should be quarantined for at least 4 wk before commingling with other horses.

The veterinarian must also consider that vaccines vary in magnitude and duration of immunity engendered, and they may need to be given more often than the label directions. Typically, modified live vaccines (MLV) produce a stronger immunologic response than killed vaccines that rely on adjuvants for enhanced immune-system signaling. This is a significant advantage for preventing disease. However, the disadvantage of some MLVs is a greater risk of side effects, including shedding of the vaccine strain. For most killed vaccines, short-lived humoral responses to vaccination create a need for strategic timing of administration or repeated dosing. The killed vaccines for EEE, WEE, and West Nile virus are ideally given shortly before the onset of mosquito season when the virus can be transmitted from birds to horses. However, maximal risk of these diseases is often months later in late summer and early fall in most endemic areas, which leads to the recommendation for a late-summer booster [18]. This is particularly important in years with heavy mosquito populations and viral activity in either sentinel birds, humans, or other horses. IM influenza vaccines have been advocated as frequently as every 2 - 3 mo in high risk environments. Even then, not all horses will have high titers that are predictably protective [19,20].

The shortcomings of many killed vaccines, including weak initial and anamnestic responses, are being addressed by advances in vaccine technology such as local antibody production with intranasal vaccines (Flu Avert I.N. [a] and Pinnacle I.N. [c]), recombinant vaccine vectors to stimulate immune recognition (Recombitek West Nile Virus [d]), subunit vaccines, and in the near future, perhaps DNA vaccines using bacterial plasmids. Hopefully, vaccine innovations in the future will lead to safer vaccines with fewer vaccine failures and ideally, mechanisms of action that circumvent blockage by passively acquired immunity in foals.

3. Adverse Effects

Every vaccine has the potential to cause an adverse reaction, including anaphylaxis and death. This is a very important reason to avoid unwarranted vaccinations. Fever and localized muscle soreness from IM injections can be significant. At least two vaccine manufacturers recommend that specific vaccines be given in the hindquarters (Potomacguard [c], Pneumabort K [c], and Strepvax II [b]), and most IM product inserts recommend mild exercise post-IM vaccination [21]. After IM vaccination in the neck, a recently vaccinated horse may appear ataxic or have impaired performance because of neck pain; thus, common sense should be used in timing these vaccinations. IM vaccines carry the rare risk of clostridial myositis, which can be fatal, even with aggressive therapy. Purpura hemorrhagica can be provoked by any of the strangles vaccines, and the intranasal vaccine can cause abscesses at the site of concurrent IM vaccinations [14].

Freezing of vaccines can increase the reaction rate and should be avoided by maintaining a professional chain of custody of vaccine stocks. Virtually all commercial vaccines recommend storage between 35°F and 45°F (2 - 7°C) in the dark. This narrow temperature range poses a challenge in extreme climates and must be stressed to veterinary staff as well as to clients that choose not to have the veterinarian handle and administer their horse's vaccines.

Multidose vials carry an increased risk of bacterial contamination, particularly if the contents are not used within a short time period, because vial reentry or surface contamination may introduce microbes. Fortunately, most products contain color indicators to signal microbial growth. Veterinarians should not administer vaccines that have changed color. Most vaccines also contain antibiotics, preservatives, and antifungal agents to minimize the risk. Antibiotic inclusion is one reason why horses should not be slaughtered for human consumption for 21 days post-vaccination. A 60-day withdrawal is required after anthrax vaccination [21].

Specific adjuvants or preservatives, not vaccine antigens, may be responsible for adverse reactions in individuals. For example, thimerosal and ethylmercurithiosalicylate are present in many vaccines and are known contact irritants. Some vaccines also specifically state that the product should not be mixed with other vaccines; therefore, veterinarians should read the product inserts carefully.

Centralized reporting of adverse effects associated with biologics needs to be more widely practiced to develop better data on frequency of these events. Some vaccines have higher reported rates of complications than others, and the types of complications are usually listed on the technical bulletin for the product (Table 1). Technical failures in vaccine production are fortunately rare but still happen. Veterinarians should keep records of vaccine lot numbers, and when administering multiple vaccines on a single visit, they should record the site in which each product was injected. This type of recording is also important to follow one manufacturer's directions to avoid administering initial boosters at the same injection site as the original vaccination for two vaccine lines (Calvenza and Cephalovac [b]).

Case-specific information on adverse reactions should be immediately relayed to the vaccine manufacturer and the federal government's Center for Veterinary Biologics. The latter can be done through their web site (www.aphis.usda.gov/vs/cvb/ic/adverseeventreport.htm). If an allergic reaction is linked to vaccination, the veterinarian should discuss future coadministration of antihistamines and/or non-steroidals (unknown effect on vaccine response), alternative vaccines, or other strategies for disease prevention with the client. The choice of a non-adjuvanted vaccine, such as Recombitek [d] or Flu Avert I.N. [a], could have a safety advantage for horses in which an adjuvant triggered the adverse reaction. Another rare source of reaction to vaccinations is the coating on disposable needles. Horses with sensitivity to this silicone coating typically develop hard lumps at the site of any injection [22]. When recognized, these animals can be managed with silicone-free needles [f].

4. Controversies in Vaccine Use

Strangles

Agreement on vaccination recommendations for strangles prevention is hard to achieve. Recently, the American College of Veterinary Internal Medicine (ACVIM) released a consensus statement developed by large-animal, internal-medicine diplomates [14]. Three vaccine choices are currently available: two M-protein-based inactivated vaccines for IM use (Strepguard [a] and Strepvax II [b]) and one modified, attenuated live bacterial intranasal vaccine (Pinnacle I.N. [c]). Strepguard requires two initial vaccinations, and Strepvax II requires three doses. Although label recommendations suggest that only yearly boosters are needed, many farms with endemic strangles found that semiannual vaccination was more effective in reducing disease morbidity when using the IM vaccines. Both IM vaccines reduce clinical manifestations of strangles [23,24]. All three vaccines reduce the severity of the disease in experimental challenge with more significant reduction in clinical scores observed with the intranasal product. This suggests that Pinnacle I.N. is the most effective vaccine [g]. In many areas of the country, clinical experience has echoed this laboratory challenge data, because a switch to the intranasal vaccine has reduced the number of strangles cases requiring veterinary intervention. Because all three vaccines have the potential to produce purpura hemorrhagica, strangles vaccines should be avoided if not warranted by risk of exposure.

Because Pinnacle I.N. induces localized immunity in the nasopharynx, a significant rise in serum titer is not expected after vaccination. For breeding farms that rely on transfer of passive immunity to strangles through the colostrum, perhaps IM vaccines that stimulate high serum titers are a valid choice. It has been shown that pre-partum, IM vaccination of the mare significantly increases SeM-specific IgGb isotype antibodies in mucosal washes of their foals during the first 2 mo of life [14]. Alternatively, practitioners on endemic farms can choose to use only the intranasal vaccine in the broodmare and then choose to vaccinate foals as early as 2 wk of age with the intranasal product. This can be successful if the foal is not exposed to strangles in the first 2 wk. The safety and efficacy of this practice needs to be determined beyond observations of individual practitioners. A trial assessing strangles protection derived from colostral antibodies is warranted in nursing foals as well as determination of colostral antibody interference with either IM or intranasal vaccination.

The risks associated with intranasal strangles vaccination have discouraged its use by some veterinarians. These risks include abscess formation at the site of concurrent injections, mild clinical disease, and immune-mediated complications such as purpura or thrombocytopenia. Even minimal vaccine contamination of the needle or skin at the site of IM injections after Pinnacle I.N. administration has led to marked abscess formation at the injection site. Veterinarians quickly learned to administer all IM injections first to all horses on the premises before even mixing up the intranasal vaccine; this eliminates the risk of contamination. Unfortunately, even with this precaution, a very small number of horses still developed an abscess at the site of an IM vaccination, and the attenuated Streptococcus equi var. equi was isolated [h]. Presumably, in these animals, the vaccine strain was able to overcome natural pharyngeal barriers and create a transient bacteremia, which led to the colonization at inflamed muscle sites. After experiencing this adverse reaction, a veterinarian may choose to avoid administering any other vaccines on the same visit as the intranasal vaccine or may resort to using the IM products for protection against strangles.

Horses vaccinated with the intranasal strangles vaccine may very occasionally develop mild signs of strangles including depression, lymphadenopathy, and fever. Rarely, mandibular lymph nodes may progress to rupturing. The vaccine strain can also purportedly be transmitted to other horses in the herd, causing similar signs, but this has not been verified by polymerase chain-reaction (PCR) testing of the isolate. This possibility can be of particular concern in boarding stables where multiple owners can independently choose the vaccines that they wish their horse to receive. If a veterinarian chooses to use the intranasal vaccine at the beginning of a strangles outbreak, clients should be advised of this risk. The veterinarian must also be cautious not to inadvertently spread the wild type S. equi during administration of the vaccine to multiple horses [14]. Clinical experience suggests that early use of the intranasal vaccine in minimally exposed horses may be effective in reducing outbreak severity. However, the vaccine should be avoided in horses that have developed early signs of the disease, such as depression or fever. The recent ACVIM consensus statement takes a more conservative approach and recommends that the use of the intranasal vaccine during an outbreak should be avoided, except in horses with no known contact with infected or exposed animals. However, it also states that no published data show that use in the face of exposure is detrimental [14].

Immune-mediated complications, such as purpura hemorrhagica, can develop with any of the strangles vaccines, but no data is available to suggest that there is a difference in the frequency of severe complications between products. One horse presented to the University of Minnesota developed fatal immune-mediated thrombocytopenia after administration of the intranasal vaccine.

Influenza

The leading role of equine influenza virus in upper respiratory disease in the horse has been well established, and its economic impact on the racing industry has prompted the most extensive research on efficacy for equine vaccines. Killed IM vaccine failures were repeatedly documented in the 1990s; these vaccines did not reliably produce measurable titers in all horses nor did they exert a significant protective effect without very frequent administration [11,19,20,25]. Poor protection has also been attributed to antigenic drift in influenza strains, leading to periodic updates in the vaccine strains incorporated in the products [26]. More recently, the intranasal vaccine Flu Avert I.N. [a], a cold-adapted, modified live virus vaccine, has been widely adopted and shown to provide more effective protection after experimental challenge. Additionally, it greatly reduced viral shedding for 6 mo and still had significant effect on the reduction in shedding for as long as 12 mo post-vaccination [16]. Consequently, in high-risk environments, administration is recommended every 6 mo in contrast to every 2 - 3 mo for the older IM killed vaccines. The only shortcoming of the intranasal route of administration is for broodmares, because the vaccine does not produce high-circulating serum titers; thus, colostrum-derived protection for foals may be minimal. Practitioners need to consider incorporating an IM product into the mare's schedule if her foal has a significant risk of influenza exposure, or they should use a product such as Calvenza EIV [b], which can be administered intramuscularly or intranasally after the initial two vaccinations. This product has also performed well in a live-virus challenge for yearlings 4 mo post-IM vaccination. It provided more significant diminution in clinical signs and viral shedding compared with other IM vaccines, and it gave similar protection to that achieved with the intranasal Flu Avert I.N. [a] [10].

Vaccination in the Face of Exposure

When an infectious disease is first identified in a group of horses, vaccination is one tool amongst many to control the spread of disease. Conventional IM vaccines usually require at least 1 wk for measurable humoral responses to a booster or a second dose and similar time period in naïve animals. This time lag has discouraged the use of vaccines in exposed animals, yet vaccination has been successful in protecting adjacent groups not yet exposed.

In the recent West Nile epidemics, many horses were vaccinated during exposure with no recognized ill effects other than incomplete protection; additionally, there was the benefit of lower mortality than in unvaccinated horses [27]. In Minnesota's 2002 West Nile epidemic, horses that had received two doses of killed vaccine [c] at least 2 wk before the onset of clinical signs were 2.9 times less likely to die from the disease or be euthanized. None of the vaccinated horses suffered residual deficits from the infection, which contrasts with those not vaccinated [28]. In a laboratory challenge model, the recombinant West Nile vaccine has shown protection from viremia 26 days after a single dose in 8 of 9 horses [29]. This rapid response led to more widespread use of this vaccine in previously unvaccinated horses on the West coast when the first equine cases of West Nile virus were diagnosed.

The more rapid development of local immune responses after intranasal vaccination has led some veterinarians to use these two vaccines in animals that have not yet have become exposed. This technique of disease control is most effective in well-managed facilities where frequent observations are made of the horses, and there is reason to believe that the index case of either influenza or strangles was very quickly identified. Vaccination of exposed animals that may be subclinically infected carries the risk of iatrogenic passage of the wild type organism, which is highly undesirable [14]. However, better studies of this type of vaccine usage are needed.

The neurologic form of EHV1 presents a different conundrum for the veterinarian. There is often a farm history of antecedent mild respiratory disease or less often, abortion. In a California outbreak of neurologic EHV1 infection, horses vaccinated with either type of vaccine within the previous year were 9 - 14 times more likely to develop neurologic signs than non-vaccinated horses [30]. Because the vasculitis associated with the neurologic form is immune mediated, vaccination after exposure raises concerns of producing a more severe disease. Consequently, vaccination in the face of a confirmed outbreak of EHV1 neurologic disease has been controversial. This contrasts with recommendations to immediately booster pregnant mares that have been exposed to a confirmed EHV1 abortion, yet neither Pneumabort K [c] or Rhinomune [i] produce detectable mucosal antibody in nasal washes [31]. A newly released study from Cornell University reports protection from clinical signs of EHV1 neurologic disease using the modified live EHV1 vaccine, Rhinomune [i], in an experimental respiratory challenge model. Both control horses and those vaccinated with a killed EHV1 vaccine (Pneumabort K [c]), five horses per group, developed neurologic signs and fever post-challenge, and the neurovirulent challenge isolate could be recovered from nasal swabs. This virus, at a very low level, was recovered from only 1 of 5 of the modified live-vaccinated group, and none of the MLV vaccinates showed neurologic deficits. This encouraging new data suggests that Rhinomune [i] could offer a degree of protection against this strain of EHV1. However, this falls short from recommending it in the face of exposure, because the experimental challenge was delivered 4 wk after the second dose of the vaccine [32]. Hopefully, protection from clinical signs can be similarly evaluated for Calvenza EHV1 [b] in the future.

Vaccination After Natural Infection

Little information is available on the duration of immunity after recovery from natural infections in horses. Virtually all natural infections engender much stronger antibody responses than corresponding vaccines, but the half-lives of those antibodies, degree of individual variation, and definition of protective titer are unknown. This information void has led to general recommendations to resume normal vaccination schedules in the year after the diagnosis.

Several diseases merit discussion of whether this presumption should be applied. Active strangles infection seems to confer very long lasting immunity in the majority of horses, because few develop clinical disease more than one time in their life. A prospective study in foals suggested that previous exposure to strangles greatly reduced attack rates from 86% to 17%, and it highlighted the importance of mucosal M protein-specific IgG [33]. An exception to the long-lasting immunity observations is the geriatric horse with Cushing's disease. Overall, these horses seem to have a slightly increased susceptibility to infectious diseases because of cortisol-induced immunosuppression. Some recovered strangles cases may develop signs of purpura after vaccination, leading to vaccine avoidance. A recent paper recommends waiting 12 mo after natural disease before resuming vaccination [14].

Both EHV1 and EHV4 can establish latency, recrudescing after periods of stress or immunosuppression. This may arguably be a reason to choose not to revaccinate them, yet infection with new strains can happen. A modified live herpes vaccine, no longer on the market, produced neurologic signs attributable to vasculitis, but the herpes status of these horses was unknown before vaccination [34]. Clearly, a greater understanding of the pathogenesis of neurologic signs in EHV1 cases is needed to answer the question of revaccination.

Equine influenza presents a contrasting picture that promotes vaccination post-infection, because antigenic mutation in equine strains leads to renewed susceptibility in horses as in other species [26]. Studies have shown differences in definition of protective titer for homologous versus heterologous challenges. Higher titers are required to prevent clinical signs and significant viral shedding for heterologous strains [25,26].

Currently marketed vaccines against N. risticii, the causative agent of Potomac Horse Fever, contain a single strain of this organism. Studies have suggested that multiple antigenic strains exist, which results in the relatively low efficacy of the vaccines [35]. Natural or experimental infection engenders very high titers, but it is unknown if these would be cross protective if the horse was exposed. Until further research is done or more strains are incorporated into available vaccines, this question of revaccination will remain unanswered.

Overvaccination Versus Titers

Public attention to the risks of vaccination has been raised by the link of adjuvanted vaccines to sarcomas in cats as well as concerns about mercury-containing preservatives (e.g., Thimerosal) in humans and animals [36-38]. Multiple epidemiologic studies produced conflicting data linking infant exposure to thimerosal-containing vaccines to neurodevelopmental defects such as autism and speech delays. This has led to a recommendation to remove this preservative from childhood immunizations [39,40]. This potential link and subsequent vaccine change have eroded public confidence in the safety of vaccinations and have led to fewer uses of vaccines [41]. Holistic approaches to infectious disease prevention have also influenced some horse owners' perceptions of vaccines. Additionally, there are several websites that tout the evils of vaccination [42].

Owner perceptions, combined with economic considerations and risk of adverse reactions, have led some veterinarians to designate certain vaccines as "core" and others as optional, based on regional risks. For example, tetanus, EEE, WEE, and West Nile virus protection would be considered "core" for virtually all horses in the United States; perhaps, this may be all that is needed for a closed herd (Table 2). Yet, even the need for annual vaccination for tetanus is frequently questioned because of the lower frequency of vaccination advocated for humans. Practitioners need to remind clients of the high-exposure risk in horses and the increased sensitivity of horses to this toxin. Guidelines from the American Veterinary Medical Association advocate yearly vaccination with tetanus toxoid and a booster if the horse sustains a wound 6 mo or longer after immunization [43]. The choice to diminish vaccination frequency for tetanus could be considered substandard care unless titers are first evaluated. Written agreement that the client understands the risks of less-frequent vaccination may have some legal merit.

Serologic titers are a commonly advocated method to determine need for vaccination in companion animals, and occasionally, in humans. However, for many equine infectious diseases, the definition of a "protective" titer is non-existent. Similarly, in a number of challenge models, serologically negative animals have resisted challenges that produced clinical disease or viremia in naïve animals [29,44]. Because of the paucity of data for definition of protective titers for most equine infectious diseases, laboratories that run vaccine panels for dogs and cats are reluctant to do so for horses. For example, the New York State Veterinary Diagnostic Laboratory at Cornell University offers three different panels for guiding vaccination decisions for dogs and cats, complete with titer interpretation. For horses, the only combined panel is for diagnosing the etiology of an equine abortion and offering serology for leptospirosis, equine arteritis virus, herpes, and influenza without interpretation, unless paired samples are run. For most diseases, lack of detectable titer might suggest vaccine failure or need for boostering. It does not guarantee susceptibility should the horse become exposed. A very high titer in a healthy animal could suggest either natural exposure or excellent response to vaccination, and consequently, it could be interpreted as evidence that revaccination at that time may not be necessary. However, the client should never be left with the impression that a high titer is a guarantee of disease resistance. For strangles, vaccination in the face of a very high titer (1:1600 on SeM enzyme-linked immunosorbent assay [ELISA]) has been discouraged as well as vaccination within 12 mo of natural infection [14].

Vaccination During Pregnancy

Long-term safety studies of potential adverse effects of vaccine use during pregnancy have not been done for broodmares. Concerns regarding possible teratogenic or abortigenic effects of equine vaccines used in the first trimester have no published scientific basis to date. The recommendation to avoid all unnecessary vaccine use during this time period seems to stem from precautions advised by a subset of physicians for pregnant women. Metals, such as the mercury in thimerosal, have been associated with teratogenic and carcinogenic effects in laboratory animals and in specific neural cell lines in vitro. Over 50% of the approved vaccines for horses contain thimerosal, making construction of a thimerosal-free vaccination program challenging (Table 3). Claims that the killed West Nile virus vaccine was responsible for pre-natal fetal loss have not been substantiated. A retrospective study of 595 broodmares in Texas showed no detrimental effects of vaccination on live foal rates [45].

The majority of available vaccines do not discourage use in pregnant mares, but most decline to make safety claims for this population. An exception is the modified live vaccine for equine arteritis (Arvac [c]). The manufacturer and experts on the disease warn against its use in the last 2 mo of gestation [46]. Early pregnancy in many mares coincides with the greatest risk periods for arthropod-borne diseases. A broodmare is most likely to be exposed to EEE, WEE, or West Nile virus during her first trimester of pregnancy, and therefore, she should be protected from this known cause of preventible mortality. Strategically vaccinating the mare post-foaling and before breeding is an option, but it might result in insufficient protection in the face of overwhelming challenge in the late summer or early fall if the mare foaled in the early spring. This reasoning also prompts the question of whether the mare that receives an EEE, WEE, or West Nile virus booster pre-foaling may diminish her "reserve" of humoral antibody by transferring significant levels to the colostrum and thereby increasing her susceptibility. Further studies on this topic are certainly warranted.

Extralabel Vaccine Use

Lacking approved products for use in horses, veterinarians may consider using vaccines manufactured for other species or custom-manufactured vaccines when an infectious agent causes morbidity in a group of horses. Lyme disease, leptospirosis (uveitis and abortion), and clostridial enteritis are examples of use of this avenue for disease prevention. Again, good field data is scant to support this practice, but it may emerge in the future. The veterinarian should ensure that the client fully understands the risks of "unproven" biologicals in case adverse reactions occur or the vaccinations fails to protect the animal [47].

Lyme disease, caused by Borrelia burgdorferi, is recognized as the most important arthropod-borne bacterial disease in humans, and it is linked to a reasonably well-described syndrome in dogs and cats. In horses in endemic areas, seroconversion is common, but proven clinical disease is relatively rare. Anecdotal reports suggest use of canine vaccines in horses. Recently, an experimental model of Lyme disease in ponies was described in which eight specific pathogen-free ponies were exposed to infected Ixodes ticks. All developed detectable antibody at 5 - 6 wk, and the organisms could be consistently recovered in the skin at sites of tick attachment. Pre-scapular lymphadenopathy was the only gross abnormality noted, but microscopically, most ponies had evidence of lymphohistiocytic nodules in the dermis, perivascular and perineural lymphocytic reactions, and lymphocytic infiltration of the facial nerve. An experimental recombinant outer-surface protein A (rOspA) subunit vaccine, similar to the canine rOspA vaccine, blocked all signs of infection in ponies similarly challenged after administration three times before challenge 30 days post-third vaccine. The authors suggest that antibodies engendered by the vaccine inhibited B. burgdorferi within the tick, because no serologic conversion could be documented. This contrasts with the non-vaccinated controls, which developed titers. Further safety studies are recommended before using the canine rOspA vaccine outside of an experimental setting [2,48,49]. This precautionary statement is timely, because adverse reactions have been associated with rOspA vaccines in humans and other mammals where the vaccine was linked to Lyme-like symptoms via immune-mediated mechanisms [50,51].

High antibody titers to serovars of Leptospira and organisms have been found in horses with recurrent uveitis and aborted fetuses [52-56]. These observations have prompted the extralabel use of leptospiral vaccines marketed for cattle or swine, in horses by some veterinarians and horse owners. The higher frequency of recurrent uveitis in Appaloosas in one study has sparked particular interest in this extralabel vaccine use [57]. The impact of leptospiral vaccination using a multivalent, inactivated swine vaccine was recently published [3]. In a randomized, blinded, prospective clinical trial in 41 horses with recurrent uveitis, the investigators found that a half-dose (1 ml) of vaccine, administered four times during the course of the study, significantly increased days to recurrence. However, vaccination did not significantly slow the progression of the disease in the overall comparisons. In 5 of 6 horses that had active uveitis at the time of first vaccination, the clinical signs failed to progress after this first dose compared with only 2 of 14 horses in the placebo-vaccinated group. No aggravation of uveitis was noted after vaccination, and only 1 of 41 horses developed very slight, transient swelling at the site of vaccination in the neck.

Leptospiral vaccines labeled for use in horses are used in other countries to protect against leptospiral abortion, but studies on the efficacy of swine or cattle vaccines to prevent abortions in the United States are not available. Pilgrim and Threlfall [58], after studying the high seroprevalence (36%) for leptospirosis in 669 broodmares sampled in Ohio, hypothesized that natural exposure may be more efficient at providing immunity than any vaccine. Should a practitioner choose to use leptospiral vaccines, it is important that the leptospiral species and serovars in the vaccine are appropriate for that farm.

Enteritis in foals may be caused by both Clostridium difficile or Clostridium perfringens [4,59,60]. Widespread diarrhea and mortality in a foal crop have prompted use of cattle vaccines or custom-made bacterins in some breeding farms where improvement in management factors was not sufficient to minimize this disease. East et al. [4] recommended vaccination with C. perfringens type C and D toxoid for broodmares between 3 and 6 wk pre-partum. There are anecdotal reports of administration of clostridial antitoxins at birth as well, but there are no reported efficacy studies to date. Because these products are derived from hyperimmunized horses, it is theorized that fillies could have an increased risk of later producing a foal with neonatal isoerythrolysis.

5. Conclusion

Veterinarians use information regarding vaccine efficacy and safety to guide vaccination choices for horse owners to minimize adverse reactions or vaccine failures. Many questions regarding vaccine efficacy and safety still need to be addressed, including potential impact on the developing fetus, duration of immunity, and effects of administering or combining multiple vaccinations in one session. Discerning the optimal choices for protecting horses requires integration of information from many sources as well as clinical experience. Wherever possible, the practitioner should try to match expected duration of immunity with risk period for both individual and farm vaccination programs. Ideal product selection can best be made when unbiased, vaccine comparison studies that are performed in the field are available for scrutiny. This type of information is essential for veterinarians, particularly as new vaccines come on the market. Ongoing client education on the advantages and shortcomings of vaccines is essential for horse owners to make informed decisions. Owners must also understand that vaccines are tools, not panaceas, and they must be combined with other preventive measures to reduce the risk of infectious diseases in their horses.

Footnotes

1. Intervet, Millsboro, DE 19966.
2. Boehringer Ingelheim Pharmaceuticals Inc., Ridgefield, CT 06877.
3. Fort Dodge Animal Health, Fort Dodge, IA 50501.
4. Merial Limited, Duluth, GA 30096.
5. Companion Animal Technical Report No. 18.Intervet Inc., Veterinary Technical Services, 2001.
6. MWI Veterinary Supply, Meridian, ID 83642.
7. Hustead DR.Fort Dodge Animal Health Technical Information, February 19, 1998.
8. Ames TR, Al-Ghamdi G.Personal communication, 2003.
9. Pfizer Animal Health, Exton, PA 19341, a division of Pfizer Inc., New York, NY 10017.
10. Colorado Serum Co., Denver, CO 80216.