Vol. 23, No. 4 April 2001 375 Email comments/questions to compendium@medimedia.com CE Article #6 (1.5 contact hours) Refereed Peer Review KEY FACTS Large numbers of horses in the more eastern parts of the northeastern and mid-atlantic United States have at some time been infected with Borrelia burgdorferi. Indirect fluorescent antibody and kinetics-linked immunosorbent assay are both highly sensitive and specific for B. burgdorferi exposure. Although not commercially available, an outer surface protein A antigen vaccine was shown to prevent infected ticks from transmitting B. burgdorferi to horses. Lyme Disease in Horses Cornell University T.J. Divers, DVM Y. F. Chang, DVM, PhD R.H. Jacobson, PhD S.P. McDonough, DVM, PhD ABSTRACT: Equine exposure to Borrelia burgdorferi appears to be very common in some areas of the United States. The clinical impact of this exposure is uncertain, although experimentally infected ponies have microscopic evidence of dermatitis, lymphadenitis, and neuritis. Lyme disease is best diagnosed by considering clinical signs, serologic testing, polymerase chain reaction testing of skin or synovial membranes, and response to therapy. Other tickborne diseases should be ruled out, especially equine granulocytic ehrlichiosis and other disorders that may cause disease of the skin, connective tissue, nerves, and joints. Although tetracycline and doxycycline are the two antimicrobials used to treat equine Lyme disease, their efficacy is not proven. Lyme disease is the most important arthropod-borne bacterial infection in humans in the United States. 1,2 Lyme disease is caused by the spirochete Borrelia burgdorferi, which is maintained in a 2-year enzootic cycle involving ticks (Ixodes spp.) and mammals (Figure 1). Deer and the white-footed mouse (Peromyscus leucopus) are the mammals most commonly involved in maintaining the life cycle. Infection in mammals generally occurs from larval or nymph bites in the spring and summer or adult (female) ticks feeding in summer, fall, or winter. In most instances, the ticks must be attached to the mammal for at least 24 hours for B. burgdorferi transmission to occur. 1 Lyme disease has a high prevalence in some areas of the United States but may be nonexistent in other areas (Figure 2). 2 Some East Coast states, Wisconsin, and Minnesota have an especially high incidence of infection in humans, horses, and dogs. In horses, a broad spectrum of clinical signs has been attributed to Lyme disease (see Clinical Signs of Equine Lyme Disease, p. 376) but cause and effect have been difficult to document. 3 6 PREVALENCE AND DIAGNOSTICS A large percentage of adult horses in the more eastern parts of the northeastern and mid-atlantic United States are or may have been infected with B. burgdorferi. This has been documented by serologic surveys that demonstrate that a very high percentage of adult horses in some of these areas are seropositive. 3 Seroprevalence in other parts of the country has not been reported but would be expected to reflect that documented in the human form of the disease. 2 Indirect fluorescent antibody and kinetics-based, enzyme-linked im-
376 Equine Compendium April 2001 Figure 1 Two-year enzootic cycle of Borrelia burgdorferi and distinguishing features of Ixodes species ticks (female ventral abdomen; inset). munosorbent assay (KELA) are both highly sensitive and specific for B. burgdorferi exposure. In one study, horses in endemic regions that had clinical signs suggestive of Lyme disease (e.g., lameness, behavior change) were more likely to have higher KELA titers, positive Western blots, and a higher incidence of spirochetemia and spirochetosis than horses in the same region lacking suggestive clinical signs. 4 Although this does not prove a cause and effect between B. burgdorferi infection and the observed clinical signs, the inclusion of a control group from the same geographic areas makes the postulated association stronger. At our hospital, we have detected B. burgdorferi DNA (evaluated by polymerase chain reaction [PCR] testing) in synovial membranes in cases involving similar clinical signs, strengthening the proposed association between lameness and the presence of B. burgdorferi. We have most commonly diagnosed Lyme disease in horses with chronic stiffness, multiple-limb lameness, hyperesthesia, and/or thoracolumbar pain. Some of the stiff and hyperesthetic horses have had muscle wasting over the dorsum. Only rarely have the affected horses had noticeable joint swelling. The diagnosis of Lyme disease has been based on the horse being housed in an endemic area; clinical signs; and ruling out other causes for these signs (see Clinical Signs of Equine Lyme Disease) with thorough Clinical Signs of Equine Lyme Disease 3 6 Low-grade fever Swollen joints and/or limbs Stiffness Lameness Muscle tenderness Lethargy Neurologic disease Panuveitis Behavior change Hyperesthesia Figure 2 Areas (yellow) of the United States with the highest reported incidence of Lyme disease in humans. 2 lameness and neurologic examinations, radiographs, scintigraphy, muscle biopsy, cerebrospinal fluid collection, serum concentrations of muscle enzymes (before and after exercise), a high titer (generally greater than 300 KELA units) plus a strong positive Western blot for B. burgdorferi, and B. burgdorferi DNA in the synovial membrane of a painful joint. Common disorders that we attempt to rule out with these tests include hock osteoarthritis, osteochondrosis, polysaccharide storage myopathy, chronic intermittent rhabdomyolysis, thoracic spinous process osteoarthritis, and equine protozoal myelitis. Experimental infection of naïve ponies yielded further diagnostic clues as detailed in the Experimental Infections, p. 377. 7,8 TREATMENT The two most commonly used drugs for treating Lyme disease in horses are intravenous (IV) tetracycline and oral doxycycline. Tetracycline has been used with great success most commonly for acute cases with stiffness, limb edema, and fever. Although these animals generally seroconvert to B. burgdorferi after treatment, it is possible that these clinical signs are a result of concurrent Ehrlichia equi infection rather than B. burgdorferi infection. 9 Horses with what we believe to be more typical signs of Lyme disease (e.g., chronic stiffness, lameness, hyperesthesia) are most frequently treated with doxycycline (5 to 10 mg/kg PO q12h) for 1 month. Horses treated with doxycycline should be observed for a change in stool consistency because diarrhea occurs in a low
Compendium April 2001 Equine 377 Experimental Infections We successfully infected naïve ponies (11 of 11) with B. burgdorferi by placing infected Ixodes scapularis ticks on their backs; control ponies receiving noninfected I. scapularis ticks remained uniformly uninfected. 7,8 Once ponies were infected, the area of the skin surrounding the tick bite remained both PCR and culture positive for B. burgdorferi for at least 4 months. When the ponies were euthanized 9 months after tick exposure, live B. burgdorferi and/or its DNA was most commonly found in synovial membranes (66 of 70), regional lymph nodes (12 of 12; DNA only), muscle (12 of 14), or antebrachial fascia (12 of 14). The next most frequent sites for detection were the pericardium, myocardium, and urinary system. Lesions were limited to the skin and regional lymph nodes in most ponies. Lymphocytic infiltrates around nerves or vessels were present in some ponies, mostly involving nerves supplying muscle and joints on the side of the body to which the ticks were attached. In addition, mononuclear perivascular reactions were found around the facial, tibial, and lumbar nerves and subsynovial connective tissues. Joint effusion was not noted either clinically or on necropsy of infected ponies, and fevers were absent or rare throughout the experimental infection. Clinical signs of lameness, hyperesthesia, and/or changes in behavior were difficult to evaluate because the experimental ponies were previously unhandled and limited to housing in a percentage of treated horses. Tetracycline (6.6 mg/kg IV sid) for 1 week prior to beginning doxycycline treatment may provide a more rapid clinical response. We have only limited experience conducting PCR tests and/or cultures for B. burgdorferi on synovial biopsy following treatments. Two horses have been PCR positive and two others PCR negative on synovial biopsies taken following 1 month of treatment with tetracycline or doxycycline. One other horse was treated with ceftiofur (2 mg/kg IV q12h) for 30 days, and follow-up PCR was negative on the same joint that had been positive 1 month earlier. We had one experience involving a presumed farm outbreak of Lyme disease in horses with severe tick infestation. In those animals, the common signs were low-grade fever, lameness, and mildly swollen joints. There was a mild decline in KELA antibody after 1 month of treatment with doxycycline. Supportive treatments that can be considered include Lymphohistiocytic aggregate around a small arteriole within the cutaneous trunci muscle in a pony experimentally infected with Borrelia burgdorferi. confinement facility. Infected ponies were seropositive (determined by ELISA) by 5 to 6 weeks after exposure, reached 200 to 300 KELA units by 3 months, and remained at this level until they were euthanized. Infected ponies became positive on the Western blot between 2 and 3 months. B. burgdorferi was cultured from all infected ponies at the time of euthanasia despite high circulating antibodies. No outer surface protein A (OspA) antibody was produced during the infection. Ticks not infected with B. burgdorferi were placed on control ponies, and all controls remained seronegative and PCR and culture negative during the 9 months before euthanasia. chondroprotective agents, NSAIDs, and acupuncture. Acupuncture may be of value in treating horses with evidence of hyperesthesia/perineuritis that show little response to NSAIDs. PREVENTION Prevention of Lyme disease in endemic areas involves preventing tick exposure or prolonged (more than 24 hours) attachment, prompt treatment after Ixodes exposure; or vaccination. Various insecticidal sprays are used to prevent tick infestation, but most are not approved for use on horses and the efficacy of such use is unproven. We are not aware of any adverse effects in horses caused by the more common canine tick sprays. Spraying is most commonly done when adult ticks are noticeable in summer, fall, and early winter. Larval/ nymph stages are more common earlier in the year, but whether horses are infected with B. burgdorferi by these stages is unknown. If present on the horse, ticks should
378 Equine Compendium April 2001 be identified to determine whether they are Ixodes species (Figure 1), which might transmit B. burgdorferi. We evaluated a recombinant OspA-antigen vaccine, which was capable of preventing infected ticks from transmitting B. burgdorferi to horses. 8 This vaccine is not commercially available and has not been extensively evaluated for safety. DISCUSSION Although B. burgdorferi infection is very common in horses in areas known to have a high incidence of Lyme disease in humans, we believe the percentage of horses that demonstrate clinical signs is low. The relative paucity of clinical signs was also reflected in our experimental studies (see Experimental Infections, p. 377) in which only 1 of 11 infected ponies had obvious lesions involving the perineural and synovial tissues. We now know that the organism seems to prefer to reside/move in skin, fascia, perineural tissue (especially that of muscle and synovial membranes), and synovial membranes. 7,8 Movement of the organism is largely within connective tissue, which may protect the organism from humoral antibodies. The severe lymphocytic/plasmacytic perineuritis of the synovium that occurred in one experimental pony 8 may offer an explanation for the lameness seen in many cases of Lyme disease. Most of the horses with positive PCR synovial biopsies that we have diagnosed with Lyme disease had minimal changes in their synovial membranes and minimal effusion. The pain from the disease may be more a result of synovial perineuritis than synovitis in horses. We do not know why certain horses develop clinical signs with Lyme disease and others with similar titers do not. The most plausible explanation would be differences in host response to the organism. Many of the horses we have diagnosed with Lyme disease have been upper-level event horses. Hyperesthesia and mild stiffness are more likely to be noticed in these horses than in pleasure horses, which are not evaluated as closely each day. Therefore, the incidence of Lyme disease may be underestimated in the entire equine population. The hyperesthesia noted in many affected horses has been called myofascial syndrome and is probably a result of movement of the B. burgdorferi organism in the larger fascial planes of the neck, axillae, and thorax. Some acupuncture points can be used to provide almost instant relief of this myofascial pain. 10 If this is the case, the horse should be serologically tested for Lyme disease and, if positive, treated with appropriate antibiotics. Horses with ELISA values of 440 KELA units or more have a high probability (greater than 95%) of being infected. When the ELISA values are 110 KELA units or less, the chance that the horse is infected is less than 1%. The range of values between 110 and 440 KELA units represents about 52% of all horses tested a ; disease in these horses should be confirmed by Western blot. These data indicate that considerable care must be exercised in evaluating ELISA data. If the usual single cutoff is used in the equine Lyme ELISA, there is a fairly high probability of misclassifying the horse as infected due to false-positive reactivity in the assay. Many of the horses successfully treated with tetracycline for presumed acute Lyme disease (fever, leg edema, stiffness) may actually have E. equi infection because many of the Ixodes species ticks are dually infected with both E. equi and B. burgdorferi. Dual infection would explain the acute clinical signs due to E. equi, dramatic response to tetracycline treatment, and seroconversion to B. burgdorferi. 9 Treatment for E. equi with tetracycline might also prove highly effective in preventing more chronic signs of Lyme disease. If we assume that infection with both organisms might occur at the same time and that clinical signs of E. equi occur within 10 to 14 days of infection, then B. burgdorferi organisms might still be in the skin or less-protected (connective tissue) areas and thus short-term treatment would be highly effective. Since either horses do not develop skin rash from Lyme disease or these lesions are not observed, this is a major hindrance in recognizing early infection in horses. Proper duration of treatment for chronic infections is unknown. We recommend 30 days of doxycycline but do not know how successful this is in ridding synovial membranes of B. burgdorferi, if indeed that is possible. The positive PCR after treatment in some cases could reflect residual DNA in the synovial membranes rather than live organisms. Persistent immune responses against B. burgdorferi DNA in synovial membranes may play a role in the progressive reactive arthritis that can occur with Lyme disease. 11 Sterilization of synovial membranes with any treatment may be difficult. Clinical improvement may occur and titers may decrease, but even when reinfection is completely prevented, clinical signs may recur weeks or months after antimicrobial treatment is discontinued. Although not proven to be the best drug for treating Lyme disease in horses, doxycycline is the only drug evaluated for oral administration in horses that has in vitro efficacy against B. burgdorferi. In acute cases, short-term IV therapy with tetracycline prior to long-term doxycycline therapy may offer some advantages because IV tetracycline yields much higher tissue levels than oral doxycycline and may have immediate antiinflammatory effects unrelated to its antimicrobial effect. 12,13 a Jacobson RH: Personal communication. Ithaca, NY, Cornell University, 2000.
Compendium April 2001 Equine 379 Vaccination with recombinant B. burgdorferi OspA protected ponies from infection when experimentally challenged with B. burgdorferi infected ticks. The challenge with infected ticks was at a single time point (30 days after the third recombinant OspA vaccination); thus the duration of protection could not be determined. If a Lyme vaccine becomes commercially available for use in horses, some points seem noteworthy from our experimental studies. Anti-OspA antibodies appear to be critical to inhibition of B. burgdorferi. Natural infection does not produce significant amounts of this antibody, suggesting that reinfection could commonly occur in horses. In addition, as ticks feed on horse plasma, the effect of the blocking antibody is within the tick and not the horse or pony and in our study apparently prevented exposure of the challenged ponies to B. burgdorferi. Therefore serologic separation of vaccinated versus infected animals should not be a problem with this vaccine. It should be emphasized that the OspA vaccine would be expected to be a protective vaccine and not a therapeutic vaccine. Further studies are needed to determine the protective duration and safety of the vaccine. CONCLUSION Experimental studies and field experience have significantly improved our understanding of Lyme disease in horses. A large percentage of horses in several regions of the United States are currently or have been previously infected with B. burgdorferi, but the incidence of clinical disease is undetermined. Additional experimental studies and clinical case investigations are needed to better define the clinical disease in the horse. Therapeutic trials using experimentally infected animals at different times after infection are needed to determine the most effective treatment for each stage of the disease. Finally, the development of what appears to be a potentially effective and safe vaccine offers the hope of preventing infection in horses in endemic areas. REFERENCES 1. Thanassi WT, Schoen RT: The Lyme disease vaccine: Conception, development, and implementation. Ann Intern Med 132(4):661 667, 2000. 2. Orloski KA, Hayes EB, Campbell GL, Dennis DT: Surveillance for Lyme Disease United States 1992 1998. Morb Mortal Wkly Rep CDC Surveill Summ 49(3):1 11, 2000. 3. Mangnarelli LA, Anderson JF, Kaufman AF: Borreliosis in equids in the Northeastern United States. Am J Vet Res 49: 359 362, 1988. 4. Manion TB, Bushmick SL, Mittel L, et al: Lyme disease in horses: Serological and antigen testing differences. Proc AAEP 44:144 145, 1998.
380 Equine Compendium April 2001 5. Burgess EC, Gillette D, Pickett P: Arthritis and panuveitis as a manifestation of Borrelia burgdorferi in a Wisconsin pony. JAVMA 190:1340 1342, 1986. 6. Burgess EC, Mattison M: Encephalitis associated with Borrelia burgdorferi infection in a horse. JAVMA 191:1457 1458, 1987. 7. Chang Y F, Novosol V, McDonough SP, et al: Experimental infection of ponies with Borrelia burgdorferi by exposure to Ixodid ticks. Vet Pathol 37:68 76, 2000. 8. Chang Y F, Novosol V, McDonough SP, et al: Vaccination against Lyme disease with recombinant Borrelia burgdorferi outer-surface protein A (rospa) in horses. Vaccine 18:540 548, 2000. 9. Chang Y F, McDonough SP, Chang C F, et al. Human granulocytic ehrlichiosis agent infection in a pony vaccinated with a Borrelia burgdorferi recombinant OspA vaccine and challenged by exposure to naturally infected ticks. Clin Diagn Lab Immunol 7(1):68 71, 2000. 10. Schoen AM: Veterinary Acupuncture. Goleta, CA, American Veterinary Publications, Inc., 1994, pp 561 562. 11. Franz JK, Priem S, Rittig MG, et al: Wien Klin Wochenschr 111(22 23):981 984, 1999. 12. Kloppenburg M, Dijkmans BA, Breedveld FC: Antimicrobial therapy for rheumatoid arthritis. Baillieres Clin Rheumatol 9(4):759 769, 1995. 13. Whiteman M, Kaur H, Halliwell B: Protection against peroxynitrite dependent tyrosine nitration and alpha 1 antiproteinase inactivation by some anti inflammatory drugs and by the antibiotic tetracycline. Ann Rheum Dis 55(6):383 387, 1996. About the Authors Dr. Divers is affiliated with the Department of Clinical Sciences and Cornell University Hospital for Animals; Drs. Chang and Jacobson are affiliated with the Department of Population Medicine and Diagnostic Sciences; and Dr. McDonough is affiliated with the Department of Biomedical Sciences; Cornell University, Ithaca, New York. Dr. Divers is a Diplomate of the American College of Veterinary Internal Medicine and American College of Veterinary Emergency and Critical Care. ARTICLE #6 CE TEST The article you have read qualifies for 1.5 contact hours of Continuing Education Credit from the Auburn University College of Veterinary Medicine. Choose the one best answer to each of the following questions; then mark your answers on the test form inserted in Compendium. 1. Which mammals are most commonly involved in maintaining the B. burgdorferi life cycle? a. dog, cat b. horse, pony
Compendium April 2001 Equine 381 c. deer, white-footed mouse d. opossum, raccoon e. skunk, squirrel 2. Which geographic area(s) of the United States have the highest prevalence of Lyme disease in humans? a. Texas b. northeastern states, Wisconsin, Minnesota c. southwestern states d. northwestern states e. southeastern states 3. Equine experimental studies indicate that B. burgdorferi has a predilection for all of the following sites except a. synovial membranes. d. the liver. b. regional lymph nodes. e. perineural tissue. c. regional fascia. 4. Experimentally infected ponies were seropositive (determined by ELISA) by weeks after exposure. a. 2 to 3 b. 5 to 6 c. 8 to 10 d. 12 to 14 e. not all exposed ponies seroconverted 5. Which of the following statements regarding transmission of B. burgdorferi to mammals is false? a. Transmission in North America is only by Ixodes species ticks. b. Adult larval or nymph stages may transmit the infection. c. Adult males are more likely than adult females to transmit the infection. d. The infected tick must be attached to the mammal for several hours for B. burgdorferi transmission to occur. e. Transmission by adult ticks generally occurs in the summer or fall. 6. The Western blot test to detect antibody against specific B. burgdorferi antigens is best indicated a. when the KELA is less than 110 units. b. within the first week after exposure. c. when the KELA is greater than 440 units. d. within the first 2 weeks after exposure. e. when the KELA is between 110 and 440 units. 7. The two most commonly used antimicrobials for treating equine Lyme disease are a. tetracycline and doxycycline. b. ceftiofur and chloramphenicol. (continues on page 391)
Compendium April 2001 Equine 391 Lyme Disease in Horses (continued from page 381) c. chloramphenicol and doxycycline. d. tetracycline and enrofloxacin. e. gentamicin and ceftiofur. 8. Ticks infected with B. burgdorferi may be dually infected with, another equine pathogen. a. Ehrlichia risticii b. Ehrlichia equi c. Pythium insidiosum d. Eperythrozoon wenyoni e. Eperythrozoon equi 9. An experimental OspA vaccine was able to a. prevent clinical signs but not infection in ponies. b. prevent short-term infection in ponies. c. produce antibodies in the vaccinated ponies that were indistinguishable from those seen in natural infections. d. prevent long-term (at least 1 year) infection in ponies. e. prevent infection but had noticeable side effects. 10. Detection of B. burgdorferi DNA by PCR testing has been successful in clinical cases on a. liver biopsy. b. renal biopsy. c. synovial fluid. d. synovial biopsy. e. cerebrospinal fluid.