Managing Orthopedic Infections
Orthopedic infections in horses can be devastating to soundness, athletic careers, and even life. In the short term, the inflammation and pain of sepsis disrupt normal function, and if sepsis is unsuccessfully treated, long-term function is affected because of degenerative joint disease, fibrous tissue restriction of joint capsule motion, or failure of bone to properly heal. Equine orthopedic infections may originate from endogenous or exogenous routes and can involve bone (osteitis, osteomyelitis) or synovial compartments such as joints, tendon sheaths, and bursae. Injuries to the distal part of the limbs in horses are quite common because of the anatomy and exposed nature of the horse’s limbs and frequently result in orthopedic infections. However, timely diagnosis and aggressive treatment often result in a satisfactory return to normal function. This chapter reviews the diagnosis and therapeutic options for the management of orthopedic infections.
Etiopathogenesis and Clinical Signs
An accurate medical history is essential for the veterinarian to completely understand the etiopathogenesis of septic synovitis or osseous infections. Younger foals are typically affected with infection of the joints, physis, or epiphyses after systemic distribution of bacteria from a site distant to the affected structure. The vascular anatomy of the metaphyseal and epiphyseal plate predisposes the foal to these types of infections because circulating blood tends to sludge at these sites and bacteria can become established. The most common pathogens isolated in foals with septic arthritis are the Enterobacteriaceae, most notably Escherichia coli. Salmonella spp and Klebsiella spp are also commonly cultured. Gram-positive bacteria, including Streptococcus spp, Staphylococcus spp, and Rhodococcus equi can also cause infection. Foals are presented for veterinary evaluation because of lameness, synovial effusion, or localized heat and edema in one or multiple limbs.
Bacteria most commonly reach the synovial compartments of adult horses through a wound or because of poor aseptic technique during fracture repair, joint surgery, or intraarticular injection. Horses with septic arthritis usually develop severe lameness 3 to 21 days after direct trauma to a synovial structure. Occasionally, extension of contamination into a synovial compartment is delayed if the capsule is damaged such that necrosis and sloughing of overlying capsular tissue occurs several days after a tissue injury. Horses with wounded synovial structures may be comfortable if the synovial compartment is draining or the horse is receiving high doses of nonsteroidal antiinflammatory drugs but may become painful when the wound heals by second intention, closing and entrapping septic, inflamed fluid.
Osteomyelitis and osteitis in adult horses require inoculation of bacteria and disruption of the blood supply to the cortical bone or medullary cavity, and this usually occurs after direct trauma to the bone (e.g., lacerations or fractures). Traumatic injury that strips the periosteum exposes the bone surface to the environment and compromises the bone’s vascular supply. The consequence is formation of a piece of dead bone called osseous sequestrum. The clinical signs associated with this type of infection include localized swelling, heat, pain on palpation, mild to severe lameness, or a nonhealing wound with purulent discharge that responds to antimicrobial treatment but drains again when antimicrobials are discontinued.
Traumatic bone infection and septic arthritis in adult horses are usually caused by mixed bacterial infections, but frequently one pathogen predominates. Isolates are usually aerobic and include Enterobacteriaceae, Streptococcus spp, Staphylococcus spp, and other gram-negative and gram-positive bacteria. Infections following surgery or iatrogenic infections are often caused by staphylococci. Fungal organisms can also cause orthopedic infections and should be suspected in horses that do not respond to antimicrobials.
The physical examination of a horse suspected of having orthopedic infection should be thorough and systematic. Careful local examination of wounded or suspected infected tissues is indicated. Palpation of the affected structure reveals swelling, which can be from both edema in perisynovial tissues and effusion within the synovial compartment. After aseptic preparation, manual palpation of the wound with a sterile gloved hand can help to assess the depth of a wound as well as quickly identify synovial tissue exposure. If any doubt exists, a sterile malleable probe can be used to investigate punctures, smaller wounds, and the depths of larger wounds inaccessible to fingers. If uncertainty remains, radiographs can be taken with a probe in place to better determine the tissues involved.
The most convincing evidence that a synovial compartment is septic or has been penetrated in association with a wound is to collect a fluid sample for culture and inject a sterile solution into the synovial space and observe for seeping to confirm communication with the wound. Synoviocentesis is performed through healthy skin on the side of a joint opposite from that of the wound, with care taken to avoid areas of severe inflammation. A volume of sterile irrigation fluid sufficient to distend the synovial compartment should be injected under pressure. If the injected fluid distends the synovial compartment without communicating with the wound site, the joint or tendon sheath can be assumed to be closed at the time of examination. However, if the injected fluid becomes evident at the wound site, it must be interpreted that the synovial compartment is open, contaminated, and, for treatment purposes, infected. If synovial fluid can be obtained, nucleated cell count and total protein can be used to confirm infection. Total protein concentrations of 3.5 mg/dL or greater (normal, ≤2.5 mg/dL) and white cell counts higher than 30,000/µL in synovial fluid are associated with synovial sepsis. However, lower cell counts may be seen if substantial deposits of fibrinous material are present. Cytologic evaluation usually reveals more than 90% neutrophils, with most having degenerative changes. Cytology can also help identify the bacteria in the synovial fluid and within synovial white cells. Culture and identification of the infecting microorganism or microorganisms is the definitive confirmation of sepsis and is also tremendously helpful in selecting appropriate therapy. Most often there is an initial period of empirical antimicrobial treatment while synovial fluid cultures are pending.
Radiography must be used in any case in which there is suspicion of septic synovitis, an open wound extending into a joint, or osteomyelitis. Gas densities in a joint space is not a normal finding and should be interpreted as air that has entered a joint space. Radiographs are also used to assess extension of infection into subchondral bone or cartilage damage because septic osteomyelitis may precede or follow septic arthritis. This information helps determine prognosis and modify treatment strategies. Radiographic signs of osteomyelitis include soft tissue swelling, periosteal new bone formation, radiolucent areas of bone lysis surrounded by sclerotic margins, and osseous sequestrum. Serial radiographic examination every 10 to 14 days is recommended to determine the extent and status of the infection and assess the effectiveness of treatment. In addition, positive contrast radiography can also be of great assistance. Contrast material can be injected into a wound to determine whether the underlying synovial space has been entered, or the contrast media can be placed directly in a joint, bursa, or tendon sheath in which penetration is suspected and communication between the wound track and synovial structure determined.
Ultrasonography of joint spaces, bone, and tendon sheaths can also provide useful information. When necessary, ultrasonographic guidance allows accurate aspiration of suspected infected sites for testing. Normal synovial fluid has a uniform anechoic appearance. With sepsis, synovial fluid can contain echogenic particles and strands of material, likely consistent with accumulation of fibrin, inflammatory debris, and foreign material. With experience, ultrasound can be used to detect early changes of osteomyelitis that is not radiographically apparent in cortical or subchondral bone.
Other imaging techniques that have been used to diagnose orthopedic infection in horses include nuclear scintigraphy, magnetic resonance imaging, and computed tomography. These imaging modalities provide better definition and localization of the infection, but their availability is limited to select referral institutions, and in some situations they are cost prohibitive for clients. Laboratory tests such as complete blood count and plasma fibrinogen determination are helpful for diagnosing orthopedic infections but are not specific. High plasma fibrinogen concentration and leukocytosis are common findings, but are not always present. In foals with orthopedic infection, the primary site of infection (i.e., the umbilicus or lungs) must be cultured because the synovial or bone infection will likely result from the same pathogen.
The aim in treating orthopedic infections is eradication of the bacterial load, removal of foreign material, elimination of inflammatory mediators and free radicals, pain relief, and restoration of the normal synovial environment. These objectives are achieved by appropriate administration of antimicrobial drugs, joint lavage, surgical debridement, antiinflammatory drugs, and a postinfection rehabilitation program. In the past decade, various means of local delivery of antimicrobials have been described and have enhanced the success of treating orthopedic infections in horses.
The goal of antimicrobial therapy is to achieve concentrations above the minimum inhibitory concentration in the infected tissues. Therefore combinations of systemic and local administration of antimicrobials are essential. The clinician should administer broad-spectrum antimicrobials initially, and then adjust according to clinical progression, culture, and antimicrobial sensitivity. Intravenous administration of broad-spectrum antimicrobials shold be initiated as soon as the diagnosis of a septic condition is presumed. The intravenous route provides a faster onset of action than the oral or intramuscular route and maximizes drug penetration into the synovial structures. Traditionally, the most common antimicrobials used to treat horses with orthopedic infections are a β-lactam agent combined with an aminoglycoside. However, many other antimicrobials enter the synovial fluid at therapeutic concentrations and are clinically effective. Oral antimicrobial administration is generally preferred when long-term administration is required, but is not recommended in the acute phase of infection because gastrointestinal tract absorption may be erratic, resulting in lower concentrations. The author’s group routinely prescribes oral antimicrobials for 10 to 14 days after resolution of clinical signs and discharge from the hospital.
Systemic administration of antimicrobials is sometimes unreliable because penetration into devitalized and necrotic tissue can be erratic and unpredictable, and this can result in persistent infections. Local delivery of antimicrobial into the area of interest is therefore an essential part of the treatment because it results in high antimicrobial concentration in infected tissues. Methods of local delivery include regional limb perfusion, direct intraarticular infusion of antimicrobials, and use of biodegradable and nonbiodegradable materials that elute antimicrobials.
Intravenous and Intraosseous Regional Perfusion
Regional perfusion can be used to achieve high local concentrations of antimicrobials in a selected region of the limb, particularly in areas of ischemic tissue. The antimicrobial is injected into a superficial vein proximal to the site of infection (intravenous regional perfusion) or into the medullary cavity of a bone in a similar location (intraosseous regional perfusion). For both techniques, a tourniquet is placed proximal and distal to the site of infection to occlude the superficial venous system. As the perfusate is infused, it distends the venous system and enables antimicrobial to diffuse into the tissues. Amikacin and gentamicin are usually the drugs of choice to infuse by regional perfusion, but almost any other antimicrobial can be used (Table 186-1). Antifungals such as amphotericin B can also be infused by regional perfusion. Typically, antimicrobial doses of one third of the systemic dose are diluted with saline solution to a volume of 35 to 60 mL. The more proximal in the limb the infusion, the larger the volume used. After aseptic skin preparation, a 23-gauge butterfly catheter is inserted into the selected vein, and the antimicrobial solution is slowly infused, over a period of 5 to 15 minutes, while the tourniquet is maintained in place for 20 to 30 minutes (Figure 186-1). The procedure is repeated every 24 hours for 3 to 5 days or as needed. Two components of this procedure are necessary to prevent loss of the antimicrobial into the systemic circulation: (1) an Esmarch’s tourniquet with a wide rubber cuff or a pneumatic tourniquet should be used; and (2) sedation and local anesthesia should be used to alleviate patient discomfort during the procedure and prevent limb motion. Complications of regional limb perfusion are rare and self-limited and include hematoma at the injection site, phlebitis or thrombosis of the vein, extravasation of the perfusate, and difficulty injecting the solution into the medullary cavity.
Range of Antimicrobial Doses Commonly Administered by Regional Limb Perfusion*
|Drug||Dosing Information (Minimum to Maximum)|
|Amikacin||125 mg to 1 g|
|Amphotericin B (antifungal)||50 mg|
|Cefazolin||1 g to 2 g|
|Ceftiofur||1 g to 2 g|
|Gentamicin||100 mg to 3 g|
|Imipenem||500 mg to 1 g|
|Potassium penicillin||106 units to 10 × 106 units|
|Ticarcillin||125 mg to 1 g|
|Vancomycin||300 mg to 1 g|
* Antimicrobials are diluted with physiologic saline to a total solution of 35 to 60 mL. Doses are derived from published studies in the literature.