Synovial sepsis is associated with substantial morbidity and mortality (Martens et al, 1986); it is defined as infection of the synovial joint by pathogens and can occur with or without multisystem disease. In neonatal foals, synovial sepsis is frequently associated with haematogenous spread of infection. Once infection is established in a synovial joint the resulting inflammatory response can result in deleterious effects on the articular cartilage and joint. Rapid diagnosis and initiation of suitable antimicrobial treatment is vital for a positive outcome. The mainstays of treatment are broad spectrum systemic antimicrobial therapy and lavage of the infected joint. In neonatal foals, treatment is further complicated by physiological differences and comorbidities that affect the distribution and elimination of drugs.
Aetiology and pathogenesis
Septic arthritis can occur as the result of a periarticular infection, inoculation of the joint by a puncture wound, iatrogenic causes or by haematogenous spread (Hardy, 2006). In foals, haematogenous dissemination of bacteria is the most common route of transmission. The presence of bacteraemia and septicaemia is an important risk factor, with septic arthritis reported to occur in 14–38% of neonatal foals with sepsis (Gayle et al, 1998; Corley at al, 2007). Another factor that plays an important role in the development of joint infection is the foal's immune defence, which relies upon passively acquired immunity. Failure of transfer of passive immunity is the highest risk factor for the development of bacteraemia in foals, and therefore plays an important role in the development of septic arthritis (Robinson, 1993).
Haematogenous dissemination of bacteria can result in inoculation of infection within the joint or the bone. Lesions are classified according to their location and types include: type S (involving the synovial membrane), type E (involving the epiphyseal subchondral bone), type P (involving the metaphyseal side of the physis) and type T (infection of the small tarsal or carpal bones in premature or dysmature foals). Haematogenous inoculation of the joint is facilitated by the highly vascular nature of the synovium and epiphyseal subchondral bone. Transphyseal vessels are present in neonatal foals and allow communication of the metaphysis and epiphysis. Consequently, infections in young foals primarily involve the synovium or epiphyseal bone (Firth, 1983). Blood flow through these vessels is slow which leads to enhanced bacterial adhesion and invasion of adjacent structures (Firth, 1983). In older foals, type P lesions are more common. This difference is associated with the regression of the transphyseal vessels which is reported to occur anywhere from 7–45 days of age (Firth, 1983).
Once localised within the joint, the presence of microorganisms initiates an inflammatory response. This results in hyperaemia and acute inflammation of the synovium, with polymorphonuclear neutrophils infiltrating the synovium and the lumen of the joint (Alderson and Nade, 1987). The release of proinflammatory cytokines and proteolytic enzymes contributes to further vascular congestion and neutrophilic inflammation (Annear et al, 2011). This local inflammatory response has negative effects on joint health and destruction of articular cartilage rapidly occurs.
Diagnosis
History and clinical signs of synovial sepsis in neonates A complete history should be obtained as this often indicates risk factors for failure of transfer of passive immunity (such as the mare running milk before foaling or the foal failing to ingest adequate colostrum). Foals frequently present with a recent or concurrent history of septicaemia or a localised septic focus, therefore the history should include information relating to the foal's overall health status.
In any foal presenting with lameness, septic arthritis, osteomyelitis or physitis should be suspected. Clinical signs of septic arthritis include heat over the joint, effusion, peri-capsular swelling, pain on palpation of the joint and lameness. The anatomical location can often be established based on heat, pain and/or joint effusion, although for certain joints (such as the shoulder, elbow or hip) this may be more difficult to detect (Figure 1). Arthrocentesis of these joints should therefore be performed in foals with unidentified lameness (Hardy, 2006). It is also important to consider the risk of infection in multiple joints, which is relatively common in young foals. Lameness may be variable early in the disease process, therefore palpation of all joints is recommended as part of any neonatal foal examination (Annear et al, 2011).
In cases of septic arthritis, the systemic condition of the foal should also be carefully evaluated to identify any concurrent disease. This is particularly important in foals with risk factors such as failure of passive transfer or sepsis. Evidence of septicaemia or focal infection are commonly identified, with the umbilicus, gastrointestinal tract or respiratory tract most frequently implicated. In addition to the clinical examination, basic evaluation should also include ultrasound examination of the thorax, abdomen and umbilical structures, with additional diagnostics dependent on the presenting clinical signs. The foal should also be monitored closely for the development of infection at any additional sites.
Laboratory findings
Blood evaluation typically reveals evidence of an inflammatory response, including increased serum amyloid A, hyperfibrinogenaemia and a neutrophilic leucocytosis. Serum amyloid A is a non-specific indicator of inflammation and has been shown to be significantly higher in neonatal foals with sepsis than in healthy foals or sick non-septic foals (Barr and Nieman, 2022). IgG concentrations should be also assessed as septic arthritis is often a consequence of failure of passive transfer. Blood cultures should be obtained at the time of initial evaluation given the importance of bacteraemia as a risk factor.
Arthrocentesis
The collection and analysis of synovial fluid is a fundamental part of the assessment of joint sepsis. This is usually straightforward from a distended joint, although it can be more challenging in chronic cases. The joint should be prepared aseptically and an 18- or 20-gauge needle directed through the joint capsule. The gross appearance of synovial fluid from an infected joint will be serosanguinous, cloudy and will have decreased viscosity. Increased total protein (>4g/dl), total nucleated cell counts (greater than 10 000–30 000 cell/ml) and percentage of neutrophils greater than 90% are strongly suggestive of joint sepsis (Glass and Watts, 2017). Neutrophils may degenerate and bacteria can occasionally be seen on a Gram stain of the fluid, although this is not required to confirm sepsis.
In addition to cytological evaluation, a sample should also be submitted for bacterial culture and sensitivity. Other synovial parameters that have been investigated include synovial lactate, glucose and pH (Dechant et al, 2011). Reports suggest that synovial lactate may be useful, but cutoff values have yet to be established (Glass and Watts, 2017).
Diagnostic imaging
Owing to the high risk of bone involvement, radiographic examination is an essential part of the assessment in foals with suspected septic arthritis (Figure 3). However, there is often a lag between the initial infection and the development of osseous changes which makes early recognition challenging (Paradis, 2010).
Computed tomography (CT) can demonstrate osteomyelitis in some foals before radiographic changes become evident, and is therefore helpful in identifying bony involvement at an earlier stage. Alternatively, if bony involvement is suspected but no abnormalities were noted on initial evaluation, then repeated radiography (every 3–5 days) is recommended.
As well as aiding the early identification of bony lesions, advanced imaging such as magnetic resonance imaging (MRI, as well as CT) can also be useful for identifying lesions in the proximal limbs, where overlying soft tissues and incomplete ossification of subchondral bones limits evaluation (Glass and Watts, 2017) (Figures 4 and 5). While MRI is the modality of choice in human patients, it is rarely used in the diagnosis and evaluation of septic arthritis in foals (Annear et al, 2011). CT is more commonly used because of the speed of image acquisition, the larger gantry which allows the entire body of the foal to be imaged, and the superior contrast resolution for imaging bone (Glass and Watts, 2017).
Ultrasound may also provide useful additional information and can be used to assess the peri-articular soft tissues and the echogenicity and appearance of joint effusion. It offers the advantage of being readily available, but provides limited information regarding the amount of bone or cartilage involvement (Paradis, 2010).
Treatment
Successful treatment of neonatal synovial sepsis depends upon early diagnosis and initiation of appropriate therapy (Vos and Ducharme, 2008; Wright et al, 2017). Mainstays of treatment, as in adults, include antimicrobial therapy and surgical removal of inflammatory and infectious material from the joint (Annear et al, 2011; Glass and Watts, 2017). Owing to the importance of rapid treatment, initial antimicrobial therapy is usually empirical while bacterial culture and sensitivity results are pending.
Neonatal considerations
There are several pharmacological and physiological differences between adults and neonates that need to be considered when deciding on appropriate antimicrobial treatment. Neonatal foals have a higher body water content, resulting in the need for higher dosages of hydrophilic drugs, while lipophilic drug dosages are normally unaltered (Carlson et al, 1979; Andrews et al, 1997; Fielding et al, 2003, 2011; Corley and Hollis, 2009). As a foal matures the rate of hepatic metabolism and renal excretion increase, making it difficult to infer dosages based on adult pharmacokinetics (Adamson et al, 1991; Swain O'Fallon et al, 2021). The treatment of critically ill neonatal foals with synovial sepsis is often further complicated by the presence of multisystem disease and sepsis. Factors such as organ dysfunction, hypovolaemia and the administration of other medications can all alter the bioavailability of drugs (Corley and Hollis, 2009; Floyd et al, 2022).
Commonly involved bacteria and guidelines for antimicrobial selection
Empirical selection of antimicrobials is necessary in neonatal foals with synovial sepsis, while the results of culture and sensitivity are pending, to ensure rapid treatment. Broad spectrum intravenously administered bactericidal antimicrobials should be selected pending culture results (Magdesian, 2017). The foal's immature immune system necessitates the use of bactericidal drugs. The intravenous route is favoured as sick foals frequently have hypoperfusion of their gastrointestinal tract and muscle, leading to reduced absorption of antimicrobials via the intramuscular or oral routes (Magdesian, 2017).
In neonatal foals with synovial sepsis, the bacteria most commonly infect the joint secondary to bacteraemia and haematogenous spread. Several studies have summarised the most frequently identified bacteria in septic neonatal foals (Brewer and Koterba, 1990; Marsh and Palmer, 2001b; Corley and Hollis, 2009, Theelen et al, 2014, 2019; Floyd et al, 2022). Both Gram-positive and Gram-negative bacterial isolates are commonly identified and polymicrobic infections occur frequently (Brewer and Koterba, 1990; Gayle et al, 1998). Analysis of previously published studies of septic neonates suggests the most commonly identified isolates from blood cultures are Escherichia coli, Actinobacillus, Enterobacter, Enterococcus, Streptococcus and Klebsiella in descending order of frequency isolated (Corley and Hollis, 2009). Historically, Gramnegative bacteria have been the most common isolates in neonatal foals, with the prevalence of Gram-positive isolates increasing in older foals (Wilson and Madigan, 1989; Marsh and Palmer, 2001a, b; Russell et al, 2008). One study tracking bacterial isolates and sensitivities in neonatal septic foals over 30 years revealed that while Gram-negative bacteria continue to be the most frequently detected, the incidence of Gram-positive bacteria is increasing (Theelen et al, 2014). It is clear from several published studies that bacterial prevalence and susceptibility are both geographic and hospital dependent. Monitoring and reviewing previous bacteriology and sensitivity results can therefore help guide empirical selection of antimicrobials in neonatal foals with septic arthritis (Marsh and Palmer, 2001a, 2001b; Theelen et al, 2014).
Ampicillin coupled with an aminoglycoside such as amikacin is usually considered a reasonable first line treatment for neonatal sepsis (Theelen, 2014). Ampicillin is often preferred to penicillin because of its better Gram-negative coverage, increased sensitivity to Enterococcus spp. and better bone penetration (Theelen 2014; Magdesian, 2017; Floyd et al, 2022). Historically, cephalosporins have been used frequently in cases of neonatal sepsis and septic arthritis because they are bactericidal, broad spectrum, easily administered and have few side effects. However, cephalosporins are classified by the World Health Organization as high priority critically important antimicrobials, hence attempts are being made to reduce their use in equine practice (Floyd et al 2022). Tetracyclines are lipophilic and have good bone penetration and are therefore sometimes used to treat septic arthritis or physitis. While they are a suitable choice for foals with localised bacterial infections, they are not considered an ideal choice for neonatal foals with generalised sepsis because they are bacteriostatic and lack efficacy against Enterococcus and some vital Gram-negative bacteria (Floyd et al, 2022). Adequate length of treatment is important and it is recommended that antimicrobial therapy is continued until clinical signs have resolved and bloods have normalised. Cases of osteomyelitis frequently require treatment for several months to adequately resolve the infection. Dosages of commonly used antimicrobials in foals are summarised in Table 1.
Table 1. Dosages of commonly used antimicrobials in foals
| Drug | Dosage | Route of administration | Frequency of administration (hours) |
|---|---|---|---|
| Amikacin sulphate | <1 week old 25 mg/kg, 2–4 weeks old 20–25 mg/kg | Intravenous | 24 |
| Ampicillin sodium | 20 mg/kg | Intravenous | 6 |
| Amoxicillin clavulanate | 30 mg/kg | Per os | 8 |
| Azithromycin | 10 mg/kg | Per os | Every 24 for 5 days then every 48 |
| Ceftiofur sodium | 5–10 mg/kg | Intravenous, intramuscular, subcutaneous | 12 |
| Cefotaxime | 40 mg/kg | Intravenous | 6 |
| Ceftriaxone | 25 mg/kg | Intravenous | 12 |
| Clarithromycin | 7.5 mg/kg | Per os | 12 |
| Doxycycline | 10 mg/kg | Per os | 12 |
| Gentamicin | <14 days old 12 mg/kg | Intravenous | 36 |
| Marbofloxacin | 2 mg/kg | Intravenous | 24 |
| 4 mg/kg | Per os | 24 | |
| Metronidazole | 10 mg/kg | Intravenous, per os | 12 |
| Oxytetracycline | 5–10 mg/kg | Intravenous | 12 |
| Piperacillin tazobactam | 30 mg/kg* | Intravenous | 6 |
| Procaine penicillin G | 20 | Intramuscular | 12 |
| Sodium penicillin | 20 000–44 000 iu/kg | Intravenous | 6 |
| Trimethoprim-sulfonamide | 30 mg/kg | Per os | 12 |
| Rifampin | 5 mg/kg | Per os | 12 |
Potential toxicities
In foals, arthropathy and nephrotoxicity are the antimicrobial side effects of most concern. The aminoglycosides and oxytetracycline are most frequently associated with nephrotoxicity (Vivrette et al, 1993). In neonatal foals with sepsis, the likelihood of acute injury kidney is increased as a result of the high incidence of hypovolaemia and hypoperfusion. Therapeutic drug monitoring is recommended when aminoglycoside antibiotics are used to reduce the chance of toxicities occurring. Suggested target peak and trough concentrations for aminoglycosides are included in Table 2. Fluoroquinolones such as enrofloxacin have been associated with arthropathy (Vivrette et al, 2001; Giguére et al, 2013). However, marbofloxacin has been used in foals without reports of similar musculoskeletal side effects.
Table 2. Therapeutic drug monitoring
| Drug | 30-minute peak (µg/ml) | 8-hour level (µg/ml) | 20-hour trough (µg/ml) |
|---|---|---|---|
| Amikacin | >53–60 | 15–20 | <2 |
| Gentamicin | >30–40 | 3–5 | <1 |
Therapeutic drug monitoring involves measuring peak and trough drug concentrations to ensure drug efficacy and adequate excretion.
Regional treatment
Local antimicrobial therapy can be used in addition to systemic antimicrobial therapy but is not adequate alone for treatment owing to the high incidence of systemic sepsis in neonates. Regional therapies including intrasynovial injections, intravenous regional limb perfusion (Figure 6), antibiotic-impregnated beads or intrasynovial constant rate infusions can be used to achieve high drug concentrations in the synovial fluid (Hall et al, 2012; Richardson and Stewart, 2019). The potential for systemic absorption of intra-articular anti-microbials is an important consideration in neonates. For example, foals receiving systemic and intra-articular aminoglycosides concurrently may have a higher potential for toxicity and adjustment of the dosages or therapeutic drug monitoring may be warranted.
Surgical joint lavage and bone debridement
Joint lavage is recommended to help remove the debris and bacterial products associated with synovial sepsis. It can be performed via needle lavage, arthroscopy or arthrotomy (Hall et al, 2012; Barceló Oliver et al, 2017; Richardson and Stewart, 2019). In a simple, acute septic joint without bone or physeal involvement a simple through-and-through needle lavage may be adequate (Figure 7). This can be performed with the foal sedated or lightly anaesthetised. Lavage is performed aseptically with isotonic, crystalloid fluids via large gauge needles (14–16 gauge) placed to allow flushing across the entire joint. Following lavage and drainage of fluid from the joint, intra-articular antimicrobials are instilled. Arthroscopic lavage is preferred in cases that are more chronic, have bone involvement or have failed to respond appropriately to needle lavage. Arthroscopy allows visualisation of the articular surfaces, and more effective removal of infectious and inflammatory debris, as well as providing more prognostic information. The presence of concurrent osteomyelitis or physitis usually necessitates a longer duration of treatment.
Anti-inflammatory and analgesic treatment
Cartilage damage can develop secondary to the inflammatory cascade caused by infection of the joint. Anti-inflammatories are therefore necessary to preserve joint health and ensure the foal's comfort while the infection is cleared. Neonatal foals with synovial sepsis often develop varying levels of lameness, stiffness, fever and depression secondary to the inflammatory response. In some cases they may struggle to stand, ambulate and nurse sufficiently. Non-steroidal anti-inflammatory drugs (NSAIDs) are therefore necessary to combat the inflammatory cascade and provide pain relief. Neonates with sepsis are susceptible to gastrointestinal and renal toxicities so NSAIDs must be used judiciously. If NSAIDs are necessary at higher doses or for longer durations, then monitoring of renal function is advised. While NSAIDs such as flunixin meglumine can be used, COX-2 selective NSAIDs such as meloxicam are preferred to reduce the likelihood of toxicities occurring. Phenylbutazone is usually avoided in neonatal foals because of the risk of side effects. Multimodal analgesia can be used to reduce the dosages of NSAIDs required. Topical application of an opioid, such as transdermal fentanyl patches, is easy in neonatal foals and well tolerated with few side effects (Table 3).
Table 3. Dosages of anti-inflammatories and analgesia in neonatal foals
| Drug | Dosage | Route | Frequency of administration (h) |
|---|---|---|---|
| Flunixin meglumine | 1.1 mg/kg | Intravenous (IV) or per os (PO) | Every 24 |
| Meloxicam | 0.6 mg/kg | IV or PO | Every 12 dosing in foals <6 weeks of age |
| Fentanyl | Patches deliver 25 g/hour. Dose 1–2 patches per 50 kg foal | Transdermal patches | Patches last 72 hours, change 12 hours before next patch is due |
Prognosis
Several studies have investigated the factors impacting both survival and athletic performance of foals with septic arthritis. The prognosis for short-term survival from synovial sepsis in neonates is usually good, but the prognosis for long-term survival and athletic performance is often less optimistic (Steel et al, 1999; Smith et al, 2004; Vos and Ducharme, 2008; Neil et al, 2010; Wright et al, 2017). In neonates the prognosis is usually worse with multiple joint involvement, intra-articular Gram-negative infections, mixed bacterial infections, concurrent osteomyelitis, multisystem disease, and prolonged joint disease (Raisis et al, 1996; Smith et al, 2004; Vos and Ducharme, 2008; Annear et al, 2011; Hepworth-Warren et al, 2015). Multiple joint involvement has repeatedly been associated with a poorer prognosis, with one study finding the probability of a foal surviving to hospital discharge decreased by 50% for every additional joint that was affected (Hepworth-Warren et al, 2015). The median number of joints affected per foals was also significantly less for older foals between 31–180 days compared with foals <30 days of age (Hepworth-Warren et al, 2015).
Several studies have indicated that the presence of concurrent osteomyelitis decreases the likelihood of survival as well as affecting future athletic performance compared with unaffected foals (Schneider et al, 1992; Raisis et al, 1996; Steel et al, 1999; Smith et al, 2004). Furthermore, critically ill neonates and foals <30 days of age at the time of osteomyelitis development were less likely to survive to discharge (Neil et al, 2010). This is often attributed to the high prevalence of failure of passive transfer of immunity, multisystem disease and multiple joint/bone involvement in neonatal foals (Neil et al, 2010).
One study of Thoroughbred foals with septic arthritis found that affected foals were less likely to race than unaffected foals from the same dam, and that foals that survived to discharge took significantly longer to start in their first race compared to the sibling population. Multisystem disease was associated with a decreased likelihood of survival, but this did not affect the likelihood of the foal racing if the foal survived to discharge (Smith et al, 2004).
Conclusions
Septic arthritis and osteomyelitis are frequently recognised sequelae to systemic sepsis and bacteraemia in neonatal foals. Partial or total failure of transfer of passive immunity is often concurrently present and owing to the foal's immature immune system, intravenous broad-spectrum, bactericidal antimicrobials are preferred. Dosages of antimicrobials used should consider age-dependent physiological differences that alter the metabolism and excretion of drugs. The presence of multi-system disease increases the likelihood of drug toxicities occurring and careful monitoring is recommended. Overall, while there is often a good prognosis for resolution of infection with rapid and appropriate treatment, the prognosis for the foal's future athletic performance is more uncertain.
KEY POINTS
- In neonatal foals synovial infections are usually the result of haematogenous spread from systemic sepsis.
- Owing to their immature immune systems synovial sepsis in neonates should be treated with intravenous, broad-spectrum, and bactericidal antimicrobials.
- Multisystem disease is often concurrently present making the risk of side effects from treatment higher and close monitoring necessary
- Although the prognosis for resolution of joint infection is generally good, the prognosis for future athletic performance is often more uncertain. Rapid diagnosis and treatment of synovial sepsis is important to reduce the morbidity and mortality associated with the condition.