Hematogenous Septic Arthritis in the Dog: Results of Five Patients Treated Nonsurgically With Antibiotics

Introduction

Few reports exist evaluating treatment outcomes in dogs with septic arthritis.^1–4 Due to the lack of reports in the dog, many treatment recommendations have been extrapolated from other species.^5–10 Extrapolation from other species, in particular the horse, combined with the concern of potential devastating consequences of unsuccessful treatment of joint sepsis, has led to recommendations of invasive surgical therapies including arthrotomy and synovectomy.⁸⁹ However, recent reports of septic arthritis in humans suggest better outcomes with minimally invasive techniques (e.g., needle aspiration or arthroscopy) than with surgical drainage and lavage.^11–15 One of these studies reported full recoveries in 67% of patients with needle aspiration compared to 42% of patients treated surgically.¹¹

Two common etiologies exist for septic arthritis: infection following a penetrating wound (from surgery or trauma) and hematogenous spread.¹ When infection involves a foreign body or infectious nidus, as is common with penetrating wounds, surgery is often required to resolve the infection.³ However, only one report in the dog has attempted to separate treatment and outcome based on underlying etiology, in which it was found that hematogenous septic arthritis had favorable outcomes when treated only with antibiotics.³ The purpose of this study was to evaluate the effectiveness of nonsurgical treatment (i.e., antibiotics and needle joint lavage) for hematogenous septic arthritis in the dog.

Materials and Methods

Five cases were selected with the diagnosis of hematogenous septic arthritis from hospital records spanning 1994 to 2000 from Colorado State University and Louisiana State University. For case inclusion in this study, patients must not have had any previous or preexisting regional surgery or trauma to the affected limb and have evaluation of clinical outcome beyond 1 year. Diagnosis was based on clinical signs, joint fluid analysis, radiography, microbiology, and response to treatment. Patients were fully evaluated for presence of coexisting infections (e.g., dermatitis, otitis, heart murmur, etc.). Data retrieved included signalment, duration of lameness, limb and joint affected, rectal temperature, joint fluid cytopathology, joint fluid culture and sensitivity, antibiotic therapy, and treatment outcome. For synovial collection, patients were sedated, regionally shaved, aseptically prepped, and had arthrocentesis performed. Some patients received needle lavage with sterile saline using one ingress needle and one egress needle. Two 18-gauge needles were placed, and the joint was lavaged with 1 to 3 liters of sterile saline. Flexion and extension of the joint were performed during the later stages of lavage, with the intent to improve effectiveness of lavage by diminishing canalization between the ingress and egress needles. An empirically chosen intravenous (IV) antibiotic was administered in many cases for the first 24 hours; then, pending clinical response, oral antibiotics were administered for the duration of treatment. Antibiotic choice was reassessed following evaluation of synovial fluid culture and sensitivity.

Results

All dogs were sexually intact; four were male and one was female [see Table]. All dogs were immature from 2 to 9 months of age (median, 5 months), and there were three giant breeds, one large breed, and one medium-sized dog. Anatomical distribution of infected joints consisted of elbow (4/5) and tarsus (1/5). Multiple joint involvement was present in one dog (case no. 4, both elbows). Time from onset of clinical signs to treatment ranged from 1 day to 1 month; however, three dogs were treated within 3 days of clinical onset. Rectal temperature was elevated (>103°F) in four of five dogs. Radiographic evaluation was performed in four of five patients, which showed soft-tissue swelling and joint effusion without radiographic changes at the articular surface at the time of diagnosis. Marked synovial effusion and arthralgia with distal limb edema were frequent findings. Arthrocentesis with decompression of the joint was performed in all dogs. Synovial fluid evaluation showed changes consistent with sepsis on gross evaluation, including increased volume, opaque clarity, color changes, and loss of viscosity. Two-needle joint lavage was performed in two (case nos. 2, 3) of five dogs with 1 to 3 liters of saline, but it did not appear to affect long-term outcome in these cases.

Organisms cultured from synovial fluid included Streptococcus B-haemolytic spp. (1/3), Pasteurella multocida (1/3), and Staphylococcus intermedius (1/3). Synovial fluid culture was declined by two owners. All synovial fluid cultures obtained (3/3) showed positive growth. Antibiotics utilized included cephalosporins (3/5), amoxicillin-clavulanic acid (3/5), and enrofloxacin (1/5). Duration of antibiotic administration ranged from 10 to 56 days, with four of five patients receiving antibiotics for at least 21 days (mean, 36 days). Case no. 1 was only treated for 10 days initially and had recurrence of clinical signs within 10 days, requiring further antibiotic treatment. Case no. 4 had cultured Pasteurella multocida from synovial fluid but was continued on only amoxicillin-clavulanic acid with a poor response to treatment. Both of these patients have persistent lameness. Long-term follow-up (>1 year) by owner interview and questionnaire was achieved in five cases; four owners noted improvement within 3 days of initiating antibiotic therapy, and clinical signs resolved within 1 month.

Discussion

In this case series, an effort was made to identify dogs with suspected hematogenous septic arthritis by identifying patients with septic arthritis but without an overt underlying cause of infection (e.g., previous surgery, penetrating wound, dermatitis, gingivitis, otitis, etc.). In humans, an extra-articular nidus of infection was found in approximately 50% of patients with suspected hematogenous septic arthritis.¹⁶ Therefore, great care must be taken in fully evaluating patients with suspected hematogenous septic arthritis. A complete physical examination should be performed, with special attention given to skin, ears, mouth, urinary tract, and chest.

Bacteria isolated in this study and previous studies included Staphylococcus spp., Streptococcus spp., and Pasteurella, which suggests that the primary sources are likely skin, oral cavity, and the environment.¹³⁴ Although synovial biopsy culture has been reported to increase culture sensitivity,¹ the results of the authors’ study and another recent study showed 100% positive cultures with synovial fluid.³ Synovial biopsies require surgical invasion and greater expense, which may not be justified or feasible in all cases. Furthermore, synovial fluid incubated 24 hours in blood culture medium (e.g., nine parts medium to one part blood by volume) was found to be significantly more reliable than synovial membrane biopsy.¹⁷

The most common site of suspected hematogenous septic arthritis in this case series was the elbow (4/5), but the tarsus was also represented. The site prevalence of hematogenous septic arthritis is difficult to compare, because few reports clearly separate hematogenous septic arthritis; however, nearly all major joints are represented in the literature.¹³⁴ Severe, acute, nearly nonweight-bearing lameness was a frequent finding, although chronic lameness has been reported even in hematogenous cases.³ Marked synovial effusion and arthralgia with distal limb edema were frequent findings in the authors’ cases and are consistent with previous reports.¹³ Arthrocentesis with decompression of the joint was performed in three dogs, and two-needle joint lavage was performed in two dogs; but long-term outcome in these cases was not affected. This lavage technique has been shown to only partially flush debris from the joint because of the canalization that occurs between the ingress and egress needles.¹³ To fully achieve thorough joint flushing with a less invasive technique, arthroscopy has been found to be the technique of choice in humans.^18–22

The therapeutic spectrum of empirically chosen antibiotics reflected the reported bacterial isolates in the published literature¹³⁴ and appears to be appropriate based on the authors’ outcomes. Cephalosporins and penicillin-type antibiotics were the most common first choice, with enrofloxacin included in the treatment regimen when Pasteurella was cultured (case no. 4); this dog had continued resistance problems, persistent lameness, and was never treated with an antibiotic to which the bacterium was susceptible. Ideally, this case should have more aggressive monitoring, including repeat joint fluid analysis with culture and sensitivity, to better direct therapy; however, because of the retrospective nature of this study, the reasons these were not done are unavailable for study. The authors’ preferred treatment regimen consisted of IV antibiotics for the first 24 hours, followed by oral antibiotics and hospitalization until culture results were evaluated. The intent of the IV antibiotics with IV fluids is to ensure proper hydration with aggressive antibiotic treatment, with the expectation of a more rapid clinical response that can be closely monitored. Interestingly, case no. 2 was placed only on oral antibiotics and showed rapid recovery with an excellent outcome, as has been reported for dogs with hematogenous septic arthritis in a previous study in which IV antibiotics were not used.³

Duration of antibiotic administration in the authors’ study ranged from 10 to 56 days. Case no. 1 initially received only a 10-day course of antibiotics. Although this dog showed initial dramatic improvement, clinical signs recurred within 10 days following discontinuation of antibiotics. Despite long-term aggressive treatment with antibiotics, this dog has permanent functional joint deficits with pain and loss of range of motion. Reported cases of hematogenous arthritis also have a wide range in duration of antibiotic administration, with 10 to 56 days reported and mostly good outcomes. However, some dogs that were treated only 2 weeks with antibiotics had poor outcomes.³ The authors’ findings are in agreement with the literature recommendation of antibiotic administration for 4 to 6 weeks.¹⁴⁸

Conclusion

Despite the recommendations in textbooks for aggressive surgical management,⁸⁹ the results of this study and a previous study³ suggest that suspected hematogenous septic arthritis in the dog can be successfully managed by less invasive treatments.