- Top of page
- Materials and Methods
Background: Extraintestinal infections caused by multidrug-resistant (MDR) Escherichia coli and Enterobacter are becoming more common in veterinary medicine.
Objective: To generate hypotheses for risk factors for dogs acquiring extraintestinal infection caused by MDR E. coli and Enterobacter, describe antimicrobial resistance profiles and analyze treatment and clinical outcomes.
Animals: Thirty-seven dogs diagnosed with extraintestinal infection caused by MDR E. coli and Enterobacter spp. between October 1999 and June 2006.
Methods: Retrospective case series assembled from hospital records data, including clinical history before 1st MDR isolation and treatment outcome. Identity and antimicrobial susceptibility profiles were confirmed by standard microbiological techniques for 57 isolates.
Results: Most dogs had an underlying disease condition (97%), received prior antimicrobial treatment (87%), were hospitalized for ≥3 days (82%), and had a surgical intervention (57%). The urinary tract was the most common infection site (62%), and urinary catheterization, bladder stasis, or both were common among dogs (24%). Some dogs were treated with high doses of co-amoxyclavulanate (n = 14) and subsequently recovered even though the isolates showed in vitro resistance to this antimicrobial. Other dogs were successfully treated with chloramphenicol (n = 11) and imipenem (n = 2).
Conclusion and Clinical Importance: Predisposing disease condition, any prior antimicrobial use rather than a specific class of antimicrobial, duration of hospitalization, and type of surgical procedure might be risk factors for acquiring MDR extraintestinal infections. Whereas culture and sensitivity results can indicate use of last-resort antimicrobials such as imipenem for MDR infections, some affected dogs can recover after administration of high doses of co-amoxyclavulanate.
Escherichia coli and Enterobacter spp. are common causes of extraintestinal opportunistic infections in humans.1 Isolates that are resistant to 3rd-generation cephalosporins and ciprofloxacin are becoming more common in health care facilities and the general community.1,2 Multidrug-resistant (MDR) E. coli and Enterobacter have a similar role in opportunistic infections in dogs.3,4 Recently, there has been an increased frequency of infections in veterinary settings,3–9 possibly because of greater use of intensive care facilities, newer generation antimicrobials, and longer hospitalization periods.4
A cluster of extraintestinal infections caused by MDR E. coli occurred in hospitalized dogs at The University of Queensland Veterinary Teaching Hospital (UQVTH) between 2000 and 2001. Clonal expansion of 2 distinct genetic groups of E. coli was confirmed, with each clonal group showing differences in plasmid carriage and resistance profile. However, they shared a common plasmid-mediated AmpC gene blaCMY7, responsible for resistance to 3rd-generation cephalosporins and β-lactam/β-lactamase inhibitor combinations.7 MDR Enterobacter were also isolated at the same time and were more genetically diverse, with 3rd-generation cephalosporin resistance attributed to the presence of plasmid-mediated blaSHV−12 extended-spectrum β-lactamase in 9 isolates and an AmpC blaCMY−2 in the remaining isolate.5
There are few published studies describing risk factors for nosocomial infections caused by MDR bacteria in veterinary medicine.3,4 In human medicine, identified risk factors include underlying condition, age (pediatric or geriatric), immunosuppression, concurrent chemotherapy or radiation therapy, introduction of medical and surgical devices, prolonged hospitalization, and antimicrobial treatment. These could also be clinically relevant factors for MDR nosocomial infections in veterinary hospitals, which may be less common because companion animals are not kept in aged care facilities and immunocompromised/chronically ill animals are often euthanized.3,4
The objective of this case series was to generate hypotheses for risk factors for dogs acquiring extraintestinal infection caused by MDR E. coli and Enterobacter by describing the frequency of exposure to putative risk factors for 37 dogs diagnosed with extraintestinal infection. The study also aimed to describe antimicrobial resistance profiles of the isolates and treatment and clinical outcomes of the cases.
- Top of page
- Materials and Methods
Extraintestinal infections in both humans and dogs occur through direct patient-to-patient transfer or contact with a contaminated environment, iatrogenically or via endogenous transfer of bacteria that predominate in the patient's fecal microbiota.4,14,15 Previous characterization of rectal and clinical MDR E. coli isolates suggested that endogenous transfer is the major route of infection in this case series, because a high rate of gastrointestinal carriage of MDR E. coli was also demonstrated in hospitalized dogs over the same time period.7,16
Of the major published risk factors for acquiring MDR extraintestinal infection in animals and humans,4 severe underlying illness (97% of dogs), hospitalization for ≥3 days (82%), and surgical intervention (57%) were common in the present study. More specific risk factors could not be established as animal movements within the hospitals, and the number and nature of interactions with veterinary staff could not be clearly determined from the case records. In addition, these factors are confounded by concurrent antimicrobial use.
Antimicrobial treatment selects for resistance in both pathogenic and commensal Enterobacteriaceae, and is considered the most important risk factor for acquiring extraintestinal infection with MDR strains.17 The results from our study suggest that prior antimicrobial treatment with β-lactam/β-lactamase inhibitors, fluoroquinolones, and 1st-generation cephalosporins, either singularly or in combination, may be a risk factor. Interestingly, the predominant E. coli resistance profile changed over time, with resistance profile 1 and 2 strains becoming less prevalent and resistance profiles 3, 4, and 5 more so in later years. Resistance profile 2 isolates correspond to previously described clonal group 2 isolates6 and contain a 93 kb blaCMY−7 plasmid responsible for resistance to 3rd-generation cephalosporins and β-lactam/β-lactamase inhibitors. Resistance profile 3, 4, and 5 strains are probably closely related to resistance profile 2, except that they are sensitive to β-lactams and may have lost or not acquired the AmpC β-lactamase-bearing 93 kb plasmid.5,6 Molecular typing would be required to confirm whether they indeed belong to the same clonal group.18
Surprisingly, MDR infection could be held directly responsible for the euthanasia of only 1 dog that developed sepsis. For the remaining dogs, treatment varied according to the antimicrobial sensitivity profile of the isolate and the site of infection, with the majority of dogs being UTIs. Risk factors for acquiring UTI include poor host defenses, paresis, or urinary stasis requiring catheterization, such as would be expected in cases of intervertebral disk disease.19–21 In these dogs, antimicrobial administration during the period of catheterization is a major risk factor for development of UTI.21 In the current study, all dogs yielding an MDR urine culture were treated by the clinician as UTIs on the basis of clinical history and physical examination, including 3 dogs from which only <103 MDR bacteria per milliliter urine were obtained from samples collected by cystocentesis. Similarly, the remaining samples collected by voiding or catheter yielded moderate to heavy growth of MDR bacteria in pure culture, indicating UTI rather than asymptomatic bacteruria (Table 4).
Fourteen UTIs resolved in association with co-amoxyclavulanate treatment, even though the isolates showed in vitro resistance to this agent. This may be a result of co-amoxyclavulanate undergoing renal elimination; therefore, urine concentration is high. The minimum inhibition concentration (MIC) range of amoxicillin/clavulanic acid for resistance profile 1 and 2 isolates was 32–128 mg/L (J. Gibson, unpublished data). To treat UTIs, it is necessary to maintain urine antimicrobial concentrations at 4 times the MIC of the pathogen.22 The mean urine concentration of co-amoxyclavulanate at 12.5 mg/kg PO q8h is 201 mg/L.22 This indicates that some resistance in profiles 1 and 2 UTIs would not be expected to respond to treatment with co-amoxyclavulanate at the recommended dosage, but as documented in our study, they may resolve at the higher dose of 25 mg/kg.
Interestingly, the postsurgical dogs of MDR osteomyelitis resolved or were improving after antimicrobial treatment and supportive care, even when the prescribed antimicrobial showed in vitro resistance. Three of these dogs were treated with chloramphenicol at 50 mg/kg PO q8h for >20 days, and the bacteria isolated from these dogs were all sensitive to this agent. Chloramphenicol attains therapeutic concentrations in most tissues,23 but it is not an ideal choice because it is bacteriostatic24 and may trigger bone marrow suppression and development of immune-mediated aplastic anemia in some patients.25 The remaining 2 orthopedic dogs yielded isolates with an amoxicillin/clavulanic acid MIC of 64 mg/L (J. Gibson, unpublished data). These were treated with co-amoxyclavulanate at 25 mg/kg PO q12h for >21 days. Whereas this antimicrobial has excellent penetration through bone,26 it could not be determined whether it maintained a therapeutic drug concentration in this site that exceeded the MIC of the MDR isolates between doses. It must also be considered that complete resolution of infection may in part be attributed to supportive treatment and/or the host response.
Imipenem, a carbapenem for IV use, was used only in 2 dogs because it is expensive, requires IV administration, and affects renal and hepatic function. In addition, the carbapenems are a last-line treatment in human medicine for serious MDR Gram-negative infections, and their use in veterinary medicine should be restricted to an absolute last resort. Other dogs resolved after surgical intervention and treatment of the primary condition, including draining of the abscess, castration, and ovariohysterectomy, without the concurrent use of antimicrobials.
Although the dogs' histories suggested that the majority of infections were nosocomial in origin, this could be established only for dog 11. This dog had a rectal swab taken on admission that was negative for MDR E. coli. After bilateral tibial plateau levelling osteotomy, it had a long hospitalization period, was treated with fluoroquinolones for a Pseudomonas postsurgical infection, and, 2 weeks after the negative rectal swab, became colonized with MDR E. coli. Insertion of a urinary catheter presumably predisposed to development of an MDR UTI (profile 2) caused by the same endogenous E. coli strain isolated from the rectal swab. The UTI resolved after treatment with chloramphenicol.
Case series usually provide only limited evidence when assessing putative risk factors for disease development. However, results from case series can be useful as supportive evidence for findings from other study types and for generating new hypotheses. The potential risk factors for dogs acquiring MDR extraintestinal infection suggested by this study were severe underlying illness, prolonged hospitalization, surgical intervention, and prior antimicrobial treatment with β-lactams, fluoroquinolones, or both, which maintains selection pressure for colonization of the hospitalized animal. In terms of treatment options for veterinary clinicians confronting MDR opportunistic infections, determining the relationship between antimicrobial pharmacokinetics/pharmacodynamics and the MIC of the MDR pathogen may be more judicious than reaching for expensive, top-shelf drugs usually reserved for treating human infections.
- Top of page
- Materials and Methods
aMedvet Diagnostics, Thebarton, SA, Australia
bPanalog Ointment®, Novartis Animal Health Australasia Pty Limited, North Ryde, NSW, Australia
cOtamax®, Schering-Plough Animal Health, Omaha, NE
dCanaural Compositum®, Boehringer Ingelheim Pty Limited, Ingelheim, Germany