Clinical and in vitro efficacy of amoxicillin against bacteria associated with feline skin wounds and abscesses

Abstract

A clinical trial involving 122 cats with infected skin wounds or abscesses presented to 10 veterinary clinics was conducted to evaluate the efficacy of 2 oral amoxicillin drug products (a paste and a suspension). A 2nd objective of the study was to identify bacteria involved in such infections and verify their in vitro sensitivity to amoxicillin. Samples of wound exudate were harvested at the time of presentation and submitted for aerobic and anaerobic culture. The sensitivity to amoxicillin of isolates thought to be infecting agents was tested, using a standard minimum inhibitory concentration method. Pasteuralla multocida and obligate anaerobes of the genera Prevotella, Fusobacterium, and Porphyromonas were the most frequently isolated pathogens. Overall, their in vitro susceptibility to amoxicillin was very good. Both drug products were clinically efficacious with a global success rate of 95.1% for cats administered oral amoxicillin at 11–22 mg/kg bodyweight (mean 13.8 mg/kg bodyweight) twice daily for 7 to 10 days.

Introduction

Cats are commonly presented to veterinary clinics with abscesses or infected wounds resulting from trauma, most frequently due to a fight with another animal. Typically, veterinarians treat this condition by cleansing the wound, surgically or not, and prescribing an oral, broad spectrum antimicrobial agent.

Recognizing the importance of emerging antimicrobial resistance in both human and veterinary medicine, today’s practitioners are encouraged to use antimicrobial agents prudently (1). When selecting an antimicrobial agent empirically, the veterinarian should opt for an agent that has a narrow spectrum covering the target pathogens and is not classified as critical for human medicine. Consequently, knowing those pathogens involved in a specific condition and their susceptibility to antimicrobials is crucial for the selection of an appropriate antimicrobial agent. Administration of the selected antimicrobial agent at an optimum dose is also required to minimize the risk of the development of resistant bacteria.

The most common bacteria infecting feline wounds and abscesses are reported to be Pasteurella multocida and obligate anaerobes, for which the predicted efficacy of amoxicillin is excellent (2). Although oral amoxicillin products are currently approved in Canada for once daily administration in cats, most veterinary textbooks recommend that amoxicillin be prescribed at least twice daily because of its relatively short half-life and time-dependent pharmacodynamics (3–5).

The purpose of this study was to identify the bacteria involved in feline skin wounds and abscesses, evaluate their in vitro susceptibility to amoxicillin, and verify the clinical efficacy of a new amoxicillin oral paste and an approved amoxicillin suspension, both administered within a target dose range of 11–22 mg/kg bodyweight (BW), q12h for 7 to 10 d.

Materials and methods

From August 2004 to May 2005, 122 cats presented to 10 Canadian veterinary clinics were enrolled in a clinical trial evaluating a new amoxicillin oral paste and an approved amoxicillin oral suspension for the treatment of skin and soft tissue infections. Any cat with soft tissue infection for which antibiotherapy was indicated was considered for entry into this study. The eligible cases were those presented with identifiable lesions from which swabs for bacterial culture could be taken.

Bacteriology

Prior to any wound treatment, a sample of exudate was taken. The area around the entry point was clipped and disinfected, and the excess antiseptic was then removed with sterile gauze. By applying gentle pressure, the exudate was expressed from the lesion and cleaned away from the opening. A sterile swab was then inserted and rotated deep into the wound to collect a sample for culture. The swab was inserted into a vial containing a commercial transport medium adequate for anaerobic and aerobic specimens (BD BBL Port-A-Cul Envelope; Becton, Dickinson and Company, Sparks, Maryland, USA) and sealed. Samples in their transport medium were sent accompanied by an ice-pack (not in direct contact) by overnight courier to the bacteriology laboratory of the Faculté de médecine vétérinaire, Université de Montréal.

At the laboratory, the samples were streaked onto Columbia base agar (Difco, Detroit, Michigan, USA), supplemented with 5% blood, and onto trypticase soy agar (BBL; Becton, Dickinson and Company, Sparks, Maryland, USA) base medium, supplemented with vitamin K, yeast extract, hemin, and gentamicin. Brain heart infusion (BHI) broth (Difco) was also inoculated for enrichment. Plates were incubated in aerobic conditions, enriched with 5% CO₂, and in anaerobic conditions at 35°C. Plates were evaluated after 24 and 48 h for aerobic cultures and after 48 h and 5 d for anaerobic cultures. Standard bacteriological techniques were used to identify aerobic bacteria (6). Anaerobic bacteria were identified by using the API 20A system (BioMérieux, Marcy l’Étoile, France). Isolates were frozen in trypticase soy broth (TSB, BBL Becton, Dickinson and Company) + 10% glycerol at −70°C until antimicrobial susceptibility tests had been performed.

Antimicrobial susceptibility of the bacterial isolates to amoxicillin was determined by using minimum inhibitory concentrations (MIC). A stock solution was prepared by dissolving 1280 μg of amoxicillin powder (Fersina Gb, Ramos Arizpe Coah., Mexico) into each milliliter of PBS 0.1 M, pH 6.0 and stored in 1-mL aliquots at −70°C until use.

For aerobic bacteria, MICs were determined by using a microdilution method (7). Briefly, the MIC panels consisted of a 96-well microtiter plate containing serial dilutions of amoxicillin. Tested concentrations ranged from 0.0625 μg/mL to 64 μg/mL. Bacterial suspensions used for inoculation of the MIC panels were prepared according to guidelines (7). Using a sterile loop, 3 to 5 colonies were removed from the surface of an 18-hour culture of the organism and suspended in 4.5 mL of saline. This bacterial suspension was adjusted to a 0.5 McFarland standard (6) and 0.5 mL of this suspension was transferred to 4.5 mL of sterile saline. Of this suspension, 0.66 mL was transferred into 6 mL of cation-adjusted Mueller-Hinton broth (CAMHB, Difco). For streptococcal isolates, 2.5% of lysed horse blood was added to the CAMHB. Fifty microliters of the standardized inoculum was dispensed into each well of the MIC panel. Escherichia coli, American Type Culture Collection (ATCC) isolate 25922 and Enterococcus faecalis, ATCC isolate 29212 were used as controls. After inoculation, MIC panels were incubated aerobically at 35°C for 18 h. The microplates were evaluated by naked eye, always by the same person. The lowest concentration with no visible growth was considered to be the MIC for that isolate.

For anaerobic isolates, MICs were determined by using an agar dilution method (8). Briefly, 17 mL of supplemented brucella agar medium (Difco) mixed with 1 mL of laked sheep blood and 2 mL of 10 times the required final concentration of amoxicillin in PBS was poured into sterile petri dishes for each concentration tested. The range of tested concentrations was from 0.0625 μg/mL to 64 μg/mL. Bacterial suspensions used for inoculation of the agar plates were prepared according to Clinical and Laboratory Standards Institute (CLSI) guidelines (8). Three to 5 colonies were removed from the surface of a 48-hour culture of the organism by using a sterile loop and suspended in 4.5 mL of saline. This bacterial suspension was adjusted to a 0.5 McFarland standard (8). Two hundred micro-liters of each bacterial suspension was transferred into a 96-well microtiter plate, 1 isolate per well. A replicator tool was used to inoculate the agar plates with a maximum of 48 isolates per plate. Bacteroides fragilis ATCC 25285 was used as the control strain. After inoculation, the agar plates were incubated under anaerobic conditions at 35°C for 48 h. Control agar plates were also incubated aerobically at 35°C. The lowest concentration with no visible growth was considered to be the MIC for that isolate.

Clinical efficacy

Enrolled cats were prescribed either amoxicillin oral paste (Vetomoxin; Vétoquinol N-A, Lavaltrie, Quebec) or amoxicillin oral suspension (Moxilean-50; Bimeda-MTC, Cambridge, Ontario) at a target dose of 11–22 mg/kg BW q12h. Treatment was randomly assigned to each eligible cat on presentation. A list of 22 lines, each representing a case to be presented to the clinic, was made and lines were labeled 1 to 22. By using software, a random number was assigned to each line (“=RAND” function of Microsoft Excel, Microsoft, Redmond, Washington, USA). Each 2 lines in succession was considered a block and within each block, each line was ranked according to its random number; the 1st ranked line was then assigned as treatment 1 and the 2nd line assigned as treatment 2. A unique randomized list, printed as a form onto which identifying case information would be recorded, was prepared for each participating clinic. As each eligible case was presented, it was assigned treatment as prescribed by the next available line. Charts were prepared for each treatment, giving dose instructions for each potential BW range, allowing for achievable dose intervals (Table 1).

Table 1.

Dosing instructions to participating clinics for each potential bodyweight range

Bodyweight	Amoxicillin dose	Amoxicillin dose rate
1.4–1.8 kg (3–4 lbs)	20 mg	11.1–14.3 mg/kg
1.9–3.6 kg (4.1–8 lbs)	40 mg	11.1–21.1 mg/kg
3.7–5.4 kg (8.1–12 lbs)	60 mg	11.1–16.2 mg/kg
5.5–9.1 kg (12.1–20 lbs)	100 mg	11.0–18.2 mg/kg

If judged appropriate by the attending veterinarian, wound preparation, such as abscess drainage, surgical debridement of devitalized tissue, insertion of a drain, or wound flushing with sterile water or saline could be performed. Irrigation of the lesion with an antiseptic solution or application of topical antiseptic or antimicrobial agent products was not permitted.

Pet owners were instructed to administer the test product for 7 d (± 2 d) and to return for a reexamination at that time. If the veterinarian found that the infection had not resolved, the cat was prescribed an additional 3 d of treatment with the same product. A 2nd reexamination of unresolved cases was done 10 d (± 2 d) after the initial visit. To conclude that the infection had resolved, the following criteria had to be achieved: the cat was afebrile, there was no exudate escaping from the wound, swelling was significantly decreased, any previously open wounds were dry and could not be reopened, there was no pocket under the skin in which pus could continue to accumulate, and the site was not painful.

Results

Bacterial culture

Exudate from the lesions was collected from 120 cats at the time of enrolment. In only 8 cases (6.7%) was there no bacterial growth detected by either aerobic or anaerobic culture.

Both aerobic and obligate anaerobic bacteria were isolated concomitantly from 54 cases (45%), whereas only aerobic bacteria were isolated from 37 samples (30.8%), and only obligate anaerobes from 21 other samples (17.5%).

Forty-one samples yielded only 1 bacterial species (34.2%). Growth of 2 bacterial species was observed in 36 samples (30%), 3 bacterial species in 27 samples (22.5%), and > 3 bacterial species in 8 samples (6.7%).

Aerobes and facultative anaerobes

Members of the Pasteurellaceae family were isolated the most frequently, in more than half of the cases. Of the 66 Pasteurellaceae isolated, 61 were Pasteurella multocida. Streptococci were isolated from 17 samples, with Streptococcus canis (6 samples) being the predominant species of this family. Staphylococci were isolated from 14 samples, Staphylococcus intermedius and other Staphylococcus spp. being equally represented. Other aerobic bacteria isolated were Corynebacterium spp. from 8 samples and Enterococcus spp. from 5 samples.

Obligate anaerobes

When grouped, the 3 main anaerobic genera isolated included Prevotella, Fusobacterium, and Porphyromonas, which were isolated from 77 samples. Prevotella spp. isolated from 40 samples were Pr. melaninogenica/oralis (25 samples), Pr. bivida (10 samples), and Pr. intermedia (1 sample), and undetermined Prevotella spp. (4 samples). Members of the genus Fusobacterium were isolated from 25 samples; F. necrophorum/nucleatum (9 samples), F. varium (3 samples), and undetermined Fusobacterium spp. (13 samples). Porphyromonas asaccharolytica was isolated from 12 samples. In addition, bacteria of the genus Bacteroides were isolated 8 times (B. uniformis, B. stercoris, B. vulgatus, B. fragilis, and 1 unidentified Bacteroides sp.).

Minimum inhibitory concentrations

The MIC of amoxicillin was determined for 186 bacterial isolates. The MIC values for bacteria judged to be of interest by the investigators in feline wounds/abscesses are presented in Table 2. Of 60 P. multocida isolates, 55 had MIC values of ≤ 0.5 μg/mL. Prevotella spp. showed slightly less sensitivity with 25 of 40 isolates with MIC values of ≤ 0.5 μg/mL. Together there were 32 Fusobacterium and Porphyromonas species, of which 30 had MIC values of ≤ 0.125 μg/mL.

Table 2.

Minimum inhibitory concentrations (MIC) of amoxicillin for bacteria tested

	Number of isolates per MIC
	MIC (μg/mL)
Bacteria	0.0625	0.125	0.25	0.5	1	2	4	16	32	64	> 64	Number tested
Pasteurella multocida	3	33	16	3	4	1	0	0	0	0	0	60
Porphyromonas asaccharolytica	8	2	0	0	0	0	0	0	0	0	0	10
Prevotellae	16	2	5	2	2	4	2	4	2	0	1	40
Pr. melaninogenica/oralis	10	2	1	1	2	3	2	3	1	0	0	25
Pr. bivida	4	0	3	1	0	0	0	1	0	0	1	10
Pr. intermedia	0	0	1	0	0	0	0	0	0	0	0	1
Prevotella spp.	2	0	0	0	0	1	0	0	1	0	0	4
Fusobacterium	11	9	0	0	0	1	0	0	0	0	1	22
F. necroforum/nucleatum	3	3	0	0	0	1	0	0	0	0	0	7
F. varium	3	0	0	0	0	0	0	0	0	0	0	3
Fusobacterium spp.	5	6	0	0	0	0	0	0	0	0	1	12
Staphylococci	1	2	2	0	0	0	1	3	0	0	0	9
Streptococci	9	0	0	0	0	0	0	0	0	0	0	9
Enterococcus spp.	0	1	0	1	1	0	1	0	0	1	0	5
Bacteroides	0	2	2	0	0	0	0	2	2	0	0	8

The MIC for amoxicillin for the 9 streptococci tested was 0.0625 μg/mL. Within the 9 staphylococci tested, 5 isolates appeared sensitive to amoxicillin (MIC from 0.0625 to 0.25 μg/mL), whereas the other 4 appeared resistant (MIC of 4 or 16 μg/mL). The sensitivity of Bacteroides isolates to amoxicillin was also variable with 4 of 8 showing MIC values of 0.125 or 0.25 μg/mL, while the remaining 4 had MIC values of 16 or 32 μg/mL.

Clinical efficacy

There was no apparent difference between the efficacy of the 2 types of treatment and consequently all treatment results are combined. Of the 122 cats enrolled, 119 were returned for the 1st reexamination. One cat was excluded at this time because the pet owner applied the treatment topically instead of orally. The mean age of the 118 eligible feline patients was 5.1 y (range 0.4–16 y) and the mean weight was 5.2 kg (range 1.8–9.8 kg). More males (75 cats) than females (43 cats) participated in this study and the vast majority were domestic shorthair or longhair. Overall, the mean dose of amoxicillin administered was 13.8 mg/kg BW (s ± 2.4).

Of the 118 cats returned for the 1st recheck, 12 were excluded from the overall analysis because they were returned beyond the required 9 d after the initial examination. Evaluation of the success rate at the 1st recheck, therefore, included 106 cats. There was a single treatment failure (no improvement in the clinical condition) at the 1st reexamination, while the infection was considered to be resolved in 70 cats (66.0%). Infection was resolved in 10 of the 12 cats presented late for reexamination.

The remaining 35 cats presented with some clinical improvement but were prescribed an additional 3 d of treatment because the infection was not completely resolved. Thirty-one of the 35 were presented for the 2nd recheck. Treatment was considered a success in 27 cats; a failure in 4.

Overall, of the 102 cats that were presented for reexamination as requested and according to the protocol, in only 5 cats was the infection considered to be unresolved. This represents a 95.1% success rate.

Discussion

The rate at which bacteria were recovered in this study was very good, with growth observed in 93.3% of samples. Bacterial species isolated were relatively similar to those reported by other authors for feline abscesses or anaerobic infections (2,9–12).

One not inconsiderable difference between our results and those from 2 other groups is the isolation rate of peptostreptococci. Representatives from this genus were isolated only twice in the present study, whereas Love et al (8,9) reported peptostreptococci in 14.9% to 17.4% of the anaerobic isolates, and Jang et al reported 20% of anaerobes to be peptostreptococci (11).

Aminopenicillins, the group of penicillin to which amoxicillin and ampicillin belong, are sensitive to beta-lactamase destruction. In a study on the frequency and antimicrobial susceptibility of Staphylococcus spp. from feline skin lesions, resistance to ampicillin was reported to be common for S. aureus and S. intermedius (13). In the present study, staphylococci were isolated in only 14 of the 120 samples (12%). Furthermore, only 4 of the 9 Staphylococcus isolates for which sensitivity was tested had an MIC for amoxicillin above 0.25 μg/mL. Some species of Prevotella, particularly Pr. melaninogenica, have been reported to produce beta-lactamase (14,15). This probably explains the variable sensitivity of the Prevotellae spp. in our study.

The present study demonstrated a very good pattern of susceptibility to amoxicillin for organisms most frequently involved in feline abscesses and skin wounds. If we assume that a MIC ≤ 1 μg/mL is associated with a good susceptibility to aminobenzylpenicillins (16), then 98.3% of P. multocida, 67.5% of Prevotella spp., 90.9% of Fusobacterium spp., 100% of Streptococcus spp., and 100% of Porphyromonas spp. tested in the present study were susceptible to amoxicillin. Unfortunately, the number of isolates of each bacterial species for which the MIC was tested was insufficient to calculate a MIC₅₀ or MIC₉₀.

Two clinical trials from the late 1970s supported the efficacy of amoxicillin for the treatment of feline infections when administered once daily (17,18). However, recommended doses in recent veterinary textbooks vary from 10 to 22 mg/kg BW q8 or 12h, but not q24h (3–5). In this study, although some strains of bacteria appeared resistant to amoxicillin when tested in vitro, this did not correlate with a lack of clinical efficacy when the drug was administered to cats at the dose of 11–22 mg/kg BW, q12h for an overall success rate of 95.1%. Furthermore, of the 5 cases that were classified as treatment failures, only 1 required a change of antimicrobial agent after 7 d of treatment and a Bacillus sp. was the only bacterium isolated from this cat’s wound. The 4 other cases eventually resolved following treatment with amoxicillin for more than 10 d.

In conclusion, amoxicillin administered at 11–22 mg/kg BW, PO, q12h is an excellent empirical choice to treat cats with skin wounds or abscesses usually infected by P. multocida and obligate anaerobes.