IntroductionUrinary tract infections (UTIs) are among the most com-

mon conditions requiring diagnostic and therapeutic inter-vention in small animal veterinary practices. They typicallyinvolve bacteria and are estimated to affect approximately14% of all dogs during their lifetime ; female dogs are moreoften affected than male dogs [15]. Bacterial UTIs are morecommon in older cats than in younger cats, and the incidenceincreases with age [12]. They occur not only in companionanimals such as dogs and cats, but are also frequently obser-ved among humans [2, 8, 9]. The causative pathogens indogs suffering from UTIs of bacterial origin are diverse.Escherichia coli is the most important microorganism res-ponsible for UTIs of dogs. Staphylococcus aureus andProteus mirabilis were also found to be among the mostcommon infecting bacteria in UTI cases in dogs. Alpha-hae-molytic streptococci, Klebsiella pneumoniae, Pseudomonasaeruginosa, Enterobacter spp., other Proteus spp., beta-hae-molytic streptococci, and some other miscellaneous bacterialspecies can be occasionally documented [14]. However, theprevalence of the various species varies considerably fromstudy to study.

Factors thought to be important in preventing UTIs includenormal micturition, mucosal defense barrier, antibacterialproperties of urine, specific anatomic structures, and syste-mic immunocompetence. UTIs occur when there is a breach(either temporary or permanent) in host defense mechanismsand sufficient numbers of virulent microbe are allowed to

adhere, multiply, and persist in a portion of the urinary tract.The alterations in host defenses may be intrinsic to thepatient (i.e., diabetes mellitus) or iatrogenic (i.e., corticoste-roid administration). UTIs caused by etiological bacterialagents have been associated with urogenital disease such ascystitis, nephritis, metritis and prostatitis in dogs and cats[24]. Furthermore, some bacterial strains isolated from dogsand cats with UTI were similar to strains isolated fromhuman UTI [12, 16]. These findings suggest that similarstrains might be able to cause the infection both in compa-nion animals and humans.

Therapeutic management of UTIs is based on the type ofinfection, duration of symptoms and aetiology. Surveillanceof the susceptibility of major pathogens to a range of antimi-crobials can aid clinicians in prescribing the most appro-priate antimicrobial agents for infections and is an essentialpart of monitoring and detecting any increase in resistance.Despite this, empirical treatment of UTIs must be often ini-tiated before availability of laboratory results. Because UTIsare frequently treated with empirically selected antimicro-bial agents, assessment of changes in antimicrobial suscepti-bility patterns over time is important to ensure that the orga-nisms remain susceptible to empirically selected antimicro-bial agents. In addition, treatment of these infections is oftenchallenging because some of the isolates are frequently resis-tant to numerous antimicrobial agents, including those usedempirically. Because of the multi-drug resistant profile ofsome of the isolates, in vitro antimicrobial susceptibility tes-ting is advised to ensure appropriate antibiotic selection in

Survey on bacterial isolates from dogs with uri-nary tract infections and their in vitro sensitivityR. PAPINI1*, V.V. EBANI2, D. CERRI2 and G. GUIDI1

1 Dipartimento di Clinica Veterinaria, Viale delle Piagge 2, 56124 Pisa, Italy2 Dipartimento di Patologia Animale, Profilassi e Igiene degli Alimenti, Viale delle Piagge 2, 56124 Pisa, Italy

* Corresponding author : Dr Roberto Papini, Dipartimento di Clinica Veterinaria, Facoltà di Medicina Veterinaria, Viale delle Piagge, 2 - 56124 Pisa, Italia.Phone +39/050/2216801; Fax +39/050/2216813; e-mail: rpapini@vet.unipi.it

RÉSUMÉ

Enquête sur les souches bactériennes isolées et détermination de leursensibilité in vitro aux antibiotiques chez des chiens atteints d’infectiondu tractus urinaire. Par R. Papini, V.V. Ebani, D. Cerri et G. Guidi.

Dans une région géographique de l’Italie centrale, les bactéries isolées dechiens atteints d’infections urinaires ont été : Escherichia coli (17.48%),Proteus mirabilis (16.59%), Pseudomonas aeruginosa (7.17%),Staphylococcus aureus (6.72%), streptocoques α-hemolytiques (2.24%) etKlebsiella sp. (0.89%). Les résultats de la sensibilité in vitro ont montré quetoutes les souches étaient très résistantes aux agents antibactériens testés,incluant amikacine, amoxycilline, amoxycilline-acide clavulanique, ampi-cilline, céfalexine, céfotaxim, doxicycline, enrofloxacine, erythromycine,gentamycine, kanamycine, netylmicine, rifampicine, streptomycine, sulpha-méthoxazole/triméthoprime et tétracycline.

Mots-clés : Chien - Infections du tractus urinaire -Sensibilité in vitro - Agents Antibactériens.

Revue Méd. Vét., 2006, 157, 1, 35-41

SUMMARY

Bacterial isolates from urine specimen of dogs with urinary tract infec-tions in a geographical region of Central Italy were : Escherichia coli(17.48%), Proteus mirabilis (16.59%), Pseudomonas aeruginosa (7.17%),Staphylococcus aureus (6.72%), α-haemolytic streptococci (2.24%), andKlebsiella sp. (0.89%). In vitro antibacterial agent sensitivity showed thatnone of them was effectively susceptible to any of the antibacterial agentstested, including amikacin, amoxicillin, amoxicillin-clavulanic acid, ampi-cillin, cefalexin, cefotaxin, doxicyclin, enrofloxacin, erythromycin, genta-mycin, kanamycin, netilmycin, rifampicin, streptomycin, sulphamethoxazo-le/trimethoprim, and tetracyclin.

Keywords : Dog - Urinary tract infection - In vitro sensiti-vity - Antibacterial agents.

Revue Méd. Vét., 2006, 157, 1, 35-41

addition to allowing for monitoring the development ofresistant pathogens. Therefore, clarity with regard to the cur-rent susceptibility to antimicrobial agents among urinary iso-lates is important and benefit to veterinary clinicians for pro-viding clinically appropriate and cost-effective therapy forUTIs.

During the last years, there is considerable and growingglobal concern about the increasing levels of antimicrobialresistance in human and veterinary medicine [9, 22] suchthat antimicrobial resistance surveillance studies are now anecessity. Moreover, given that a transition in therapy mayeasily occur or be imminent, it has become important todefine not only national and international rates of resistancefor a range of pathogens but local rates as well, since it hasbeen shown that differential resistance rates even exist forspecific pathogens and antimicrobial agents recovered fromrestricted geographical areas [2, 8]. As already said, antimi-crobial treatment of UTIs caused by the most common uro-pathogens in animals and humans may become complicated,since it may be compromised by increasing antimicrobialresistance. Continued concern about antimicrobial resistanceis warranted as are studies on the activities of new agents.However, whereas the development of totally new antimi-crobial agents takes several years, the re-evaluation of olderantibiotics can be conducted more rapidly. Therefore, thepresent study was undertaken to assess the prevalence ofuropathogens isolated from uncomplicated UTIs of dogs in ageographical region of Italy, to evaluate their in vitro suscep-tibility to antibiotics commonly used for the treatment ofsuch infections, and to compare the difference in antibioticsusceptibility among different isolates, so that optimal empi-rical antibiotic therapy in dogs could be determined.

Materials and methodsA survey was carried out on 223 adult dogs (range from 3

to 15 years of age, mean ± standard deviation = 8.8 ± 3.7).There were 116 female dogs and 107 male dogs. All the dogswere from different areas of Tuscany, a geographical regionof Central Italy. They were presented with a history of clini-cal signs suggestive of UTIs at our Veterinary MedicalTeaching Hospital (Department of Veterinary Clinic,University of Pisa, Italy) during a 6-month period, beginningin January 1 and ending in June 30, 2004. Such signs inclu-ded inappropriate urination, dysuria, stranguria, haematuria,or pollakiura. None of the dogs had been given antimicrobialagents, corticosteroids, or other immunosuppresive agentswithin 15 days before sampling. However, the possibility ofprior treatments with antibiotics during the entire course oftheir lives can not be fully excluded. Since the aim of ourstudy was to have an overview, we did not attempt to selectby gender, ages or specific breeds of dogs.

Urine samples were collected by antepubic cystocentesis.An area of skin on the central midline, caudal to the umbeli-cus, was cleaned and disinfected using 2 per cent iodine w/vsolution in 70 per cent alcohol. The bladder was immobilisedby digital palpation and a one-inch long sterile needle wasinserted into the bladder. Urine specimens were drawn into a

sterile syringe. This method of urine collection was chosenin order to avoid or reduce, as much as possible, the speci-men contamination originating in the prepuce and urethra ofmales, and in the urethra and vagina of females [5].

Following collection, urine samples (8-10 mL) were eitherprocessed for bacteriological investigation immediately orrefrigerated at 4°C within a few minutes from sampling,until inoculations of bacteriologic media were completed(not more than 4 hours later). Once arrived at the clinicalmicrobiology laboratory (Department of Animal Pathology,Prophylaxis and Food Hygiene, University of Pisa, Italy),well-mixed urine samples were divided into two aliquots insterile tubes. The approximate number of colony-formingunits/millilitre (CFU/mL) of urine was determined for eachspecimen. For this purpose, an aliquot of 1.0 mL urine samplewas diluted in sterile saline solution from 10-1 to 10-5. One mlof 10-3, 10-4, and 10-5 dilutions were inoculated in Plate-Count agar plates and incubated at 37°C for 72 hours. Thecounts were determined by the standard method. Fourgroups were identified : a) <1,000 CFU/mL of urine b) bet-ween ≥1,000 and <10,000 CFU/mL c) from ≥10,000 to<100,000 CFU/mL d) ≥100,000 CFU/mL. According toCARTER [4], counts ≥10,000 to <100,000 CFU/mL can beconsidered as suggestive of infection and a presence≥100,000 CFU/mL can be considered as clinically importantbacteriuria. Therefore, the identity of bacterial isolates andtheir sensitivity to antibacterial agents were only determinedfrom specimens yielding bacterial isolations ≥10,000 to<100,000 CFU/mL or ≥100,000 CFU/mL, while all the othercultures were discarded.

For this purpose, an aliquot of 5.0 mL urine specimen wascentrifuged for 20 minutes at 3,000 x g at room temperature.After the supernatant was decanted, cultures were madefrom the deposit. As indicated by LING et al. [15], bloodagar and MacConkey (Oxoid, Unipath S.p.A., GarbagnateMilanese, Milano, Italy) agar plates were inoculated, using asterile loop from the urine sediment for each plate. All plateswere incubated aerobically at 37°C and were examined forbacterial growth after overnight incubation, at which timecolony morphology was observed. Single bacterial colonieswere picked for identification of organisms by standard bac-teriologic and biochemical procedures, as reported by CAR-TER [4].

Antibacterial sensitivity tests were performed using thedisc method. A bacterial colony was seeded from the agarplate to Serum Broth and incubated. Samples from the brothcultures with a turbidity corresponding to 0.5 McFarlandstandard suspension were evenly smeared over the entiresurface of Mueller-Hinton agar plates [3]. After allowing thesurfaces of the plate to dry for approximately 2 minutes,selected antibacterial discs (Oxoid, Unipath S.p.A.,Garbagnate Milanese, Milano, Italy and Bayer, AGLeverkusen, Germany) were placed on the inoculated sur-face. Discs were stored at 4°C until their use, according tomanufacturers’ instructions. The following antibacterialswere tested: amikacin (AK), 30 µg ; amoxicillin (AML),10 µg ; amoxicillin-clavulanic acid (AMC), 20 µg + 10 µg ;ampicillin (AMP), 10 µg ; cefalexin (CL), 30 µg ; cefotaxin

36 PAPINI (R.) AND COLLABORATORS

Revue Méd. Vét., 2006, 157, 1, 35-41

SURVEY ON BACTERIAL ISOLATES FROM DOGS WITH URINARY TRACT INFECTIONS 37

(CTX), 30 µg ; doxicyclin (DO), 30 µg ; enrofloxacin(ENR), 5 µg ; erythromycin (E), 15 µg ; gentamycin (CN),10 µg ; kanamycin (K), 30 µg ; netilmycin (NET), 30 µg ;rifampicin (RD), 30 µg ; streptomycin (S), 10 µg ; sulpha-methoxazole/trimethoprim (SXT) 23.75 µg + 1.25 µg, andtetracyclin (TE), 30 µg. Results were evaluated after incuba-tion at 37°C for 24 hours. Based on the size of the zone ofbacterial growth inhibition, bacteria were classified as sensi-tive, intermediate or moderately sensitive, and resistant,according to manufacturers’ instructions (Table I). As repor-ted by the Oxoid manual, the category moderately sensitiveindicates a possible therapeutic use of antibiotics under par-ticular clinical circumstances. For example, when antibioticscan be used at high doses, or concentrate in particular sites ofinfection such as the urinary tract due to their pharmacokine-tic, or when an association of antibiotics can be employed.

ResultsAll of the 223 urine specimens yielded bacterial growth.

However, 99 of them (44.39%) showed bacterial counts<1,000 CFU/mL of urine specimen, 11 (4.93%) counts bet-

ween ≥1,000 and <10,000 CFU/mL, 32 (14.34%) from≥10,000 to <100,000 CFU/mL, and finally 81 (36.32%)≥100,000 CFU/mL. Therefore, the identity of bacterial iso-lates and their sensitivity to antibacterial agents were deter-mined from 114 (51.12%) urine specimens yielding bacterialcounts suggestive of infections or clinically significant. Onlyone species was found in each sample. Altogether, 6 bacte-rial species were identified in culture-positive specimens.Overall, E. coli was isolated from 39 (17.48%) specimens, P.mirabilis from 37 (16.59%), P. aeruginosa from 16 (7.17%),S. aureus from 15 (6.72%), α-haemolytic streptococci from5 (2.24%), and Klebsiella sp. from 2 (0.89%). Percentagesreferred to the totality of cases where bacteriuria was sug-gestive of infections or clinically important are reported inTable II.

In the present study, none of the bacterial isolates waseffectively susceptible to any of the antimicrobial agents tes-ted. Their sensitivity to antibacterial agents is shown in TableIII. E. coli isolates were completely resistant to AK, CL, DO,E, K and S (39/39, 100%). Likewise, great numbers of themwere resistant to AML, AMP and RD (37/39, 94.87%), TE(36/39, 93.30%), SXT (33/39, 84.61%), AMC (32/39,

TABLE I. — Diameters (mm) of bacterial growth inhibition induced by the antibacterial agents tested in vitro.

AK : amikacin ; AML : amoxicillin ; AMC : amoxicillin-clavulanic acid ; AMP : ampicillin ; CL : cefalexin ;CTX : cefotaxin ; DO : doxicyclin ; ENR : enrofloxacin ; E : erythromycin ; CN : gentamycin ; K : kana-mycin ; NET : netilmycin ; RD : rifampicin ; S : streptomycin ; SXT : sulphamethoxazole/trimethoprim ;TE : tetracyclin.

38 PAPINI (R.) AND COLLABORATORS

Revue Méd. Vét., 2006, 157, 1, 35-41

82.05%), NET (31/39, 74.48%), CN (26/39, 66.66%), CTXand ERN (21/39, 53.84%). A small part showed susceptibi-lity to ENR (13/39, 33.33%), CN (10/39, 25.64%), CTX andNET (8/39, 20.51%), SXT (6/39, 15.38%), AMC (4/39,10.25%), TE (3/39, 7.69%), AML, AMP and RD (2/39,5.12%). Consequently, very small numbers showed mode-rate susceptibility to CTX (10/39, 25.64%), ENR (5/39,12.82%), AMC and CN (3/39, 7.69%).

All the P. mirabilis isolates were highly resistant to DO, E,K and TE (37/37, 100%). A great number of the isolates wasalso resistant to AK and S (36/37, 97.29%), AMP, CL andRD (35/37, 94.59%), AML (34/37, 91.89%), ENR and SXT(32/37, 86.48%), followed by NET (22/37, 59.45%), AMC(19/37, 51.35%), CTX (12/37, 32.43%) and CN (10/37,27.02%). Decreasing numbers of them showed sensitivity toCN (19/37, 51.35%%), CTX (17/37, 45.94%), NET (13/37,35.13%), AMC (12/37, 32.43%), SXT (5/37, 13.51%), ENR(4/37, 10.81%), AK, AML, AMP, RD and S (1/37, 2.70%).In general, a smaller part of these isolates showed moderatesensitivity to CTX and CN (8/37, 21.62%), AMC (6/37,16.21%), AML, CL and NET (2/37, 5.40%), AMP, ENR andRD (1/37, 2.70%).

P. aeruginosa isolates were completely (16/16, 100%) oralmost completely (13 to 15/16, 81.25 to 93.75%) resistant toall the antibacterial agents tested in the present study, withthe exception of CN (7/16, 43.75%), to which a half (8/16,50%) of the isolates showed susceptibility. Sporadically,moderate susceptibility to CL (3/16, 18.75%), ENR (2/16,12.5%), AMC, CN and NET (1/16, 6.25%) was observed, aswell as very low susceptibility to AK, AMP, CTX, ENR andNET (1/16, 6.25%).

Similarly, the majority of the S. aureus isolates were com-pletely (15/15, 100%) or highly (10 to 14/15, 66.66 to93.33%) resistant to a great number of the antimicrobialagents tested. Approximately, half of the isolates were sus-ceptible to CN and ENR (8 and 7/15, 53.33 and 46.66 %, res-pectively). In the other cases, low susceptibility to CTX andRD (4/15, 26.66%), AMC and SXT (3/15, 20.0%), AK andNET (2/15, 13.33%), CL, E and TE (1/15, 6.66%) was detec-ted, as well as moderate susceptibility to CTX (4/15,26.66%), AMC and CL (2/15, 13.33%), AK, AMP, DO andCN (1/15, 6.66%).

The five isolates of α-haemolytic streptococci were totally(5/5, 100%) resistant to AK, AML, AMP, DO, NET, RD andTE, or almost totally (4/5, 80.0%) resistant to ENR, E, CN,K and SXT. Approximately, half (3 or 2/5, 60.0 or 40.0%) ofthe isolates were also resistant to CTX and S or AMC andCL. In decreasing order, the isolates showed susceptibility toCL (3/5, 60.0%), S (2/5, 40.0%), AMC, CTX, E, and K (1/5,20.0%), or moderate susceptibility to AMC (2/5, 40.0%),CTX, ENR, CN and SXT (1/5, 20.0%).

Finally, the two isolates of Klebsiella sp. were resistant toall the tested antimicrobial agents except one (20%) case ofmoderate susceptibility to CTX.

DiscussionThis study represents an evaluation of the species distribu-

tion of bacterial uropathogens in dogs from a geographicalarea of Italy. In addition, an evaluation was performed of thein vitro effectiveness of a representative panel of antibioticsagainst a number of isolates obtained from dogs with UTI. E.coli, P. mirabilis, P. aeruginosa, S. aureus, α-haemolyticstreptococci, and Klebsiella spp. were isolated from the urineof dogs with UTIs. Some investigators reported data on bac-terial isolates from urine specimens of dogs in many coun-tries. However, to the best of our knowledge, no data areavailable from Italy. In the United States, HOGLE [10]observed that the bacteria most frequently isolated fromselected specimens of canine urine were P. aeruginosa,Proteus spp., E. coli, and S. aureus. Later, WOLLEY andBLUE [24] showed that the most frequently cultured bacte-ria from canine urine specimens were E. coli, Proteus spp., S.aureus, and P. aeruginosa. Otherwise, E. coli, Streptococcus(Enterococcus) faecalis, Staphylococcus epidermidis andProteus spp. were the most common bacteria isolated fromdogs with UTIs in the United Kingdom [23]. OXENBORGet al. [19] found that Proteus spp., E. coli andStaphylococcus spp. were the most common isolates fromdogs affected with UTIs in Australia. PERSON et al. [20]reported that in France the most frequently isolated bacteriafrom canine urine specimens were P. mirabilis, E. coli, S.aureus, P. aeruginosa and Acinetobacter spp. Later, in fur-ther investigations carried out in the United States, E. coliwas the most common isolate, along with Stapylococcus

*Percentages are referred to the totality of cases where bacteriuria was suggestive ofinfections (≥10,000 to <100,000 CFU/mL) or clinically important (≥100,000CFU/mL).

TABLE II. — Bacterial isolates from 114 out of 223 cases of urinary tract infections indogs.

SURVEY ON BACTERIAL ISOLATES FROM DOGS WITH URINARY TRACT INFECTIONS 39

Revue Méd. Vét., 2006, 157, 1, 35-41

S : S

ensi

tive

; I/M

S : I

nter

med

iate

or

Mod

erat

ely

Sens

itive

; R

: R

esis

tant

.A

K :

amik

acin

; A

ML

: am

oxic

illin

; A

MC

: am

oxic

illin

-cla

vula

nic

acid

; A

MP

: am

pici

llin

; CL

: ce

fale

xin

; CT

X :

cefo

taxi

n ; D

O :

doxi

cycl

in ;

EN

R :

enro

flox

acin

; E

: er

ythr

omyc

in ;

CN

: ge

ntam

ycin

; K

: ka

nam

ycin

;N

ET

: ne

tilm

ycin

; R

D :

rifa

mpi

cin

; S :

stre

ptom

ycin

; SX

T :

sulp

ham

etho

xazo

le/tr

imet

hopr

im ;

TE

: te

trac

yclin

.

TAB

LE

III.

— I

n vi

tro

sens

itivi

ty to

the

test

ed a

ntib

acte

rial

age

nts

of b

acte

rial

isol

ate

from

can

ine

urin

e sp

ecim

en.

spp., Proteus spp. and Klebsiella spp. as bacterial genera [15,17]. More recently, E. coli and Streptococcus/Enterococcusspp. were the most frequently isolated microorganisms fromcases of persistent UTIs and reinfections in dogs [21].Therefore, the spectrum of bacteria observed in our survey issimilar to those previously reported by other authors.

Different percentages of bacterial isolations have beenreported. This may be due to differences of the populationstudied and to the different number of specimens examined.It is evident that E. coli and P. mirabilis in our study were themain organisms associated with UTIs in dogs accounting for17.48% and 16.59%, respectively, of the 114 cases sugges-tive of infection or with clinically important bacteriuria(Table II).

In earlier studies no bacteria were isolated from relativelyhigh percentages of canine urine specimens. For instance,WOOLEY and BLUE [24] reported that no bacteria wereisolated from 21.9% of canine urines. In our study all thesamples yielded bacterial growth, although small numbers ofbacteria were isolated from approximately a half of thesamples. This is in agreement with previous results of otherauthors. PERSON et al. [20] showed that 49% of canineurine specimens yielded not significant (<100,000 CFU/mL)bacterial growth. LING et al. [15] reported that 21.8% ofurine isolates from male dogs and 18.0% of isolates fromfemale dogs had colony counts ≤10,000 CFU/mL in samplescollected by cystocentesis.

Since the aim of our study was to have an overview, we didnot attempt to select by gender, ages or specific breeds ofdogs. However, it has been previously demonstrated thathost age, sex and breed can represent some important predis-posing factors to UTIs and even to some of the bacterial spe-cies isolated [15, 20].

E. coli, P. mirabilis, P. aeruginosa, S. aureus, α-haemoly-tic streptococci, and Klebsiella spp. are ubiquitous in theenvironment. Usually considered as contaminants they haveemerged as important pathogens. These organisms cause avariety of infections including UTI [14]. Uropathogens haveshown a slow but steady increase in resistance to severalantibiotics over the last decade and antibiotic resistance is aremarkable clinical problem in treating infections caused bythese microorganisms. For instance, E. coli and otherEnterobacteriaceae have become less susceptible to com-monly used antibiotics such as ampicillin, amoxycillin, sul-phonamides, trimethoprim and fluoroquinolones [21]. Ourresults signal the problem of multi-drug resistant microorga-nisms. Considering the totality of isolates, general resistancerates varied from 96.48% to 99.12% for tetracyclines (TEand DO), and from 49.12% to 97.36% for aminoglycosides(CN, NET, K, AK and S). These latter were the most repre-sented antibiotic group in the present survey, and the lowest(49.12%) resistance rate was recorded to gentamycin.Overall, resistance rates to beta-lactam antibiotics variedfrom 51.75% to 95.61%. In particular, this accounted for70.17% to 95.61% in the penicillin group (AMC, AMP andAML), and for 51.75% to 90.35% in the cephalosporin group(CTX and CL). To the other groups of antibiotics, recordedresistance rates in decreasing order were 98.24% to erythro-mycin (belonging to the group of macrolides), 92.98% to

rifampicin (group of rifamycins), 85.96% to the associationsulphametoxazole/trimethoprim, and finally 69.29% to enro-floxacin (group of quinolones). Therefore, all the antimicro-bial agents that we studied demonstrated very high levels ofin vitro resistance, which suggests that in this case none ofthem would have been useful as adequate therapy for dogswith UTI, and that probably they would have virtually beenuseless for treatment of other infections caused by the samebacterial isolates. As a consequence, the role of new agentsnot tested in this study should have needed to be consideredbased on available susceptibility/resistance data. Despitethis, it is worth noting that susceptibility tests may yield mis-leading results, since discs placed on the cultures approxi-mate concentrations of antibacterial achievable in blood,whereas urinary antibacterials achieve 10 to 100 times theseconcentrations. Therefore, results read out as resistant invitro may not really be resistant in vivo. However, resultsread out as susceptible should really be effective.

In the present study, P. aeruginosa, E. coli, S. aureus, α-haemolytic streptococci and Klebsiella exhibited high levelof drug resistance to all the antibiotics. Lowest resistances(27.02% and 32.43%, respectively) were shown by P. mira-bilis to CN and CTX. The resistance rates to antimicrobialshas increased over the years and vary from country to coun-try [8]. Although it is inaccurate to compare prevalence dataof studies which used different antibacterial sensitivity tests,the results we obtained are different to those described inother publications for the same pathogens in other areas ofthe world. In one study from the United States, all the iso-lates with the exception of P. aeruginosa exhibited 100%susceptibility to one, two or three of the tested antimicrobialagents [9]. Results from another study showed that suscepti-bility of the different isolates varied from 80 to 100% tosome of the tested bactericidal agents [23]. Only ten out ofone-hundred-ten isolates were resistant to all agents in a sur-vey carried out in the United Kingdom [22]. Strains of E. coliand Proteus spp. isolated from Australia were 100% sensi-tive to gentamycin, whereas E. coli was susceptible (>80%)to nitrofurantoin and nalidixic acid too [18]. In France, arepresentative panel of antibiotics was effective against anumber of isolates obtained from dogs with UTI [19]. In amore recent investigation carried out in the United States,70.5% of the bacterial isolates were susceptible to at leastone commonly prescribed antibiotic per os, and 18.5% ofthem were resistant to all commonly prescribed antibioticsadministered per os, but were susceptible to a secondaryantibiotic [20]. There is no obvious explanation for all thesedifferences. These may be due to geographical variations orto differences in the number of animals and type of popula-tion surveyed, or may be attributable to differences of theantibacterial sensitivity tests performed. In the present inves-tigation, we can not completely exclude the possibility thatthe isolates could have acquired drug resistance after expo-sure to prior treatments with antibiotics, undertaken becauseof various bacterial pathologies occurring in the course ofanimals’ life. In addition, we did not attempt to investigateand identify mechanisms of antibacterial drug resistance,since this was beyond the aim of our study. However, it canbe presumed that one or more of the most common mecha-nisms of bacterial resistance could have been occurred, as

40 PAPINI (R.) AND COLLABORATORS

Revue Méd. Vét., 2006, 157, 1, 35-41

reviewed by DEVER and DERMODY [7].

As known, a global trend of increase in antibiotic resis-tance has been shown worldwide [9]. One of the main rea-sons for this trend is the increase in antibiotic consumption,because increased usage of antibiotics in a patient populationmay affect the rate of infections caused by resistant bacteriain that population. The use of antimicrobial drugs in animalsis another of the major factors selecting for the emergence ofresistant bacterial stains. As a consequence, the veterinaryuse of antimicrobials has received much publicity recentlyand is now under close scrutiny because of the importance ofpreserving the efficacy of antimicrobial therapy. At present,investigations in veterinary species are concentrated on pro-duction animals owing to the extensive use of antimicrobialsin agriculture and the well-documented transfer of resistantbacteria from animals to human beings in food [1]. Sinceantimicrobial usage in production animals has been shown tolead to the emergence of resistant bacteria, surveillance pro-tocols have been established in many countries, and severalcountries have now banned the use of certain antibiotics asgrowth promoters. Therefore, there is a substantial pool ofinformation for livestock species. In contrast, there is com-paratively little information about antimicrobial resistance indomestic pets such as dogs and cats. The development ofantimicrobial resistance in companion animals and theimplications to human health are poorly documented.Companion animals occupy a unique position among thedomestic species because they are kept in the home. Arecent report suggests an increasing trend in the developmentof resistance in companion animal pathogens [22]. The inter-relationship of antimicrobial resistance in the flora of humanbeings and domestic pets was confirmed by DAVIES andSTEWART [6]. Therefore, since the use of antimicrobialdrugs for companion animals is far more liberal than for foodproducing animals [18], it is also very important to evaluatetheir use in companion animals because they may act asreservoirs of drug-resistant bacteria. UTI is one of the mostcommon diseases in dogs and is also observed both inhumans and other companion animals such as cats. From theresults in some studies, it has been presumed that dogs andcats may be alternative reservoirs for humans [12, 16]. In thepresent study, the fact that 49.12% to 96.49% of isolateswere resistant to all the antimicrobial agents tested is of greatimportance and means these antibiotics cannot be used asempirical therapy for UTI in dogs. Probably, the transmis-sion of strains or genetic determinants like these to peoplecould be a threat [22]. This information needs to be conside-red when empirical therapy for UTIs in dogs is chosen. Asthe magnitude and variation of antimicrobial resistance inhuman medicine grow, so does the need for continuous sur-veillance in veterinary medicine. Clearly, continued sur-veillance is essential to maintaining the safety and cost-effectiveness of empirical therapy as a management strategyfor UTI in dogs.

References1. — ANONYMOUS : The Path of Least Resistance. Standing Medical

Advisory Committee : Sub-Group on Antimicrobial Resistance.London : Department of Health, 1998.

2. — BARISIC Z., BABIC-ERCEG A., BORZIC E., ZORANIC V.,

KALITERNA V., CAREV M. : Urinary tract infections in SouthCroatia : aetiology and antimicrobial resistance. Int. J. Antimicrob.Agents, 2003, 22, Suppl 2, 61-64.

3. — BAUER A.W., KIRBY W.M., SHERRIS J.C., TURCK M. :Antibiotic susceptibility testing by a standardized disk method. Am.J. Clin. Pathol., 1966, 45, 493-496.

4. — CARTER G.R. : Isolation and identification of bacteria from clinicalspecimen. In : G.R. CARTER and J.R. COLE Jr. (ed.) : Diagnosticprocedures in veterinary bacteriology and mycology, 5th ed.,Academic Press, Inc., San Diego, 1990, 19-40.

5. — COMER K., LING G.V. : Results of urinalysis and bacterial cultureof canine urine obtained by antepubic cystocentesis, catheterization,and the midstream voided methods. J. Am. Vet. Med. Ass., 1981, 179,891-895.

6. — DAVIES M., STEWART P.R. : Transferable drug resistance in manand animals : genetic relationship between R-plasmids in entericbacteria from man and domestic pets. Austr. Vet. J., 1978, 54, 507-512.

7. — DEVER L.A., DERMODY T.S. : Mechanisms of bacterial resistanceto antibiotics. Arch. Intern. Med., 1991, 151, 886-895.

8. — DROMIGNY J.A., NABETH P., PERRIER GROS CLAUDE J.D. :Distribution and susceptibility of bacterial urinary tract infections inDakar, Senegal. Int. J. Antimicrob. Agents, 2002, 20, 339-347.

9. — GUPTA K., HOOTON T.M., STAMM W.E. : Increasing antimicro-bial resistance and the management of uncomplicated communityacquired urinary tract infections. Ann. Int. Med., 2001, 135, 41-50.

10. — HOGLE R.M. : Antibacterial-agent sensitivity of bacteria isolatedfrom dogs and cats. J. Am. Vet. Med. Ass., 1970, 156, 761-764.

11. — JOHNSTON A.M. : Use of antimicrobial drugs in veterinary prac-tice. Brit. Med. J., 1998, 317, 665-667.

12. — KURAZONO I., NAKANO M., YAMAMOTO S., OGAWA O.,YURI K., NAKATA K., KIMURA M., MAKINO S., BALAKRISHNAIR G. : Distribution of the usp gene in uropathogenic Escherichiacoli isolated from companion animals and correlation with serotypesand size-variations of the pathogenicity island. Microbiol. Immunol.,2003, 47, 797-802.

13. — LEKCHAROENSUK C., OSBORNE C.A., LULICH J.P. :Epidemiologic study of risk factors for lower urinary tract diseases incats. J. Am. Vet. Med. Assoc., 2001, 218, 1429-1435.

14. — LING G.V., BIRBERSTEIN E.L., DWIGHT C.H. : Bacterial asso-ciated with urinary tract infections. Vet. Clin. North Am. Small Anim.Pract., 1979, 9, 617-631.

15. — LING G.V., NORRIS C.R., FRANTI C.E., EISELE P.H., JOHNSOND.L., RUBY A.L., JANG S.S. : Interrelations of organism preva-lence, specimen collection method, and host age, sex, and breedamong 8,354 canine urinary tract infections (1969-1995). J. Vet.Intern. Med., 2001, 15, 341-347.

16. — LOW D.A., BRUATEN B.A., LING G.V., JOHNSON D.L., RUDYA.L. : Isolation and comparison of Escherichia coli strains fromcanine and human patients with urinary tract infections. Infect.Immun., 1988, 56, 2601-2609.

17. — NORRIS C.R., WILLIAMS B.J., LING G.V., FRANTI C.E., JOHN-SON D.L., RUBY A.L. : Recurrent and persistent urinary tract infec-tions in dogs : 383 cases (1969-1995). J. Am. An. Hosp. Ass., 2000,36, 484-492.

18. — ODENSVIK K., GRAVE K., GREKO C. : Antibacterial drugs pres-cribed for dogs and cats in Sweden and Norway 1990-1998. Acta Vet.Scand., 2001, 42, 189-198.

19. — OXENFORD C.J., LOMAS G.R., LOVE D.N. : Bacteriuria in thedog. J. Small. Anim. Pract., 1984, 25, 83-91.

20. — PERSON J.-M., DE RIEUX-MEUNIER C., QUINTIN-COLONNAF. : L’infection urinaire chez le chien et le chat. Analyse des résultatsde 459 urocultures réalisées entre 1977 et 1981 à l’Ecole NationaleVétérinaire d’Alfort. Rec. Méd. Vét., 1989, 165, 45-53.

21. — SEGUIN M.A., VADEN S.L., ALTIER C., STONE E., LEVINEJ.F. : Persistent urinary tract infections and reinfections in dogs(1989-1999). J. Vet. Intern. Med., 2003, 17, 622-631.

22. — WARREN A.L., TOWSEND K.M., KING T., MOSS S., O’BOYLED., YATES R,M., TROTT D.J. : Multi-drug resistant Escherichiacoli with extended-spectrum β-lactamase activity and fluoroquino-lone resistance isolated from clinical infection in dogs. Austr. Vet. J.,2001, 79, 621-623.

23. — WEAVER A.D., PILLINGER R. : Lower urinary tract pathogens inthe dog and their sensitivity to chemoterapeutic agents. Vet. Rec.,1977, 101, 77-79.

24. — WOOLEY R.E., BLUE J.L. : Bacterial isolations from canine andfeline urine. Mod. Vet. Pract., 1976, 57, 535.538.

25. — WILSON R.A., KEEFE T.J., DAVIS M.A., BROWING M.T.,ONDRUSEK K. : Strains of Escherichia coli associated with uroge-nital disease in dogs and cats. Am. J. Vet. Med. Res., 1988, 49, 743-746.

Revue Méd. Vét., 2006, 157, 1, 35-41

SURVEY ON BACTERIAL ISOLATES FROM DOGS WITH URINARY TRACT INFECTIONS 41