Antimicrobial Testing of Selected Fluoroquinolones Against Pseudomonas aeruginosa Isolated From Canine Otitis

Introduction

Canine otitis externa is a common dermatological problem with multiple underlying causes. Common primary causes include hypersensitivity disorders (e.g., adverse food reaction or atopic dermatitis), ectoparasites, foreign bodies, and keratinization disorders.¹ Otitis externa is often complicated by secondary infection with bacterial or fungal organisms that perpetuate the otitis once inflammation has been established.

Pseudomonas aeruginosa (P. aeruginosa) is a gram-negative aerobic bacterium that was present in 26% to 35% of cases of canine bacterial otitis, for which bacterial culture was performed, classically when rod bacteria were seen on cytology.^2–4 In one study, P. aeruginosa was the sole bacterial pathogen in 33% of cases in which it was isolated.⁴

Treatment of P. aeruginosa otitis is challenging, because some strains of the bacterium have developed multidrug resistance. Moderate to high levels of resistance to some commonly used antibiotics have been reported for P. aeruginosa isolates recovered from cases of chronic canine otitis.^2–4

Fluoroquinolones are a class of antibiotics with known activity against Pseudomonas.⁵ These bactericidal drugs prevent bacterial replication by inhibiting deoxyribonucleic acid (DNA) gyrase. Fluoroquinolones are excellent antibiotics against multidrug-resistant bacteria (such as P. aeruginosa) because of their broad-spectrum activity against both gram-positive and gram-negative bacteria, good tissue penetration, and long serum half-lives.

The fluoroquinolones approved for veterinary use in the United States (US) include enrofloxacin, marbofloxacin, and orbifloxacin. Enrofloxacin^a has been marketed in the US since 1989, and orbifloxacin^b has been marketed since 1997. Marbofloxacin^c was released to US markets in 1999, and it has gained popularity and widespread application within the last few years. High fluoroquinolone usage and inappropriate dosage regimens in dogs may increase the risk of resistant strain development over time.^2–4 Previous studies have reported susceptibility of P. aeruginosa to marbofloxacin and enrofloxacin,^2–4,6–11 but the authors are not aware of clinical studies quantifying sensitivity to orbifloxacin. However, the breakpoint minimum inhibitory concentration (MIC) data released by the orbifloxacin manufacturer suggest only marginal activity against P. aeruginosa.⁵

Both qualitative and quantitative methods are used for antimicrobial susceptibility testing. The disk-diffusion susceptibility test (DDST) is qualitative in nature, and it is currently the most common testing method used by veterinary microbiology laboratories.¹² Many previous P. aeruginosa susceptibility studies have used this method.^2–4,6–9 The DDST is performed by applying filter-paper disks impregnated with a known concentration of an antibiotic onto the surface of a Mueller-Hinton agar plate that contains a standardized inoculum of bacteria.After the plate has been incubated, the zone of inhibition is measured and correlated with the antibacterial activity of the drug.

Although DDST is widely used, dilution antimicrobial susceptibility testing is considered the gold standard.¹³ This quantitative method involves serial dilution of antibiotics in broth media. The lowest concentration of an antibiotic that prevents visible growth after incubation is known as the MIC. The quantitative results obtained with MIC testing (MICT) are considered more useful and reliable than results of DDST for determining therapeutic dosing regimens,¹³ although the degree of reliability is currently uncertain. In two studies, DDST underestimated P. aeruginosa sensitivity to enrofloxacin (when compared with MICT),^3,10 whereas in another study, DDST overestimated enrofloxacin susceptibility.¹¹

The authors hypothesize that marbofloxacin may have greater efficacy against P. aeruginosa than enrofloxacin, because enrofloxacin has been in veterinary use in the US for a longer period of time. The authors also hypothesize that P. aeruginosa isolates have low sensitivity to orbifloxacin, in accordance with information provided by the manufacturer. The purpose of the current study was to test these two hypotheses and to compare MICT and DDST sensitivity testing.

The susceptibilities of P. aeruginosa to marbofloxacin, enrofloxacin, and orbifloxacin were compared using MICT. Such updated surveys provide valuable information for veterinarians selecting antibiotics for animals. The authors also assessed agreement between MICT and DDST regarding P. aeruginosa sensitivity to enrofloxacin or marbofloxacin. This should help determine whether DDST overestimates or underestimates the susceptibility of P. aeruginosa isolates.

Materials and Methods

Pseudomonas aeruginosa isolates were collected from two sources: 1) clinical cases of canine otitis observed at two specialty private practices and 2) canine otitis specimens submitted to the clinical microbiology laboratory at the University of Florida. All samples were collected between February 2005 and September 2006. Strains were identified using Gram stain, colony morphology, and oxidase testing; they were confirmed with a commercially available biochemical test kit consisting of a nonenteric identification panel with 20 biochemical assays.^d These assays were incubated at 37°C for 24 to 48 hours, after which results were assessed colorimetrically.

Microtiter Sensititre plates^e containing titrated concentrations of marbofloxacin, enrofloxacin, and orbifloxacin were obtained from the manufacturer and stored at 25°C. Minimum inhibitory concentrations were determined by the microdilution broth method in accordance with the principles of the Clinical Laboratory Standards Institute (CLSI; formerly the National Committee for Clinical Laboratory Standards).¹⁴ For each clinical isolate of P. aeruginosa, three to five colonies from the primary agar plate were selected and transferred to a tube containing 4 mL of demineralized water; then 50 μL was transferred to 10 mL of Mueller-Hinton broth. This resulted in a final inoculum concentration of approximately 5 × 10⁵ colony-forming units per mL. Finally, 50 μL of this suspension was added to each well of the Sensititre susceptibility plate and incubated at 35°±2°C for 18 hours. A quality control test using P. aeruginosa reference strain ATCC 27853 was performed upon receipt of the plates, as recommended by the manufacturer.^e In addition, each microtiter plate had its own positive and negative control; these controls were run when the isolate was tested.

The MIC was defined as the lowest concentration of antibiotic that inhibited visible growth. Susceptibility breakpoints were based on CLSI recommendations:¹⁴

• Marbofloxacin: Susceptible strains have an MIC ≤1 μg/mL; intermediate strains have an MIC >1 but <4 μg/mL; and resistant strains have an MIC ≥4 μg/mL.

• Enrofloxacin: Susceptible strains have an MIC ≤0.5 μg/mL; intermediate strains have an MIC >0.5 but <4 μg/mL; and resistant strains have an MIC ≥4 μg/mL.

• Orbifloxacin: Susceptible strains have an MIC ≤1 μg/mL; intermediate strains have an MIC >2 but ≤4 μg/mL; and resistant strains have an MIC ≥8 μg/mL.

Each isolate was also tested using the DDST for marbofloxacin (5 μg) and enrofloxacin (5 μg); DDST for orbifloxacin was not performed, because this method was not routinely used in the microbiology laboratory at the University of Florida, and resources were limited. For each clinical isolate, P. aeruginosa colonies were transferred to a tube of sterile deionized water to make a standardized suspension (0.5 McFarland turbidity standard). The suspension of P. aeruginosa was applied to a Mueller-Hinton agar plate, and individual filter paper disks containing known concentrations of the individual antibiotics were placed on the inoculated surface. The plate was incubated for 18 to 24 hours at 35°±2°C. The zone of inhibition around each disk was measured, and the isolate was categorized as sensitive, intermediate, or resistant according to CLSI specifications:¹⁴

• Enrofloxacin: Susceptible if the zone of inhibition was ≥21 mm, intermediate if the zone of inhibition was 16 to 20 mm, and resistant if the zone of inhibition was ≤15 mm.

• Marbofloxacin: Susceptible if the zone of inhibition was ≥20 mm, intermediate if the zone of inhibition was 15 to 19 mm, and resistant if the zone of inhibition was ≤14 mm.

• Pseudomonas aeruginosa reference strain ATCC 27853 was used as a control. The data for each isolate were then coded (0=sensitive, 1=intermediate, and 2=resistant) and recorded in Microsoft Excel.

Statistical Analysis

All statistical tests were conducted using Statistical Package for the Social Sciences (SPSS), version 13. To evaluate susceptibility differences among the three fluoroquinolones, three paired-sample Wilcoxon z-tests were computed on the data from isolates subjected to MICT. Paired-sample Wilcoxon z-tests were also computed on the data from isolates subjected to DDST for marbofloxacin and enrofloxacin. The differences between the two methods (MICT and DDST) were analyzed using separate paired sample Wilcoxon z-tests for marbofloxacin and enrofloxacin. Statistical significance was set at α=0.05.

Results

Of the 100 P. aeruginosa isolates collected, MICT for marbofloxacin identified 50 strains that were sensitive, 49 that were resistant, and one that had intermediate sensitivity. Enrofloxacin MICT identified 16 sensitive strains, 53 resistant strains, and 31 intermediate strains. Orbifloxacin MICT identified 15 sensitive strains, 65 resistant strains, and 20 intermediate strains. Based on MICT, P. aeruginosa isolates were significantly more sensitive to marbofloxacin than to enrofloxacin (z = −4.57; P<0.05) or orbifloxacin (z = −5.02; P<0.05). The isolates were also significantly more susceptible to enrofloxacin than to orbifloxacin (z = −2.25; P<0.05).

Based on DDST results, the isolates were also significantly more susceptible to marbofloxacin than to enrofloxacin (z = −4.8; P<0.05), with 47 strains sensitive to marbofloxacin (with 46 resistant strains and seven intermediate strains) and 28 strains sensitive to enrofloxacin (with 50 resistant strains and 22 intermediate strains).

For marbofloxacin, results of MICT and DDST were in 87% agreement, with no statistically significant disagreement between the two methods (z = −0.07; P>0.05). This comparison identified eight minor errors and five major errors [Table 1]. A minor error was defined as a discrepancy of 1° of susceptibility, such as a strain demonstrating susceptibility on DDST but showing only intermediate sensitivity on MICT. A major error was defined as a discrepancy of 2° of susceptibility, such as a strain demonstrating resistance on MICT but showing susceptibility on DDST. In six instances, DDST overestimated the number of marbofloxacin-susceptible strains compared with MICT. In seven instances, DDST underestimated the number of marbofloxacin- susceptible strains compared with MICT.

For enrofloxacin, results of MICT and DDST were in 74% agreement, with a statistically significant difference between the two tests (z = −2.79; P<0.05). Twenty-five minor errors and one major error were identified. In 20 instances, DDST overestimated the number of enrofloxacin-susceptible strains compared with MICT; in only six instances did DDST underestimate the number of enrofloxacin-susceptible strains [Table 2].

Discussion

In a study by Meunier et al., the MICT method was used to determine marbofloxacin susceptibility among P. aeruginosa strains; canine otic isolates obtained from cases in several European countries maintained high (86.7%) susceptibility to marbofloxacin over a 7-year period (1994 to 2001).¹⁵ Marbofloxacin sensitivity against otic P. aeruginosa was also high (89.5%) in a European study using the DDST method.⁷ The rates of marboflaxacin susceptibility found in this current study were lower than those found in either of the prior studies. The authors’ findings are in closer agreement with a recent US study in which only 66.7% marbofloxacin sensitivity via DDST was reported.⁹ The variability in results among studies suggests regional differences among P. aeruginosa isolates from Europe and the US.

Temporal differences may also account for the lower percentages of marbofloxacin-sensitive strains reported here. The European studies were performed between 1994 and 2001,^7,15 while the previous US study⁹ and the authors’ study were performed between 2005 and 2006—approximately 5 years after marbofloxacin became commercially available in the US. Increased use of fluoroquinolones over time tends to promote resistance,⁸ although it is difficult to estimate the amount of time that must elapse before resistance might be expected. Meunier et al. reported that canine otic isolates maintained a high susceptibility to marbofloxacin over a 7-year period,¹⁵ while Hariharan et al. reported increased enrofloxacin resistance in paired studies over an 11-year period.^6,8

The authors speculated that there would be more resistance to enrofloxacin than to marbofloxacin, because marbofloxacin was released into the US veterinary market 10 years after enrofloxacin. Results of this study support this hypothesis, given the low enrofloxacin sensitivity compared with that of marbofloxacin. The authors also speculated that enrofloxacin resistance in the current study would be higher than that reported in previous US studies. Three US studies that used DDST have reported enrofloxacin sensitivity ranging from 12.5% to 51% among P. aeruginosa isolates from canine otitis.^2,3,9 The results of this study most closely agree with those of Cole et al., who reported a sensitivity of 12.5% to enrofloxacin.² Two studies from the same region of Canada reported that P. aeruginosa resistance to enrofloxacin increased from 9% to 38% over an 11-year period (1995 to 2006).^6,8 Such results demonstrate that enrofloxacin resistance can develop over time within the same clinical population.⁸

Manufacturer testing with orbifloxacin produced an MIC⁹⁰ of 12.5 μg/mL for P. aeruginosa, suggesting that most strains will likely be resistant under current MIC breakpoints.⁵ This study was in general agreement with the manufacturer’s documentation, given the relatively low orbifloxacin sensitivity (15%). These findings suggest that orbifloxacin does not have high activity against Pseudomonas.

Resistance in Pseudomonas spp. developed through mutations in DNA gyrase or through the presence or size of porins in the outer membrane of the bacterium. These mutational changes can occur rapidly, even within a single course of antimicrobial therapy. Although plasmid-mediated resistance against fluoroquinolones is not currently thought to occur, indiscriminate use could lead to plasmid-mediated resistance.⁵ Empirical use of any antimicrobials for treatment of suspected P. aeruginosa infections is not recommended; therapy should be guided by assessment of antimicrobial susceptibility testing.

The authors’ study compared the qualitative DDST method with the quantitative MICT method of antimicrobial susceptibility testing. The results obtained with MICT are considered more accurate and useful for establishing therapeutic dosages.¹³ An 87% agreement was found between MICT and DDST for marbofloxacin, and a 74% agreement was found between MICT and DDST for enrofloxacin. The two methods were not significantly different (z = −0.07; P>0.05) for marbofloxacin, but they were significantly different (z = −2.79; P<0.05) for enrofloxacin. It is unclear why greater agreement between testing methods was found for marbofloxacin than for enrofloxacin, given that similar protocols were used with each antibiotic. The results of this study agree quite well with those of DeBoer et al., who tested 100 P. aeruginosa isolates by both DDST and MICT and reported 88% agreement for marbofloxacin and 79% agreement for enrofloxacin.¹⁰ However, Cole et al. and Colombini et al. reported less agreement (30%to 54.3%) when these two methods were compared in enrofloxacin susceptibility testing.^3,11

This study showed that DDST tends to overestimate the number of enrofloxacin-susceptible Pseudomonas strains when compared with MICT. However, approximately equal numbers of overestimation and underestimation errors were found for marbofloxacin, so any general trends should be similar using either DDST or MICT.

DeBoer et al. and Colombini et al. found that enrofloxacin DDST underestimated the susceptibility determined by enrofloxacin MICT, so DDST provided a more conservative interpretation.^3,10 In contrast, the authors’ study and that of Cole et al.¹¹ showed enrofloxacin DDST to overestimate the number of susceptible strains, so reliance on DDST might occasionally result in ineffective treatment. The reason for this discrepancy is uncertain. Many aspects of DDST can introduce variability, but results should be comparable when these variables are standardized. In human medicine, laboratories have reported 95% reproducibility among DDST results when methods were standardized.¹⁶ However, no regulatory oversight of veterinary diagnostic laboratories occurs in the US, and a recent study found that not all veterinary laboratories follow current CLSI guidelines.¹² Thus, variability in DDST methodology may account for at least some of the discrepancies between DDST and MIC reported among veterinary laboratories. Laboratory procedures in the current study were performed in accordance with CLSI guidelines, so reported differences are likely caused by inherent qualitative differences between DDST and MICT.

Conclusion

Results of this study showed that canine otic isolates of P. aeruginosa are more sensitive to marbofloxacin than to either enrofloxacin or orbifloxacin. Based on these results, the use of DDST to evaluate the enrofloxacin susceptibility of P. aeruginosa may result in an overestimation of susceptibility and the potential for ineffective treatment. Because the authors investigated only fluoroquinolone antibiotics licensed for veterinary use in the US, and nonveterinary products were not included, it would be useful in the future to obtain similar antimicrobial testing data on other fluoroquinolone antibiotics (such as ciprofloxacin) that are frequently used off-label for treatment of veterinary Pseudomonas infections.