Bacterial species distribution and antimicrobial susceptibility of canine ocular isolates in Japan: retrospective analysis of a diagnostic laboratory database

Abstract

This retrospective study investigated the bacterial species distribution and antimicrobial susceptibility of canine ocular isolates submitted to a Japanese diagnostic laboratory between October 2022 and March 2026. A total of 737 isolates were recovered, with Staphylococcus pseudintermedius (248/737, 33.6%) and Streptococcus canis (116/737, 15.7%) the most common isolates. Notably, 100 (40.3%) of the S. pseudintermedius isolates were methicillin-resistant. Cefotaxime demonstrated high activity against major Gram-positive pathogens, except for methicillin-resistant S. pseudintermedius, and Escherichia coli. In contrast, susceptibility to gatifloxacin varied markedly among the species, whereas the spectrum of gentamicin and polymyxin B was limited due to intrinsic resistance among common Gram-positive pathogens. These findings provide useful information for empirical antimicrobial selection in canine ocular infections in Japan.

Bacterial infections are a common cause of ocular surface disease in dogs, and they may result in conjunctivitis, keratitis, corneal ulceration, and vision-threatening complications if left untreated [1, 6]. Because topical antimicrobial therapy is often initiated empirically before culture and antimicrobial susceptibility results become available, knowledge of the species associated with canine ocular infections and their antimicrobial susceptibility patterns is essential for appropriate therapeutic decision-making [1, 6].

In veterinary ophthalmology, a variety of topical antimicrobial agents are available for the treatment of bacterial ocular infections. Commonly used preparations include aminoglycosides (e.g., gentamicin [GEN] and tobramycin), fluoroquinolones (e.g., ofloxacin, levofloxacin, and gatifloxacin [GAT]), cephalosporins such as cefmenoxime (CMX), and chloramphenicol (CHL) [6, 20]. These agents differ substantially in their antimicrobial spectrum and activity against Gram-positive and Gram-negative pathogens. Therefore, knowledge of local bacterial epidemiology and antimicrobial susceptibility patterns is essential for selecting appropriate empirical therapy.

Previous studies have identified Staphylococcus spp., Streptococcus spp., Pseudomonas aeruginosa, and other opportunistic pathogens as important causes of canine ocular infections [1, 7]. Additionally, increasing antimicrobial resistance among canine ocular pathogens, particularly methicillin-resistant Staphylococcus pseudintermedius (MRSP), has raised concerns regarding the continued effectiveness of empirical antimicrobial therapy [14, 15]. Because bacterial species’ distribution and antimicrobial resistance patterns vary geographically and over time, regional surveillance is necessary to support evidence-based antimicrobial selection [17].

Although several studies have reported bacterial isolation frequencies and antimicrobial susceptibility patterns in canine ocular infections [1, 7, 8], comparable data from Japan remains limited. The objectives of the present study were to investigate the bacterial species distribution of canine ocular isolates submitted for diagnostic testing in Japan and to evaluate the antimicrobial susceptibility profiles of major bacterial pathogens.

This retrospective study analyzed bacterial isolates recovered from canine ocular samples submitted to the FUJIFILM VET Systems Co., Ltd. (Tokyo, Japan) from October 2022–March 2026. Samples were collected from client-owned dogs presenting to veterinary clinics and hospitals throughout Japan for the diagnosis of ocular diseases. The isolates were obtained from veterinary facilities located in 32 prefectures spanning all major geographic regions of Japan, from Hokkaido to Okinawa, thereby providing broad geographic representation. Ocular specimens were submitted for bacterial culture as part of routine clinical investigations at the discretion of the attending veterinarians. Duplicate isolates of the same bacterial species recovered from the same animal were excluded whenever possible to minimize individual animals’ overrepresentation. When multiple distinct strains were isolated from a single sample, each strain was considered an independent isolate and included in the analysis. All data were obtained retrospectively from a diagnostic laboratory database and generated as part of routine clinical diagnostic procedures. No animals were sampled specifically for the purposes of this study.

Clinical samples were cultured using standard microbiological procedures routinely employed at the diagnostic laboratory. Bacterial isolates were identified to the species level using matrix-assisted, laser desorption/ionization, time-of-flight mass spectrometry (MALDI-TOF MS) with the MALDI Biotyper (Bruker Daltonics, Bremen, Germany). In general, a single spot was analyzed for each isolate. Identification scores of ≥2.0 were considered reliable for species-level identification, whereas scores of 1.7 to <2.0 were considered reliable for genus-level identification and were reported as sp. Isolates with scores <1.7 were considered as unidentified.

Antimicrobial susceptibility testing was performed using the broth microdilution method following the Clinical and Laboratory Standards Institute (CLSI) guidelines [3, 4]. The antimicrobial agents evaluated included representative agents from antimicrobial classes commonly used in veterinary ophthalmology. The tested drugs included a third-generation cephalosporin (cefotaxime [CTX]), an aminoglycoside (GEN), a fluoroquinolone (GAT), a polymyxin (polymyxin B [PMB]), and phenicol (CHL). Minimum inhibitory concentrations (MICs) were determined using custom-made microdilution panels (Frozen Plate, Eiken Chemical Co., Ltd., Tokyo, Japan) and the results were interpreted using the IA20 MIC Pro system (Koden Industry Co., Ltd., Tokyo, Japan). Susceptibility breakpoints for GEN and CHL were based on CLSI supplement VET01S [3]. Because veterinary-specific breakpoints have not been established for CTX, GAT, or PMB, interpretive criteria from CLSI supplement M100 were applied [4]. For statistical analyses, isolates categorized as intermediate were grouped with resistant isolates and considered non-susceptible. Methicillin resistance in staphylococci was determined based on oxacillin MICs according to the CLSI guidelines [3].

The frequency of isolated bacterial species (Table 1) was expressed as the number and percentage of total isolates. Antimicrobial susceptibility rates were calculated for major species as the proportion of susceptible isolates among all isolates tested for each antimicrobial agent. Corynebacterium spp. were excluded from susceptibility analyses because antimicrobial susceptibility testing was performed using laboratory-specific criteria rather than CLSI guideline M45 [2]. Descriptive statistics were used to summarize species distribution and antimicrobial susceptibility profiles.

Table 1. Bacterial species distribution among canine ocular isolates collected in Japan.

Genus		Number of isolates (%)
Staphylococcus spp.		335 (45.5)
	Sta. pseudintermedius	248 (33.6)
	Sta. schleiferi	44 (6.0)
Streptococcus spp.		156 (21.2)
	Str. canis	116 (15.7)
Corynebacterium spp.		61 (8.3)
Pseudomonas spp.		56 (7.6)
	P. aeruginosa	55 (7.5)
Escherichia coli		35 (4.7)
Moraxella canis		15 (2.0)
Pasteurella spp.		11 (1.5)
Enterococcus spp.		8 (1.1)
Proteus spp.		8 (1.1)
Klebsiella spp.		6 (0.8)
Rothia nasimurium		5 (0.7)
Stenotrophomonas maltophilia		5 (0.7)
Acinetobacter spp.		4 (0.5)
Serratia spp.		4 (0.5)
Frederiksenia canicola		4 (0.5)
Other speciesa		24 (3.3)

Data are presented as number of isolates (% of total isolates). ^a)Other species include Enterobacter spp. (n=3), Citrobacter koseri (n=3), Neisseria dumasiana (n=2), Haemophilus spp. (n=2), Psycrobacter spp. (n=2), Achromobacter xylosoxidans (n=1), Arcanobacterium sp. (n=1), Bacteroides pyogenes (n=1), Capnocytophaga cynodegmi (n=1), Chryseobacterium indologenes (n=1), Elizabethkingia miricla (n=1), Lactobacillus sp. (n=1), Micrococcus luteus (n=1), Morganella morganii (n=1), Ochrobactrum sp. (n=1), Pseudarthrobacter oxydans (n=1), and unidentified bacteria (n=1).

A total of 737 ocular bacterial isolates were recovered from 409 dogs during the study period. Staphylococcus spp. were the most frequently isolated bacteria (335/737, 45.5%), followed by Streptococcus spp. (156/737, 21.2%), Corynebacterium spp. (61/737, 8.3%), Pseudomonas spp. (56/737, 7.6%), and Escherichia coli (35/737, 4.7%) (Table 1). Staphylococcus spp. being the most frequently isolated bacteria was consistent with previous studies of canine ocular infections [1, 7]. Among them, Sta. pseudintermedius was the predominant species, accounting for approximately one-third (248) of all isolates, confirming its importance as a major canine ocular pathogen across geographic regions. Of the 248 Sta. pseudintermedius isolates, 100 (40.3%) were identified as MRSP. This proportion of MRSP is comparable to that reported in Korea (36.0%) [9], but much higher than in Germany (7.9%) [18], suggesting that MRSP is widely distributed among canine ocular pathogens in Japan, as well as other countries.

A notable finding was the relatively high percentage of Str. canis, which accounted for 116/156 (74.4%) of streptococcal isolates and 15.7% of all bacterial isolates. This organism was recovered more frequently than Pseudomonas spp. in the present study, which is consistent with a study in the United States [7], although staphylococci and Pseudomonas spp. were the predominant bacterial pathogens in Brazil [1]. The high frequency of Str. canis observed in the present study suggests that this β-hemolytic Lancefield group G streptococcus may play a more important role in canine ocular infections in Japan, consistent with the recent report by Kurita et al. [10]. Given its recognized pathogenic potential and association with invasive and occasionally rapidly progressive infections in dogs [11], this finding may have important clinical implications.

As for antimicrobial susceptibility (Table 2), CTX exhibited good activity against major Gram-positive pathogens, with susceptibility rates of 96.8% in methicillin-susceptible Sta. pseudintermedius (MSSP), 77.8% in Sta. schleiferi, and 100% in Str. canis, in addition to E. coli (76.9%). Nevertheless, when MRSP isolates are taken into account, the overall coverage of Gram-positive cocci is substantially reduced due to the high prevalence of MRSP, suggesting that CTX activity against Gram-positive pathogens is not consistently high. Because CMX, a third-generation cephalosporin widely used in veterinary ophthalmology, has been reported to exhibit an antimicrobial spectrum and activity comparable to those of CTX [19], these findings suggest that CMX may retain activity against common Gram-positive pathogens and E. coli with canine ocular infections in Japan. Nevertheless, the clinical utility of CMX as an empirical treatment should be interpreted with caution because detailed clinical information, including prior antimicrobial exposure, was not available for many cases in this study. In addition, the limited activity of CMX against MRSP and P. aeruginosa, attributable to intrinsic or acquired resistance [3, 4], should also be considered when selecting empirical antimicrobial therapy.

Table 2. Antimicrobial susceptibility profiles of major bacterial species isolated from canine ocular samples in Japan.

Antimicrobial class	Drug	Gram-positive cocci				Gram-negative rod
		MSSP (n=148)	MRSP (n=100)	Sta. schleiferi (n=44)	Str. canis (n=116)	P. aerginosa (n=55)	E. coli (n=35)
Third-generation cephalosporins	CTX	96.8 (92/95)	NA	77.8 (28/36)	100 (80/80)	NA	76.9 (20/26)
Aminoglycosides	GEN	35.0 (42/120)	6.6 (5/76)	82.9 (29/35)	NA	100 (43/43)	79.3 (23/29)
Fluoroquinolones	GAT	35.9 (28/78)	6.1 (3/49)	37.5 (12/32)	63.9 (39/61)	100 (33/33)	31.6 (6/19)
Polymixyns	PMB	NA	NA	NA	NA	100 (27/27)	15.8 (3/19)
Phenicols	CHL	40.0 (52/130)	20.2 (18/89)	100 (40/40)	95.1 (98/103)	NA	90.3 (28/31)

Data are presented as percentage susceptible, with the number of susceptible isolates/number of isolates tested shown in parentheses. NA, not applicable due to intrinsic resistance according to CLSI criteria. MSSP, methicillin-susceptible Sta. pseuintermedius; MRSP, methicillin-resistant Sta. pseudintermedius; CTX, cefotaxime; GEN, gentamicin; GAT, gatifloxacin; PMB, polymyxin B; CHL, chloramphenicol.

Interestingly, susceptibility to CHL differed markedly among Gram-positive pathogens. Sta. schleiferi (100%) and Str. canis (95.1%), as well as E. coli (90.3%), remained highly susceptible, and susceptibility among MSSP and MRSP was considerably lower (40.0% and 20.2%, respectively). CHL susceptibility rates in our study were lower than those reported in previous studies [1, 7]. Reasons for the lower CHL susceptibility observed in MSSP, in addition to MRSP, remain unclear, but reduced CHL susceptibility has also been reported in MSSP isolates from other infected sites in dogs [5]. This suggests that CHL resistance determinants may be relatively common in canine ocular S. pseudintermedius populations irrespective of methicillin resistance status [16].

Although GAT is generally considered a broad-spectrum ophthalmic antimicrobial, the present study revealed substantial species-specific differences in susceptibility, ranging from 6.1% in MRSP to 100% in P. aeruginosa. The low fluoroquinolone susceptibility of MRSP isolates is consistent with previous reports [14, 18], possibly due to its multidrug-resistant phenotype [15]. However, the reason for the low susceptibility rate observed in other staphylococcal isolates remains unclear but may reflect the large use of fluoroquinolone drugs for staphylococci in Japan. In contrast, the complete susceptibility of P. aeruginosa is consistent, although lower susceptibility rates have been reported in Brazil (75%) [13] and the United States (45.5%) [7]. Collectively, these findings suggest that fluoroquinolone susceptibility was variable among common canine ocular pathogens in Japan, with particularly high activity observed against P. aeruginosa. However, because detailed clinical information was unavailable for many isolates, these results should not be interpreted as direct evidence for empirical treatment recommendations.

GEN maintained excellent activity against Gram-negative pathogens, including P. aeruginosa (100%) and E. coli (79.3%), while PMB exhibited complete activity against P. aeruginosa but limited activity against E. coli (15.8%). These findings are consistent with previous reports indicating susceptibility of canine ocular Pseudomonas isolates to aminoglycosides and polymyxins [1, 7, 13]. Therefore, these antimicrobial classes remain appropriate empirical choices when Pseudomonas infection is strongly suspected clinically. Notably, Str. canis is intrinsically resistant to aminoglycosides, and Gram-positive bacteria are intrinsically resistant to polymyxins [3, 4, 12]. Given the relatively high prevalence of Gram-positive organisms, particularly Str. canis, in the present study, this intrinsic resistance would substantially reduce the overall empirical coverage provided by aminoglycosides and polymyxins.

Several limitations should be acknowledged. Clinical diagnoses, disease severity, previous antimicrobial treatment, and treatment outcomes were unavailable due to the retrospective, laboratory-based design. Additionally, susceptibility interpretations were based on systemic CLSI breakpoints and may not directly predict efficacy following topical ocular administration. Moreover, full MIC distributions were not determined because susceptibility testing was designed for categorical interpretation. Therefore, the relationship between observed susceptibility patterns and drug concentrations achievable in ocular tissues could not be evaluated. Nevertheless, the large number of isolates analyzed in this study provides valuable epidemiological information regarding canine ocular bacterial pathogens and their antimicrobial susceptibility profiles in Japan.

In conclusion, Sta. pseudintermedius and Str. canis were the predominant bacterial species isolated from canine ocular samples in Japan. Third-generation cephalosporins demonstrated excellent activity against major Gram-positive pathogens and E. coli, while the empirical coverage provided by fluoroquinolones, aminoglycosides, and polymyxins was substantially influenced by species-specific resistance patterns. To the best of our knowledge, this is the first large-scale study describing bacterial species distribution and antimicrobial susceptibility profiles among canine ocular isolates in Japan. These findings highlight the importance of regional surveillance for optimizing empirical antimicrobial therapy in canine ocular infections.

CONFLICTS OF INTEREST

There are no conflicts of interest to declare.