Trends in Antibiotic Susceptibility/Resistance in Ocular Infections

Penny A. Asbell, MD, FACS, MBA

Due to the fact that serious eye infections are rare, studies of infections are published as case reports. Most reports in the literature are from a single center with a small sample size. They are retrospective and offer no longitudinal data. Identification of trends through these studies is difficult if not impossible. In addition, the data may not be reflective of the whole country’s trends or even those in an individual practice. For example, what may be relevant to ophthalmologists in Florida, where there is high humidity, may not be relevant to ophthalmologists in Nevada.

The Surveillance Network

A larger sample size can provide more information. The Surveillance Network (TSN) sampled antibiograms from more than 200 laboratories.¹ These antibiograms revealed the trend that the percentage of Staphylococcus aureus isolates sensitive to methicillin decreased between 2000 and 2005 while the percentage of methicillin-resistant S aureus isolates (MRSA) increased from 29.5% in 2000 to 41.6% in 2005. MRSA ocular isolates were multidrug resistant (ie, resistant to > 3 antibiotics, including all fluoroquinolones tested). Because S aureus continues to be one of the most common organisms causing endophthalmitis, MRSA continues to be a growing problem in ophthalmology.

Ocular TRUST

Tracking Resistance in the United States Today (TRUST) is an initiative that began when fluoroquinolones were first introduced. Researchers were concerned that widespread use of these agents for treating community-acquired pneumonia would make antibiotic resistance an increasing problem.

In 2006, Ocular TRUST was initiated. The goal of Ocular TRUST is to provide annual, longitudinal, nationwide surveillance of antibiotic resistance in ocular pathogens. Ocular TRUST is the largest surveillance program specific to isolates from non-endophthalmitis ocular infections. By the third year, Alaska and Wyoming were the only states not participating. The Ocular TRUST network comprises 104 institutions and 10 eye centers. To date, 80,000 specimens have been collected. There are large sample sizes (> 30 isolates per species). Isolates are submitted to a centralized independent laboratory for susceptibility testing by broth microdilution according to Clinical and Laboratory Standards Institute (CLSI) methodology. Isolates are subject to a comprehensive antimicrobial panel. Each isolate is tested twice to avoid bias and other errors that can make data collected less valuable.

The antimicrobial panel test included representatives of 6 classes of drugs: fluoroquinolones (ciprofloxacin, gatifloxacin, levofloxacin, moxifloxacin); dihydrofolate reductase inhibitors (trimethoprim); macrolides (azithromycin); aminoglycosides (tobramycin); polypeptides (polymyxin B); and ß-lactams (penicillin). Besifloxacin was not tested in Ocular TRUST as it was not FDA-approved at the initiation of the study.

Minimum inhibitory concentrations (MICs) were interpreted as susceptible, intermediate, or resistant according to CLSI 2008 published interpretive criteria (M100-S18), where available. Other interpretive criteria were used for staphylococci susceptibility to polymyxin B (CLSI breakpoints for Pseudomonas aeruginosa), Streptococcus pneumoniae susceptibility to ciprofloxacin (FDA breakpoints), polymyxin B (CLSI P aeruginosa breakpoints), tobramycin (package insert), and trimethoprim (CLSI Staphylococcus spp. breakpoints). Staphylococci were classified as methicillin-resistant or methicillin-susceptible based on susceptibility to oxacillin.

Staphylococcus aureus

Data gathered from 2006 through 2008 have been analyzed.² In the methicillin-susceptible S aureus (MSSA) isolates, there were similarly high percentages (80% to 93%) of isolates sensitive to fluoroquinolones. Tobramycin and trimethoprim were also effective against high percentages of the MSSA isolates, with at least 90% of isolates susceptible to these agents. S aureus was less susceptible to macrolides, as represented by azithromycin, which was effective against 54% to 62% of MSSA isolates in 2006-2008. Penicillin was effective against low percentages (10%-17%) of isolates.

When the antibiograms for MRSA collected from 2006 through 2008 were examined (Figure 1), significantly lower percentages of isolates (15%-30%) were susceptible to the fluoroquinolones compared to the MSSA isolates. Azithromycin was only effective against 6% to 8% of MRSA isolates. Tobramycin was effective against 36% of MRSA isolates in 2006, 50% in 2007, and 55% in 2008. Trimethoprim was effective against approximately 95% of MRSA isolates during this time period.² Thus trimethoprim was still effective against S aureus isolates that were methicillin-resistant.

Figure 1. Ocular TRUST: MRSA

Resistance of MRSA to fluoroquinolones is high, while trimethoprim is effective against this pathogen.
Source: Asbell P, et al. Presented at: Annual Meeting of the American Society of Cataract and Refractive Surgery; April 3-8, 2009; San Francisco, CA.

Click here for a larger view of this image.

Coagulase-negative Staphylococci

A similar pattern can be seen looking at coagulase-negative staphylococci (CoNS). Almost all samples of CoNS taken were Staphylococcus epidermis isolates. S epidermis is actually a normal inhabitant of the ocular surface. CoNS, however, is a significant cause of endophthalmitis and permanent vision loss if it penetrates the eye. Of the antimicrobials tested, only azithromycin, penicillin, and polymyxin B were not highly active against methicillin-susceptible CoNS. However, < 50% of methicillin-resistant CoNS isolates were susceptible to the fluoroquinolones. Tobramycin and trimethoprim were still somewhat active against methicillin-resistant CoNS, but the percentage of isolates susceptible to these antibiotics was significantly reduced.²

Streptococcus pneumoniae

S pneumoniae was another pathogen tested. The susceptibility rates of this organism against the fluoroquinolones tested were high and virtually identical, except for modestly lower susceptibility rates for ciprofloxacin (Figure 2).² Polymyxin B and tobramycin were not active against S pneumoniae isolates, and S pneumoniae isolates were less susceptible to azithromycin, penicillin, and trimethoprim than to fluoroquinolones.

Figure 2. Ocular TRUST: S Pneumoniae

Fluoroquinolones, azithromycin, penicillin, and trimethoprim were effective against S Pneumoniae, while resistance to polymyxin B and tobramycin was high.
Source: Asbell P, et al. Presented at: Annual Meeting of the American Society of Cataract and Refractive Surgery; April 3-8, 2009; San Francisco, CA.

Click here for a larger view of this image.

ARMOR

Another surveillance study, Antibiotic Resistance Monitoring in Ocular Microorganisms (ARMOR), has data from 2009 that reinforce the findings of Ocular TRUST.³ ARMOR also collected isolates from throughout the United States. The difference in the studies lies in the methodology. Both studies test for the MICs, the antibiotic concentration required to kill 90% of organisms overnight. Ocular TRUST combined that data with the available systemic data on the effectiveness of each antibiotic. Because that information is not available for all antibiotics, ARMOR used only the MICs to determine the sensitivity of a particular antibiotic. All topical fluoroquinolones available in the United States were tested, including besifloxacin.

As in TSN,¹ approximately 39% of the S aureus isolates were methicillin-resistant.³ In the ARMOR study, samples were separated into a group of community-acquired isolates (CA) and a group of hospital-acquired isolates (HA). As expected, CA S aureus isolates displayed much different characteristics than the HA S aureus isolates. Approximately 63% of the isolates in the study were HA and 26% were CA. The genetic code of all of the HA isolates contained staphylococcus cassette chromosome mec Type II (SCCmec Type II), a mec operon that codes for methicillin resistance. The isolates were all multi-drug resistant and over 95% were resistant to 4 or more classes of antibiotics. Finding the appropriate antibiotic to treat an HA ocular infection can, therefore, be difficult. The HA isolates, however, did not contain the genetic code for the toxin Panton-Valentine leukocidin (PVL), which renders S aureus more virulent. The isolates of the CA group were more likely to express the gene for PVL, causing tissue necrosis. However, the CA group did display less resistance to antibiotics. Many isolates were resistant to fluoroquinolones but contained SCC mec Type IVa, which does not confer methicillin resistance. Future studies are required to further discern the differences between HA and CA isolates.

In summary, across 3 years of surveillance in the Ocular TRUST program, antimicrobial susceptibility rates have fluctuated from year to year but have remained relatively stable. Polymyxin B and penicillin have the most limited spectra of activity of all antibiotics tested. Methicillin resistance in staphylococci is a marker of diminished susceptibility (ie, multi-drug resistance) to ophthalmic antimicrobials. If MRSA or methicillin-resistant CoNS are suspected, treatment with tobramycin or trimethoprim should be considered.

Fluoroquinolones should be taken into account when treating infections caused by S pneumoniae. Additional longitudinal data are needed to determine whether a change in S pneumoniae azithromycin susceptibility from 2006 to 2008 represents a true decline in azithromycin susceptibility or a year-to-year fluctuation. Susceptibility patterns for the 3-year period (2006-2008) are virtually identical for all fluoroquinolones and do not show a decline.

Large surveillance studies like Ocular TRUST and ARMOR are able to give the physician a starting point for antibiotic selection for a given infection. The physician cannot be sure the antibiotic chosen is optimal until the results of the culture and antibiotic sensitivity from the patient become available. Not every antibiotic will work for every infection. Thus, every physician should stay up-to-date on the data that Ocular TRUST and ARMOR provide as well as future sources of surveillance data on antibiotic sensitivities.

References

Asbell PA, Sahm DF, Shaw M, Draghi DC, Brown NP. Increasing prevalence of methicillin resistance in serious ocular infections caused by Staphylococcus aureus in the United States: 2000 to 2005. J Cataract Refract Surg. 2008;34(5):814-818.
Asbell P, et al. Presented at: Annual Meeting of the American Society of Cataract and Refractive Surgery; April 3-8, 2009; San Francisco, California.
Asbell P, et al. Presented at: Association for Research in Vision and Ophthalmology 2010 Annual Meeting; May 2-6, 2010; Fort Lauderdale, FL.