Bacterial Resistance: Causes and Consequences

Our evolving understanding of the relationship between antibiotic exposure, bacterial virulence and clinical outcome.

By Joseph Capriotti, MD, and Jesse S. Pelletier, MD

There has been an unequivocal increase in the isolation of antibiotic-resistant organisms over the last few decades.¹ The literature is overflowing with reports, statistics, trends, case series, in vitro data and laboratory studies documenting the ascendancy of resistant organisms in almost every human infectious disease that has been examined. There is a clear and constant reminder in the medical literature and the popular press that we are under assault from MRSA, VRE, penicillin-resistant Strep, multi-resistant salmonellae, MDR-TB, chloroquine-resistant malaria and hyper-resistant strains of HIV. Most recently, we have seen the emergence of carbapenem-resistance in the enterobacteriaceae.²

While most of these alarms are less urgently sounded in ophthalmology, there is a steady and pervasive message that even the newest and best antibiotics are losing efficacy against many ocular pathogens. And this much is certainly true: resis tant organisms are on the rise as antibiotic sensitivity patterns continue to evolve. Ocular specimens are in creasingly resistant to the common, inexpensive antibiotics and the more expensive third- and fourth-generation fluoroquinolones.^4-6 Rather than revisit this trend, we will instead analyze the phenomenon of resistance from a slightly different perspective. We will describe the general properties of anti biotics and antibiotic resistance, highlight some interesting studies from other specialties and finally examine the relationship between resistance and outcome in ophthalmic infections.

Categories of Antibiotics and Mechanisms of Resistance

Bacteria cause infections and antibiotics kill bacteria. If bacteria come in contact with antibiotics, they have two choices: adapt or die. This lies at the heart of antibiotic resistance and forms the basis of selection pressure.

To understand the phenomenon of antibiotic resistance more completely, we must first define what we mean by antibiotic. It typically describes a substance produced by a microorganism that can inhibit the growth of bacteria in dilution. This description of antibiotic (in contradistinction to antiseptic) excludes many naturally occurring substances that kill bacteria but are not produced by microorganisms — acids, steam, radiation, metal cations, iodine, ozone, etc. — and also historically excluded synthetic antibacterial compounds, many of which had been identified during the nascent aniline dye industry in late 19th century Germany.

Paul Ehrlich had developed a synthesis of arsephenine as early as 1907 that was commercialized as an anti-syphilitic by Hoechst in 1910.⁷ Though his work in immunology and chemotherapy earned him a well-deserved Nobel Prize in 1908, Erlich's most unsung contribution may be the singlehanded refinement of the modern pharmaceutical research model: synthesis of novel compound, chemical modification, lead optimization and clinical evaluation. The commercialization of arsephenine is the culmination of a systematic investigation into trivalent arsenic chemistry, derivativization, in vitro screening, animal studies and finally human trials.

Alexander Fleming took a different approach that focused on the isolation of antibiotic substances from living organisms, most famously the identification of penicillin from Penicillium notatum. With the supporting chemistry and clinical work of Edward Chain and Howard Florey, the penicillin era was set to change medicine forever, and all three shared the Nobel Prize in 1945 for their penicillin work. Joining the systemic synthetic methodology of Ehrlich to the natural antibiotics of Fleming yields the modern complement of naturally occurring, purely synthetic and semi-synthetic antibiotics.

Common to all antibiotics — again in contrast with antiseptics — is a precise efficacy at one or a few key sites within bacteria. This allows them to specifically damage bacteria without harming the host. Antibiotics are targeted inhibitors of some very specific cellular process necessary for microbial survival. It is precisely this exquisite specificity that makes them such excellent human drugs but also allows them to be neutralized with relatively small changes in bacterial biochemistry. Small, survivable modifications in bacterial architecture and electronic structure can completely circumvent the activity of almost all antibiotics. The specific mechanisms employed by any given bacteria to survive any given antibiotic challenge have been well described in countless reviews.^8,9 For our purposes, we will review only the general categories of resistance into which all mechanisms fall: (1) innate, (2) mutational or (3) acquired (conjugation, transduction, transformation, transposon-mediated).

Innate or natural resistance exists in some organisms to specific antibiotics. These are usually structural properties that either limit the access or quench the activity of the antibiotic. If a drug can't access the target, or penetrate the cell wall or bind to the receptor active site, it is essentially useless.

These organisms persist in the environment, as they are predisposed towards survival, particularly in selection environments like hospitals, rife with antibiotics that eliminate the less robust colonies. Pseudomonas aeruiginosa and Acinetobacter baumannii are notorious for well-known innate resistance, though there is often a component of acquired resistance coincident in both species.^10,11 The important feature of innate resistance is that it can exist without any prior exposure to or selection pressure from any existing antibiotic.

Resistance can be conferred by very subtle changes at any point on the bacteria/antibiotic interaction landscape. These mutational changes could confer resistance to an individual drug, an entire class of antibiotics or even resistance to multiple classes of drugs.^12,13 Point-mutations in active sites that only minimally change the drug-binding kinetics may be enough to confer resistance. When you consider that a bacterial genome will experience spontaneous mutations about once in every 10⁷ progeny, that a typical infection may contain 10⁵ CFUs and that a bacterial "generation" can be as few as 20 minutes, it is not surprising that spontaneous mutations occur within clinically relevant timescales. If a mutation arises that confers some survival advantage, it will be preferentially expressed through selection pressure.

Though not all mutations will confer antibiotic resistance, and not all resistant mutations will be stable, there is a high likelihood that some stable, resistance-conferring spontaneous mutations will arise in some infections. It is not an unreasonable proposition that given enough time, and the right selection pressure, any given bacterial population may be able to develop resistance to any given antibiotic. These can go on to survive further through a variety of exchange mechanisms with progeny or with adjacent bacterial neighbors.

Viewed from this perspective, it may actually be surprising that we see such relatively infrequent resistance given the sheer number of infectious colonies, imperfectly reproducing, spontaneously mutating and struggling for survival at any given time. Indeed there are species that are particularly poor at developing resistance (e.g., beta-hemolytic Strep isolated from the pharynx) despite repeated exposure to conditions that would seem to favor its presence.

Bacteria have the ability to pass on resistance genes through replication and also through a variety of exchange mechanisms. Conjugation (exchange of plasmids through ad jacent contact), transformation (assimilation of naked DNA from degenerating cell fragments) and transduction (phage-mediated DNA transfer) are all important mechanisms for the diversification of the bacterial genome. All these mechanisms can confer resistance-encoding genomic information to unrelated bacteria. Much work has been done on describing which mechanisms, plasmids and phages are important in development of resistance to a variety of agents. Finally, transposons (short gene segments that can "hop" from one site to another) can also be effective resistance transfer agents and can persist in the genomic population of bacteria long after any selection pressure has been removed.

The key thing to consider is that all of these exchange mechanisms can produce survivable mutations that lead to an increase in the resistant population. All adjacent exchange pathways can rapidly transmit the successful mutations through a population within a clinically significant time scale. Bacterial resistance can quickly spread through a single bacterial population and genetic diversification can rapidly be achieved.

Resistance Before and After Exposure

Bacteria have been around for at least three billion years, while the modern antibiotic era is only about 100 years old. Though the "Penicillin Era" marks the beginning of the commercial human use of systemic antibiotic therapy, it is not the first exposure of bacteria to penicillin. Along with macrolides, tetracyclines and in fact all naturally occurring antibiotics, crude penicillin has been used for millennia. There are reported examples in Europe of the use of molds for the treatment of infections as early as 1640 and evidence that the Egyptian physician Imhotep used moldy bread to treat surface infections as early as 2600 BC.¹⁴ There is evidence of tetracycline consumption in early Sudanese Christian populations that appears to correspond to a relatively low prevalence of infectious disease.¹⁵

So even though we have dramatically changed the selection landscape, at least for the naturally occurring antibiotics, it is incorrect to suggest that there was no exposure before 1945. It is entirely expected that some low-level acquired resistance to natural antibiotics has existed in some form for at least the 5000 years of human use and probably as long as the 10⁹ years that yeasts themselves have harbored them. Fleming himself describes both innate and acquired resistance during the course of many of his earliest experiments with penicillin.

While innate resistance to natural antibiotics could possibly be enhanced by prior ecological or human exposure, a more curious story unfolds when we examine the earliest experiments describing resistance with new synthetic antibiotics. Here, the argument for prior exposure and selection pressure leading to acquired resistance fails. Trimethoprim, for example, is a fully synthetic DHF reductase inhibitor first available for use in 1968. Trimethoprim resistance, however, was surprisingly discovered in Klebsiellae samples that were freeze-dried in 1964.¹⁶ This pre existing immunity to an unencountered chemotherapeutic agent debunks the popular misconception that resistance requires some kind of individual or group exposure. It also underscores the intrinsic difficulty of the battle against resistance: that resistance can exist even before the antibiotic itself is introduced.

Exposure — particularly sub-therapeutic and improperly administered — undeniably leads to increased prevalence of resistant organisms. The indiscriminate use of antibiotics for non-bacterial infections, through inappropriate treatment regimes by untrained consumers, has certainly created a selection environment that favors the development of widespread acquired resistance. The problem is especially acute in worldwide tuberculosis management, where multidrug resistance (MDR-TB) is complicated by a long, costly therapeutic course with poor compliance. In other infections, like Salmonella and Camplyobacter, antibiotics in the food supply provide a near chronic sub-therapeutic exposure for livestock that has contributed to the rise of resistant human infections.¹⁷ There is undeniably a relationship between the development of resistance and antibiotic exposure, but one is not required for the other.

Resistance, Susceptibility and Outcomes

The outcome of any infection is the result of myriad complicated interactions between host, pathogen and therapy. Common reasons for treatment failure include poor host response, inadequate drug dose, inadequate or improper drug administration, inadequate contact between drug and pathogen, sequestration of infectious agent in cysts or cavities, misidentification of pathogen or misdiagnosis of infection.¹⁸

In some infections (MDR-TB is probably the clearest example), it is unquestionably drug resistance that leads to increased morbidity and mortality. It is incorrect, however, to assume that bacterial resistance is always the cause of treat ment failures. In the exhaustive literature of drug-resistant Strep pneumoniae respiratory infections, there is very little clarity on the relationship between susceptibility of isolated organism and ultimate treatment outcome.¹⁹ There are convincing reports that resistant Strep pneumo causes a less virulent clinical presentation,²⁰ multiple reports that outcome is unrelated to organism susceptibility²¹ and conflicting reports that resistance is a marker for poor outcome.²² Mutations that confer resistance to one drug may also cause a less severe clinical presentation, while resistance to another drug may provoke a variant immune response. The Strep pneumo story highlights a very critical failure of resistance to accurately or completely predict outcome in respiratory infections. It is a lesson we should remember when we think about resistance in ophthalmology.

Bacterial "resistance" and "susceptibility" are properties described for clinical isolates grown in vitro. The distinctions are derived from growth characteristics of isolates exposed to systematic dilutions of test antibiotics. Standards have been established for reporting of susceptibilities by the National Committee for Clinical Laboratory Standards, based on antibiotic serum concentrations that are safely achievable for a given drug. They are not designed to be the only determinant in clinical decision making. They have less relevance in treatment decisions made where high local concentrations are achievable through direct administration of drug to infection site. Any topical application to the skin, external auditory canal, nasal mucosa, oropharynx or ocular surface is expected to achieve relatively higher concentrations than would normally be seen at the serum breakpoint. The resistant Strep pneumo literature offers a guide to the complex relationship between susceptibility and outcome. The meaning of resistance in ophthalmology is even further complicated by the imperfect use of serum standards for a therapy that is (nearly) always dosed topically.

Outcomes in Ophthalmology

Many factors affect the incidence and etiology of ocular infections. There is a balance on the ocular surface between com mensal organisms, host immune response and the multicomponent barrier formed by the tear film. We know that disruptions to this balance can rapidly lead to pathogenic infection, most frequently derived from host colonizers. We are similarly aware that in the last decade, organisms isolated from these infections are increasingly resistant to many antibiotics, including the 8-methoxy fluoroquinolones.²³

There is a wealth of reports in the ophthalmic literature that attempt to categorize resistance trends over time. Reports have illuminated the chronology of increasingly resistant MRSA endophthalmitis^24,25 and have described similar trends in a variety of other superficial infectious settings.^26,27 We have yet, however, to define the relationship between this increasing resistance and the clinical outcome of the corresponding ocular infections.

The anti-infective workhorse of ophthalmology for the past several years has undoubtedly been the fluoroquinolone. Advances in fluoroquinolone chemistry have been paralleled by better understanding of systemic pharmacokinetic and pharmacodynamic concepts.

Currently, ophthalmologists use concepts such as Cmax (peak concentration at the site of action), AUC (area under the drug concentration curve) and MIC (minimum inhibitory concentration) to describe behaviors of antiinfective agents and quantify resistance. Optimal ratios that predict success in equations involving these variables are currently derived from the aforementioned systemic parameters, which fail to accurately describe ocular pharma cokinetics. Moreover, demonstrating clinical efficacy for competing fluoroquinolones in varied ocular infectious settings remains challenging.

Work at the Charles Campbell Laboratory in Pittsburgh has demonstrated that serum MIC standards as described by the NCLSI guidelines for antibiotic sensitivities may not be ideally suited for analysis of ocular isolates.²⁸ Though new ocular standards may not be on the horizon, consideration should be made when reviewing lab MIC data.

For instance, it may be possible to achieve local antibiotic concentrations 1700 times higher than serum levels by direct topical application to the ocular surface. Conversely, it may be impossible to achieve breakpoint concentrations in the posterior vitreous or even in the anterior chamber, which may further complicate interpretation of in vitro data. In a recent study with experimental keratitis in rabbits, high local dose of levofloxacin was able to overcome in vitro resistance in a levo-resistant infection.²⁹ Similar results were seen in levo-resistant Pseudomonas infections in the same model. This highlights the difficulty of predicting in vivo outcome from in vitro resistance data.

Some studies have tried to correlate outcome with in vitro susceptibility in ophthalmic infections.^30-32 Comparisons between studies and conclusions from all reported data are difficult due to the heterogeneity of the investigations, limitations on collected data and the inability to control for confounding variables in incompletely collected data sets. We can learn from the Strep pneumo experience in respiratory infections that precise definitions of the study problem are as important as robust data collection that can later be controlled for a host of clinical parameters.

For now, in ophthalmology, clinical response is not predictable from in vitro data and even the suggestion of a strong correlation remains elusive. A systematic examination of clinical course, treatment, microbiology and outcome in a narrowly well-defined disease state is needed to more thoroughly analyze and control for all the variables that affect our ability to treat these infections. A multivariate analysis of a welldefined, extensively collected and rigorously reviewed database may be a useful resource for unraveling the web of factors that currently obscure the relationship between resistance, exposure, therapy and outcome in ocular infections. OM

References

1. Livermore DM. Bacterial resistance: origins, epidemiology, and impact. Clin Infect Dis 2003;36(Suppl 1):S11–23.
2. Kumarasamy K, Toleman M, Walsh T, et al. Emergence of a new antibiotic resistance mechanism in India, Pakistan, and the UK: a molecular, biological, and epidemiological study Lancet Infect Dis 2010; published online Aug11.
3. Kowalski RP, Karenchak LM, Romanowski EG. Infectious Disease: Changing antibiotic susceptibility. In: Foulks GN Gardler RG, ed. Ophthalmology Clinics of North America, Cornea and External Disease. Philadelphia:WB Saunders Company; 2003:1-9.
4. Blondeau JM. Fluoroquinolones: mechanism of action, classification, and development of resistance. Surv Ophthalmol. 2004;49(Suppl 2):S73–S78.
5. Bertino J. Impact of antibiotic resistance in the management of ocular infections: the role of current and future antibiotics. Clin Ophthalmol 2009;3:507–521.
6. Kowalski RP, Dhaliwal DK, Karenchak LM, et al. Gatifloxacin and moxifloxacin: an in vitro susceptibility comparison to levofloxacin, ciprofloxacin, and ofloxacin using bacterial keratitis isolates. Am J Ophthalmol, 2003;136:500-5.
7. Sörgel F: World Conference on Dosing of Anti-infectives: ‘Dosing the Magic Bullets’. Celebrating the 150th Birthday of Paul Ehrlich, the ‘Founder of Chemotherapy’, September 9–11, 2004, Nürnberg, Germany. Chemotherapy, 2004;50:1–5.
8. Lambert P.A. Bacterial resistance to antibiotics: modified target sites. Adv Drug Deliv Rev 2005; 57, 1471-1485.
9. Wise R. A review of the mechanisms of action and resistance of antimicrobial agents. Can Respir J, 1999; 6 Suppl A, 20A-22A.
10. Centers for Disease Control and Prevention (CDC) (2004). "Acinetobacter baumannii infections among patients at military medical facilities treating injured U.S. service members, 2002-2004". MMWR Morb. Mortal. Wkly. Rep. 53 (45): 1063–6.
11. Cornelis P (editor). (2008). Pseudomonas: Genomics and Molecular Biology (1st ed.). Caister Academic Press.
12. Cohen SP, McMurry LM, Hooper DC, Wolfson JS, Levy SB. Cross-resistance to fluoroquinolones in multiple-antibiotic resistant (Mar) Escherichia coli selected by tetracycline or chloramphenieol: decreased drug accumulation associated with membrane changes in addition to OmpF reduction. Antimicrob Agents Chemother, 1989; 33:1318-1325.
13. Liñares J, Pallarés R, Alonso T, et al. Trends in antibiotic resistance of clinical isolates of Streptococcus pneumoniae in Bellvitge Hospital, Barcelona, Spain. Clin Infect Dis 1992;15:99–105.
14. Wainright M, Swan HT. The history of the therapeutic use of crude penicillin. Med Hist, 1987;31:41-50.
15. Basset EJ, Keith MT, Armelagos SJ, Martin DL. Tetracycline labeled human bone from Sudanese Nubia. Science, 1980; 209:1532-1534.
16. Hamilton-Miller JMT, Grey D (1975) Resistance to trimethoprim in Klebsiellae isolated before its introduction. J Antimicrob Chemother 1:213-218.
17. Division of Emerging and Other Communicable Disease Surveillance and Control. The medical impact of the use of antimicrobials in food animals: report and proceedings of a WHO meeting: Berlin, Germany, 13-17 October 1997. Geneva: World Health Organization, 1997. (Document no. WHO/EMC/ZOO/97.4.)
18. Brumfitt W, Hamilton-Miller JMT (1987) Principles and practice of antimicrobial chemotherapy. In: Speight TM (ed) Avery's drug treatment, 3rd edn. ADIS Press, Auckland, pp 1207-1235.
19. Garau J. Treatment of drug-resistant pneumococcal pneumonia. Lancet Inf Dis, 2002;2:404-415.
20. Sanchez C, Armengol R, Lite J, Mir I, Garau J. Penicillin-resistant pneumococci and community acquired pneumonia. Lancet 1992;339:988.
21. Moroney JF, Fiore AE, Harrison LH, et al. Clinical outcomes of bacteraemic pneumococcal pneumonia in the era of antibiotic resistance. Clin Infect Dis 2001;33:797–805.
22. Kelley MA, Weber DJ, Gilligan P, Cohen MS. Breakthrough pneumococcal bacteremia in patients being treated with azithromycin and clarithromycin. Clin Infect Dis 2000;31:1008–1011.
23. Hwang DG. Fluoroquinolone Resistance in Ophthalmology and the Potential Rolefor Newer Ophthalmic Fluoroquinolones. Surv Ophthalmol, 2004; SUP. 2.
24. Deramo VA, Lai JC, Winokur J, et al. Visual outcome and bacterial sensitivity after methicillin-resistant Staphylococcus aureus related acute endophthalmitis. Am J Ophthalmol. 2008;145:413-7.
25. Major JC, Engelbert M, Flynn HW Jr, et al. Staphylococus aureus endophthalmitis; antibiotic susceptibilities, methicillin resistance, and clinical outcomes. Am J Ophthalmol. 2010;149:278-83.
26. Woodward M, Randleman JB. Bilateral Methicillin-resistant Staphlococcus aureus keratitis after photorefractive keratectomy. J Cat Refract Surg. 2007;33:316-9.
27. Moshirfar M, Mirzaian Z, Feiz V, et al. Fourth generation fluoroquinolone-resistant bacterial keratitis after refractive surgery. J Cat Refract Surg. 2006;32:515-8.
28. Kowalski RP, Yates KA, Romanowski EG, Karenchak, L, Mah F, Gordon YJ. An Ophthalmologist's guide to understanding antibiotic susceptibility and minimum inhibitory concentration (MIC) data. Ophthalmol 2005:112:1987–1991.
29. Kowalski RP, Romanowski EG, Mah F, Shanks RM, Gordon YJ. Topical levofloxacin 1.5% overcomes in vitro resistance in rabbit keratitis models. Acta Ophthalmol. 2010:88:e120–e125.
30. Kim RY, Cooper KL, Kelly LD. Predictive factors for response to medical therapy in bacterial ulcerative keratitis. Graefes Arch Clin Exp Ophthalmol. 1996;234:731-738.
31. Coster DJ, Badenoch PR. Host, microbial, and pharmacological factors affecting the outcome of suppurative keratitis. Br J Ophthalmol. 1987;71:96-101.
32. Wilhelmus KR, Abshire RL, Schlech BA. Influence of fluoroquinolone susceptibility on the therapeutic response of fluoroquinolone-treatedbacterial keratitis. Arch Ophthalmol. 2003.

	Jesse Pelletier, MD is a founding partner and director of cornea, cataract and refractive surgery at the Ocean Ophthalmology Group in Miami, voluntary assistant professor of ophthalmology at Bascom Palmer Eye Institute, and an attending ophthalmologist at the Miami VA.
	Joseph Capriotti, MD is an ophthalmologist and research scientist with an interest in ocular microbiology, infection and inflammation. He is associate research director of the Ocean Ophthalmology Group and an adjunct scientist in the department of chemistry at Columbia University.