Dr. John,
I read with some interest your article on the efficacy of old antibiotics compared with the new fourth-generation fluoroquinolones. I would like to offer several points to clarify your article's broad and surprising conclusions.
In your introduction, you allude to the lack of literature comparing old antibiotics, including tobramycin or ciprofloxacin, to the new antibiotics gatifloxacin and moxifloxacin. My laboratory and others have published studies comparing gatifloxacin and moxifloxacin to various antibiotics, including vancomycin, cefazolin, tobramycin, ofloxacin, ciprofloxacin, and levofloxacin in vitro, as in your study, but also in vivo.1-17
My first criticism is that your study does not mention or compare the agents to atypical mycobacterial species. It is well documented that up to half of all post-LASIK cases of infectious keratitis involve these nontuberculous species.17-19 When comparing these agents in an in vitro study, such as that of Shah et al,1 gatifloxacin and moxifloxacin were found to be superior to agents such as amikacin and ciprofloxacin. Vancomycin does not have significant activity against atypical mycobacteria. The newer fluoroquinolones, gatifloxacin and moxifloxacin, not only have better activity against atypical pathogens than all the agents tested, but, with these single commercially available agents (Zymar [gatifloxacin 0.3%]; Allergan, Inc., Irvine, CA, and Vigamox [moxifloxacin 0.5%]; Alcon Laboratories, Inc., Fort Worth, TX) I can cover gram-positive and gram-negative bacteria as well as atypical mycobacteria, and thereby treat more than 95% of the causes of bacterial conjunctivitis and keratitis.20 If I were being treated for conjunctivitis or keratitis, I would want a commercially available medication that could cover all pathogens, instead of having to go to a specialty pharmacy every 4 days to fill prescriptions for vancomycin to treat gram-positive pathogens and gentamicin to treat the gram-negative pathogens, and still having to worry about atypical pathogens.
Another major defect of your study is its purpose. Its underlying premise is that patients see many physicians for a red eye secondary to an ocular infection, including ER physicians, family practitioners, internists, or pediatricians, and that the efficacy of topically applied antibiotics has not been compared because a patient may be taking many different antibiotics, including those tested. I seriously doubt that any general practitioner, including a pediatrician, would even occasionally use vancomycin for the management of a red eye secondary to an ocular infection. I also am confident that an ophthalmologist, let alone a generalist, would probably not use topical ceftriaxone or trovafloxacin even on an occasional basis for the management of conjunctivitis or keratitis. If the premise of your study is testing those agents that are commonly used for the management of a red eye secondary to an ocular infection by the frontline physicians, this list of antimicrobial agents is flawed due to the inclusion of antibiotics that even super cornea and external disease specialists do not use in the management of conjunctivitis and keratitis.
Dr. John, you also reported that vancomycin and gentamicin show effective activity against ocular gram-positive and gram-negative bacterial isolates, respectively. Our group, The Charles T. Campbell laboratory at the University of Pittsburgh, concurs and has documented its conclusions in the peer-reviewed literature.14,15 However, due to topical toxicity as well as pharmacodynamic and pharmacokinetic properties, the results of our in vitro laboratory study should be corroborated with in vivo and clinical experience. You reported that vancomycin and gentamicin have limited systemic utilization due to systemic toxicity, and you are partially correct. All hospitals, including tertiary care centers, limit the use of systemic vancomycin to methacillin-resistant, gram-positive, cocci culture-proven or suspected infections. Vancomycin, when administered properly, has amazingly few side effects, but its use is reserved because of the concern of resistance. Both vancomycin and gentamicin cause toxicity when administered topically. Vancomycin has a pH value of 4. In our rabbit keratitis models, the clinical appearance of the rabbits' eyes is indistinguishable from the untreated control keratitis eyes. Obviously, in vitro susceptibilities are one piece of the puzzle when choosing an agent to treat an ocular infection, just like toxicity is a concern. I am confident that you would agree.
Not only do toxicity and in vitro activity play a role in the treatment of ocular infections, but pharmacokinetics, or the ability of a drug to penetrate into the ocular tissues, is also crucial to the resolution of conjunctivitis and keratitis. Unlike vancomycin and gentamicin, which have been documented to have poor penetration into ocular tissues including the cornea, no other class of topically used antibiotic can compare with fluoroquinolones in terms of concentrations in ocular tissue after topical application.14 Among the fluoroquinolones, gatifloxacin and moxifloxacin have been shown to be at least as good if not superior to the older generation of fluoroquinolones.15
Pharmacodynamics involves an agent's ability to inhibit the growth of bacteria. Fluoroquinolones are rapidly bacteriocidal, meaning they are able to eradicate bacteria in a short period of time (usually less than 2 hours). They are also concentration-dependent killing agents, meaning the more fluoroquinolone present, the more efficacious it is. Vancomycin is a time-dependent killing agent, meaning that no matter how much drug is present, it will still take about 12 hours for it to kill the gram-positive bacteria. Aminoglycosides, like gentamicin, are more time- than concentration-dependent and require approximately 8 hours to kill gram-negative bacteria. If I were being treated for a central corneal ulcer, I would want something that works in 2 rather than 8 to 12 hours.
In conclusion, I agree with you that the fourth-generation fluoroquinolones are easy to use, and in the vast majority of cases, they will be more than enough for the management of conjunctivitis and keratitis. I also agree that vancomycin and gentamicin provide excellent coverage, but care must be taken when offering a blanket statement such as “old antibiotics covering both gram-positive and gram-negative organisms can be safely used for the effective management of ocular infections.” Dr. John, your study did not address the management of ocular infections using these or any other topical antibiotics; it looked at in vitro susceptibilities, not efficacy in the in vivo or clinical settings, which involves other properties such as spectrum of activity, pharmacokinetics, pharmacodynamics, toxicity, and ease of access.
Francis S. Mah, MDPittsburgh
1. Shah MK, Ritterband DC, Terraciano AJ, et al. Will the topical fourth generation fluoroquinolones become the antibiotics of choice for treating atypical mycobacteria related eye disease? Poster presented at: The ARVO Annual Meeting; May 5, 2003; Fort Lauderdale, FL.
2. Long M, Jensen HG, Allergan Gatifloxacin Study Group. Ocular bacteria from conjunctivitis patients: susceptibility to gatifloxacin and older fluoroquinolones. Paper presented at: The ARVO Annual Meeting; May 6, 2003; Fort Lauderdale, FL.
3. Silver LH, Burkey R, Montegomery D, et al. Safety of ophthalmic moxifloxacin in the treatment of newborns, infants and toddlers, children, and adolescents with bacterial conjunctivitis. Poster presented at: The ARVO Annual Meeting; May 4, 2003; Fort Lauderdale, FL.
4. Mather R, Karenchak LM, Romanowski EG, et al. Fourth generation fluoroquinolones: new weapons in the arsenal of ophthalmic antibiotics. Am J Ophthalmol. 2002;133:463.
5. 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;136:500-505.
6. Yates KA, Kowalski, PR, Romanowski EG, et al. The in vitro evaluation of the ophthalmic fluoroquinolones against bacterial conjunctivitis isolates. Paper presented at: The ARVO Annual Meeting; May 6, 2003; Fort Lauderdale, FL.
7. Shamie N, Sarayba M, Reiser B, et al. Fluoroquinolone therapy in an animal model of Mycobacterium chelonae keratitis after lamellar keratectomy. Paper presented at: The ARVO Annual Meeting; May 8, 2003; Fort Lauderdale, FL.
8. Donnenfeld ED, Solomon R, Doshi S, et al. Prophylaxis of Streptococcus pneumoniae keratitis with gatifloxacin in a rabbit LASIK model. Poster presented at: The ARVO Annual Meeting; May 8, 2003; Fort Lauderdale, FL.
9. Tungsiripat T, Sarayba MA, Kaufman MB, et al. Fluoroquinolone therapy in multiple-drug resistant staphylococcal keratitis after lamellar keratectomy in a rabbit model. Am J Ophthalmol. 2003;136:76-81.
10. Mah FS, Romanowski EG, Yates KA, et al. The successful treatment of gatifloxaxin-resistant Staphylococcus aureus keratitis with gatifloxacin (Zymar) in a NZW rabbit model. Paper presented at: The 2003 OMIG Meeting; November 15, 2003; Anaheim, CA.
11. Aliprandis ET, Ciralsky J, Lai H, et al. Comparison of topical moxifloxacin 0.5% to ciprofloxacin 0.3% in the treatment of experimental Pseudomonas aeruginosa keratitis in rabbits. Poster presented at: The ARVO Annual Meeting; May 8, 2003; Fort Lauderdale, FL.
12. Thibodeaux BA, Dajcs JJ, Caballero AR, et al. The effectiveness of moxifloxacin for gram-negative bacterial keratitis. Poster presented at: The ARVO Annual Meeting; May 5, 2003; Fort Lauderdale, FL.
13. Terai K, Joo M-J, Hyon J-Y, et al. Comparative efficacy of topical moxifloxacin, an expanded spectrum fluoroquinolone, versus topical ofloxacin, penicillin G and tobramycin in the treatment of experimental S. pneumoniae and P. aeruginosa keratitis in rabbits. Paper presented at: The ARVO Annual Meeting; May 6, 2003; Fort Lauderdale, FL.
14. Rhee MK, Kowalski RP, Romanowski EG, et al. A laboratory evaluation of antibiotic therapy for ciprofloxacin-resistant Pseudomonas aeruginosa. Am J Ophthalmol. 2004;138:226-230.
15. Mah FS. New ocular antibiotics. Ophthalmol Clin North Am. 2003;16:11-27.
16. Mah FS. Fourth-generation fluoroquinolones. Curr Opin Ophthamlol. 2004;15:316-320.
17. Abshire R, Cockrum P, Crider J, Schlech B. Topical antibacterial therapy for mycobacterial keratitis: potential for surgical prophylaxis and treatment. Clin Ther. 2004;26:191-196.
18. Chang MA, Jain S, Azar DT. Infections following laser in situ keratomileusis: an integration of the published literature. Surv Ophthalmol. 2004;49:269-280.
19. Solomon R. Donnenfeld ED, Azar DT, et al. Infectious keratitis after laser in situ keratomileusis: results of an ASCRS survey. J Cataract Refract Surg. 2003;29:2001-2006.
20. Kowalski RP, Karenchak LM, Romanowski EG. Infectious disease: changing antibiotic susceptibility. Ophthalmol Clin North Am. 2003;16:1-9.
Dr. Mah,
The majority (58%) of the references that you quote in your letter to the editor are abstracts, and you refer to them as “published abstracts.” All abstracts are usually published in a program/abstract book for annual meetings (eg, by the AAO, ASCRS, and ARVO). However, abstracts are very brief. They are not peer-reviewed publications in ophthalmic journals, nor do they carry the same merit.
Additionally, it is my understanding that most drug-related studies at your laboratory, the Charles T. Campbell Ophthalmic Laboratory, are supported by ophthalmic drug companies. Furthermore, according to a personal communication I received from a staff member at your laboratory in 2003, microorganisms are not shared usually with other centers. Although corporate-supported studies are usually good, they are primarily focused on drugs from the sponsoring company or on their competitors' products. This limits the number of ophthalmic drugs studied at any given time (Table 1). Dr. Mah, you state that there is a major flaw in the purpose of our study. Is the flaw that the study is not supported by a corporation? The purpose of our study was to be an noncorporate-supported investigation not supported by industry and to include a large number of drugs (n=12) in the study.
I would like to comment further on specific issues you raise. You state that your laboratory has published studies comparable to our study, which is false (Tables 1 and 2). None of the studies you quote in your letter to the editor is similar to our study, which included 12 antibiotics and nine microorganisms. All of the studies to which you refer have a lower total number of antibiotics studied. One study1 that used 11 antibiotics included only one type of microorganism, namely, ciprofloxacin-resistant Pseudomonas aeruginosa. More than 75% of the articles that you list as references are corporate-supported studies, and fewer than 20% were not. Furthermore, none of the studies to which you refer includes the use of vancomycin, and only one included the use of gentamicin.
In response to your criticism that our study does not mention or compare agents to atypical mycobacterial species, this group of organisms has recently gained importance with regard to post-LASIK corneal infections. The purpose of our study was not to look at the efficacy of antibiotics against atypical mycobacteria. Additionally, there are other medications that may be preferable to or in combination with the fourth-generation fluoroquinolones.2 Only two of your referenced studies3,4 deal with atypical mycobacteria, and neither includes other gram-positive or gram-negative bacteria. None of the studies that deals with microorganisms included mycobacteria. Clearly, your criticism is invalid.
It may be fallacious to assume that one antibiotic will effectively treat atypical mycobacterial keratitis; multidrug therapy is typically required for the effective treatment of the aforementioned infection.2
Dr. Mah, it appears that you are unaware that currently there are pharmacies in the US that will make vancomycin eye drops and other such medications that are not commercially available and have them shipped for next-day delivery at the patient's residence. Also, the shelf life is not 4 days! The shelf life of vancomycin and gentamicin eye drops is 14 and 30 days, respectively. Vancomycin eye drops are commonly prepared by hospital pharmacies and are often used by nonophthalmic physicians for methicillin-resistant Staphylococcus aureus conjunctivitis.
Dr. Mah, you state that tobramycin and ciprofloxacin are old antibiotics. These drugs received FDA approval between 1990 and 1998 (Table 2). The truly old antibiotics include vancomycin and erythromycin, which received initial approval in 1964 (Table 2).
We have received calls from ophthalmologists in the US thanking us for our study because it is the only one that looked at such a large range of antibiotics and organisms.
In my article, I never implicated the pediatrician as using vancomycin for the management of red eyes but stated that a wide range of antibiotics may be used for ocular infections depending on who the initial treating physician is. Obviously, no nonophthalmic physician in the US will treat a serious ocular infection such as keratitis or corneal ulceration.
Dr. Mah, you state in your letter that your group has published data on vancomycin as being effective, but none of the studies you have listed appears to have used vancomycin. Only one study1 used gentamicin against ciprofloxacin-resistant P. aeruginosa. You also failed to document the efficacy of gentamicin against gram-positive and gram-negative organisms, as we did in our study.
Our in vitro study does not superimpose in vivo situations but only offers discussion within the limits of the in vitro study. Factors regarding pharmacokinetics, penetration, bioavailability, minimum inhibitory concentrations, toxicity, etc., are not new concepts; they are well established and well known.
With regard to treating a corneal ulcer, getting a favorable effect is multifactorial. The type of organism and the type of antibiotic used are important factors, in addition to the host's status. Obviously, arguments about epithelial barrier and drug penetration may be negated when one has a frank corneal ulcer with a large epithelial defect.
With reference to the type of bacteria, ophthalmologists know that gram-positive bacteria progress much less rapidly than gram-negative bacteria, such as P. aeruginosa, which can cause rapid melting and corneal perforation. Hence, the time of kill is much more significant when dealing with gram-negative versus gram-positive bacteria. Vancomycin is primarily active against gram-positive organisms, and gentamicin is primarily active against gram-negative bacteria.
Vancomycin is time-dependent in its killing of gram-positive bacteria, but 12 hours to make a kill may not be catastrophic when dealing with gram-positive bacterial keratitis. To be bacteriocidal, there has to be a 3-log reduction of the organisms. In addition to the infection, the amount of organisms will also play a role. Additionally, with a corneal infection, we initially use drugs that will cover a wide range of organisms. When vancomycin is used, it is in combination with other antibiotics.
Furthermore, you are incorrect in your reference to aminoglycosides' being more time-dependent in terms of killing bacteria. Aminoglycosides are concentration-dependent antibiotics, and, consequently, they do have the rapid kill potential of microorganisms. The basis of once daily dosing for aminoglycosides for systemic treatment is based on the drug's being a concentration-dependent antibiotic.8
Vancomycin provides effective coverage of the usually slower progressing gram-positive organisms with a time-dependent antibiotic. Gentamicin is a concentration-dependent aminoglycoside for the more rapidly progressing gram-negative organisms. A combination of vancomycin and gentamicin covers both gram-positive and gram-negative organisms. Both aminoglycosides and fluoroquinolones are concentration-dependent antibiotics.
My “blanket statement” has been misinterpreted. To clarify, according to our study's results, old antibiotics (vancomycin and gentamycin) are more effective against gram-positive and gram-negative organisms isolated from eye infections when compared with the new antibiotics, including fourth-generation fluoroquinolones. With that said, properly chosen old antibiotics covering both gram-positive and gram-negative organisms can be safely used for the effective management of ocular infections. And, as I stated in my article, clinicians should rely on the patient's medical response to treatment and change antibiotics as needed based on clinical findings.
I appreciate the initiation of such an interesting discussion regarding these ocular antibiotics.Thomas John, MD
Chicago n
1. Rhee MK, Kowalski RP, Romanowski EG, et al. A laboratory evaluation of antibiotic therapy for ciprofloxacin-resistant Pseudomonas aeruginosa. Am J Ophthalmol. 2004;138:226-230.
2. John T, Velotta E. Nontuberculos (atypical) mycobacterial keratitis following LASIK: Current status and clinical implications. Cornea. In press.
3. Mather R, Karenchak LM, Romanowski EG, Kowalski RP. Fourth generation fluoroquinolones: new weapons in the arsenal of ophthalmic antibiotics. Am J Ophthalmol. 2002;133:463-466.
4. Donnenfeld ED, Solomon R, Doshi S, et al. Prophylaxis of Streptococcus pneumoniae keratitis with Gatifloxacin in a rabbit LASIK model. Poster presented at: The ARVO Annual meeting; May 8, 2003; Fort Lauderdale, FL.
5. 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;136:500-505.
6. Yates KA, Kowalski PR, Romanowski EG, et al. The in vitro evaluation of the ophthalmic fluoroquinolones against bacterial conjunctivitis isolates. Paper presented at: The ARVO Annual Meeting; May 6, 2003; Fort Lauderdale, FL.
7. Kowalski, PR, Romanowski EG, Mah FS, et al. The prevention of bacterial endophthalmitis by topical moxifloxacin in a rabbit prophylaxis model. Poster presented at: The ARVO Annual Meeting; May 5, 2003; Fort Lauderdale, FL.
8. McLean AJ, Ioannides-Demos LL, Li SC, et al. Bactericidal effect of gentamicin peak concentration provides a rationale for administration of bolus doses. J Antimicrob Chemother. 1993;32:301-305.
