Dear Editor

We thank Dr Doft, et al. for their useful and expert opinion. The choice of which agent to use to empirically treat gram-negative organisms implicated in endophthalmitis remains controversial. As amikacin has been proven to cause macular infarction, we feel one should look at viable alternatives. Ceftazidime is already in widespread use in the UK and appears not only to have an excellent safety profile but also good clinical effect. Unfortunately until we have proper in vivo and in vitro “head-to-head” comparison studies, it is difficult to know which is the more efficacious agent. As far as synergism is concerned, vancomycin and ceftazidime are usually not tested together because vancomycin acts on gram-positive organisms and ceftazidime is used primarily for gram-negative infections. However, there is however one study that reported synergy between vancomycin and ceftazidime against gram-positive organisms.[1]

The study by Kwok and colleagues raises a concern that ceftazidime precipitation, as assessed by in-vitro studies, may affect its action in vivo.[2] The authors of our study have noticed temporary precipitants in vivo without apparent alteration of clinical effect. (AR) Previous animal models do show that ceftazidime reaches intravitreal minimal inhibitory concentrations for common gram-negative microbes after a single intravitreal injection.3 Perhaps assay at the time of repeat injection, non-invasive confocal Raman spectroscopy of the anterior chamber, or further animal models may provide additional insight into ceftazidime pharmacokinetics and the phenomenon of ceftazidime precipitation so as to guide future therapeutic choice. Ultimately the decision lies with the treating surgeon who should be aware of both the efficacy and safety profiles of the agents available. We still believe, with the evidence presented in our article, that ceftazidime currently represents the best agent for the treatment of gram-negative microbes in endophthalmitis.

References

1. Simon C, Littschwager G. In vitro activity of ceftazidime in combination with other antibiotics. Infection. 1985 Jul-Aug;13(4):184-9.

2. Kwok AK, Hui M, Pang CP, et al. An in vitro study of ceftazidime and vancomycin concentrations in various fluid media: implications for use in treating endophthalmitis. Invest Ophthalmol Vis Sci. 2002 Apr;43(4):1182-8.

3. Mochizuki K, Yamashita Y, Torisaki M, et al. Intraocular kinetics of ceftazidime (Modacin). Ophthalmic Res. 1992;24(3):150-4.

Dear Editor

We write in reference to the letter by Galloway et al. "Macular infarction after intravitreal amikacin: Mounting evidence against amikacin".[1]

The authors report a single case of macular infarction in a patient who had been given intravitreal amikacin for endophthalmitis. They cite that single case plus some prior literature as reason to support a change in the choice of antibiotic for intravitreal injection from the treatment guidelines based on the results of the Endophthalmitis Vitrectomy Study (EVS).

While aminoglycoside induced retinal toxicity certainly can occur, we disagree with their statement that there is good evidence that aminoglycosides should not be primary drugs of choice in this disease. There are several theoretical and practical advantages of aminoglycosides over ceftazidime. Amikacin provides concentration dependent killing (so that the higher concentration of drug the more rapid the kill) which is not true for ceftazidime.

This is an important issue since high concentrations of drug are administered by intravitreal injection, thus possibly allowing for more rapid kill with amikacin. Amikacin is considered to be synergistic with vancomycin for certain gram positive species, so its use provides benefit against gram positive organisms, not just for gram negatives. Gram positive organisms make up the overwhelming majority of cases of endophthalmitis. In addition, there has been a recent report that ceftazidime may precipitate in the vitreous at normal body temperature,[2] possibly making it less available than one might wish in the vitreous cavity.

Finally, and very important, is the fact that amikacin has been found to be effective in a clinical trial but there is no such evidence yet available on ceftazidime. The only apparent advantage to ceftazidime is that it may be a somewhat safer drug in the sense that macular toxicity has not been reported. Even so, the incidence of macular toxicity is extremely rare (only 1 in 420 eyes in the EVS suffered macular toxicity possibly from the drug). In a very severe disease such as endophthalmitis a risk this low is worth tolerating when there may be substantial potential advantages.

References

(1) Galloway G, Ramsay A, Jordan K, et al, Macular infarction after intravitreal Amikacin: mounting evidence against Amikacin. Br J Ophthalmol 2002;86:359-360.

(2) Kwok, AK., M. Hui, et al. An in vitro study of ceft)azidime and vancomycin concentrations in various fluid media: implications for use in treating endophthalmitis." Invest Ophthalmol Vis Sci 2002,43(4): 1182-8.