Aminoglycoside drugs in clinical practice: an evidence-based approach--authors' response

Abstract

Sir, We thank Keuleyan and Kirilov for their comments and beg to differ with most of them. The question of the advantages and disadvantages of aminoglycosides compared with other drugs is discussed in our review. Based on observational studies, we believe that treatment with an aminoglycoside is less effective than b-lactam treatment in septic patients with Gram-negative infections in sites other than those in the urinary tract. Randomized controlled trials on aminoglycosides as single treatment (for infections other than the urinary tract) are few and included a small number of septic patients. To the best of our knowledge, Acinetobacter spp. are not inherently resistant to aminoglycosides. Resistance is mediated through several mechanisms, mainly aminoglycoside-modifying enzymes, and its prevalence is dependent on local epidemiology and strain distribution. In our centre, aminoglycosides are the fourth most active antibiotic class against Acinetobacter baumannii, following colistin, carbapenems and ampicillin/ sulbactam. Other Acinetobacter spp. are susceptible to aminoglycosides. In the SENTRY survey, more than 50% of clinical Acinetobacter spp. strains resistant to ceftazidime were susceptible to amikacin and tobramycin, which were second only to imipenem. In the MYSTIC survey, 58% and 53% of nosocomial Acinetobacter spp. in Europe in 2007 were susceptible to gentamicin and tobramycin, respectively. The data shown for the ‘usual’ in vitro spectrum of activity of aminoglycosides against Acinetobacter spp. in textbooks should probably be adapted to current epidemiology to avoid confusion among clinicians and microbiologists. Our respect for guidelines is expressed in the quote from the review (‘Practitioners would be well advised to use published guidelines, but within their limits, to take into account the fact that combination treatment is not supported by evidence’.). The worrisome point about the management of staphylococcal and streptococcal endocarditis with aminoglycosides is not lack of evidence but weak evidence that points in the opposite direction to the guidelines. Our point is well emphasized by a recent report about the nephrotoxicity of aminoglycosides in Staphylococcus aureus endocarditis and by the accompanying editorial. We would certainly advise clinicians not to steer their practice by apprehension regarding future ‘medico-legal investigation’. We think that Oxman et al. offer good advice: ‘. . . clinicians need guidelines and use them all the time, but they should not accept recommendations uncritically. To serve their patients well, they must be able to make informed judgments about which guidelines are appropriate’. Keuleyan and Kirilov ask ‘. . . why the superiority of antibiotic combinations for infective endocarditis and sepsis, already proven in vitro and in animal experiments, has not been confirmed in the studies cited by the authors’ (i.e. in a synthesis of all randomized controlled trials addressing this question and performed on humans). This is indeed a major concern. Mechanisms or drugs that are allegedly efficacious in vitro and in animal models but do not work in humans – 14 should encourage us to examine whether the models that have been used were appropriate, and whether the results were interpreted correctly.