Article révisé par les pairs
Résumé : Combinations of antibiotics (namely penicillins and aminoglycosides) have been advocated in the 1970s for the empirical therapy of FN in cancer patients in order to take advantage of the possible synergism between these agents and to extend the potential antimicrobial spectrum of empirical therapy.Later, with the development of potent broad spectrum antibiotics, the need for combinations became less obvious as monotherapy with these new agents appeared as effective and less toxic than previously used combinations.However, today we are facing a major challenge through the emergence of multi-resistant microrganisms. With such bacteria, we might be coming back to the pre-antibiotic era when no active agents were available. This situation is due, in part, by the excessive use of antibiotics, namely as a prophylaxis for infection, and is complicated by the fact that very few new effective antibiotics are being developed by the pharmaceutical industry.Under these circumstances, it is likely that we will have to resort to "old timers" such as the polymyxins. It is also possible that combination therapy will come back in favor to take advantage of the synergism and extend the spectrum of coverage, just as it has been the case for the management of resistant tuberculosis.At the same time, the development of multidisciplinary antimicrobial stewardship is mandatory for efficient infection control and minimizing emergence of antimicrobial resistance.