Fluoroquinolone prophylaxis: a word of caution

Abstract

The recent audit of treatment outcomes of fluor-oquinolone (FQ) prophylaxis from a tertiary hospitalin Singapore [1] highlights important considerationsfor the use of FQ prophylaxis in patients withhematologic–oncologic disease. A hospital preva-lence of FQ resistance in Escherichia coli isolatedfrom internal medicine patients above 20% has beenreported to be associated with significant loss of FQprophylaxis efficacy in patients with hematologic–oncologic disease, such that the protective effect ofFQ prophylaxis on reducing febrile episodes anddocumented infections is severely blunted [2,3]. Thishas been well demonstrated in the study by Ng et al.,where the rate of resistance in E. coli was 34.4% [1].As the authors indicate, the role of FQ prophylaxisrequires further evaluation in areas of the worldwhere FQ resistance is prevalent amongst gram-negative isolates, such as parts of South and LatinAmerica, and South East Asia [4].Although not all countries have such rates of FQresistance, no appropriately designed time-seriesanalyses have effectively evaluated the relationshipbetween FQ prophylaxis and rates of infection byFQ-resistant bacteria in oncology patients. Attemptsto do so [5] have been limited by study design,inadequate duration of surveillance, and incompletereporting. There is concern that widespread use ofFQ prophylaxis may result in increased rates ofresistance, and this has recently been demonstratedin a single center in allogeneic transplant recipientsreceiving levofloxacin prophylaxis [6,7].FQ prophylaxis alters colonizing flora, particularlyof the gastrointestinal tract. This may occur byindividual exposure to FQs or by horizontaltransmission of biologically fit, FQ-resistant bacteria(such as E. coli, Pseudomonas aeruginosa, and coagu-lase-negative staphylococci) such that, over time, themajority of a hospital’s patient population maydevelop colonization with FQ-resistant organisms.Support for these concepts comes from ecologicalstudies that have shown selection of FQ-resistantstrains results from selective pressure of antibiotics,mainly (but not limited to) excessive use of fluor-oquinolones, and this may take many months (up toa year) to manifest within a hospital setting [8].Further, FQ prophylaxis has been consistentlyassociated with FQ resistance in (but not limitedto) gram-negative bacteria (GNB), particularly E. coli[3,9]. There also appears to be an association withFQ use and FQ resistance in viridans groupstreptococci, coagulase-negative staphylococci, Sta-phylococcus aureus, and outbreaks of Clostridiumdifficile, including the hypervirulent strain [10].Although FQ prophylaxis is advocated in someclinical practice guidelines [11], the evidence for amortality benefit is not conclusive. Gafter-Gvili et al.,in a large systematic review (SR) (which includedpseudo-randomized controlled trials), demonstratedreduced overall mortality [5]. However, many of theincluded studies were small, were of variable quality(i.e. not always blinded), and spanned over a longtime period (e.g. 1987–2005), during which practicesrelated to empiric antibiotic therapy are likely to havechanged. No mortality benefit could be shown in asub-analysis of good-quality randomized controlledtrials (RCTs). In contrast, a number of RCTs andSRs have consistently shown significant reductions infebrile neutropenia episodes and microbiologically