Fluoroquinolone resistance of Escherichia coli at a cancer center: epidemiologic evolution and effects of discontinuing prophylactic fluoroquinolone use in neutropenic patients with leukemia.

The aim of the present study was to investigate the epidemiologic evolution of fluoroquinolone resistance of E. coli clinical isolates from patients admitted to a hematology-oncology service where fluoroquinolone prophylaxis during neutropenia was recommended as the standard of care for many years but was then discontinued in a trial conducted in patients with acute leukemia. Fluoroquinolones had been shown to decrease the incidence of gram-negative bacteremia in cancer patients with neutropenia, yet it was thought that the emergence of resistance in Escherichia coli and other gram-negative bacteria may have caused a progressive lack of efficacy of fluoroquinolone prophylaxis. Epidemiologic surveillance of fluoroquinolone resistance of E. coli clinical isolates at our cancer center since 1992 showed a continuing influx of new clones not previously observed in the population of cancer patients, an increase in the number of cancer patients per year colonized and/or infected by fluoroquinolone-resistant E. coli (1992-1994, 10-16 patients; 1995-1997, 24-27 patients), and a resistance rate of >50% among E. coli bloodstream isolates of hematology-oncology patients. A 6-month fluoroquinolone prophylaxis discontinuation intervention trial in 1998 suggested that despite increasing resistance among E. coli isolates, fluoroquinolone prophylaxis in acute leukemia patients was still effective in the prevention of gram-negative bacteremia (incidence rates, 8% during the pre-intervention period vs. 20% after discontinuation; p<0.01). The resumption of fluoroquinolone prophylaxis in acute leukemia patients thereafter decreased the incidence of gram-negative bacteremia to the pre-intervention level (9%; p=0.03), while the proportion of in vitro fluoroquinolone resistance in E. coli bacteremia isolates again increased (from 15% during the intervention period to >50% in the post-intervention period). Relative rates of resistance thus were a poor indicator of the potential clinical benefits associated with fluoroquinolone prophylaxis in cancer patients.