We read with interest the recent article by Bow and colleagues [1] on quinolone prophylaxis. The authors showed that the patients who received ofloxacin plus rifampin had fewer documented infections than patients who received other prophylactic regimens. However, these results were hampered by an increased rate of fevers of unknown origin. The authors also noted a high rate of yeast colonization (71%) in their patients. Only 6 of 31 patients were colonized by gram-negative rods; data on the susceptibility of these organisms are not available. The rate of febrile neutropenia and mortality were similar among the three groups.

The authors stated that reducing the rate of bacteriologically documented infections is beneficial, but large randomized trials have not shown this to be true in terms of global mortality or reduction in the use of different antibiotics or antifungal agents. We disagree with Bow and colleagues that many unexplained fevers are not bacterial infections but rather manifestations of absorption of toxins from the gut. Pizzo and colleagues [2] showed that stopping antibiotic therapy too soon, even in stable patients who have no documented infection but who are profoundly neutropenic, was associated with a high rate of documented infections and mortality. This finding suggests that many unexplained fevers are in fact deep infections.

We are also concerned about the high rate of yeast colonization, which has been shown to precede invasive infection [3]. The fact that few fungal infections were documented does not mean that these infections did not occur; it is well known that the rate of documentation of fungal infections is low.

Our multicenter (but not prospective or randomized) trial [4] showed that the rate of gram-negative infections was the same in different centers, regardless of the use of quinolones. More concerning was the finding that in some centers that used quinolones, the rate of quinolone resistance was high (14% to 18%). One center that used quinolones had a high incidence of infections with Pseudomonas organisms other than Pseudomonas aeruginosa, all of them fatal. On the basis of these results and those of many other studies, quinolone-based prophylaxis is no longer used in some centers that participated in our study.

Quinolones are very useful drugs for low-risk febrile patients with neutropenia, but we do not advocate the widespread use of quinolone prophylaxis.

Javier Altclas, MD

Instituto de Criopreservacion y Transplante de Medula Osea IMA; Buenos Aires, Argentina