Antibacterial Prophylaxis for Patients with Cancer: Should We Use It or Not?
Abstracts & Commentary
By J. Peter Donnelly, PhD, Clinical Microbiologist, University Hospital, Nijmegen, The Netherlands, Section Editor, Microbiology, is Associate Editor for Infectious Disease Alert.
Dr. Donnelly is a consultant for Ortho Biotech, and does research for Janssen, Merck, Novartis, Numico, Pharmacia, and Pfizer.
Synopsis: A large, double-blinded, randomized, placebo-controlled trial done by the Italian GIMEMA group shows that prophylaxis with once daily levofloxacin significantly reduces febrile episodes and infectious complications but not mortality.
Sources: Bucaneve G, et al. Levofloxacin to Prevent Bacterial Infection in Patients with Cancer and Neutropenia. N Engl J Med. 2005;353:977-987; Cullen M, et al. Antibacterial Prophylaxis After Chemotherapy for Solid Tumors and Lymphomas. N Engl J Med. 2005;353:988-998; Baden L R. Prophylactic Antimicrobial Agents and the Importance of Fitness. N Engl J Med. 2005;353:1052-1054; Gafter-Gvili A, et al. Meta-Analysis: Antibiotic Prophylaxis Reduces Mortality in Neutropenic Patients. Ann Intern Med. 2005;142:979-995.
The Ginema study reported by Bucaneve and colleagues recruited for a formal, double-blind, randomized placebo controlled trial of 500 mg/d levofloxacin among 760 patients who were admitted to the hospital and who were at risk of developing neutropenia resulting from treatment of acute leukemia lymphoma or solid tumors. The primary end point was the development of fever during neutropenia that required empirical therapy. Secondary end points included the sort and number of defined infectious complications, as well as survival. Prophylaxis was started 1-3 days before chemotherapy was initiated, and was continued until the patient was no longer neutropenic. Empirical antibacterial therapy was started to manage fever during neutropenia. Infectious complications were classified as microbiologically defined, clinically defined, or unexplained fever. At the end of treatment, fever had affected 243/375 (85%) of those given placebo and 243/375 (65%) of those given levofloxacin, resulting in a relative risk reduction of 24%, an absolute reduction of 20% and, therefore, 5 patients needed to be treated to prevent each episode of fever. This difference was accounted for by the reduction in bacteremia, involving Gram negative and Gram positive bacteria (see Figure 1). The rates of clinically defined infection and unexplained fevers were similar for both treatment groups, as were the rates of mortality and tolerability. There were more isolates resistant to levofloxacin in the group given the fluoroquinolone for prophylaxis [41 of 47 (87%), with 10 of the 13 Gram-negative species being resistant than in the placebo group (32/68 (47%), in which only 4 of the 24 isolates were resistant. Bucaneve et al, therefore, concluded that prophylaxis was effective, well tolerated and cost-effective although there was no effect on the risk of death.
The study of Cullen and colleagues differed from that of the GIMEMA group in several important respects. It focused exclusively on patients receiving chemotherapy for lymphoma and solid tumors (see Table 1), which led to shorter periods of neutropenia and started a 7-day course of prophylaxis around the time the neutropenia was anticipated. The occurrence of probable infections formed the primary outcome measure (defined as a febrile episode, signs attributed to systemic response to infection, signs of a focus of infection, or the use of antibacterial therapy) and affected 109 (14%) of the 781 given levofloxacin and 152 (19.4%) of the 784 given placebo. Fifty (96.7%) patients given the fluoroquinolone were admitted to hospital because of infection, compared with 81 (10.3) of those given the placebo. At first glance, the relative risk reduction brought about by levofloxacin for the primary outcome measure, bacteremia, and mortality were similar for both studies, the absolute risk reduction and, hence, the numbers needed to treat to achieve this were radically different (see Table 2), indicating that prophylaxis would benefit significantly patients admitted to the hospital for treatment of their cancer with intensive chemotherapy, as in the study of GIMEMA group in terms of reducing fever and bacteremia, whereas, it would benefit only a very few of those given less intensive chemotherapy for solid tumors as an outpatient.
In an editorial, Baden acknowledged that these 2 studies provided evidence of the significant benefit of levofloxacin as prophylaxis but questioned whether the costs of prophylaxis in the broadest sense of the word, namely financial, side-effects, emergent resistance among the endogenous bacteria and, theoretically at least, susceptibility to enteric infections would be outweighed by the benefits. In particular, there have been several reports on fluoroquinolone-resistant Escherichia coli in centers using these drugs for prophylaxis of their neutropenic patients. In order to achieve the benefits whilst minimizing the risks, Baden suggested restricting prophylaxis to those at highest risk, but then went on to state that this group has yet to be well defined. There is also not enough known about the period of increased risk nor about the likelihood of resistant organism emerging. She concluded that "if prophylactic antimi-crobial therapy is to be adopted at a cancer center, it should be accompanied by vigorous infection-control practices and careful monitoring for the emergence of resistant bacteria."
The feeling that we have been here before is confirmed by the meta-analysis of Gafter-Gvilli and colleagues on antibacterial prophylaxis for neutropenic patients. They collected all the trials of prophylaxis reported from 1980 to 2004, and identified 95 that were of sufficient quality to be included in the meta-analysis. Fourteen of these trials involved a fluoroquinolone versus either a placebo or no intervention. The results of the GIMEMA study, but not those of the study of Cullen et al correspond very closely to those reported for the meta-analysis, including the prevention of fever (see Figure 2). However, there are some noticeable differences. Whereas the meta-analysis found evidence in favor of fluoroquinolone prophylaxis leading to reduced mortality with a reduction of the risk by 49% from 10% to 5%, the GIMEMA study did not, as the mortality rates for prophylaxis and placebo were both 5%. Next, the risk of developing clinically defined infections and adverse events was higher in the GIMEMA study than found in the meta-analysis. Nevertheless, the results of the GIMEMA study fall broadly in line with those of the meta-analysis.
Commentary
There is a distinct feeling of déjà vu concerning this topic, not least because the pros and cons have been rehearsed many times during the last 2 decades, leading the IDSA to conclude that "Use of antibiotic prophylaxis is not routine because of emerging antibiotic resistance, except for the use of trimethoprim-sulfamethoxazole to prevent Pneumocystis carinii pneumonitis." (IDSA Guidelines: Hughes WT, et al. 2002 Guidelines for the Use of Antimicrobial Agents in Neutropenic Patients with Cancer. Clin Infect Dis. 2002;34:730-751. www.journals.uchicago.edu/CID/journal/issues/v34n6/011605/011605.html
However, never before has there been a study which met all the criteria for a first class clinical trial. Here we are treated to 2 such reports for which the authors deserve our appreciation and editors of the New England Journal of Medicine deserve our gratitude for bringing this topic firmly back into the limelight to help us decide the issues once and for all. Indeed, if decisions about prophylaxis are to be based on evidence rather than emotion, the conclusions seem pretty clear. Reserve the practice for those who really are at high risk of developing infection during the neutropenia-induced intensive chemotherapy. Certainly the numbers needed to treat to prevent death seem way too high to justify prophylaxis for this purpose, despite the meta-analysis. On the other hand, a relatively small numbers of patients would need to be treated to prevent fever and help stay the hand against giving empirical antibiotics. However, the question remains—what do we want to achieve from prophylaxis. If bacteremia is to be reduced, than only patients neutropenic after intensive chemotherapy would fit the bill of being high-risk. Such patients are characterized by having neutropenia of at least a week’s duration and are invariably already in the hospital and are being managed by central venous catheters. Closer inspection of the cause of bacteremia in Figure 3 shows that the principle benefit is seen by a reduction in Staphylococcus aureus, E. coli, and polymicrobial bacteremia, and not the rest. Therefore, a further refinement would be to restrict prophylaxis to those who harbor S. aureus in their nose or E. coli in their gut. This could be known on admission by taking a nasal swab for S. aureus and a fecal sample for E. coli. Although we are getting to the heart of the matter, life is also becoming somewhat complicated, not the least because it would require an active surveillance program which many have long since abandoned for being too costly. The fact remains that prophylaxis with fluoroquinolones had become routine practice in many centers for recipients of hematopoietic stem cell transplants, as evidenced by the Italian participants in the GIMEMA study. Intuitively, it seems better to prevent E. coli sepsis than to wait until it happens, as this can lead to an untimely death. These tragedies do still occur despite prompt administration of empiric therapy, as every hematologist can attest and are all the more bitter because they are considered events that can be prevented. I suspect that North America, just as in Europe, lip service is being paid to guidelines that do not recommend the use of fluoroquinolones for prophylaxis, and that patients in and out of hospitals are being given, and will continue to be given, these useful drugs to tilt the odds in favor of a good outcome. Evidence comes in many guises, and though formal, randomized, controlled trials provide the road map for modern medicine, it is experience that forms the compass to guide the physician from the general to the particular.
The Ginema study reported by Bucaneve and colleagues recruited for a formal, double-blind, randomized placebo controlled trial of 500 mg/d levofloxacin among 760 patients who were admitted to the hospital and who were at risk of developing neutropenia resulting from treatment of acute leukemia lymphoma or solid tumors.You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content