Guidelines for the Use of Antimicrobial Agents in Neutropenic Patients with Unexplained Fever--Reply

Abstract

EDITOR-Our article [1] referred to by Kalman and Barriere [2] is a consensus report and not a literature review.However, an in-depth study of the world literature was undertaken and an effort was made to base our conclusions on published scientific data when possible, attempting to avoid certain pitfalls.Individual biases of committee members were guarded against, and recommendations based on anecdotal experiences and selected unilateral literature references were discouraged.The most treacherous pitfall was misinterpretation of published research on the use of antibiotics in granulocytopenic patients with fever.Such factors as experimental design, criteria for enrollment, definitions of therapeutic success and failure, and critical analysis of data vary greatly from study to study, often not permitting valid comparison of studies with the same drug or drug combinations.An important companion paper to the consensus report in the same issue of the Journal [3] provides guidelines for more uniform and valid clinical trials.We believe Kalman and Barriere have experienced the above pitfalls.They recommend that physicians "avoid the use of double {3-lactams whenever possible" and support the statement with selected references dealing with only the negative aspects of the combination, and they have misinterpreted some of the published articles.In the committee's report we acknowledged the disadvantages of {3-lactam combinations including the occasional emergence of resistant organisms, high cost, possible antagonism with certain bacterial infections, and possible limited effects against staphylococci, enterococci, and anaerobic bacteria.We commented on the first European Organization for Research on Treatment of Cancer (EORfC) study [4], which demonstrated the greater efficacy of {3lactam plus aminoglycosides than of two {3-lactams, but we also pointed out that the {3-lactams (cephalothin and carbenicillin) used in this 1978 study are obsolete and not comparable to more modern {3-lactams in current use.The fourth EORTC study [5] is not a study of double {3-lactams; it compares single {3-lactams with amikacin and is not relevant to the argument.The study by Winston et al. [6] is misinterpreted by Kalman and Barriere.This prospective randomized study of 212 patients at the University of California Los Angeles (UCLA) showed the overall therapeutic response rates for moxolactam plus piperacillin (77 %) and moxolactam plus amikacin (79 %) were similar.There was no difference in toxicity.Furthermore, Kalman and Barriere did not mention a more recent study of 243 febrile granulocytopenic patients, also at UCLA, treated with piperacillin plus cefoperazone [7].The authors concluded that the combination "provides adequate coverage for most bacterial pathogens and is safe and effective therapy for febrile granulocytopenic patients:' These studies were influential in the committee's decision to include a statement on double {3-lactams in the report.It is therefore surprising that an argument against double {3-lactams comes from the same center where the combination was found successful.