| Objective To describe the clinical and microbiological epidemiologic characteristics of patients with acute leukemia who suffered febrile episodes. The clinical benefits and the impacts on distribution and antibiotic resistance of pathogenic bacterium associated with fluoroquinolone prophylaxis during neutropenia were investigated. Another aim of the present study was to evaluate the choice and the results of different initial empirical antibiotic therapy schemes.Methods A total of 309 infection episodes occurred in patients with acute leukemia were retrospectively analyzed. These patients were hospitalized during the period June 2003-May 2004. They were belonging to two different independent hematology wards in the same hematopathy-special hospital. The patients admitted in ward A (group A) received both oral ofloxacin as antibacterial prophylaxis and oral fluconazol as antifungal prophylaxis during the neutropenia phase, and only Oral fluconazol was administered as antifungal prophylaxis to the patients in ward B(group B). The influences of prophylactic ofloxacin were determined by comparative evaluation on the clinical characteristics and the microbiological profile in both groups' patients. The initial and final clinical responses were assessed among four different initial empirical antibiotic therapy schemes. Four different empirical antibiotic schemes were as follows: penicillins/quinolones, third-generation cephalosporin, third-generation cephalosporin combined withβ-lactamase inhibitor and carbapenem. Results Almost all enrolled patients experienced severe neutropenia. The median durations of ANC (absolute neutrophil count, ANC)<0.5×10~9/L and<0.2×10~9/L were 14(range 0-50) days, 11(range 0-43) days, respectively. The median persistence time of fever was 8(range 1-46) days. Gram-negative bacilli (51.5%) still were the predominant causes, gram-positive cocci and fungi accounted for 31.9%, 16.6%, respectively. The frequently isolated microorganisms were Escherichia coli (17.6%), Staphylococcus aureus (10.9%), Enterococcus faecalis (6.7%), Klebsiella pneumoniae (6.3%), and Candida albicans (5.5%). The positive rates of ESBLs in Escherichia coli and Klebsiella pneumoniae were 33.3% and 8.5%, respectively. Most gram-negative bacilli were no longer sensitive to penicillins antibiotics; the susceptibility rate was less than 60%, except for piperacillin-tazobactam. They were still sensitive to the third-generation cephalosporins such as ceftazidime and cefoperazone/sulbactam with a susceptibility rate of more than 80%. Unfortunately, the sensitive rate of gram-positive bacteria to erythromycin and gentamycin was merely 0-57.5%. Vancomycin-resisted gram-positive cocci were not found. The common infection sites included the gingivitis, dental ulcer, upper respiratory infections /tonsillitis, pneumonia, gastrointestinal tract infections, perianal infections, while origins of infection couldn't be identified in 58(18.8%) episodes of fever. The incidence of septicemia was 18.5%(37/200) and the most microorganism isolated from blood was Escherichia coli (38.4%). The prophylaxis of ofloxacin decreased the proportion of normal flora (hemolytic streptococcus A/neisseria) (p=0.00). The frequency of Escherichia coli was higher in group A than group B (23.9%, 12.5%, respectively, p=0.00). Exposure to ofloxacin didn't affect the distributions of other pathogenic bacteria. No difference in the rate of ESBLs (+) Escherichia coli was discovered between ofloxacin and no ofloxacin prophylaxis groups (31.9%, 35.4%, respectively, p=0.61). Commonly used antibiotics excluding piperacillin-tazobactam, cephalothin and amikacin had similar activity against the major strains isolated from two groups. Patients who received antibacterial prophylaxis showed a lower incidence (10.7%) of upper respiratory infections/tonsillitis and a relative higher incidence (14.4%) of gastrointestinal tract infections than those without intervention measure. The prophylactic use of ofloxacin during neutropenia couldn't cut down the risks of perianal infections. The overall initial clinical response rate to empirical antimicrobial treatment schemes was 63.1% and the best initial efficacy (92.9%) was acquired in carbapenem-therapy group. Final efficacy in patients who received four different empirical antimicrobial schemes was similar, with response rate of 93.9% and mortality rate of 0.9%. Nevertheless, carbapenem-therapy group included more severe infections that means more frequency of chills, higher initial temperature, lower ANC and higher peaked body temperature throughout fever episodes. Conclusion Symptoms and signs of infection may be atypical because of minimal or absent inflammatory reaction in severely neutropenic patients with acute leukemia, and the infections often show a promptly progressing course. Most third-generation cephalosporins keep excellent activity against major gram-negative bacteria. The quinolones prophylaxis may not influence the distribution(except Escherichia coli) and susceptility of main pathogens, don't decrease the incidence of infection in gastrointestinal tract and perianal tissue. Early individual empirical antimicrobial therapy can predominantly consist of third-generation cephalosporins according to the epidemiology and sensitivity of bacteria. The empirical carbapenem-therapy is essential for patients with more severe infections. |