Levofloxacin for the treatment of pneumonia caused by Streptococcus pneumoniae including multidrug-resistant strains: pooled analysis.

OBJECTIVE: To determine the clinical and microbiologic efficacy of levofloxacin for the treatment of subjects with pneumonia caused by multidrug-resistant (MDR) Streptococcus pneumoniae (MDRSP) and non-MDRSP strains. RESEARCH DESIGN AND METHODS: A pooled analysis was conducted using data from ten clinical studies in pneumonia: five comparative studies and five noncomparative studies conducted from 1992 to 2002. This analysis included data from levofloxacin-treated subjects with S. pneumoniae isolated at study entry. Susceptibility of S. pneumoniae isolated from subjects at study entry was determined against representative agents from five antimicrobial classes: tetracyclines, sulfonamides, second-generation cephalosporins, penicillins, and macrolides. Isolates were classified as MDRSP (based on resistance to two or more antimicrobial classes) or non-MDRSP (intermediate resistance or susceptible to all classes or resistant to 1 antimicrobial class). Clinical and microbiologic efficacy of levofloxacin (i.v., p.o., or i.v./p.o. for 5 to 14 days) in the microbiologically evaluable population was determined at post-therapy; a test for homogeneity of the odds ratio of the difference in clinical success for comparative versus noncomparative studies was performed. MAIN OUTCOME MEASURES AND RESULTS: The main outcome measures were clinical success rates and microbiologic eradication rates of 419 microbiologically evaluable levofloxacin-treated subjects with MDRSP or non-MDRSP. Clinical success rates were 96.3% (52/54) and 95.1% (347/365), respectively (difference -1.2; 95% confidence interval [CI]: -6.7, 4.3). Similarly, per pathogen microbiologic eradication rates for MDRSP and non-MDRSP were 96.3% (52/54) and 95.6% (350/366), respectively (difference -0.7; 95% CI: -6.1, 4.8). Study limitations include the use of data from comparative and noncomparative studies. A test for homogeneity of the odds ratios for clinical success in comparative versus noncomparative studies showed no significant difference (p = 0.27). CONCLUSIONS: These data support the use of levofloxacin for patients with community-acquired pneumonia caused by S. pneumoniae, including MDR strains.