The integrated use of pharmacokinetic and pharmacodynamic models for the definition of breakpoints.

For fluoroquinolones AUC/MIC ratios are known to correlate with clinical outcomes for patients suffering from respiratory tract infections (RTI) and complicated skin and skin structure infections (cSSSI). This paper describes the results of a population PK/PD analysis followed by Monte Carlo simulations to estimate clinical outcome and the microbiological breakpoints for a 400 mg once-daily moxifloxacin (MFX) treatment schedule. Based on PK data from 416 subjects, a non-compartmental population PK model was developed first to describe the expected exposure (AUC) distribution in humans. Height and gender were the main population covariates with moderate influence on PK variability. Albumin, bilirubin, and creatinine clearance (as derived from serum creatinine according to Cockroft and Gault) had a mild effect on AUC. Residual unexplained variability of AUC was low (13.1%). To describe the PD function the MIC distribution pattern of more than 3,000 isolates of S. pneumoniae as the representative pathogen for RTI (MIC90, range: 0.125; 0.006-4 mg/l) was built into the population PK/PD model for RTI, while 126 isolates of methicillin-susceptible Staphylococcus aureus strains (MIC90, range: 0.125; 0.03-4 mg/l) were the basis for the PD function in cSSSI. Simulations for 20,000 (RTI) and 4,000 (cSSSI) subjects were performed to evaluate the AUC/MIC characteristics for moxifloxacin for these two diseases. Overall, a target hit rate was THR = 99% for RTI, while it amounted to THR = 97.5% for cSSSI when applying a threshold of AUIC > 30 [h] as the PK/PD surrogate parameter which is predictive for a positive clinical outcome. A target hit rate of THR = 93.6 % (RTI) and 97.3% (cSSSI), respectively, was predicted when assuming that an AUIC of > 125 [h] is indicative of clinical success (as shown for ciprofloxacin and severe RTIs due to gram-negative infections). In clinical trials with patients receiving 400 mg moxifloxacin once daily for the treatment of community-acquired pneumonia (CAP) success rate was approximately 93.5%. From the simulations performed for RTI an analysis of the overall likelihood of therapeutic failure broken down according to MICs suggests that the risk of a negative clinical outcome at a MIC = 1 mg/l is approximately 0.25% (for MIC = 2 mg/l: predicted likelihood approximately 0.5%) assuming that a cutoff of AUIC = 30 [h] is applicable. Likewise, for cSSSI the probability to fail is predicted as 1.6% at a MIC = 2 mg/l (no strains with MICs between 0.5 and 1 mg/l available from the clinical isolates). These findings are in line with the breakpoint definition of the former National Committee for Clinical Laboratory Standards (NCCLS) for MFX (=1 mg/L to differentiate between susceptible and intermediately susceptible microorganisms; = 2 mg/l to separate intermediate from resistant pathogens). The results of the investigation indicate that the noncompartmental PK/PD model for MFX is suitable to predict clinical outcomes in CAP and cSSSI caused by gram-positive aerobe pathogens. They confirmed that a 400 mg once-daily dosing regimen is suitable to treat these diseases successfully. They are in agreement with the microbiological breakpoints determined by independent methods by the Clinical and Laboratory Standards Institute (CLSI) (former NCCLS).