© 2003 American Thoracic Society
Bronchoscopic Sampling of Drug ConcentrationsPenetration to Tissue Is the IssueDepartment of Clinical Pharmacy University of Colorado Health Sciences Center Denver, Colorado When selecting antibiotics for treatment of severe infections, the traditional approach has been to eradicate pathogens by achieving high serum drug concentrations relative to the minimum inhibitory concentration for infecting organisms. Reliance on serum pharmacokinetics for establishing susceptibility breakpoints and appropriate drug dosing regimens has largely ignored the fact that most infections do not occur in the serum and that penetration of antibiotics to the actual site of infection is not always accurately reflected by serum concentrations (1). The study of antimicrobial pharmacodynamics attempts to integrate in vitro activity of antibiotics with pharmacokinetic characteristics of the drugs to predict and optimize clinical and bacteriological responses to therapy. Antibiotic penetration to tissues and fluids at the specific site of infection, rather than concentrations in serum, has been increasingly recognized as being much more valuable in predicting response to drugs. Epithelial lining fluid (ELF) has become a well-accepted medium in which to study intrapulmonary penetration of drugs because it provides a good model for the ability of drugs to cross from blood through various tissue barriers and is representative of the extracellular environment in which pulmonary pathogens are often located (2). In this issue of the Journal (pp. 1304–1307), Yamazaki and colleagues (3) describe a novel method of sampling intrapulmonary drug concentrations using bronchoscopy with a bronchoscopic microsampling probe. This microsampling method uses a polyethylene sheath containing an inner polyester fiber probe that immediately adsorbs fluid. The probe is advanced into a distal airway and ELF is directly sampled through adsorption of fluid onto the probe; the volume of epithelial lining fluid and corresponding drug concentrations are easily determined through simple laboratory procedures. When compared with standard methods using bronchoalveolar lavage, levofloxacin concentrations determined by the microsampling method were approximately one-half of the concentrations determined through lavage (3). The microsampling method described by Yamazaki and colleagues is intriguing for a number of reasons related to pharmacodynamics research and implications for antibiotic use. Although bronchoalveolar lavage has been a very widely used method for determining intrapulmonary drug concentrations, this method is subject to important limitations (2). First, it is difficult to accurately determine the volume of ELF recovered because of the relatively large volumes of lavage fluid used during the bronchoscopy procedure. Although endogenous markers such as urea are typically used to make corrections for ELF dilution and more accurately calculate ELF volume, bronchoalveolar lavage methods have been associated with overestimation of ELF volume by as much as 100–300% (2). Because calculation of drug concentration relies on precise determination of ELF volume, results of studies using this methodology may not be very accurate. Another limitation of bronchoalveolar lavage is that it does not readily allow for repeated sampling of ELF from the same portion of the lung due to the potential for residual lavage fluid which dilutes both ELF and drug concentrations. Characterizing the time course of drug penetration thus becomes problematic because different portions of the lung must be sampled for serial determinations of drug concentrations, and actual distribution of drug to these sites may differ due to anatomical or disease-state induced considerations (2, 4). The microsampling method described by Yamazaki and coworkers (3) has the potential to overcome both of these important limitations. Since it does not rely on installation of lavage fluid for ELF recovery, it does not involve the same error-prone calculation of ELF volumes to determine drug concentrations. In addition, the same site may be repeatedly sampled without concern over residual lavage fluids or variability in drug penetration to different portions of the lung. Why is more accurate determination of intrapulmonary penetration of antibiotics so important? Pharmacodynamic research has yielded information that has greatly improved our understanding of the pharmacologic actions of antibiotics and relevance to clinical outcomes. Increased understanding of antibiotic pharmacodynamics has helped explain the discrepancy between high rates of in vitro penicillin and macrolide resistance among Streptococcus pneumoniae (5, 6) and the apparent lack of clinical treatment failures in patients treated with ß–lactam or macrolide drugs for community-acquired respiratory tract infections (1). Pharmacodynamic research has also helped explain why antibiotics with quite different in vitro activities and pharmacokinetic profiles often have similar clinical and microbiological efficacies (7, 8), or why drugs with apparently similar activities may be quite different in clinical outcomes related to the treatment of certain pathogens (9–11). In addition, pharmacodynamic research has led to new treatment strategies such as high-dose, short-course therapies designed to achieve rapid bacterial eradication and limit the development of resistance while still maintaining good clinical efficacy (12, 13). There is still is a lack of studies, however, in which clinical efficacy has been prospectively evaluated in relation to intrapulmonary drug concentrations determined within the same group of patients (4). Indeed, with notable exceptions (14), clinical studies evaluating pharmacodynamic parameters and their relation to patient outcomes have largely been either retrospectively conducted or extrapolated from pooled data generated in separate studies (15, 16). Although our understanding of antibiotic pharmacodynamics has become much more advanced, continuation of this progress depends on the performance of studies that yield appropriate and accurate data. The bronchoscopic microsampling method described by Yamazaki and coworkers (3) appears promising, but this technique has not yet been sufficiently validated to make it a method of choice for sampling of ELF. In this study, bronchoscopic microsampling was compared with standard bronchoalveolar lavage with regard to levofloxacin concentrations determined by each method. It is not yet known, however, whether the microsampling technique yields results that are truly more accurate and reliable; inaccurate results that do not reflect actual drug distribution to the ELF could likewise have been produced by both methods. More work remains to be done to fully validate this technique. Development of new methodologies, such as the one described by Yamazaki and colleagues (3), however, have the potential for increasing the accuracy and scientific validity of pharmacodynamic studies, thus allowing for further application of antimicrobial pharmacodynamics in an appropriate and clinically relevant manner. FOOTNOTES Conflict of Interest Statement: D.N.F. has no declared conflict of interest. REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
