ABSTRACT
Although the influence of protein binding (PB) on antibacterial activity has been reported for many antibiotics and over many years, there is currently no standardization for pharmacodynamic models that account for the impact of protein binding of antimicrobial agents in vitro. This might explain the somewhat contradictory results obtained from different studies. Simple in vitro models which compare the MIC obtained in protein-free standard medium versus a protein-rich medium are prone to methodological pitfalls and may lead to flawed conclusions. Within in vitro test systems, a range of test conditions, including source of protein, concentration of the tested antibiotic, temperature, pH, electrolytes, and supplements may influence the impact of protein binding. As new antibiotics with a high degree of protein binding are in clinical development, attention and action directed toward the optimization and standardization of testing the impact of protein binding on the activity of antibiotics in vitro become even more urgent. In addition, the quantitative relationship between the effects of protein binding in vitro and in vivo needs to be established, since the physiological conditions differ. General recommendations for testing the impact of protein binding in vitro are suggested.
INTRODUCTION
Binding to plasma proteins plays a major role in drug therapy as this binding provides a depot for many compounds, affects pharmacokinetics (PK) and pharmacodynamics (PD) of drugs, and may influence the metabolic modification of ligands (34, 104). Protein binding (PB) of antibiotics may affect the efficacy of antimicrobial therapy in two ways. First, only the non-protein-bound fraction of a drug in plasma can penetrate into and equilibrate with the extravascular space (13). Penetration into the extravascular space is highly important for antimicrobial therapy, as the majority of bacterial and fungal infections occur in the interstitial fluid of tissues or in other body fluids than blood (113). PB also affects drug clearance from the body. For antibiotics eliminated by tubular secretion or hepatic metabolism, high PB is associated with lowered drug elimination. Additionally, PB negatively correlates with glomerular filtration, since only the free drug is filtered (63). On the other hand, small fluctuations in protein binding usually have no effect on average unbound concentrations, and this is especially the case for drugs with a low extraction ratio (11). Changes in protein binding can also alter the volume of distribution and, hence, the half-life, which may, in turn, modify the time above MIC. Extensive reviews on the impact of PB on pharmacokinetics of drugs are available (12, 13, 27, 29, 82, 86).
The second effect of PB is demonstrated by the impact on the antibacterial activity. Over many decades, various methods were used to show that only the non-protein-bound fraction of an antibiotic is microbiologically active (10, 28, 70, 111). For important antibiotic classes, such as penicillins and quinolones, clear relationships between the percentage of protein binding and the impairment of antimicrobial action by protein-containing growth media have been specified (54, 117). Although the influence of PB on antibacterial activity has been reported for many antibiotics and over many years, there is currently no standardization for pharmacodynamic models that account for the impact of PB of antimicrobial agents in vitro (23). The consequent use of an assortment of methods and test conditions has led to the current set of diverging results regarding the impact of PB on antimicrobial efficacy (10, 84, 117). This diversity of models, methods, and results foment ongoing debate regarding the relevance of protein binding, especially with regard to novel antibiotics where marketing motivation plays an undeniable role.
With these important concerns in mind, the present article focuses on the current situation regarding the investigation of PB in vitro and its impact on antimicrobial dosing and therapy. Unsolved problems and questions regarding standardization of in vitro testing will be discussed with the goals of encouraging appropriate research and of stimulating discussion regarding the advantages and disadvantages of currently used methodologies for the investigation of the impact of protein binding in PD models.
CURRENTLY AVAILABLE PD MODELS USED IN DETERMINING PB IMPACT
In vitro models. (i) Media.The first point that has to be realized when studying the effect of protein binding in vitro is that the mass balance as measured in virtually all studies in vitro is significantly different from that measured in vivo. In an in vivo infection, the ratio between drug binding receptors on the microorganisms at a specific site and the available drug molecules is different and might be much more variable than the ratio in an in vitro test tube where the impact of protein binding is studied in the same compartment where the bacteria reside. Studies that look at the impact of protein binding in a separate bacterial compartment, however, would more closely resemble the in vivo situation. Conclusions with respect to the impact of protein binding based on in vitro studies should therefore be viewed with the utmost caution.
Irrespective of this concern, in vitro models are, by far, the more frequently used models when studying the influence of PB on antimicrobial activity (20, 23, 57, 79, 84, 91, 114, 117, 118). However, the antimicrobial activity of drugs with comparable degrees of PB has been shown to be differentially affected by the presence of proteins in vitro (20, 23, 79, 91, 102, 117, 118). The wide range of microbiological methods and media employed for in vitro PB models clearly jeopardizes the interpretation of results due to various factors influencing PD (20, 23, 57, 79, 84, 91, 114, 117, 118).
To investigate the impact of PB on the antibacterial activity of antibiotics, bacterial susceptibility is often determined in Mueller-Hinton broth (MHB) where MICs are consecutively compared to those determined in a protein-rich medium (76, 91, 117, 118). A standard broth, such as MHB, provides optimum conditions required for bacterial growth and closely resembles human serum in terms of pH, Na+, K+, and Cl− content and osmolality (83). To mimic in vivo conditions as closely as possible, an ideal test medium would achieve a level of PB comparable to that present at the infection site. On the other hand, compared to protein-free growth media, the chosen test medium should not influence bacterial growth to allow for quantification of the effect of PB independent of other factors. Various approaches have been taken to account for the lack of protein in MHB.
(ii) Adding serum to test media.Although applying to the conditions in blood only, using 100% serum for antimicrobial testing would best mimic in vivo conditions, including pH and protein concentration. However, human serum is less optimal for bacterial growth than microbiological standard media (74, 117). When bacterial strains are exposed to antibiotics in active serum, even subinhibitory antibiotic concentrations can exert a bactericidal effect, due to the synergistic effect of antibiotics and serum on bacterial growth (17, 23, 40). There have been attempts to acclimate test strains to growth in human serum, but these strains lost their serum resistance after several passages in microbiological media, precluding their use as reference strains (74).
A wide variety of serum concentrations in test medium, ranging from 20% to 100% serum, have been used by investigators to study the changes in antimicrobial activity in relation to PB (7, 9, 32, 61, 75, 80, 88, 102, 114). In order to minimize the inhibiting effects of serum on bacterial growth, most studies have limited the human serum concentration in test media to 70%. Nevertheless, a cutoff value for the impairment of bacterial growth by the addition of serum has not been established, as serum may inhibit bacterial growth even at concentrations below 50% (74). Components in native serum, such as complement factors properdin and transferrin, might induce complex and unpredictable inhibition of bacterial growth (37). Therefore, heat inactivation of serum, but in a way that does not destroy the functionality of binding proteins, is crucial for all antimicrobial models investigating PB (50). Even with heat inactivation and without the alteration of protein binding, serum could still contain factors that enhance the antimicrobial activity of some antibiotics against specific bacteria (59). The use of bacterial strains that grow well in the chosen medium is a prerequisite for any model investigating impact of PB. Therefore, control strains growing in standard and serum-supplemented media are necessary.
(iii) Adding albumin to test media.As an alternative to using serum, broth containing human albumin as a protein supplement has been utilized (23, 76, 118). An albumin concentration of 4 g/dl has been widely used, as this is equivalent to the albumin concentration in human serum (20, 23, 75, 79, 91, 100, 103). However, due to the possibility of binding to nonalbumin serum proteins, a major limitation of using human albumin is that it does not necessarily provide the same PB capacity as found with human serum (76, 117). It is well-known that the binding capacity of albumin is influenced by conformational changes of the protein that is induced by the presence or absence of fatty acids and electrolytes or pH (15, 31, 67). Indeed, many commercial preparations of human albumin are fatty acid free and therefore, have decreased binding of some antibiotics (Fig. 1). In this context, a more appropriate test medium with regard to PB capacity might be obtained by titrating the concentration of albumin up until a PB level equal to that in pure serum is achieved.
In vitro mean protein binding of moxifloxacin in Mueller-Hinton broth (MHB), pure serum, and MHB containing various amounts of albumin or serum. The broken line indicates protein binding (PB) achieved in pure serum. PB was corrected for nonspecific membrane binding during ultrafiltration. Data are presented as means plus standard deviations (SD) (error bars) (n = 6). The albumin was fatty acid free. (Reprinted from reference 117 with permission of the publisher.)
Other media used to simulate PB in vitro have included ultrafiltrate of serum in comparison with pure serum as well as in vitro investigations from samples obtained in vivo (56, 68, 117).
Efficacy measurement.In general, two main PD approaches for determining the effect of PB on bacterial activity in vitro can be distinguished: those based on MIC testing and those based on a kill-curve approach (72). MIC testing is easy to do and provides a fast way of screening for the influence of PB on antimicrobial activity. However, standard MIC testing uses only 2-fold dilution steps and detects only visible growth, i.e., a 100-fold increase in bacterial counts after 18 to 24 h of incubation, but MIC testing cannot distinguish between less pronounced growth and bacterial killing. It may therefore be inappropriate for investigation of antibiotics displaying a moderate PB of less than 50%. Time-killing curves (TKC), on the other hand, provide information about the interaction between pathogens and antimicrobial agents as a function of concentration and time (96, 118), thus allowing for an exact determination of CFU/ml. For fluoroquinolones, for example, the impact of PB has been repeatedly overlooked when using MICs, while a clear impact has been consistently observed when using TKCs (24, 117). Often, only antibiotic concentrations equaling those achieved in serum after standard dosing are used to detect impact of protein binding on antimicrobial activity (20). While including a range of different clinically achievable concentrations will help to assess potential impact of PB on clinical efficacy, concentrations of the antibiotic equal to (or near) the MIC should be included in the analysis when the main purpose of the experiment is the investigation of the overall impact of PB on antimicrobial activity.
Animal models.PK parameters corrected for PB have been shown to be the most predictive for efficacy in recent animal models for novel antibiotics (30, 85). In order to assess the relevance of an individual factor on clinical cure, information on dose response of the individual factor must be available. This information is clearly more difficult to discern in a living organism than ex vivo (111). As far back as 1983, a study by Merrikin et al. (70) examined the efficacy of various penicillins against the test strain of Staphylococcus aureus in a peritonitis model. The penicillins were selected for comparable in vitro antimicrobial activity and administered in identical doses of 10 mg/kg of body weight. This study design was later repeated for cephalosporins and Streptococcus pneumoniae (36, 70). Although these studies found a clear inverse correlation between efficacy and PB, they could not distinguish the extent of impairment of activity from various factors such as differences in drug penetration into the site of infection, elimination half-life, or favorable PK/PD characteristics (e.g., ceftriaxone).
A meaningful study design for investigation of the impact of protein binding on efficacy in vivo would examine the activity of antibiotics in infection models in various animal species by choosing species that exhibit differences in PB of the respective drug. However, other differences between species that could impact antibiotic activity, for instance interactions of the immune system with the antibiotic or differences in natural tolerability toward the studied infections would still have to be taken into account when analyzing study results.
Clinical data.Breakpoints are an important tool for clinical decision-making. Not correcting experiments that are used to set breakpoints for PB may lead to overestimating the activity. In the case of highly protein-bound antibiotics such as fusidic acid (around 98%) (41) or ceftriaxone (around 96%), this overestimation of activity may contribute to clinical failure (14, 16, 111). In the case of ceftriaxone, the high protein binding may explain the high microbiologic failure rate of a single dose of ceftriaxone in group A streptococcal tonsillopharyngitis based on population modeling techniques and Monte Carlo simulation (14). In a more recent example, the early clinical development of daptomycin (PB > 90%) at low dosage resulted in clinical failure when used in bacteremia despite favorable in vitro data obtained in protein-free media (38, 58). As it is not possible to modify protein binding in vivo, clinical confirmation regarding the impact of protein binding on antimicrobial efficacy is difficult to obtain. Additionally, antibiotics are commonly administered at rather high doses, often resulting in considerably higher concentrations than necessary for elimination of the bacterium. Therefore, even with antibiotics that display protein binding as high as 90% (resulting in a 10-fold decrease of activity), the clinical impact of PB might be overlooked in cases involving highly susceptible bacteria.
Last, but also of vital clinical importance, the success of antibiotic therapy depends on a range of characteristics of and interactions between the antibiotic, host, and pathogen(s). The complexity of these interactions has often been noted in studies where PK/PD indices indicate promising predictive value for microbiological success, yet these predictions fail to correlate with clinical response (4).
Mathematical models.Mathematical models have been used to describe the impact of PB on the pharmacokinetics and pharmacodynamics of antimicrobials (86). In this context, data obtained from a number of in vitro experiments can be used to estimate the impact of PB on antimicrobial killing for various antibiotics and bacteria (116).
METHODS FOR DETERMINING THE DEGREE OF PB
When evaluating the influence of PB on the PD of an antibiotic, it is necessary to measure the free antibiotic fraction precisely.
Currently used methods for free fraction measurement include equilibrium dialysis, ultrafiltration, microdialysis, ultracentrifugation, and fluorescence spectroscopy as well as chromatography and capillary electrophoresis (1, 39, 42, 45, 46, 48, 55, 60, 62, 66, 78, 89, 90, 95, 99, 101, 106–108, 115). We will briefly describe the most commonly used methods below, as each has both advantages and limitations (10). Independent of the specific method used to determine the free antibiotic fraction, factors that can impact PB (see below) should be maintained within physiologic conditions in order to mimic the in vivo situation (55, 107).
Protein binding in vitro.Although there is no standard method for PB measurements in vitro, equilibrium dialysis (ED) is often regarded as the “reference method” for determining the PB profile of a drug (78, 89). ED is relatively labor-intensive but is also quite precise (90, 107, 115). One concern related to ED is that, depending on the membrane material, drug concentration, and degree of ionization, a fraction of the drug may be adsorbed by the dialysis membrane (55). Thus, in addition to consideration of other factors, nonspecific drug adsorption must be determined and corrected for in the final calculation of the free drug ratio when using ED (10, 90).
In ultrafiltration (UF), another widely used method for determination of PB, centrifugal forces are usually employed as the driving force for the passage of plasma water across a filter membrane.
Adsorption of drug by ultrafiltrate membranes can be problematic but can be compensated for by taking into account measurements obtained from conducting preliminary UF experiments in a protein-free medium (90, 108, 115). Additionally, as the protein concentration in the plasma sample is increased during the filtration of plasma water, only a small volume of ultrafiltrate should be collected, since the protein concentration in the upper reservoir rises during the UF process.
Fluorescence spectroscopy, chromatography, and capillary electrophoresis are nowadays rarely used in this field (42, 55, 66, 78, 89).
Protein binding in vivo.Microdialysis (MD) offers the significant advantages of in vivo measurement of PB. Commonly used for determining unbound drug concentrations in the interstitial fluid (ISF) of various tissues (48), in vivo MD is also employed to determine unbound drug concentrations in the blood compartment (87, 106). MD examines the diffusion of compounds along their concentration gradient from ISF or blood into dialysate (48). For the purpose of defining PB, a MD probe containing a dialysis membrane is surgically implanted into a blood vessel (115). As a dialysate buffer is pumped through the probe, the unbound drug in the plasma diffuses across the membrane into the probe. According to the molecular weight cutoff of the semipermeable membrane, large molecules like proteins will be retained by the membrane (95). Microdialysate samples can then be collected over time for subsequent analysis of the free fraction of an antibiotic (106). MD offers the significant advantages of in vivo measurement of PB. However, due to the small volumes of dialysate, sensitive analytical techniques are required to measure drug concentrations in MD experiments (78). Microdialysis can also be employed in vitro in complex cell culture medium studies where prediction of PB is difficult and other methods are not applicable (87).
PROBLEMS AND CHALLENGES OF CURRENT MODELS
PB is a rapid and reversible process (8). Many individual factors can modify the degree of protein binding and therefore should be considered when establishing an appropriate in vitro model for protein binding of antibiotics.
Animal versus human proteins.Proteins and sera from various animals, including bovine animals, rodents, horses, and others have been used to investigate PB of antimicrobials (64). However, animal proteins often differ from those of humans in various ways (112). First, the relative proportions of individual proteins differ significantly between mammalian species. For example, albumin concentration in the plasma of mice, rats, and horses is as much as 40% lower than that found in human plasma (71). Nonetheless, the degree of protein binding among animals and humans is often quite similar.
Second, comparative studies between rodent and human albumin indicate that there are structural differences in the binding sites of the two albumins as well as quantitative differences with respect to the extent of drug binding (6, 65). In general, protein binding for most antimicrobials is lower in rodents than it is in humans (29). However, a few antibiotics display similar or even higher binding in rodents than in humans.
In some cases, significantly lower PB has been found in commercially available bovine albumin than is found in human albumin. However, as both the number of binding sites and the affinity for binding may vary between human and bovine albumin, comparison of total binding lacks predictability (51, 87). For instance, the protein binding of ceftriaxone has been shown to differ significantly between commercially available bovine and human serum albumin with protein binding levels of 20 and 57%, respectively (Fig. 2) (87). Thus, while studies in animal serum are valuable, it is essential to corroborate findings with human serum.
In vitro mean protein binding on different media. In vitro mean protein binding of ceftriaxone (CRO) and ertapenem (ERT) in Todd-Hewitt broth (THB), THB with bovine serum albumin (BSA) at a protein concentration of 40 g/liter, THB with human serum albumin (HSA) at a protein concentration of 40 g/liter, pooled adult bovine serum (ABS), and pooled human plasma (HP). The HSA was fatty acid free. (Reprinted from reference 87 with permission.)
Different classes of human proteins.Human serum albumin, the most abundant protein in plasma, binds different classes of ligands at multiple domains (34). As the PB of most antibiotics can be mostly ascribed to albumin, the addition of albumin at physiological concentrations to a test medium is commonly expected to model binding in pure serum. However, there is evidence that some antibiotics may bind to a range of other proteins including transferrin, lactoferrin, and alpha-1-acid glycoprotein (13, 21, 33, 47, 53). As a consequence, protein binding in artificial media, for instance MHB containing albumin at physiological concentrations, will not always model PB in pure native serum (76, 87, 117). For example, clindamycin which binds to alpha-1-acid glycoprotein, displays 40% protein binding in MHB containing 40 g/liter of human albumin but displays 85% binding in human serum (29). Moreover, whole blood cells, fractions of blood cells, and hemoglobin itself bind and possibly inactivate antibiotics in vivo or in vitro when these are used to supplement bacterial growth media (5, 52, 98).
Impact of temperature, pH, electrolytes, and supplements on protein binding.A range of factors influences the binding capacity of serum and albumin. Human albumin consists of three homologous domains that undergo several well-described proton-induced conformational changes (22, 31). One of these pH-dependent transformations occurs in the pH range of 6 to 9 and is referred to as the neutral-to-base (N-B) transition (25). Indeed, a clear pH dependence of PB has recently been shown for tigecycline, thus stressing the importance of pH measurement and reporting in in vitro studies (72). It has been shown that the presence of calcium ions in concentrations physiologically found in plasma leads to the N-B transition proceeding at lower pH values, thereby shifting the observed midpoint of transition from pH 8.4 to 7.7 (43, 119). While many compounds have a stronger affinity to the N isoform, others show significantly higher affinity to the B isoform of albumin (67, 110). The binding of ciprofloxacin to these conformational states of human serum albumin has been previously investigated. The free fraction of ciprofloxacin was found to be approximately 80% higher for B isomers compared to N isomers of human serum albumin (2). Although the impact of the N-B transition on the binding of other antibiotics is unclear, a higher prevalence of the B isoform in serum could be one explanation for observed differences in the binding properties of albumin in MHB compared to albumin in a physiologic environment (2, 76, 117).
Importantly, albumin binding capacity is reduced in commercially available human serum albumin preparations that have added stabilizers or that lack fatty acids (49). While fatty acids modify the conformational change of albumin, the direction on the effect on binding of drugs is heterogeneous (69, 94). For example, free fatty acids increase the albumin binding of benzylpenicillin, cephalothin, and cefoxitin but decrease the binding of dicloxacillin, cefamandole, and sulfamethoxazole. Chloramphenicol is not affected by free fatty acids. All commercially available albumin preparations contain either caprylate (octanoate) or N-acetyl-tryptophanate, both of which bind to albumin binding site II and thus impair the binding capacity in comparison to pure albumin (77). This is a concentration-dependent effect. Likewise, the anticoagulant used to obtain serum, i.e., citrate or EDTA, may impact drug binding of human serum (105).
As the impact of temperature on PB of drugs is well described, all models investigating any aspect of PB of antimicrobials should be performed at physiological temperature (25, 92).
Concentration dependence of protein binding.Concentration-dependent protein binding has been observed in many drugs (30, 85, 105). Certain antibiotics, for instance, some macrolides and β-lactams, have been shown to display saturable concentration-dependent PB within the concentration range that can result from a normal therapeutic dose (10, 19, 21, 78, 81). As the concentration of a drug in plasma increases, binding sites on proteins are increasingly saturated, resulting in higher percentages of unbound drug in plasma (examples are ceftriaxone and cefazolin) (3, 81). However, nonlinear protein binding can also occur in the opposite direction with lower unbound fractions at high drug concentrations. Tigecycline is an example for this atypical pattern of concentration-dependent PB (18). Many theories have been suggested to explain this phenomenon including the ability of tigecycline to form metal ion complexes (73). Thus, a wide range of clinically achievable drug concentrations should be considered in binding analysis studies.
Free drug hypothesis and areas of future research.While in vitro experiments are static, living systems are dynamic with regard to target binding, PB, metabolism, transport processes, and diffusion between compartments (93). Therefore, the results of properly designed in vitro tests always need to be confirmed in in vivo settings. For antibiotics, discrepancies have been noted in some in vitro studies and have often been ascribed to low binding affinity of the drug to albumin. Indeed, PB is always reversible, as the on/off rate of the binding depends on both binding affinity as well as the number of binding receptors in relation to the concentration of the drug. However, to the best of our knowledge, none of the studies observing these discrepancies in activity have determined the level of PB in their specific study setting. Instead, the studies have assumed that PB will function as it does in human serum. Therefore, we believe that it is of the utmost importance to measure PB in actual study settings. This protocol is especially critical where results are found to conflict with the free drug hypothesis. If, after measuring PB in the study setting, discrepancies remain, then direct investigation of the binding affinity should be considered.
Two newer antibiotics with very high PB retain clinical effectiveness. This observation was interpreted as limited relevance of PB for these drugs. For example, telavancin, a glycopeptide with PB above 90%, was recently approved on the basis of clinical efficacy despite in vitro studies demonstrating a 10-fold reduction of activity in the presence of proteins (44, 97, 109). The chosen clinical dosages were high enough to provide appropriate fAUC/MIC ratios (fAUC is the area under the concentration-time curve for the free, unbound fraction of a drug) (75). Similarly, the in vitro activity of daptomycin is inhibited less by serum or plasma than one would predict by its high protein binding. Among several suggestions, a weak affinity of daptomycin toward proteins was discussed as possible explanation (26, 35, 38). The in vitro test conditions were not standardized in all these studies.
In light of these open questions, further research is essential for the case of highly bound antibiotics. This research should focus on attempting to objectify whether a discrepancy with the free drug hypotheses indeed exists or can be explained by other factors such as errors in the model, special pharmacokinetic properties of the drug, such as accumulation in certain cells and compartments, or the effect of the drug on the immune system. Additionally, efforts are needed to establish models that may account for dynamically changing in vivo conditions (93). The development of methods for direct quantification of affinity of PB or its reversibility in individual models as well as in vivo may lead to a better understanding of results where the free drug hypothesis is contradicted. General recommendations that should apply to all areas of investigation of PB of antimicrobials are summarized below.
KEY FINDINGS AND RECOMMENDATIONS
Currently, no standard method for the investigation of the impact of PB on the antimicrobial activity of antibiotics exists. Numerous experimental factors impact the PB and, consequently, the relevance of results of the various recognized models. This could explain the somewhat contradictory results obtained from different studies. Therefore, appropriate standards must be developed. Regarding the development of a standardized method, significant challenges await. For instance, on the basis of currently available research, we cannot assume that it will be possible to find standard media and test conditions that are appropriate for all antibiotics. Thus, both the test medium and PD model must be carefully evaluated for their appropriateness on a case-by-case basis.
Nevertheless, some general recommendations can be proposed with confidence. The following general recommendations are based on current understanding of PB as based on experimental observations and are strongly recommended for consideration for each employed model.
(i) Measurement of PB in the experimental medium as well as determination of bacterial growth compared to standard growth media is considered mandatory. Investigations without determination of protein binding must be considered highly speculative.
(ii) For animal models of PB, it is critical to determine binding in that animal species.
(iii) When PB in humans is investigated, the use of standardized and well-characterized pooled human serum and albumin is strongly preferred. The methodology should state whether the albumin preparation is fatty acid free or not.
(iv) To increase the power of an in vitro model to detect impact of PB, concentrations near the MIC of the investigated strain and, if MIC is the endpoint, a series of arithmetic dilutions should be used.
(v) Bacteria with appropriate and well-defined growth in the selected medium should be employed.
(vi) When less sensitive methods, such as MIC determination, detect no impact of protein binding, these results should be confirmed with the selected strains studied using time-killing curves.
(vii) Binding to equipment (plates, vials) should be investigated where appropriate.
(viii) Drugs with higher protein binding typically display not only higher modifications of antimicrobial action in the presence of protein but they also show higher susceptibility to all factors that may influence PB, including selection of added proteins, pH, electrolytes, temperature as well as others that have been previously discussed in this paper. In order to be able to extrapolate data from various models to in vivo situations, models should always attempt to mimic physiological conditions as closely as possible.
ACKNOWLEDGMENTS
We thank all members of the International Society of Anti-Infective Pharmacology (ISAP) and members of the ESCMID PK/PD of Anti-Infectives Study Group (EPASG) who provided valuable advice and discussions during the yearly scientific meetings.
- Copyright © 2011, American Society for Microbiology
REFERENCES
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
- 84.↵
- 85.↵
- 86.↵
- 87.↵
- 88.↵
- 89.↵
- 90.↵
- 91.↵
- 92.↵
- 93.↵
- 94.↵
- 95.↵
- 96.↵
- 97.↵
- 98.↵
- 99.↵
- 100.↵
- 101.↵
- 102.↵
- 103.↵
- 104.↵
- 105.↵
- 106.↵
- 107.↵
- 108.↵
- 109.↵
- 110.↵
- 111.↵
- 112.↵
- 113.↵
- 114.↵
- 115.↵
- 116.↵
- 117.↵
- 118.↵
- 119.↵
