THEUS This is to certify that the thesis entitled Comparison of pharmacokinetics, nehprotoxicity, and in vitro antibacterial activity of gentamicin administered once versus three times daily to adult horses presented by Leanne Mary Godber has been accepted towards fulfillment of the requirements for MS. degree inLangLAnjmal Clinical Sciences Major professor Datem 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution 1 MUCHIGAN SYATE UNIVERSITY LIBRARIES 3 1293 LIBRARY Michigan State University PLACE IN RETURN BOXto roomov this checkout from your record. TO AVOID FINES return on or before date duo. DATE DUE DATE DUE DATE DUE l l—Tl kl MSU Is An Affirmative Action/Equal Opportunity Institution cflckaS-nt COMPARISON OF PHARMACOKINETICS, NEPHROTOXICITY, AND IN VITRO ANTIBACTERIAL ACTIVITY OF GENTAMICIN ADMINISTERED ONCE VERSUS THREE TIMES DAILY TO ADULT HORSES By Leanne Mary Godber A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Large Animal Clinical Sciences 1994 ABSTRACT COMPARISON OF PHARMACOKINETICS, NEPHROTOXICITY, AND IN VITRO ANTIBACTERIAL ACTIVITY OF GENTAMICIN ADMINISTERED ONCE VERSUS THREE TIMES DAILY TO ADULT HORSES By Leanne Mary Godber Several characteristics of aminoglycoside antibiotics suggest that once daily administration may be an effective therapeutic regimen. An important characteristic is the post-antibiotic effect (PAE). defined as the persistent suppression of bacterial growth following removal of the antimicrobial agent. This study was designed to determine serum and tissue concentrations achieved following the intravenous administration of gentamicin at dose rates of 6.6 mglkg once daily and 2.2 mglkg three times daily for 10 days in six adult horses. The nephrotoxicity of these two dosing regimens was also compared. In vitro PAES were also determined with various concentrations of gentamicin using a test organism of equine origin. Results provide evidence to support the use of once daily gentamicin therapy in adult horses. Evidence of nephrotoxicity was not produced with either dosage regimen. I dedicate this work to Prof. D. R. Hutchins, my teacher, mentor and friend, whose knowledge and enthusiasm for veterinary science have always been an inspiration to me. ACKNOWLEDGMENTS I wish to thank my graduate committee members, Drs. Frederik Derksen, Robert Walker, Gary Stein, and Joe Hauptman for their guidance and assis- tance. I would also like to thank Ryan Robison and Mary Erbele for technical assistance. I also wish to thank my clinical advisor Dr. Chris Brown and my mates Drs. Jonathan Lumsden, George Bohart, and Amanda Ragon, for their guidance and friendship throughout my residency. Finally, lwould like to thank my parents, whose love and support were never far away. TABLE OF CONTENTS LIST OF TABLES ......................................... vii LIST OF FIGURES ........................................ viii INTRODUCTION ......................................... 1 CHAPTER 1 LITERATURE REVIEW .......................... 3 The Post-Antibiotic Effect ...................... 4 Factors Affecting the PAE ................... 5 Mechanisms Involved in the PAE ............. 7 Methods Of Measuring the PAE ............... 8 In Vitm PAE Of Aminoglycosides .............. 10 In Vivo PAE of Aminoglycosides .............. 12 Host-Defense Mechanisms .................. 13 Tissue Antibiotic Concentrations ................. 14 Protein Binding ........................... 14 Lipid Solubility ........................... 14 Other Factors ............................ 15 Measurement Techniques ................... 16 Antibiotic Distribution into Tissue Sites .......... 18 Nephrotoxicity .............................. 20 Pathogenesis of Nephrotoxicity ............... 21 Effect Of Dosage Regimen .................. 23 Clinical Testing for Nephrotoxicity ............. 25 Host Effects on Selection of Dosing Regimen ....... 26 lmmunocompromised Host .................. 27 Host Differences in Pharrnacokinetic Variables ...... 29 Intermittent Aminoglycoside Dosage Regimens ...... 30 Characteristics of Aminoglycosides ............ 30 MIC of Equine Clinical Isolates ............... 31 Statement of Purpose ........................ 31 CHAPTER 2 MATERIALS AND METHODS ..................... 33 Part I: Determination of Serum, Bronchial Secretion, and Tissue Cage Fluid Concentrations and Nephrotoxicity Following Two Dosage Regimens of Gentamicin in Adult Horses .................. 33 CHAPTER 3 RESULTS CHAPTER 5 CONCLUSIONS LIST OF REFERENCES vi Experimental Animals .................... Drug Administration and Sample Collection . . . . Bioassay ............................. Pharrnacokinetic Analysis and Statistical Evaluation ............................ Part II: Comparison of In vitro Antibacterial Activity of Once with Three Times Daily Gentamicin .............................. Determination of MIC .................... Determination of PAE .................... Part I: Determination of Serum, Bronchial Secretion, and Tissue Cage Fluid Concentrations and Nephrotoxicity Following Two Dosage Regimens Of Gentamicin in Adult Horses ........ Part II: Comparison of In vitro Antibacterial Activity Of Once with Three Times Daily Gentamicin .............................. CHAPTER 4 DISCUSSION ............................... 33 ..34 36 52 Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 LIST OF TABLES Pharmacokinetic values of gentamicin on day 2, trial 1 ...... 42 Pharmacokinetic values Of gentamicin on day 10, trial 1 ..... 43 Pharmacokinetic values of gentamicin on day 2, trial 2 ...... 44 Pharmacokinetic values of gentamicin on day 10, trial 2 ..... 45 Ratio of peak concentrations from day 2, trial 1 to MIC of Pseudomonas aeruginosa ........................ 46 Serum BUN, creatinine, and urine GGT/urine creatinine concentrations ................................... 47 vii LIST OF FIGURES Figure 1 Mean (3; SEM) serum concentrations of gentamicin on day 2, trial 1 ................................ 48 Figure 2 Mean (1 SEM) tissue cage fluid concentrations of gentamicin on day 2, trial 1 ...................... 49 Figure 3 Mean (1 SEM) bronchial secretion concentrations of gentamicin on day 2, trial 1 ...................... 50 Figure 4 Growth curves for Pseudomonas aeruginosa following 1 h exposure to gentamicin at 2x, 4x, and 8x MIC. The growth control (GC) and residual antibiotic control (RAC) are also shown ........................... 51 viii INTRODUCTION Conventional dosage regimens for antimicrobial agents are designed to maintain concentrations of the drug in serum above the minimal inhibitory concentration (MIC) for the infecting organism for the entire dosage interval (Gilbert 1991). A post-antibiotic effect (PAE), defined as the persistent suppression of bacterial growth after removal of the antimicrobial agent, has been demonstrated for aminoglycosides. The presence Of a PAE raises the possibility of less frequent dosing, as the PAE inhibits bacterial regrth when drug levels decrease below the MIC. The PAE is concentration dependent, with a higher aminoglycoside concentration producing a longer PAE (Kapsunik et al 1988). Gerber and Feller-Segessenmann (1985) also demonstrated that the higher the aminoglycoside concentration, the more rapid the kill of susceptible organisms. Nephrotoxicity is a major concern with aminoglycoside therapy. Aminoglycoside uptake occurs on the brush border of proximal tubular cells, and this cortical uptake is saturable (Verpooten etaI1989). Sustained trough levels should therefore result in greater nephrotoxicity, than high peak levels of aminoglycosides. Aminoglycoside dosing regimens that result in high peak levels and low-to-undetectable trough levels in serum may therefore maintain 2 efficacy and attenuate the risk of toxicity (Gilbert 1991). High peak concentra- tions and low trough concentrations are characteristics of once-daily therapy. The first part of this study was designed to compare the pharmacokinetic variables associated with once-daily gentamicin therapy with the more conventional three times daily therapy, administered to adult horses. These variables were measured by determining gentamicin concentrations in serum, tissue cage fluid, and bronchial secretions. Gentamicin was administered for 10 days in order to compare the nephrotoxicity of once with three times daily dosing. In the second part of this study, the length of PAES produced with different concentrations of gentamicin against an equine clinical isolate was determined in vitro. Measurement of the PAE in vitro allowed assessment of antibacterial activity, without the effects of host—defense mechanisms. From the peak concentrations determined in vivo, the ratio of peak concentration to MIC of the test organism could be calculated. These ratios allowed some comparison with the in vitro PAES. By measuring the length of time that tissue gentamicin concentrations were maintained above the MIC of a Clinical test organism, plus an approximate duration of the PAE, data were generated to support once-daily therapy with gentamicin administered to adult horses. CHAPTER 1 LITERATURE REVIEW Current dosing regimens for antimicrobial drugs are designed to maintain drug levels above the minimum inhibitory concentration (MIC) for a particular pathogen, for most of the dosing interval (Gilbert 1991). This design is based on observations made over forty years ago, that the therapeutic action of penicillin was determined primarily by the total length of time that it remained at concentrations above the MIC for a particular organism (Eagle 1949). This Observation has been confirmed more recently, when it was shown that for the beta lactams, the time the serum levels exceeded the MIC was the most significant parameter determining efficacy against various pathogens (Vogelman et al 19883). For aminoglycosides, however, area under the concentration versus time curve (AUC) was the most important pharmacokinetic parameter determining efficacy. This difference is due to the different intrinsic microbiologic activities of these classes of antimicrobial agents. In this literature review, I will first discuss the PAE, including the factors that affect its presence and duration, the mechanism of production of the PAE with aminoglycosides, and methods of measuring the length of the PAE. Previous studies measuring the PAE in vitro and in vivo will be summarized, and 4 factors leading to a difference between these two environments will be discussed, particularly the effect of host-defense mechanisms. I will then discuss factors affecting tissue distribution of antimicrobial agents, followed by a review of tissue sampling sites. A discussion of the mechanisms of amino- glycoside nephrotoxicity, the effects of dosage regimen, and clinical testing of nephrotoxicity will follow. I will then review the effects of host-defense mechanisms and pharrnacokinetics on dosage regimen. Finally, there will be a discussion of the use of intermittent dosage regimens with aminoglycosides. The Post-Antibiotic Effect It was recognized many years ago that many bacteria failed to remultiply immediately following removal of penicillin (Eagle 1949). This phenomenon was later termed the post-antibiotic effect (PAE), and is defined as the persistent suppression of bacterial growth after removal of the antimicrobial agent. The effect has been shown to be due to prior antimicrobial exposure, rather than to persisting subinhibitory concentrations of drug (Craig & Gudmundsson 1986). The potential clinical value of the PAE for any antimicrobial agent was not pursued however, because of the assumption that it was necessary to maintain serum antibiotic concentrations above the MIC for the infecting pathogen for the entire dosing interval. It is now recognized that the PAE may have clinical significance and it has been suggested that this phenomenon will have a major impact on antimicrobial dosing regimen (Craig & Gudmundsson 1986). The PAE may allow less 5 frequent dosing, where serum and tissue drug levels fall below the MIC for considerable periods of time, without bacterial regrowth and loss of drug efficacy (Vogelman et al 1988b). The presence or absence of a PAE provides a theoretical rationale for either intermittent or more continuous dosing of antimicrobial agents (Vogelman et al 1988b). Factors Affecting the PAE Many different factors pertaining to the microorganism, the antimicrobial agent, and the experimental conditions can affect the duration or even the presence of the PAE. The most important variables are the type of microorgan- ism and the class of antimicrobial (Craig & Gudmundsson 1986). The effect of drug concentration is also important, as a measurable PAE is only observed at concentrations at or above the MIC. Increases in concentration prolong the PAE (Craig & Gudmundsson 1986) up to a point of maximal response (Bundtzen, Gerber, Cohn & Craig 1981). For most antimicrobial-organism combinations that readily exhibit a PAE, the maximal response is observed at a concentration of approximately ten times the MIC (Vogelman & Craig 1985). Lengthening the antimicrobial exposure time has also been shown to prolong the duration Of the PAE. In general, the prolongation in the PAE resulting from a doubling of the exposure time is very similar to that following a doubling in concentration (Vogelman & Craig 1985). Area under the concentra- tion versus time curve (AUC) incorporates both concentration of drug and 6 duration of exposure. The length of the PAE has also been shown to depend on the AUC (Vogelman er‘ al 1988a). Other factors that are thought to affect the duration Of the PAE include size of inoculum, type of medium, growth phase Of organism at the time Of exposure, mechanical shaking of culture, pH of medium, antimicrobial combina- tions, and host-defense mechanisms (Craig & Gudmundsson 1986; Craig & Vogelman 1987). Host—defense mechanisms have a significant effect on measured PAE and will be discussed later. The effect of change of inoculum of 1 log with erythromycin and penicillin against Staphoncoccus aureus was shown to be minimal (Craig & Gud- mundsson 1986). The precise role Of the size of the initial inoculum has not been evaluated sufficiently to allow any conclusions about its effect (MacKenzie & Gould 1993). The influence Of type of medium was investigated by Bundtzen et al (1981) who used serum as a medium, and reported that at similar multiples of the MIC the duration of the PAE was nearly identical in serum and broth. This was disputed by Davidson, Zhanel, Phillips, and Hobran (1991) who reported an increase in the PAE of gentamicin from 2.2 h in Mueller-Hinton Broth to 3.8 h in serum, an increase of approximately 70%. Heat-inactivation of serum caused some reduction in this PAE. This suggests that complement or some other heat- labile component may have an important role in the PAE, along with other more stable components (Davidson etal1991). 7 Most studies have been performed with organisms in the logarithmic phase of growth. The small amount of evidence available, however, suggests that duration of the PAE is similar in lag-phase and log-phase of growth of the organism (Gerber, WIprachtiger, Stettler-Spichiger & Lebek 1982; Craig & Gudmundsson 1986). The effects of mechanical shaking, pH of medium, and specific antimicrobial combinations on the PAE have not been studied sufficiently to draw any significant conclusions. Mechanisms Involved in the PAE The precise mechanisms by which antimicrobial agents induce a PAE is unknown. The observed differences in PAES of various drug-organism combinations suggest that multiple mechanisms are involved (Craig 8 Gud- mundsson 1986). It also appears that the PAE is not due to a shift in the population in test cultures to slower growing variants as the growth curves of the test and control cultures are parallel in the post-PAE phase. A more likely explanation is non-lethal damage produced by the antimicrobial agent (Craig & Gudmundsson 1986). Aminoglycoside antibiotics irreversibly attach to 308 and 508 bacterial ribosomal subunits resulting in cell death (Ristuccia & Cunha 1982). In susceptible bacteria this permanent attachment causes disruption of the ribosomal subunits and "misreading" of the information on the messenger RNA. This "misreading" leads to synthesis by the bacterium Of "silent" or "nonsense" proteins. These may be "enzymes" without catalytic activity or "structural 8 proteins" that do not fit anywhere in the bacterium. The inability of the bacterium to synthesize the functional proteins it requires leads to death of the bacterium (Tobin 1979; Conzelman 1980). The PAE may represent a period Of resynthesis of these ribosomal proteins (Craig 8. Gudmundsson 1986). Methods of Measuring the PAE To measure the duration of the PAE in vitro, the growth kinetics Of a drug- exposed organism are compared with that of an untreated control (Craig & Vogelman 1987). Investigators have mainly used viable counts (CFU/ml) to follow microbial growth kinetics after drug removal. The viability curves following drug removal do not always reflect a flat stationary period followed sharply by logarithmic growth. Often there is a gradual increase in viable counts until normal growth occurs. Occasionally, there can be even a further decrease in the number of organisms before growth becomes stationary. In order to account for this variation, an equation was developed for quantitation of the PAE (Bundtzen et al 1981): PAE=T-C where T is the time required for the count of CFU in the test culture to increase 1 log1° above the count observed immediately after drug removal, and C is the time required for the count of CFU in an untreated control culture to increase 1 log10 above the count Observed immediately after completion of the same 9 procedure used on the test culture for drug removal. What is really being mea- sured is the time to reach logarithmic growth. Exposure of microorganisms to a constant concentration Of an antimicrobi- al agent differs considerably from the in vivo situation in which organisms are usually exposed to fluctuating levels of drug (Craig & Gudmundsson 1986). A simple method for rapid drug removal in vitro is dilution (Bundtzen et al 1981). In this method a small volume Of the antimicrobial agent-exposed culture is added to a large volume of fresh drug-free media. The extent of dilution needs to be large enough so that the drug combination in the diluted culture fails to affect the growth of control organisms. At concentrations near the MIC, a dilution of 1 in 100 is sufficient (Bundtzen et al 1981). One potential drawback of the method Of using viable counts to monitor regrowth of antibiotic-exposed bacteria is that viable counts may exaggerate the bactericidal effect Of antibiotic and thus lead to an underestimation of the PAE. Viability depends not only on the numbers of surviving bacteria in the broth culture, but also on the ability of these surviving bacteria to form colonies on the agar plates. These bacteria may need a period of recovery before they can form colonies. This recovery is achieved more easily in broth than on the agar surface, thus there may be a PAE on the agar plates. This recovery phenome- non gives a false impression that there has been cell multiplication, and so can lead to an underestimation of the PAE in vitro (lsaksson, Nilsson, Mailer & Soren 1988) 1O lsaksson et al (1988) used an in vitro bioluminescent assay of bacterial ATP to measure the PAE. As the ATP Of dead but intact bacteria as well as surviving cells are registered by this method, estimates of post-exposure surviving bacteria are high, leading to an overestimation of the PAE. In spite of this, PAE results at low concentrations of aminoglycosides are consistent with the viable counting techniques. In Vitro PAE Of Aminoglycosides Aminoglycosides have been demonstrated to have a PAE in vitro. In many experiments with gentamicin using a viable counting procedure, maximal PAES could not be determined because the marked bactericidal activity of the drug reduced the numbers of CFU below the threshold of detection in the counting procedure. Bundtzen et al (1981) exposed Escherichia coli to various concentrations Of gentamicin. At two times the MIC, a persistent effect of 0.6 h was observed. However, values at high concentrations of gentamicin could not be determined. Bundtzen et al (1981) also observed a PAE of 1.6—2.6 h in all strains of Pseudomonas aeruginosa exposed for 30—60 min to gentamicin at five to six times the MIC. Using viable counting procedures, Gerber et aI (1982) demonstrated an in vitro PAE of 1.4-1.9 h, after exposure of P. aeruginosa to gentamicin at five times the MIC for 30 mins. In vitro PAEs for amikacin after 1 h exposure against P. aeruginosa varied from 1.3 h at two times the MIC to 3.3 h at eight times the MIC (Craig, Redington & Ebert 1991). Renneberg and Walder (1989) showed 11 a PAE of 1.2 to 1.8 h with a concentration of ten times the MIC of amikacin for 2 h against P. aeruginosa. Kapsunik et al (1988) reported a PAE with tobramycin against P. aeruginosa of 1.5 h with approximately eight times the MIC. No PAE was produced at concentrations of the MIC and one-eighth of the MIC. In a study using serum collected after netilmicin administration against strains of 5 different bacteria, including P. aeruginosa, the duration of the in vitro PAE measured by a viable counting technique depended on the strain of bacteria tested (Van der Auwera et a! 1991). After 30 min of exposure several strains were inhibited for an overnight period (15 hours), whereas others were inhibited for only a few hours or less. For P. aeruginosa the duration of PAE after 30 min exposure time was 0—2.4 h after a 6.6 mglkg dose, and 0—2.2 h with sera collected 1 h after administration Of a 2.2 mglkg dose. Again, longer PAEs could be obtained by increasing the duration of exposure. The PAES recorded by lsaksson et al (1988) using the in vitro biolumi- nescent assay for gentamicin at two times the MIC against Pseudomonas strains were 2.7 to 4.3 h. At four times the MIC, the PAES increased to between 5.1 and 6.1 h. In addition, the bioluminescent method offered the possibility of determining the PAES at high aminoglycoside concentrations, which cannot be achieved by viable counts. The reason for this is that the bioluminescence method can detect viable cells that can produce ATP but are incapable of developing into colonies on solid medium. In these experiments, a dose- 12 dependent PAE was seen up to sixteen times the MIC for gentamicin against E. coli (lsaksson et al 1988). In Vivo PAE Of Aminoglycosides Studies have shown that the PAE is an in vivo phenomenon similar to that observed in vitro, with one major exception. The aminoglycosides produce longer PAES with gram-negative bacteria in vivo than in vitro (Vogelman & Craig 1985). This may not represent a true difference but reflect the apparent decreased bactericidal activity of these drugs in vivo. This allows the bacteria to be exposed to higher concentrations for longer time intervals than can be studied in vitro (Vogelman & Craig 1985). The PAE measured in vivo is also probably artificially prolonged by the effect of low concentrations of aminoglycosides below the MIC remaining during the dosing interval (lsaksson eta/1988). Bacteria show increased susceptibility to sub-MIC concentrations of antibiotics during the PAE phase. This phenome- non has been named the post-antibiotic sub-MIC effect (Cars & Odenholt- Tornqvist 1993) and is defined as the long period of growth suppression Obtained when a low concentration (down to 0.1 x MIC) is added to bacteria previously exposed to a supra-inhibitory concentration. The mechanism behind this effect is unknown. Amikacin has been shown to cause a post-antibiotic sub- MIC effect against E. coli, but not against P. aeruginosa (Cars & Odenholt- Tornqvist 1993). This phenomenon may be significant in vivo where bacteria are exposed to gradually declining concentrations of antimicrobial agents. 13 Host-Defense Mechanisms Intact host defenses probably contribute significantly to the success of intermittent dosage regimens in vivo (Craig & Gudmundsson 1986). Host factors would likely enhance the period of time measured as the PAE, because organisms during the PAE are more susceptible to phagocytosis or killing by human neutrophils in vitro than are untreated bacteria (McDonald, Wetherall & Pruul 1981). This phenomenon has been termed post-antibiotic leukocyte enhancement. The PAE of gentamicin measured in vivo has been shown to be longer in normal mice than in neutropenic mice (Fantin et al 1991). Induction of neutropenia permits the study of microorganism-antimicrobial agent interactions in vivo, without the interference of a significant component of the host-defense mechanisms. In vitro pharmacokinetic models also simulate the treatment Of infection in the absence of host-defense mechanisms (Blaser 1991) In vivo PAES, recorded in a neutropenic mouse-thigh model, varied from up to 2.1 h with gentamicin against E. coli to 7.5 h against KIebsieIIa pneu- moniae and P. aeruginosa (Vogelman etal 1988b). Fantin et al (1991) found in vivo PAES for 15 clinical isolates of Enterobacteriaceee following a fixed dose of 8 mglkg of gentamicin in neutropenic mice to range from 1.4 to 7.3 h. The presence of neutrophils increased the PAE with 8 mglkg of gentamicin against K. pneumoniae approximately two-fold. 14 Tissue Antibiotic Concentrations The concentration Of antimicrobial drugs at the site of infection is not necessarily the same as that in serum. A variety of factors influence drug distribution including physicochemical characteristics of the drug such as protein binding, molecular size, lipid solubility, and electrical charge at pH 7.4 (Wong, Peirce, Goldstein & Hoeprich 1975; Pennington 1981). Route of drug adminis- tration, serum levels achieved, and characteristics Of the site of infection are other factors that influence drug distribution. Protein Binding A fixed proportion Of antibiotic molecules may be bound to albumin, the percentage being characteristic Of the particular drug and related to the concentration of albumin in the serum (Barza & Cuchural 1985). The capillary beds of most tissues contain pores large enough to admit substances of a molecular weight up to 1,000. Although the majority of antibiotics fall within this range, albumin is too big and is largely retained within the capillaries. Gentamicin is <30% protein bound in horses (Cummings, Guthrie, Harkins & Short 1990) and this will not markedly alter the levels of free antibiotic present or have a significant clinical effect (Peterson & Gerding 1980). Lipid Solubility Lipid solubility is the principal physicochemical property of an antimicrobial agent that influences the pattern and extent of distribution (Baggott 1980). Only 15 the non-ionized portion of the drug at pH 7.4 is lipid soluble, and antibiotics with good lipid solubility penetrate biological membranes. Ability to penetrate cell membranes allows distribution of the drug into intracellular tissue sites. Prokesch and Hand (1982) studied the uptake of radiolabeled antibiotics into peripheral blood polymorphonuclear leukocytes, representing a biological membrane. Gentamicin was found to be moderately lipid soluble with a ratio of cellular to extracellular concentration Of 0.8. Gentamicin should therefore be able to permeate cell membranes to some extent; however, its penetration is much less than an antibiotic with greater lipid solubility. Other Factors Route Of drug administration is an important factor determining tissue concentrations. As most drugs diffuse along a concentration gradient into tissues (Pennington 1981), higher peak serum concentrations may result in superior penetration into tissues (Noone et al 1974). Surface area to volume ratio of the compartment is also important, for as this ratio increases, the rate of change Of drug concentration inside the compartment in response to changes in the concentration gradient is more rapid (Barza & Cuchural 1985). Host-related factors may also affect antibiotic tissue penetration. The anatomical barriers may be damaged by inflammation and mechanical injury, leading to increased blood flow and enhanced membrane permeability to antibiotic molecules (Bergogne-Berezin 1981; Pennington 1981), or decreased 16 permeability due to excess mucus, cellular debris or later, fibrotic tissue (Pennington 1981). Measurement Techniques Direct measurement of antimicrobial concentration in interstitial fluid is seriously restricted by the inaccessibility Of this site. To circumvent this problem, an accessible interstitial fluid compartment can be created by subcutaneous implantation of a perforated tissue chamber (Clarke, Short, Usenik & Rawls 1989) Biochemical and cytological analyses have shown that tissue fluid accumulating within these chambers closely resembles interstitial fluid (Calnan, Pflug, Chisholm & Taylore 1972; Clarke et al 1989). By 40 days after implanta- tion, the cellular and Chemical constituents had stabilized enough to allow use of the model to study drug distribution (Clarke et al 1989). The major difference between interstitial fluid and tissue cage fluid is that the minute distances between capillaries and interstitial fluid enables more immediate balance between components than can exist in larger volume models constructed for fluid collection (Bergan 1981). Despite this difference, the tissue cage model is an accepted technique for measuring extravascular distribution of antibiotics (Bergan 1981). In the treatment of respiratory infections, it is important to be able to determine antibiotic levels at the site of bacterial infection in the lung. Bacteria can be present in the mucus layer covering the ciliated respiratory epithelium, 17 they may adhere to the epithelial cells, or they may invade them, leading to cell death and denudation Of the epithelium (Baldwin, Honeybourne & WIse 1992a). Bacteria may also invade further to the interstitial tissue in the lung. Potential antibiotic concentration sampling sites include sputum, bronchial secretions, bronchial mucosa, whole lung tissue, and collection of epithelial lining fluid and alveolar macrophages by use of bronchoalveolar lavage (BAL) (Baldwin et al 1992a) There are inherent methodological difficulties with measurement of antimicrobial agents at these sites. Although sputum can be collected in humans, it is usually contaminated with saliva leading to dilution of antibiotic (Pennington & Reynolds 1975; Honeyburne etal1988), or can be contaminated with blood, resulting in falsely elevated levels. Bronchial secretion samples in non-diseased animals are limited by the small volumes obtained, and blood contamination can occur following the trauma of multiple sampling. Bronchial mucosal biopsy specimens consist of at least two separate tissue compartments, the submucosa, which is composed of interstitial fluid containing capillaries, and the epithelium, in which the cell membrane Of epithelial cells constitutes an additional barrier. There is thus no differentiation between interstitial and cellular fluid in bronchial mucosal specimens (Baldwin et al 1992a). This is also true of whole lung tissue, which has the additional disadvantage of requiring post-mortem collection. Epithelial lining fluid recovered by BAL is difficult to quantify due to dilution, and quantification is necessary to determine antimicrobial agent 18 concentration. Measurement of antibiotic concentration in alveolar macrophages recovered by BAL also poses difficulties due to loss Of antibiotic during the lavage procedure, or failure of certain antibiotics to penetrate significantly into macrophages. Antibiotics may diffuse out of alveolar macrophages even if cells are immediately separated by centrifugation (Baldwin et al 1992a). Although all of these sites have disadvantages, bronchial secretion samples were used in this study as it is a technique that is non-invasive, practical to use in the horse, and samples could be obtained multiple times throughout the data collection period. Antibiotic Distribution into Tissue Sites Using a tissue cage fluid model, it has been shown that peak antibiotic concentrations in extravascular spaces are lower and accumulate concurrently or later than the peak in serum concentration (Chisholm, Waterworth, Calnan & Garrod 1973; Short et al 1987; Walker et al 1990; Halstead et al 1992). Antibiotics may also have a longer half-life in extravascular fluids, resulting in maintenance of tissue concentrations for a greater proportion of the dosing interval (Chisholm et al 1973). Gentamicin levels have been measured in interstitial fluid and it was found that persistent levels of aminoglycosides were present in interstitial fluid at times when no antibiotic was detectable in serum (Chisholm etal1973; Carbon, Contrepois & Lamotte-Berrillon 1978). In contrast, tissue concentrations of aminoglycosides in the mouse thigh appear to parallel serum concentrations closely (Vogelman et al 1988b). A 19 possible reason for this difference, may be that muscle tissue composing the mouse thigh has a high concentration of capillaries. This may result in drug concentrations in this tissue site more closely reflecting serum distribution. Elimination Of aminoglycosides from lung tissue in mice was found to be slower than serum (Leggett eta/1989). This resulted in active lung tissue levels after serum concentrations had declined to sub-MIC levels, and longer PAES being recorded in lung tissue (Craig, Redington & Ebert 1991). Taylor, Guyton, and Bishop (1965) showed that the permeability Of the overall pulmonary membrane is determined mainly by the permeability of the alveolar membrane and that this has permeability characteristics essentially the same as other cellular membranes. Ultrastructure studies show that epithelial cells are tightly apposed by numerous zonulae occludens (Murray 1986), therefore permeation of drugs through cell membranes is the dominant method of transport. Lipid-soluble drugs diffuse through the phospholipid bilayer, while water-soluble drugs enter by specific transport mechanisms. The barrier for drug penetration into bronchial secretions therefore consists of the relatively permeable bronchial capillary endothelium and also the relatively impermeable ciliated respiratory epithelium. Consequently, bronchial secretions are separated from serum by significant barriers and likely to have antibiotic concentrations very different from serum. Alveolar macrophages that reside in this environment also have a cell membrane that must be penetrated in order for drugs to reach the intracellular environment. This may be important as it is representative Of the site Of infection for intracellular pathogens. 20 The relevance of assessing drug concentrations in healthy lung has been questioned, as it is argued that concentrations in inflamed tissues may be different and that these may be more appropriate for the prediction of clinical efficacy. However, antibiotic levels in healthy lung tissue bordering these inflamed areas may be important for prevention of microbial spread. Also, measurement of antimicrobial agent concentration at inflamed sites of infection in the lung poses further methodological and ethical problems (Baldwin et al 1992b) In summary, the concentrations of antibiotic achieved in tissues should be measured as well as in serum, as tissue levels may better represent the concentrations present at the site of the bacterial infection. For clinical relevance, however, these antimicrobial concentrations must be evaluated in the context of the MIC of the drug for the infecting organism. Nephrotoxicity Traditionally, aminoglycoside daily dose has been divided into 2 or 3 doses. Such regimens were originally devised to maintain serum concentrations above the MIC for the infecting organism, and avoid high peak serum concentra- tions, which were feared to be toxic (Nordstrom etal1990). Nephrotoxicity has, however, been more closely associated with high trough concentrations, especially trough concentrations >2 pglml for gentamicin (Dahlgren, Anderson & Hewitt 1975), than with peak concentrations (Bennett et al 1979). This dissociation of transient high peak drug concentrations and nephrotoxicity is 21 consistent with the concept of aminoglycosides having a saturable transport mechanism into cells (Bennett et al 1979). Pathogenesis of Nephrotoxicity The pathogenesis of the nephrotoxicity of aminoglycosides is closely related to the renal handling, which includes an intracellular accumulation of the drug in the renal cortex (Verpooten et al 1989). Aminoglycosides are highly polar drugs, which are eliminated essentially entirely by glomerular filtration (Thompson & Powell 1980). After glomerular filtration, a small portion of the administered dose (approximately 5%) is reabsorbed from the renal proximal tubular lumen by pinocytosis (Silverblatt & Kuehn 1979). Acidic phospholipids in the brush border membrane have been identified as binding sites for aminoglycosides. The binding between gentamicin and phospholipids appears to be a charge interaction between cationic, polybasic gentamicin and the anionic, acidic phospholipids (Sastrasinh, Knauss, Weinberg & Humes 1982). Acidic phospholipids are integral components. of most other membrane systems; however, the kidney is more susceptible to toxic damage by aminoglycosides as the content of acidic phospholipids is higher in the kidney and the concentration of gentamicin exposed to the brush border membranes of the proximal tubules will be appreciably higher than serum levels (Sastrasinh et al1982) After binding to acidic phospholipids on the microvilli membranes, this area is then pinched off as a vesicle (Thompson 8: Powell 1980). Vesicles fuse 22 with lysosomes to form cytosegresomes (Thompson & Powell 1980). In the presence of aminoglycosides, Iysosomal phospholipase and sphingomyelinase are inhibited, resulting in accumulation of phospholipid material in the lysosomes and an increase in their size. Aminoglycosides accumulate to high concentra- tions in the renal cortex, several fold higher than in serum (Guiliano et al 1984). Above a threshold Of accumulation, aminoglycosides may labilise lysosomes, leading to the intracellular liberation of lysosomal hydrolases and large amounts of aminoglycoside, cell death, and tubular necrosis (Guiliano eta/1984; Tulkens 1986) In the presence of aminoglycoside, the normal cytosegresome functions are deranged, and endocytosed membranes are not digested but instead accumulate as "myeloid bodies." This was originally proposed by Kosek et al (1974) who interpreted these gentamicin-induced myeloid bodies as overloading of lysosomes with undegraded polar lipids. This has since been supported, as the development of that Iysosomal alteration in animals and humans occurs in parallel with a rise in phospholipid content Of the renal cortex (Guiliano et al 1984). Endocytosis, with the associated membrane recycling, is prominent in the same cells where aminoglycosides accumulate, therefore the inhibitory effect of these antibiotics on Iysosome-endocytic vacuole fusion may provide another plausible explanation for their cytotoxicity at the tubular level (Laurent, Kishore & Tulkens 1990). It should be stressed that the tubular damage caused by aminoglycosides only results in kidney dysfunction when it reaches a level such that it is no longer counterbalanced by renal tissue repair (Laurent et al 1990). 23 Although the damage is clearly to the proximal tubular epithelium, the clinical manifestation of importance is diminished glomerular filtration. Reduction Of glomerular filtration impairs the only excretory mechanism of aminoglycosides. Thus, toxicity results in drug retention that furthers toxicity—a positive feedback loop that probably causes most of the clinical nephrotoxicity Of aminoglycosides (Thompson & Powell 1980). Effect of Dosage Regimen If the length of exposure of the aminoglycosides to binding sites on the proximal tubule brush border is a primary determinant of renal accumulation (Aronoff, Pottraz, Brier & Walker 1983), the fraction of drug taken up by the proximal tubule cells would be higher when the drug is given by continuous infusion than by intermittent injections (Guiliano et al 1986). Gentamicin given twice daily accumulated more rapidly and to a greater concentration in the rat kidney than did the same total dose given once daily (Aronoff et al 1983). Frame, Phair, Watanakunakom, and Bannister (1977) reported that rabbits given once-daily gentamicin suffered less renal damage than if the same total dose was given three times daily. Long periods with a low Iuminal concentration of aminoglycoside may allow the cell time to remove aminoglycoside and recover from the cell injury (Wood et al 1988). Renal uptake of gentamicin, netilmicin, and amikacin is less in man when the daily dose is given in one large dose instead of divided doses. By dividing the dose, drug concentrations are maintained below the saturation level of the 24 binding sites, allowing more efficient uptake (Verpooten et al 1989; DeBroe, Verbist & Verpooten 1991). In contrast, tobramycin renal uptake is linearly related to serum concentration (De Broe et al 1991). Relatively less of the former three aminoglycosides accumulate in the kidney with increases in serum concentration (Pechere, Craig & Meunier 1991). Data from De Broe, Guiliano, and Verpooten (1986) also provides evidence that the uptake of gentamicin in the human kidney cortex is saturable and that the drug’s affinity for binding- uptake in proximal tubular cells is the highest of the four aminoglycosides (gentamicin, tobramycin, netilmicin, and amikacin) examined. Nephrotoxicity is also determined by the intrinsic propensity of the aminoglycoside to cause toxic injury to intracellular organelles (Kaloyanides & Pastoriza-Munoz 1980; De Broe et al 1986). For a given aminoglycoside, the risk of nephrotoxicity increases as the renal cortical concentration increases. The concentration of a drug in the renal cortex increases as a function of dose, frequency of drug administration, and duration of drug treatment until eventually a plateau is reached that reflects saturation of the transport process, or the achievement of a balance between uptake and efflux of drug form renal cortex (Schentag et al 1977). The present knowledge on the renal handling of aminoglycosides permits concluding that renal cortical concentrations are of predictive value for the development of nephrotoxicity only if they are measured early in the treatment, when necrosis has not yet taken place. Once cell necrosis occurs there is an abrupt decline in cortical drug concentrations that parallels cell shedding contain- 25 ing accumulated aminoglycoside into the lumen (Houghton et al 1976). This explains why several investigators have not found a correlation between cortical drug levels and the degree of nephrotoxicity (Reiner, Bloxham 8 Thompson, 1978; Bennett et al 1979). In these latter studies, the cortical drug levels were measured at the end of high-dose, prolonged treatments, a time when cortical aminoglycoside levels are of little value, as they represent the balance between uptake, storage, and drug release resulting from necrosis (Verpooten etaI1989). Clinical Testing for Nephrotoxicity Routine clinical tests of renal function are restricted to the determination of the concentration of endogenous substances in the blood known to be excreted by the kidney. Since these may remain normal even at greatly reduced glomerular filtration rate, such tests cannot be considered satisfactory (Malyusz 8 Braun 1981). An increase in blood urea nitrogen (BUN) or creatinine does not occur until at least 70% of renal function has been lost (Kohn 8 Chew 1987). Creatinine, BUN, and endogenous creatinine clearance are also indices of glomerular filtration, rather than tubular function (Finco 8 Duncan 1976). The regular turnover of tubular cells results in the presence of their enzymes in the urine. An increased rate of destruction of these cells results in increased urinary concentration of these enzymes (Brobst, Carroll 8 Bayly 1986). Enzymuria is a sensitive and reliable method of assessing acute renal tubular damage induced by gentamicin (Greco etaI1985). Gamma glutamyl transpepti— dase (GGT) is an enzyme found mainly in the brush border of proximal 26 convoluted tubule cells. It has a high molecular weight (126,000) which makes it unlikely to appear in urine as a result Of glomerular filtration (Crowell et al 1981). Increased urinary excretion of GGT may reflect accelerated turnover of brush border membrane (Kaloyanides 8 Pastoriza-Munoz 1980). Gamma glutamyl transpeptidase is the earliest marker of proximal tubular disease in equidae, an increase being seen up to 6 days before detection Of azotemia (Bayly, Brobst, Elfers 8 Reed 1986). Urine GGT elevations are primarily indicative of acute proximal tubular damage (Bayly et al 1986). Although urine GGT and creatinine may increase or decrease depending on factors affecting urine flow rate, such as water reabsorption or secretion by distal tubules and collecting ducts, the ratio is constant (Adams eta/1985). Urine GGT concentrations are expressed as a ratio of urine GGT (in IU) to urinary creatinine concentration (in mgIdL) according to the formula UGGT/UCr x 100 (Kohn 8 Chew 1987). The urine GGT/urine creatinine ratio was measured from random urine samples collected from 27 clinically normal adult horses and was found to be 10.52 i 4.78. Based on an assumed normal distribution, a ratio of < 25 (mean i 3SD) would be expected to include approximately 99% of the population Of healthy adult horses (Adams et al 1985). Host Effects on Selection of Dosing Regimen Despite the presence of a PAE with aminoglycosides there might be some therapeutic disadvantages to the long inter-dose interval and the resulting long period when serum concentrations fall below the MIC (Nordstrom et al 1990). 27 For certain strains of bacteria with short PAES, efficacy could be compromised by administering a drug too infrequently and thereby allowing organisms to resume growth after the PAE (Vogelman et al 1988a). The schedule of drug administration is likely to play an especially important role in the outcome Of serious infection in the immunocompromised host, where recovery from infection depends to a high degree on antibiotic therapy (Bakker-Woudenberg 8 Roosendaal 1988). Vogelman et al (1988a) states that doses should be given at intervals no greater than the duration of time that drug levels exceed the MIC plus the duration of the PAE, and this should be particularly pertinent in the immunocompromised host. lmmunocompromised Host The finding that aminoglycoside efficacy can be compromised when the dosing interval is too long is supported by studies Observing reduced efficacy for once-daily dosing compared with intermittent therapy against P. aeruginosa pneumonia in neutropenic guinea pigs. A single high daily dose of tobramycin was at least as effective as conventional intermittent closing in eradicating P. aeruginosa from the lungs of hosts with "normal" immune responses, that is, nonneutropenic guinea pigs (Kapsunik etal1988). However, when tobramycin was administered as a single high daily dose in neutropenic guinea pigs with Pseudomonas pneumonia, the regimen was not as effective, and significant bacterial regrth occurred at the end of the 24-h dosing interval. When intermittent therapy with mezlocillin, a beta-Iactam antibiotic, was added to the 28 tobramycin regimens, in this neutropenic model, single, high daily doses of an aminoglycoside plus mezlocillin were as effective as the conventional intermittent aminoglycoside plus mezlocillin combination therapy. Thus intermittent beta- lactam therapy allowed single, high daily doses of aminoglycosides to be used in the treatment of experimental Pseudomonas pneumonia, even in the immunocompromised host. Data from other animal experiments also argue against monotherapy with aminoglycosides given once daily in neutropenic individuals (Blaser 1991). Limited studies in neutropenic human patients show that single daily doses of aminoglycosides in combination with a beta-lactam were safe and effective (Meunier et a! 1991; VIscoli et a! 1991). The combination therapy may ensure antibiotic activity during the prolonged periods (more than 75% Of the time between administrations) when aminoglycoside concentrations are subinhibitory (\fIscoli et al 1991). While the results of some efficacy studies in non-neutropenic animals support once—daily dosing of aminoglycosides (Powell et al 1983; Herscovici et al 1988; Wood et al 1988; Campbell 8 Rosin 1992), others have not (Pechere, Letarte, Pechere 1987; Queiroz, Bathirunathan 8 Mawer 1987). This discrepan- cy may be due to the fact studies in laboratory animal models are limited in that the pharmacokinetic properties of these drugs in animals are often markedly different from those observed in man (Dudley 8 Zinner 1991). 29 Host Differences in Pharmacokinetic Variables In normal mice, amikacin has such a rapid rate of elimination that dosing intervals Of 12 and 24 h would exceed the duration of active tissue levels plus the length of PAE for strains of Enterobacten’aceae and allow for organism regrowth (Craig, Redington 8 Ebert 1991). In leukopenic mice with renal impairment, to simulate human pharmacoki- netics, duration of the PAE was increased as well as aminoglycoside half lives (Craig et al 1991). The increase in measured PAE was probably due to subinhibitory levels persisting for a longer time in renal-impaired mice, compared to those with normal renal function (Craig et al 1991). For mice with normal renal function, the length of time that serum levels exceeded the MIC was the major parameter correlating with efficacy for gram-negative organisms. For mice with impaired renal function, however, AUC and peak serum concentration correlated best with antibacterial efficacy (Craig et al 1991). The human clinical studies published to date support once-daily aminogly- coside therapy rather than conventional twice- or thrice-daily therapy in non- neutropenic patients using gentamicin (Prins et al 1993), amikacin (Maller, lsaksson, Nilsson 8 Soren 1988; Maller eta/1991; Tulkens 1991) and netilmicin (Tulkens 1991; Van der Auwera et al 1991; Sturrn 1989; Hollender et a! 1989; Ter Braak et al 1990). Most Of these trials were small (Mailer et al 1988; Sturrn 1989; Tulkens 1991; Van der Auwera et a! 1991) or more than one antibiotic active against gram-negative rods was used (Tulkens 1991; Malleretal1991; Ter Braak et aI 1990). 30 In human patients with neutropenia and various immunodeficiencies, studies published to date are inadequate to assess the efficacy of the once-daily dosage regimen (Pechere, Craig 8 Meunier 1991). Intermittent Aminoglycoside Dosage Regimens Characteristics Of Aminoglycosides Intermittent dosage regimens apply primarily to drug-organism combina- tions that exhibit a prolonged PAE (Craig 8 Gudmundsson 1986). Aminoglyco- sides produce a concentration-dependent PAE against common bacterial pathogens. Another property of aminoglycosides that supports intermittent dosage regimens is the first-exposure effect (Craig 1993). During the period of initial exposure there is down-regulation of subsequent uptake of drug in surviving organisms. These organisms are less susceptible to re-exposure to aminoglyco- sides than untreated bacteria. This effect favors a longer dosing interval where bacteria have time to recover before rechallenge with the drug. Aminoglycosides kill bacteria in a dose-dependent manner, that is, the higher the concentration the more rapid and complete the killing (Gerber 8 Feller-Segessenman 1985). Thus, high-dose, intermittent therapy should result in better bacterial killing. This is supported by a study of Moore, Lietman, and Smith (1987), who showed that the ratio of maximal peak serum concentration of an aminoglycoside to the aminoglycoside MIC of the infecting organism was a major determinant of the clinical response to therapy. 31 Less-frequent dosage administration therefore results in higher peak concentrations in serum relative to Mle for the infecting pathogens, which enhances antibacterial efficacy and provides a longer PAE (Gilbert 1991). The longer the PAE, the less frequently one would have to administer aminoglyco- sides and, in theory, the lower the risk of nephrotoxicity (Gilbert 1991). MIC of Equine Clinical Isolates Tobin (1979) states that most equine pathogens are inhibited by serum gentamicin concentrations of4 pg/ml. Baggot, Love, Stewart, and Raus (1986) using a broth dilution technique, report a range in MIC Of <0.195 to 1.56 pglml against 112 common equine pathogens. A higher value was found only with 2 Streptococcus species, with Mle of 3.125 pg/ml. The final selection of dosage regimen should be based on the microorganism susceptibility to gentamicin, that is, the MIC as well as the health status of the patient (Haddad et al 1985). Statement of Purpose From the above discussion, there are many characteristics of aminoglyco- side antibiotics that are compatible with once-daily dosing. These include the concentration-dependent PAE, rapid concentration-dependent bacterial killing, lower risk Of nephrotoxicity with less frequent administration and the first- exposure effect. Even if efficacy and toxicity associated with once-daily dosing were merely comparable with results from conventional divided dosing, the once- 32 daily regimen would be less costly with respect to drug administration, personnel time, and hospitalization, and more convenient (Nordstrom et al 1990). The purpose of this study was therefore to compare the tissue concentra- tions achieved with once-daily dosing and three times daily dosing of gentamicin in adult horses. The concentrations achieved in vivo could then be related to in vitro PAES produced by various concentrations of gentamicin. This allowed some comparison of the antibacterial activities of these two dosage regimens. Nephrotoxicity of the dosage regimens was also evaluated. The results of this study provided evidence to support once-daily therapy with aminoglycosides in adult horses. CHAPTER 2 MATERIALS AND METHODS Part I: Determination of Serum, Bronchial Secretion, and Tissue Cage Fluid Concentrations and Nephrotoxicity Following Two Dosage Regimens of Gentamicin in Adult Horses Experimental Animals Six, 3- to 10-year-old healthy horses of mixed breeding (2 Arabs, 2 Thoroughbreds and 2 Standardbreds) and sexes (1 mare and 5 geldings), weighing 455.5 1 50.9 kg (mean t SD), were used. Prior to admission to the study, all horses had a normal complete blood count, fibrinogen and Chemistry profile, urinalysis and urine gamma-glutamyl transpeptidase (GGT) to urine creatinine ratio. Vlfith horses under general anesthesia, 7 cylindrical silicone tissue cages (1 cm in diameter by 5 cm long) were evenly spaced and placed subcutaneously in the left neck region. Six days later, seven additional tissue cages were placed in the right neck region, to give a total of 14 tissue cages per horse. Tissue cages were made from silastic tubing as described by Chisholm et al (1973). Surgical wounds were allowed to heal for at least 54 days before beginning the study. On the first treatment day, an indwelling 16 g x 13.5 cm 33 34 long catheter‘ was placed in a jugular vein, for antibiotic administration. Catheters were flushed with 10 ml of heparinized saline every 6 hours throughout the dosing regimen. Drug Administration and Sample Collection The experiment used a cross-over design with gentamicin administered intravenously at a dose of 2.2 mglkg every 8 hours for 10 days to three horses, and at a dose Of 6.6 mglkg every 24 hours for 10 days to three additional horses (Trial 1). A four-week interval than elapsed, before the trial was repeated with each horse receiving the other dosing regimen (Trial 2). Serum and tissue cage fluid (TCF) gentamicin concentrations were collected at 0, 0.25, 0.5, 0.75, 1, 2, 4, 6, and 8 hours following the fourth and thirtieth doses of gentamicin when administered using 2.2 mglkg, and at 0, 0.25, 0.5, 0.75, 1, 2, 4, 6, 8, 10, and 12 hours after the second and tenth close when administered using 6.6 mglkg. The TCF samples were collected, using a separate cage for each sample time and alternating from side to side between sample collection times. Each cage was used only once in each trial. Serum was obtained by collecting 5 ml of blood from the jugular vein that was not catheterized. The blood was allowed to clot for 1 hour at room temperature, and was then centrifuged for 10 minutes at 1,000 x g at 4°C. Bronchial secretions (BS) were collected at 0, 1, 2, 4, 6, and 8 hours after the fourth and thirtieth doses when administered using 2.2 mglkg and at 0, 1, 2, ° L-cath, Luther Medical Products, Inc. USA. 35 4, 6, 8, 10, and 12 hours after the second and tenth dose when administered using 6.6 mglkg. Bronchial secretions were collected by passing a flexible plastic tube (12 mm external diameter; 8 mm internal diameter) through the nose and into the trachea. An endoscopic biopsy instrument, with absorbent cotton held in the forceps, was then passed through this tube to the distal end of the tube, but not beyond, to remove secretions encountered during installation. The biopsy instrument was then withdrawn, and the cotton removed and replaced with a pre-weighed, sterile paper disk”. The biopsy instrument was then reinserted into the plastic tube past the distal end, until resistance was encountered. The disk was allowed to remain in contact with the respiratory epithelium for 60 seconds and was then withdrawn. If the disk was still dry, it was replaced for an additional 60 seconds. At each sampling period, three disks were inserted, one immediately after the other. The disks were cleaned of any gross debris and weighed, prior to freezing. The serum, tissue cage fluid, and bronchial secretion samples were frozen at -70°C, until assayed. On days 3, 6, and 10, blood was collected for determination of BUN and creatinine serum concentrations. On these days, urine was also collected after all other sampling had been completed. Urine was collected by catheterization of the bladder and if sedation was required, 15 mg acepromazine was given intravenously. Urinalysis was performed and a urine GGT to urine creatinine ratio calculated according to the method of Kohn and Chew (1987). b Difco Laboratories, Inc., Detroit, Michigan 36 Bioassay After thawing to room temperature, serum, TCF and BS samples were assayed for gentamicin concentration by use of a microbiological agar diffusion assay (Edberg 1986). Staphylococcus simulans (ATCC 27851) was used as the test organism. All samples were assayed in triplicate against standards prepared in equine serum. The lower limit Of sensitivity Of the assay was 0.02 pglml. Pharmacokinetic Analysis and Statistical Evaluation Data were analyzed using standard computer software for phannaco- kinetic analysis“. Gentamicin serum concentrations of each individual horse were fitted to a monoexponential curve, while the TCF and BS concentrations were fitted to a biexponential curve. These curves were analyzed to determine peak concentration (CW), time to peak concentration (Tm), elimination half-life (T1 ,2), and area under the concentration versus time curve (AUC) for 0 to 8 hours for the 2.2 mglkg dose and for 0 to 12 hours for the 6.6 mglkg dose. The serum CW was defined as the gentamicin concentration measured at 0.25 h after drug administration. There were 5 factors that influenced each of these 4 variables: time (trial 1 versus trial 2), day (day 2 versus day 10), dose (2.2 mglkg versus 6.6 mglkg), site (serum, TCF and BS), and individual horse. Data were analyzed, using a 5-factor analysis of variance. ° R Strip ll, Micromath Scientific Software, Salt Lake City, Utah 37 The BUN and creatinine serum concentration and urine GGT to urine creatinine ratio were influenced by 4 factors: time (trial 1 versus trial 2), day (day 2 versus day 10), dose (2.2 mglkg versus 6.6 mglkg), and individual horse. These data were analyzed, using a 4-factor analysis of variance. The probability of type-l error (P) was reported. For factors that were found to be significant (P < 0.05), a Tukey’s test was used to compare group means. Part II: Comparison of In Vitro Antibacterial Activity of Once with Three Times Daily Gentamicin Determination Of MIC An equine clinical isolate of P. aeruginosa (#991345) was used as the test organism. The MIC of P. aeruginosa was determined by the microtitration broth- dilution technique (Jones eta/1985). For the MIC determination, 50 pL of 2-fold dilutions of gentamicin were added to each well of a microtiter plate. One well in each row contained only Mueller-Hinton (MH) broth and served as an inoculation and growth control. A 24-hour-old culture was subcultured to MH broth and incubated for 2—5 hours, then diluted in fresh MH broth to a density Of approximately 10" colony-fanning units (CFU)ImI, as compared to a 0.5 McFarland Standard. This suspension was further diluted to 10s CFU/ml in MH broth. A semiautomatic inoculator was used to deliver 50 pL Of the 105 CF Ulml suspension to each well containing 50 pL of gentamicin suspension or broth (final concentration of bacteria, 5 x 10‘ CFU/ml). The MIC was the lowest 38 concentration of gentamicin for which no visible growth was observed after 18 hours Of incubation at 37°C. Determination of PAE The PAE of gentamicin against P. aeruginosa #991345 was determined using the procedure described by Craig and Gudmundsson (1986). A culture in logarithmic growth phase was Obtained by diluting an overnight culture 1:10 into MH broth. The bacteria were allowed to grow for several hours in a shaking water bath at 60 rpm and 37°C, until the optical density was equivalent to a 0.5 Macfarland Standard (approximately 10° CFU/ml). This suspension was then diluted to 106 CFU/ml in MH broth. One ml of the 106 CFU/ml suspension was then added to 9.0 ml of antimicrobial agent-containing MH broth, resulting in a final inoculum Of approximately 105 CFU/ml. Gentamicin was evaluated at concentrations of 2x, 4x, and 8x the MIC of P. aeruginosa, and all experiments were performed in a shaking water bath set at 60 rpm and 37°C. After 1 h of exposure, antimicrobial agents were effectively removed by 1:100 dilution of the test medium into prewarrned MH broth. The following controls were included: (1) a growth control prepared and treated in a fashion similar to that for the test solution, but without exposure to antibiotic and (2) a residual antibiotic control containing 1/100 Of the test antibiotic concentration. Counts of CFU for the test cultures and growth controls were determined prior to exposure to antimicrobial agents, after 1 h of exposure, and after dilution. Thereafter, all cultures, including the residual antibiotic control, were assessed 39 for growth every hour, until marked turbidity was noticed. Serial 10-fold dilutions were made with phosphate-buffered saline solution (0.01 M, pH 7.2) and aliquots of 0.01 ml were plated by spot plate technique and incubated overnight. The PAE was determined as previously described (Bundtzen et al 1981). PAE determinations were repeated three times on separate days. In preliminary experiments it proved difficult to accurately count colonies below 102 CFU/ml. Therefore, for colony counts below 1OZCFUIml, a modified pour plate technique was used. Wrth this technique, 0.1 ml aliquots were spread over each plate. PAE determinations were repeated twice on separate days. CHAPTER 3 RESULTS Part I: Determination of Serum, Bronchial Secretion and Tissue Cage Fluid Concentrations and Nephrotoxicity Following Two Dosage Regimens of Gentamicin in Adult Horses The AUC was influenced by the effect of time, with AUC for trial 1 being less than trial 2 (P < 0.001). The effect of dose and site was also significant, with 2.2 mglkg having a lower AUC than 6.6 mglkg (P < 0.001), and AUC for bronchial secretions being less then serum or TCF (P < 0.001). The difference in AUC between serum and TCF was not significant. The CW was influenced by dose (P < 0.001), with the maximum concentration produced by a dose of 6.6 mglkg being higher than with 2.2 mglkg. Site also had an effect (P < 0.001), with serum Cm being greater than tissue cage fluid or bronchial secretions. The difference in Cm between tissue cage fluid and bronchial secretions was not significant. The Tm was affected by time (P = 0.03), with Tm for trial 1 being significantly less than trial 2. Site was the only other significant factor (P < 0.001), with Tm for serum being less than for bronchial secretions or TCF. The T for bronchial secretions and TCF was not significantly different. max 40 41 The T,,2 was influenced by site only (P < 0.001), with serum Tm, being significantly less than bronchial secretion Tm, which in turn was significantly less than TCF T,,2. The mean and SEM values of AUC, Cm, Tm, and T1,2 for each sampling day of each trial are shown in Tables 1-4. The results of the urinalyses of all samples collected were within normal limits. The BUN and serum creatinine concentrations and urine GGT to urine creatinine ratios were not significantly affected by time, day, or dose (Table 6). Data from one horse was not included because urine could not be repeatedly collected due to technical reasons. The concentration curves of gentamicin in serum, tissue cage fluid, and bronchial secretions on day 2 of trial 1 are shown in Figures 1, 2, and 3. Part II: Comparison of In Vitro Antibacterial Activity of Once with Three Times Daily Gentamicin The MIC for gentamicin against Pseudomonas aeruginosa #991345 was 0.625 pglml. The PAES for 2x, 4x, and 8x MIC (1.25, 2.5, and 5 pglml) for gentamicin against P. aeruginosa were 0.25, 0.8, and 2.35 h respectively (Figure 4). Using Cm values determined in vivo on day 2, trial 1, the peak gentami- cin concentration to MIC ratios were calculated for all sites and doses (Table 5). 42 Table 1: Pharmacokinetic values‘ of gentamicin on day 2, trial 1 AUC (pg.h/ml) 0.... rug/ml) T... (h) T1,2 (h) AUC (pg.hlml) Cmx (pg/ml) Tm... (h) T1,2 (h) AUC (ugh/ml) cm, (pg/ml) Tmax (h) T112 ('1) SERUM 2.2 mglkg 6.03 1 0.66 4.16 1 0.38 0.25 1 0 0.83 1 0.03 TCF 10.14 1 0.45 2.31 1 0.10 1.43 1 0.33 1.85 1 0.19 1.12 1 0.12 0.34 1 0.05 0.73 1 0.31 2.45 1 1.37 BS 6.6 mglkg 83.77 1 15.00 61.34 1 14.88 0.25 1 0 0.83 1 0.13 33.85 1 5.18 6.91 1 1.06 1.40 1 0.27 3.31 1 1.21 12.54 1 3.88 5.08 1 1.85 0.65 1 0.17 1.18 1 0.25 * Values represent the mean 1 SEM, for 3 horses. 43 Table 2: Pharmacokinetic values" of gentamicin on day 10, trial 1 AUC (ugh/ml) max (pg/ml) Tmax (h) T1,2 (h) C Auc (pg.hlml) Cmax (pg/ml) Tm... (h) T112 0‘) AUC (pg.h/ml) cm, (pg/ml) Tm (h) T1,2 (h) SERUM 2.2 mglkg 20.18 1 1.38 11.54 1 1.10 0.25 1 0 1.14 1 0.18 TCF 29.38 1 1.85 5.57 1 0.47 0.96 1 0.67 4.39 1 0.85 BS 9.86 1 2.43 3.63 1 1.73 1.13 1 0.34 2.17 1 0.54 6.6 mglkg 50.21 1 4.89 26.50 1 1.73 0.25 1 0 1.10 1 0.09 40.99 1 1.93 6.24 1 0.43 1.23 1 0.22 4.72 1 0.52 12.24 1 1.60 3.21 1 0.39 0.94 1 0.24 1.64 1 0.50 * Values represent the mean 1 SEM for 3 horses 44 Table 3: Pharmacokinetic values* of gentamicin on day 2, trial 2 AUC (pg.h/ml) Cm (pg/ml) Tm, (h) T1,2 (h) AUC (pg.h/ml) Cm,x (pg/ml) Tm, (h) T1,2 (h) AUC (ugh/ml) Cm (pg/ml) Tmax (h) T1/2 (h) SERUM 2.2 mglkg 28.06 1 2.00 21.05 1 1.45 0.25 1 0 0.82 1 0.07 TCF 19.00 1 1.17 3.64 1 0.27 1.33 1 0.02 4.00 11.18 BS 4.91 1 1.01 1.25 1 0.10 1.28 1 0.24 1.58 1 0.45 6.6 mglkg 110.12 110.65 59.69 1 4.29 0.25 1 0 1.13 1 0.13 93.90 1 14.67 12.88 1 2.39 2.47 1 0.37 3.49 1 0.53 21.02 1 4.10 3.67 1 0.90 1.61 1 0.59 3.15 1 1.23 * Values represent the mean 1 SEM, for 3 horses 45 Table 4: Pharmacokinetic values* of gentamicin on day 10, trial 2 AUC (ugh/ml) Cm (pg/ml) AUC (pg.h/ml) C... org/ml) Tm (h) T112 ('1) AUC (ugh/ml) cm (pg/ml) SERUM 2.2 mglkg 20.91 1 0.74 16.02 1 1.34 0.25 1 0 0.75 1 0.08 TCF 31.89 1 0.86 6.66 1 0.60 1.16 1 0.14 3.72 1 0.87 BS 13.00 1 1.27 2.92 1 0.38 2.32 1 1.09 3.02 1 1.10 6.6 mglkg 82.51 1 0.90 39.82 1 3.58 0.25 1 0 1.16 1 0.08 70.44 1 9.48 9.64 1 1.71 2.33 1 0.42 4.32 1 0.40 45.73 1 5.98 11.76 1 2.58 0.66 1 0.47 2.64 1 0.97 * Values represent the mean 1 SEM, for 3 horses. 46 Table 5: Ratio of peak concentrations from day 2, trial 1 to MIC of Pseudo- monas aeruginosa CmaJMIC ratio Serum 2.2 mglkg 6.66 6.6 mglkg 98.14 TCF 2.2 mglkg 3.70 6.6 mglkg 11.06 BS 2.2 mglkg 0.54 6.6 mglkg 8.13 Table 6: Serum BUN, creatinine, and urine GGT/urine creatinine concen- trations* 47 day 0 BUN (mgldL) Creat (mgldL) UGGTIUCr day 3 BUN (mgldL) Creat (mgldL) UGGTIUCr day 6 BUN (mgldL) Creat (mgldL) UGGTIUCr day 10 BUN (mgldL) Creat (mgldL) UGGTIUCr 2.2mglkg 24.00 11.17 1.18 1 0.12 20.94 1 4.95 19.20 1 2.52 1.14 1 0.08 48.14 1 23.43 23.20 1 2.55 1.14 1 0.10 18.48 1 4.75 23.20 1 1.14 1.16 1 0.07 21.0 1 2.41 6.6mglkg 21.2 1 1.14 1.06 1 0.07 9.40 1 1.58 19.4 1 0.78 1.20 1 0.08 15.54 1 3.19 22.00 1 1.20 1.06 1 0.04 14.68 1 2.30 24.40 1 1.08 1.12 1 0.08 28.06 1 5.24 * Values represent the mean 1 SEM for 5 horses, combining trials 1 and 2 48 ‘ 2.2 mglkg 06.6 mglkg @650 ~— 6: 3 8 '15 40 r E O O c 820— 0 c O 0 2 4 6 8 10 12 Time (hours) Figure 1: Mean (1 SEM) serum concentrations Of gentamicin on day 2, trial 1. 49 10 — 92.2 mglkg 06.6 mglkg A 8 _ E e 3’ l a. 6 _ l C 4 O .9 e E c 8 2 - l . O O O W I l n I I 0 2 4 6 8 10 12 Time (hours) Figure 2: Mean (1 SEM) tissue cage fluid concentrations of gentamicin on day 2, trial 1. 50 \I I 1 0 2.2 mglkg 0 6.6 mglkg O) I 01 .h OJ Concentration (pg/ml) re —L O Time (hours) Figure 3: Mean (1 SEM) bronchial secretion concentrations of gentamicin on day 2, trial 1. 8" +GC +RAC +2XMIC 7 “ +4XMIC +8XMIC 6— E 5— / in" / a 4— w .2 / ‘13— 0'5 3 2— 1_ O_ v -1 o 1 2 3 4 5 6 Time (hours) Figure 4: Growth curves for Pseudomonas aeruginosa following 1 h exposure to gentamicin at 2x, 4x, and 8x MIC. The growth control (GC) and residual antibiotic control (RAC) are also shown. CHAPTER 4 DISCUSSION Results of this study provide evidence to support once-daily gentamicin therapy to treat infections in tissue sites equivalent to interstitial fluid. The time that gentamicin concentrations were above the MIC, plus the estimated duration of PAE, exceeded the 24 h dosing interval. Following three times daily administration of gentamicin the length of time that tissue gentamicin concentra- tions were above the MIC was 6 h. This length of time plus the estimated duration of PAE exceeded the dosing interval. This suggests that both gentamicin dosing regimens used in this study can be recommended in the treatment of tissue infections in the adult horse. However, once-daily dosing has the advantage of convenience, and thus improved compliance, reduced hospitalization, and administration costs. Gentamicin serum concentrations are not as easily interpreted. Although serum gentamicin concentrations were not maintained above MIC for as long as tissue concentrations, the peak gentamicin concentrations in serum were almost 100 times the MIC. High gentamicin concentrations result in a very effective bacterial kill (Gerber 8 Feller-Segessenman 1985) and a very long PAE (Craig 8 Gudmundsson 1986). Although the length of PAE after 100 times the MIC 52 53 concentrations of gentamicin has never been determined because of the very rapid bacterial kill, it may be long enough to allow once—daily dosing. Additional studies need to be performed before once-daily dosing can be recommended for equine septicemias. After once-daily administration, gentamicin concentrations in bronchial secretions were maintained above MIC for a shorter period of time than in tissue cage fluids. Furthermore the peak concentrations were lower than in tissue fluids, presumably resulting in a shorter PAE. Combining our in vivo and in vitro data, we estimate that the time above MIC plus the PAE was approximately 8 h. Therefore, once-daily dosing of gentamicin for airway infections cannot be recommended. Interestingly, the peak gentamicin concentration measured with the conventional three times daily dosing regimen failed to reach the MIC for the test isolate at any time during the dosing interval. A PAE was therefore not present. This suggests that gentamicin administered three times daily at 2.2 mglkg is also ineffective in the treatment of airway infections. There are several lines of evidence to suggest that the in vivo PAE produced is much longer than that measured in vitro. First, the peak serum and tissue gentamicin concentrations measured were higher than the concentrations used in vitro. It is known that the higher the antimicrobial agent concentration, the longer the PAE (Craig 8 Gudmundsson 1986). Second, the length of exposure to gentamicin in tissue sites was longer than the one hour of exposure used in the in vitro studies. Although the peak concentration was not maintained for longer than one hour in serum and bronchial secretions, concentrations 54 above the MIC were maintained for a considerably longer period of time. This is due to drug concentrations falling slowly in vivo, in comparison to the rapid removal of drug in vitro. As long as the concentration is above the MIC, it will contribute to the production of the PAE. Other factors operating in vivo that enhance the efficacy of once-daily dosing are postantibiotic leukocyte enhancement, the sub-MIC PAE, and the first-exposure effect. Postantibiotic leukocyte enhancement refers to the increased effectiveness of host defense mechanisms during the PAE (McDonald, Wetherall 8 Pruul 1981). As the PAE is longer with the higher once-daily dose, postantibiotic leukocyte enhancement is used to maximum advantage. During the PAE, sub-MIC concentrations of antibiotics can kill bacteria effectively (Cars 8 Odenholt-Tornqvist 1993). This is termed the sub-MIC PAE. After once daily dosing, the PAE is prolonged again, making maximum use of this phenomenon. Finally, once-daily dosing has the advantage of minimizing the first exposure effect. The first-exposure effect refers to the decreased susceptibility of bacteria following re-exposure to aminoglycosides (Craig 1993). The first-exposure effeCt is a disadvantage in frequent gentamicin dosing regimens. In this study, the conclusions regarding once-daily dosing are based on the MIC of the strain of P. aeruginosa used. The MIC for this organism was 0.625 pglml, and organisms with different Mle will have a different PAE. Can conclusions of this study be applied to common equine pathogens? Baggott et al (1986) reported that 100 of 114 isolates Of common equine pathogens had MIC values less than or equal to 0.625 pglml, as determined by a broth dilution 55 technique. This suggests that the conclusions reached in this study may be applicable in the majority of clinical cases. The estimates of duration of PAE are based on ratios of peak concentra- tions to the MIC. For serum concentrations, we defined peak concentrations as the concentration measured 15 minutes after intravenous injection of a bolus of drug. As such, the true peak concentrations was probably missed and was most likely higher. We used the 15-minute value, however, because this value is likely to be more representative Of the concentration of gentamicin organisms were exposed to over the first hour, then the true peak concentrations. Gentamicin pharmacokinetics were not significantly different between day 2 and day 10 of both trials. In contrast, the AUC was found to be significantly influenced by the effect of time, that is, the AUCs found in trial 2 were significantly greater than those in trial 1. This was not expected, as a 28-day washout period was allowed between the two trials. In spite of this washout period, gentamicin accumulation could have occurred and significant amounts of gentamicin may still have been present in tissues 28 days after the first 10-day course of drug administration. Similarly, tissue accumulation of gentamicin has been previously reported in sheep following a 24-day washout period (Brown, Coppoc, Riviere 8 Anderson 1986). The authors proposed that sequestration of gentamicin at tissue-binding sites was responsible for this prolonged tissue-washout phase. When deep tissue accumulation occurs, less drug leaves the serum to go into these tissue sites, and consequently more gentamicin remains in the serum for a longer time. Alternatively, the drug 56 remaining at these tissue-binding sites could contribute substantially to the amount of drug in the serum (Brown, Coppoc, Riviere 8 Anderson 1986). To explore deep tissue accumulation Of gentamicin further, sheep were killed following gentamicin administration using daily doses similar to those used in the present study (Brown, Coppoc 8 Riviere 1986). Tissue concentrations Of gentamicin were highest in the renal cortex, followed by renal medulla, liver, lung, spleen, and skeletal muscle, and were 230.2 pglmg, 9.22 pglmg, 4.28 pglmg, 2.50 pg/mg, 2.30 pglmg, and 0.673 pg/mg, respectively, following 6 mglkg of gentamicin daily for 3 days. If similar tissue accumulations occur in the horse, these concentrations would be sufficient to explain the differences in pharmacokinetics observed between trial 1 and 2. Gentamicin tissue accumula- tions, with a distribution similar to those observed in sheep, have been reported in people (Schentag et al 1977). However, tissue accumulation of gentamicin is inconsistent with the finding that gentamicin pharmacokinetics did not differ between day 2 and day 10 of either trial. Peak serum concentration are expected to be three times higher following 6.6 mglkg dose compared to the 2.2 mglkg dose. On day 2 of trial 1 we found a peak concentration of 4.16 pg/ml following a 2.2 mglkg dose and a peak concentration of 61.34 pg/ml following a 6.6 mglkg dose, which was approxi- mately five times higher than expected. The volume of distribution of gentamicin can be calculated using the formula of dose divided by the concentration at time 0 (Co), calculated by extrapolating the monoexponential serum curve back to time 0. Using this formula, the volume of distribution in our horses was 57 calculated to be a mean of 214 L following the 2.2 mglkg dose, and 56 L following the 6.6 mglkg dose. The smaller volume of distribution following the higher 6.6 mglkg dose is possibly due to saturation of deep tissue binding sites. Uptake into these deep tissue sites is thought to result from binding to acidic phospholipids, similar to uptake into the renal proximal convoluted tubular cells. Gentamicin uptake into deep tissue sites should therefore also be a saturable process. Thus, above a saturation point, no more drug will be taken up into these tissue sites, resulting in a smaller volume into which the drug can distribute, and a higher peak concentration. The Tm values for BS and TCF were not significantly different. In contrast, the TCF T1,2 was significantly longer than the T1,2 of BS, which in turn was longer than serum Tm. Similar findings have been previously reported (Leggett et al 1989). The Tm of tissues is expected to be longer than serum because of the barriers that the drug must cross to enter the tissue sites. Half- life is a reflection of elimination from sites, and interestingly TCF T1,2 was longer than T1,2 of BS. The difference in T,,2 between TCF and BS may be due to the pH of TCF and recent work by Clarke et al (1989) showed chamber fluid pH to be consistently lower than blood pH. Basic aminoglycosides become ionized in the acidic pH of TCF and only the non-ionized drug is lipid soluble. Therefore aminoglycosides are retained in sites with a low pH, such as TCF. A longer T1,2 can also be explained by the small surface area to volume ratio, characteristic of tissue cages. However, the small surface area to volume ratio should also result in a prolonged Tm, which we did not observe. Gentamicin TCF 58 pharmacokinetics in our study were similar to those reported by Chisholm et al (1973). Using either gentamicin dosing regimen, nephrotoxicity was not detected. The BUN and serum creatinine are insensitive indicators of proximal tubular injury, however, urine GGT to urine creatinine ratio and urinalysis are earlier markers. Based on these data, it may be concluded that gentamicin can be safely administered by either dosage regimen to healthy horses. However, this data must be interpreted with caution when applied to diseased animals. CHAPTER 5 CONCLUSIONS Results of this study provide direct evidence to support once-daily gentamicin therapy in tissue sites equivalent to interstitial fluid. The evidence to support use of once-daily therapy against bacterial infections in sites equivalent to serum or bronchial secretions is not as clearly defined. However, other recognized factors, not measured as part of this study, would most likely increase the antibacterial effects of gentamicin in these sites in vivo. These factors include the effects of peak concentration and duration of exposure of the drug on the PAE produced by aminoglycosides. Also the effect of host-defense mechanisms, the sub-MIC PAE effect, and avoidance of the first exposure effect, would contribute to the success of once-daily gentamicin therapy in these sites in vivo. Nephrotoxicity was also monitored in this study and no significant difference was found between once-daily and three times daily administration of gentamicin, over the ten-day treatment period. To further determine the effectiveness of once-daily gentamicin therapy, use of this dosing regimen in an in vivo infection model in both immunocompe- tent and immunocompromised horses would be required. A tissue infection 59 60 model may be created by injection of an equine pathogen into tissue cages. The infected tissue cage model has the disadvantage of representing an infection associated with a foreign substance within the body. This infection may therefore be difficult to resolve, due to bacteria residing in the protective environment of the biofilm covering the tissue cage. The advantage of the tissue cage site is that sampling allows quantification Of bacterial numbers, as well as determination of the antibiotic concentrations. Measurement of the antibiotic concentration over time will allow determination of the pharmacokinetics of gentamicin in this site. A respiratory infection may also be created by nebuliza- tion Of the pathogen into the respiratory tract, however, infection is difficult to induce in the normal horse. Determination of the pharmacokinetics of once-daily gentamicin in a diseased horse model would also be useful. It has been shown, for example, that following induction of endotoxemia the volume of distribution of gentamicin decreases and consequently the peak serum concentration increases (Vlfilson, Moore 8 Eakle 1983). Assessment of nephrotoxicity in such a model, or in a model in which dehydration is induced by water deprivation, may also allow better comparison of the nephrotoxicity produced by once-daily and three times daily regimens. The final stage in assessment of the effectiveness of once-daily aminoglycoside therapy would be its use in clinical infections. Successful completion of these further steps would provide proof that once-daily administra- tion Of gentamicin was an effective therapeutic regimen in the horse. LIST OF REFERENCES LIST OF REFERENCES Adams R, McClure JJ, Gossett KA, et al. Evaluation of a technique for measurement of gamma-glutamyltransferase in equine urine. Am J Vet Res 1985;46:147-150. Aronoff GR, Pottratz ST, Brier ME, et al. Aminoglycoside accumulation kinetics in rat renal parenchyma. Antimicrob Agents Chemother 1983;23:74-78. Baggott JD. Distribution of antimicrobial agents in normal and diseased animals. J Am Vet Med Assoc 1980;176:1085-1090. Baggot JD, Love DN, Stewart J, et al. Gentamicin dosage in foals aged one month to three months. Equine Vet J 1986;18:113-116. Bakker-Woudenberg IAJM, Roosendaal R. Impact of dosage regimens on the efficacy of antibiotics in the immunO-compromised host. JAntimicrob Chemother 1988;21:145-147. Baldwin DR, Honeyburne D, WIse R. Pulmonary disposition of antimicrobial agents: methodological considerations. Antimicrob Agents Chemother 1992a;36:- 1171-1175. Baldwin DR, Honeyburne D, WIse R. Pulmonary disposition of antimicrobial agents: in vivo observations and clinical relevance. Antimicrob Agents Chemo- ther 1992b;36:1176-1180. Barza M, Cuchural G. General principles of antibiotic tissue penetration. J Antimicrob Chemother 1985;15 Suppl A:59-75. Bayly WM, Brobst DF, Elfers RS, et al. Serum and urinary biochemistry and enzyme changes in ponies with acute renal failure. Cornell Vet 1986;76:306-316. Bennett WM, Plamp CE, Gilbert DN, et al. The influence of dosage regimen on experimental gentamicin nephrotoxicity: dissociation Of peak serum levels from renal failure. J Infect Dis 1979;140:576-580. 61 62 Bergen T. Pharrnacokinetics of tissue penetration of antibiotics. Rev Infect Dis 1981 ;3:45-66. Bergogne-Berezin E. Penetration of antibiotics into the respiratory tree. J Antimicrob Chemother 1981;8z171-174. Blaser, J. Efficacy of once- and thrice-daily dosing Of aminoglycosides in in-vitro models of infection. J Antimicrob Chemother 1991;27 Suppl 0:21-28. Brobst DF, Carroll RJ, Bayly WM. Urinary enzyme concentrations in healthy horses. Comell Vet 1986;76:299-305. Brown SA, Coppoc GL, Riviere JE. Effects of dose and duration of therapy on gentamicin tissue residues in sheep. Am J Vet Res 1986;47:2373-2379. Brown SA, Coppoc GL, Riviere JE et al. Dose-dependent pharmacokinetics of gentamicin in sheep. Am J Vet Res 1986;47:789-794. Bundtzen RW, Gerber AU, Cohn DL, et al. Postantibiotic suppression of bacterial growth. Rev Infect Dis 1981;3z28-37. Calnan JS, Pflug JJ, Chisholm GD, et al. Pathophysiology of tissue fluid, in Proceedings. Royal Society of Medicine 1972;65:715-719. Campbell BG , Rosin E. Optimal gentamicin dosage regimen in dogs (abstr). in Proceedings. American College of Veterinary Surgeons. 1992;27:5. Carbon C, Contrepois A, Lamotte-Barillon S. Comparative distribution of gentamicin, tobramycin, sisomicin, netilmicin and amikacin in interstitial fluid in rabbits. Antimicrob Agents Chemother 1978;13:368-372. Cars O, OdenhOIt-Tomqvist l. The post-antibiotic sub-MIC effect in vitro and in vivo. J Antimicrob Chemother 1993;31 Suppl D:159-166. Chisholm GD, Waterworth PM, Calnan JS, et al. Concentration of antibacterial agents in interstitial tissue fluid. Br Med J 1973;1z569-573. Clarke CR, Short CR, Usenik EA, et al. Subcutaneously implanted tissue chambers: a pathophysiological study. Res Vet Sci 1989;47:195-202. Conzelman GM. Pharmacotherapeutics of aminoglycoside antibiotics. JAm Vet Med Assoc 1980;10:1078-1080. 63 Craig WA. Post-antibiotic effects in experimental infection models: relationship to in-vitro phenomena and to treatment of infections in man. J Antimicrob Chemother 1993;31 Suppl D:149—158. Craig WA, Gudmundsson S. The postantibiotic effect. In: Lorian V, ed. Antibiotics in laboratory medicine. 2nd ed. Baltimore: WIlliams 8 Wilkins CO, 1986;515-536. Craig WA, Redington J, Ebert SC. Pharrnacodynamics of amikacin in vitro and in mouse thigh and lung infections. JAntimicrob Chemother 1991 ;27 Suppl C229- 40. Craig WA, Vogelman B. The postantibiotic effect. Ann Intern Med 1987;106:900- 902. Crowell WA, Divers TJ, Byars TD, et al. Neomycin toxicosis in Calves. Am J Vet Res 1981;42:29-34. Cummings LE, Guthrie AJ, Harkins JD, et al. Pharmacokinetics of gentamicin in newborn to 30-day-Old foals. Am J Vet Res 1990;51:1988-1992. Dahlgren JG, Anderson ET, Hewitt WL. Gentamicin blood levels: a guide to nephrotoxicity. Antimicrob Agents Chemother 1975;8258-62. Davidson RJ, Zhanel 66, Phillips R, et al. Human serum enhances the postantibiotic effect Offluoroquinolones against Staph aureus. Antimicrob Agents Chemother 1991;35:1261-1263. De Broe ME, Giuliano RA, Verpooten GA. Choice of drug and dosage regimen. Am J Med 1986;80 Suppl 6B:115-118. De Broe ME, Verbist L, Verpooten GA. Influence of dosage schedule on renal cortical accumulation of amikacin and tobramycin in man. J Antimicrob Chemother 1991;27 Suppl 0:41-47. Dudley MN, Zinner SH. Single daily dosing of amikacin in an in-vitro model. J Antimicrob Chemother 1991;27 Suppl 0:15-19. Eagle H. The recovery Of bacteria from the toxic effects of penicillin. J Clin Invest 1949;28:832-836. Edberg SC. The measurement of antibiotics in human body fluids: techniques and significance. In: Lorian V, ed. Antibiotics in laboratory medicine. 2nd ed. Baltimore: WIlliams 8 WIIkins CO, 1986;381-476. 34 Fantin B, Ebert S, Leggett J et al. Factors affecting duration of in vivo post- antibiotic effect for aminoglycosides against Gram- negative bacilli. J Antimicmb Chemother 1991 ;27:829-836. Finco DR, Duncan JR. Evaluation of blood urea nitrogen and serum creatinine concentrations as indicators of renal dysfunction: a study of 111 cases and a review of related literature. J Am Vet Med Assoc 1976;168:593-601. Frame PT, Phair JP, Watanakunakorn C, et al. Phannacologic factors associated with gentamicin nephrotoxicity in rabbits. J Infect Dis 1977;135:952-956. Gerber AU, Feller-Segessenmann C. In-vivo assessment Of in-vitrO killing patterns of Pseudomonas aeruginosa. J Antimicrob Chemother 1985;15 Suppl A:201-206. Gerber AU, errachtiger P, Stettler-Spichiger U, et al. Constant infusions vs. intermittent doses of gentamicin against Pseudomonas aeruginosa in vitro. J Infect Dis 1982;145:554—560. Gilbert DN. Once—daily aminoglycoside therapy. Antimicrob Agents Chemother 1991;35:399-405. Giuliano RA, Paulus GJ, Verpooten GA et al. Recovery of cortical phospho- lipidosis and necrosis after acute gentamicin loading in rats. Kidney Int 1984;26:838-847. Giuliano RA, Verpooten GA, Verbist L et al. In vivo uptake kinetics of aminogly- cosides in the kidney cortex of rats. J Pharmacol Exp Ther1986;236:470-475. Greco DS, Turnwald GH, Adams R, et al. Urinary gamma-glutamyl transpepti- dase activity in dogs with gentamicin-induced nephrotoxicity. Am J Vet Res 1985;46:2332-2335. Haddad NS, Pedersoli WM, Ravis WR, et al. Combined pharmacokinetics of gentamicin in pony mares after a single intravenous and intramuscular administration. Am J Vet Res 1985;46:2004-2007. Halstead SL, Walker RD, Baker JC, Holland GE et al. Pharmacokinetic evaluation of ceftiofur in serum, tissue chamber fluid and bronchial secretions from healthy beef-bred calves. Can J Vet Res 1992;56:269-274. Herscovici L, Grise G, Thauvin C, et al. Efficacy and safety of once daily versus intermittent dosing of tobramycin in rabbits with acute pyelonephritis. Scand J Infect Dis 1988;20:205-212. 65 Hollender LF, Bahnini J, De Manzini N, et al. A multicentric study of netilmicin once daily versus thrice daily in patients with appendicitis and other intra- abdominal infections. J Antimicrob Chemother 1989;23:773-783. Honeybourne D, Andrews JM, Ashby JP, et al. Evaluation of the penetration of ciprofloxacin and amoxycillin into the bronchial mucosa. Thorax 1988;43:715- 719. Houghton DC, Hartnett M, Campbell-Boswell M, et al. A light and electron microscopic analysis of gentamicin nephrotoxicity in rats. Am J Pathol 1976;82:589-599. lsaksson B, Nilsson L, Mailer R et al. Postantibiotic effect of aminoglycosides on Gram-negative bacteria evaluated by a new method. J Antimicrob Chemother 1988;22:23-33. Jones RN, Barry AL, Gavan TL, et al. Susceptibility tests: microdilution and macrodilution broth procedures. In: Lennette EH, Balows A, Hausler WJ, et al. eds. Manual of clinical microbiology. 4th ed. Washington: Am SOC Microbiol 1 985:972-977. Kaloyanides GJ, Pastoriza-Munoz E. Aminoglycoside nephrotoxicity. Kidney Int 1980;18:571-582. Kapusnick JE, Hackbarth CJ, Chambers HF, et al. Single, large, daily dosing versus intermittent dosing of tobramycin for treating experimental Pseudomonas pneumonia. J Infect Dis 1988;158:7-12. Kohn CW, Chew DJ. Laboratory diagnosis and characterization of renal disease in horses. Vet Clin North Am Equine Pract 1987;3;585-615. Kosek JC, Mazze RI, Cousins MJ. Nephrotoxicity of gentamicin. Lab Invest 1974;30:48-57. Laurent G, Kishore BK, Tulkens PM. Aminoglycoside-induced renal phospho- lipidosis and nephrotoxicity. Biochem Phannacol 1990;40:2383-2392. Leggett JE, Fantin B, Ebert S, et al. Comparative antibiotic dose-effect relations of several dosing intervals in murine pneumonitis and thigh-infection models. J Infect Dis 1989;159:281-292. MacKenzie FM, Gould lM. Review: The post-antibiotic effect. J Antimicrob Chemother 1 993;32:51 9-537. 66 Maller R, Ahme H, Eilard T, et al. Efficacy and safety Of amikacin in systemic infections when given as a single daily dose or in two divided doses. J Antimicrob Chemother 1991;27 Suppl C:121-128. Maller R, lsaksson B, Nilsson L, et al. A study of amikacin given once versus twice daily in serious infections. J Antimicrob Chemother 1988;22:75-79. Malyusz M, Braun D. Enzymuria (the output Of gamma-glutamyl-transpeptidase and of N-acetyI-beta-D-glucosaminidase) in the course of experimental renovascular hypertension. Enzyme 1981;26:32-42. McDonald PJ, Wetherall BL, Pruul H. Postantibiotic leukocyte enhancement: increased susceptibility of bacteria pretreated with antibiotics to activity of leukocytes. Rev Infect Dis 1981;3:38-44. Meunier F, Van der Auwera P, Aoun M, et al. Empirical antimicrobial therapy with a single daily dose of ceftriaxone plus amikacin in febrile granulocytopenic patients: a pilot study. J Antimicrob Chemother 1991;27 Suppl 02129-139. Moore RD, Lietman PS, Smith CR. Clinical response to aminoglycoside therapy: importance of the ratio Of peak concentration to minimal inhibitory concentration. J Infect Dis 1987; 1 55:93-99. Murray, J.F. The Normal Lung 2nd ed. Philadelphia:W.B.Saunders CO, 1986;299-302. Noone P, Parsons TMC, Pattison JR, etal. Experience in monitoring gentamicin therapy during treatment of serious gram-negative sepsis. Br Med J 1974;1:477- 481. Nordstrom L, Ringberg H, Cronberg S, et al. Does administration of an aminoglycoside in a single daily dose affect its efficacy and toxicity? J Antimicrob Chemother. 1990;25; 1 59-173. Pechere JC, Craig WA, Meunier F. Once daily dosing of aminoglycoside: one step forward. J Antimicrob Chemother 1991;27 Suppl C:149-152. Pechere M, Letarte R, Pechere JC. Efficacy Of different dosing schedules of tobramycin for treating a murine Klebsiella pneumoniae bronchopneumonia. J Antimicrob Chemother 1987;19:487-491. Pennington JE. Penetration of antibiotics into respiratory secretions. Rev Infect Dis 1981;3:67—73. 67 Pennington JE, Reynolds HY. Pharmacokinetics of gentamicin sulfate in bronchial secretions. J Infect Dis 1975;131:158-162. Peterson LR, Gerding DN. Influence of protein binding of antibiotics on serum pharmacokinetics and extravascular penetration: Clinically useful concepts. Rev Infect Dis 1980;2z340-348. Powell SH, Thompson WL, Luthe MA, et al. Once-daily vs. continuous aminoglycoside dosing: efficacy and toxicity in animal and clinical studies of gentamicin, netilmicin and tobramycin. J Infect Dis 1983;147:918-932. Prins JM, Buller HR, Kuijper EJ, et al. Once versus thrice daily gentamicin in patients with serious infections. Lancet 1993;341:335-339. Prokesch RC, Hand WL. Antibiotic entry into human polymorphonuclear leukocytes. Antimicrob Agents Chemother 1982;21:373-380. Queiroz MLS, Bathirunathan N, Mawer GE. Influence of dosage interval on the therapeutic response to gentamicin in mice infected with Klebsiella pneumoniae. Chemother 1987;33:68-76. Reiner NE, Bloxham DD, Thompson WL. Nephrotoxicity Of gentamicin and tobramycin given once daily or continuously in dogs. J Antimicrob Chemother 1978;4 Suppl A:85-101. Renneberg J, Walder M. Postantibiotic effects of imipenem, norfloxacin and amikacin in vitro and in vivo. Antimicrob Agents Chemother 1989;33:1714-1720. Ristuccia AM, Cunha BA. The aminoglycosides. Med Clin North Am 1982;66:303-312. Sastrasinh M, Knauss TC, Weinberg JM et al. Identification of the aminoglyco- side binding site in rat renal brush border membranes. J Phannacol Exp Ther 1 982;222:350-358. Schentag JJ, Jusko WJ, Vance JW, et al. Gentamicin disposition and tissue accumulation on multiple dosing. J Phannacokinet Biopharm 1977;5:559-577. Short CR, Beadle RE, Aranas T, et al. Distribution of cephapirin into a tissue chamber implanted subcutaneously in horses. J Vet Phannacol Ther 1987;10:241-247. Silverblatt FJ, Kuehn C. Autoradiography of gentamicin uptake by rat proximal tubule cell. Kidney Int 1979;15:335-345. 68 Sturrn AW. Netilmicin in the treatment of gram-negative bacteremia: single daily versus multiple daily dosage. J Infect Dis 1989;159:931-937. Taylor AE, Guyton AC, Bishop VS. Permeability of alveolar membrane to solutes. Circ Res 1965;16:535-561. Ter Braak EW, De Vries PJ, Bouter KP, et al. Once-daily dosing regimen for aminoglycoside plus beta-Iactam combination therapy Of serious bacterial infections: comparative trial with netilmicin plus ceftriaxone. Am J Med 1990;89:58-66. Thompson WL, Powell S. Aminoglycoside nephrotoxicity in dogs: dependence on drug, dose and regimen. Scand J Infect Dis 1980;Suppl 23:91-95. Tobin T. An overview of the major antibiotics in equine medicine in Proceedings. Am Assoc Equine Pract 1979;25:491-502. Tulkens PM. Experimental studies on nephrotoxicity of aminoglycosides at low doses. Am J Med 1986;80 Suppl 6B;105-114. Tulkens PM.‘ Pharmacokinetic and toxicological evaluation Of a once-daily regimen versus conventional schedules of netilmicin and amikacin. J Antimicrob Chemother 1991;27 Suppl 0:49-61. Van der Auwera P, Meunier F, Ibrahim S, et al. Pharmacodynamic parameters and toxicity of netilmicin (6 mglkg/day) given once daily or in three divided doses to cancer patients with urinary tract infection. Antimicrob Agents Chemother 1991;35:640-647. Verpooten GA, Giuliano RA, Verbist L, et al. Once—daily dosing decreases renal accumulation of gentamicin and netilmicin. Clin Phannacol Ther1989;45:22-27. VISCOII C, Dudley M, Ferrea G et al. Serum concentrations and safety of single daily dosing of amikacin in children undergoing bone marrow transplantation. J Antimicrob Chemother 1991;27 Suppl C;113—120. Vogelman BS, Craig WA. Postantibiotic effects. JAntimicrob Chemother 1985;15 Suppl A:37-46. Vogelman B, Gudmundsson S, Leggett J, et al. Correlation of antimicrobial pharmacokinetic parameters with therapeutic efficacy in an animal model. J Infect Dis 1988a;158:831-847. Vogelman B, Gudmundsson S, Turnidge J, et al. In vivo postantibiotic effect in a thigh infection in neutropenic mice. J Infect Dis 1988b;157:287-298. 69 Walker RD, Stein GE, Hauptman JG, MacDonald DH et al. Serum and tissue cage fluid concentrations of ciprofloxacin after oral administration of the drug to healthy dogs. Am J Vet Res 1990;51:896-900. WIlson RC, Moore JN, Eakle N. Gentamicin pharmacokinetics in horses given small doses of Escherichia coli endotoxin. Am J Vet Res 1983;44:1746-1749. Wong GA, Peirce TH, Goldstein E, etal. Penetration of antimicrobial agents into bronchial secretions. Am J Med 1975;59:219-223. Wood CA, Norton DR, Kohlepp SJ, et al. The influence of tobramycin dosage regimens on nephrotoxicity, ototoxicity, and antibacterial efficacy in a rat model of subcutaneous abscess. J Infect Dis 1988;158:13-22. “11111111111111“