- Email: [email protected]
Changing Concepts and New Applications of Antibiotic Pharmacokinetics
Changing Concepts and New Applications of Antibiotic Pharmacokinetics
WILLIAM A. CRAIG, M.D. BENNETT VOGELMAN, M.D. Madison, Wisconsin
From the William S. Middleton Memorial Veterans Hospital and Department of Medicine, University of Wisconsin, Madison Wisconsin. Requests for reprints should be addressed to Dr. William A. Craig, William S. Middleton Memorial Veterans Hospital, 2500 Overlook Terrace, Madison, Wisconsin 53705.
24
July 31, 1984
Antibiotics with high protein binding have a lower percentage of free drug available for tissue penetration than antibiotics with lower protein binding. High protein binding, however, may have a beneficial effect on drug distribution. The smaller volume of distribution and reduced glomerular filtration of highly bound agents result in higher serum levels that are sustained longer. Although intermittent and continuous dosing regimens produce similar areas under the concentration-versus-time curves for serum and tissue, intermittent dosing produces higher peak and potentially earlier effective antibiotic levels at the site of infection. The excretion of certain antibiotic agents in the bile may be related to hepatic protein binding, high molecular weight, or unique structural features. Biliary excretion is important not only for bile concentrations but also for dosage modification. Antibiotics with dual elimination by the kidney and biliary tract require minimal dosage modification unless there is concomitant hepatic and renal dysfunction. The third-generation cephalosporin& provide good examples of how protein binding, tissue penetration, and excretory mechanisms can be used to alter pharmacokinetics advantageously. The prime objective of antimicrobial chemotherapy is to help eradicate invading microorganisms by delivering an adequate amount of drug to the focus of infection. The choice of antimicrobial agent not only depends on a drug's activity against the suspected pathogens, but also its ability to reach and maintain effective concentrations at sites of infection. Pharmacokinetics deals with the movement of drugs through the body-their absorption, distribution, metabolism, and excretion. Some changing concepts on the pharmacokinetics of antibiotics and their penetration into tissues and inflammatory fluids are reviewed here. Serum protein binding appears to be one of the important determinants of drug distribution in the body [1]. Only the free, unbound drug molecules can readily pass through capillary pores into tissue fluids. One would anticipate, therefore, that highly bound drugs would tend to remain within the intravascular compartment, giving rise to high serum concentrations but low tissue concentrations. A variety of clinical studies, using peripheral lymph, skin-blister fluid, skin-window fluid, and wound exudates document this inhibitory effect of protein binding [2-7]. The effect of protein binding on the area under the concentration-versus-time curve in both serum and blister fluid for six beta lactam antibiotics is illustrated in Figure 1 [3]. In serum, increasing protein binding was associated with an increase in the area under the curve of total drug but a decrease in the area under the curve of free drug. However, in blister fluid, increasing
The American Journal of Medicine
CHANGING PATTERNS OF HOSPITAL INFECTIONS SYMPOSIUM-CRAIG and VOGELMAN
200
.
E
~ 100 0
~
"' ::>, 60 40 ;:::
~
.,
·---
·~·
•
o~ •
:::>
u
....,
0
20
~~
•
"0
c: ::>
10
0
Figure 1. Relation between the area under the concentration curves and the percentage of drug bound in serum. Data from Wise et at [3] corrected tor the small differences in half-life among six antibiotics (flucloxacillin, amoxicillin, cetoxitin, cefuroxime, cetsulodin, Bay k 4999).
!!!
<[
6
0
4
...:::>
I
SERUM
BLISTER FLUID
' !D
2 20
protein binding reduced the area under the curve of both total and free drug, with the major decrease occurring at binding values over 80 percent. The inhibitory effect of serum protein binding on tissue penetration does not automatically imply that a drug that is only partly bound will be superior to one that is bound extensively. It is well known that protein binding reduces the rate of drug elimination by glomerular filtration. Thus, binding can have a positive effect on drug distribution into tissues by slowing elimination and producing higher and more sustained serum concentrations. One of the characteristics of many of the new third-generation cephalosporins, such as cefoperazone, moxalactam, ceftriaxone, cefsulodin, ceftazidime, and cefotetan, is that they are excreted by the kidney, primarily by glomerular filtration [8]. The lack of a probenecid effect with these drugs supports the relative absence of tubular secretion. As shown in Figure 2, the renal clearance of these drugs decreases as the protein binding increases [8]. With the exception of cefoperazone, which is eliminated primarily in the bile, the half-life of these drugs is directly related to the degree of protein binding. This varies from one and a half to two hours for cefsulodin and ceftazidime (low binding) to eight hours for ceftriaxone (high binding). If ceftriaxone had low binding, its half-life would be similar to that of cefsulodin or ceftazidime as its filtration clearance would increase. Without high binding, the half-life of cefoperazone would be shorter and similar to that of the low-bound cefotaxime. Figure 3 illustrates the total and free drug concentrations following a 1 g intravenous dose of cefoperazone, cefotaxime, ceftriaxone, or ceftazidime. By comparing free serum concentrations of these drugs, one can see that high protein binding is compensated for entirely by the higher, more sustained serum levels. For example, the total serum levels of cefotaxime and
40
60
80
100 20
40
60
80
100
Serum Protein Binding (%)
cefoperazone at four hours after a 1 g bolus dose are approximately 2 and 15 mg/L, respectively. As the fraction of free drug for cefotaxime is 0.7 whereas that for cefoperazone is only 0.1 , the actual free drug concentrations in serum are virtually identical (1.4 versus 1.5 mg/L). Similar calculations show that the free-drug concentrations of ceftriaxone, despite its very high binding of 95 percent, are similar to, but more sustained than, those of the low-bound ceftazidime. Thus, it is not surprising that
200 0
160
.Ec:
0
......
!.,
0
120
()
c: ~
., u 80 c
.,c:c
0:
40
0 0
20
40
60
80
Serum Protein Binding (%)
100
Figure 2. Relation between renal clearance and the percentage of drug bound in serum. From left to right (low to high binding) the filtered drugs are ceftazidime, cetsulodin, moxalactam, cefotetan, cetoperazone, and ceftriaxone; the secreted drugs are ceftizoxime, cefotaxime, and cetmenoxime. Data from [8].
July 31, 1984
The American Journal of Medicine
25
CHANGING PATTERNS OF HOSPITAL INFECTIONS SYMPOSIUM- CRAIG and VOGELMAN
Total Drug
Total Drug
:J ....._ o>
E
10
z 0
CTZ
~ 1 a:: I- 100 z w 50 u z 25
Free
Free
Drug
Drug
0
u
10 ::::> 5 a:: w 2.5 (j} ~
1 0.5 0.25 0
2
4
6
8 10 12
0
2
4
TIME (hours)
these highly protein-bound drugs are very effective in the therapy of a variety of infections, even when administered at widely spaced intervals. In contrast to glomerular filtration, tubular secretion is largely independent of protein binding. It is even possible that protein binding, by raising the total serum concentration, could increase drug elimination by the kidney. This is suggested in Figure 2 for the three third-generation cephalosporins excreted in significant quantities by tubular secretion [8]. The actual antibiotic concentrations obtained in various tissues and body fluids depend on the ratio of diffusion surface area to the volume of fluid [9]. For example, in most tissues, the ratio of the surface area of capillaries to the small volume of interstitial fluid is very large. Antibiotic levels in these fluids show little lag and closely approximate free drug levels in serum. The use of implanted subcutaneous threads have demonstrated this concept in man [6, 10]. On the other hand, there is a smaller ratio of diffusion surface area to fluid volume in inflammation producing excess pleural, peritoneal, or synovial fluids. There is also a significant lag between changes in concentration in serum and in these fluids; peak levels are generally smaller and occur later than those in serum. In addition, the rate of drug elimination from these fluids is slower than that from serum. This results in higher concentrations in
26
July 31, 1984
The American Journal of Medicine
6
8 10 12
Figure 3. Total- and free-drug serum concentrations following 1 g intravenous doses of cefoperazone (CPZ), cefotaxime (CTX), ceftriaxone (CTRX), and ceftazidime (CTZ).
the fluids than in the serum toward the end of a dosing interval. Penetration of antibiotics into human skin-blister fluid is a modef of this type of pharmacokinetics [2,3]. Many investigators have used these and other models to determine whether intermittent or continuous dosing of antibiotics produces higher tissue concentrations. Several conclusions can be drawn from these studies [11]. Firstly, different methods of drug administration will produce different concentration-versus-time curves in tissues as well as in serum. Secondly, the areas under the concentrationversus-time curves at tissue sites are similar, despite different methods of dosing with fixed amounts of drug. Thirdly, effective concentrations in tissues are obtained more rapidly when a drug is administered intermittently. Thus, intermittent or pulse dosing of antimicrobial agents produces higher peak concentrations and, potentially, earlier effective concentrations at the site of infection. For example, Van Etta et al [12] studied the effect of mode of administration on antimicrobial penetration into subcutaneous dialysis sacs in rabbits. As illustrated in Figure 4 for ampicillin and oxacillin, steady-rate concentrations occurred earlier with intermittent dosing than with continuous infusion. In addition, with continuous dosing, the more protein-bound oxacillin takes longer to reach steady-state concentrations than does the low-bound ampicillin. For this reason, and because protein binding is lower at high
CHANGING PATTERNS OF HOSPITAL INFECTIONS SYMPOSIUM-CRAIG and VOGELMAN
10
E
......
Cl'
3 z
0
~ 0::
Ampict71in
OxacH!in
5
2.5
f-
z
LLJ
1.)
z
0
1.) 1.)
0.5
o- Continuous •- Intermittent
i=
0
1!! f-
0.25
z <( Figure 4. The effect of continuous and intermittent administration on extravascular penetration of penicillins in the rabbit. Adapted from [12].
0.1
0
concentrations, intermittent dosing enhances tissue penetration of the highly protein-bound, third-generation cephalosporins [12]. The biliary excretion of several of the new third-generation cephalosporins is quite high. As much as 75 percent of cefoperazone, 45 percent of ceftriaxone, and 20 percent of cefotetan is eliminated in the bile [8]. The molecular basis of the high biliary excretion of these drugs is not known. Investigators have suggested that the ability of a drug to bind to intracellular hepatic proteins such as ligandin may be an important determinant of uptake and transport of antibiotics by the liver [13]. Although cefotetan and cefazolin bind similarly to serum proteins, cefotetan binding to human liver proteins is much greater [14]. Its biliary excretion is also much greater than that of cefazolin. In rats, biliary excretion of the cephalosporins is directly related to molecular weight above a threshold value of 450 [15]. Molecular weight also appears to be a determinant of biliary excretion in man (Table 1). Of the new third-generation cephalosporins, cefotaxime and ceftizoxime have molecular weights below 500 and have low biliary excretion, whereas cefoperazone has the highest molecular weight (646) and is eliminated predominantly in the bile. For those drugs with intermediate molecular weights in the 500s, biliary excretion is higher for those with high protein binding (ceftriaxone, cefotetan, and cefmenoxime) and decreases progressively as binding becomes intermediate (moxalactam), or low (ceftazidime and cefsulodin). Chemical structure also appears to be
2
4
6
8 10 12
..L
I
I
2 .4 24 TIME (hours)
I
6
I
I
I
8 10 12
..L
24
important for biliary excretion, as four of the five cephalosporins with significant biliary excretion have the same methylthiotetrazole group at the 3-position of the dihydrothiazine ring (Table 1). Enhanced biliary excretion provides not only high bile concentrations, even in the presence of severe hepatic dysfunction [16], but, in addition, an alternative route for drug elimination in patients with renal disease. Thus, little if any dosage modification is required for these drugs in patients with renal impairment. Several investigators have shown that the half-life of cefoperazone-the cephalosporin with the highest biliary excretion-is two- to four-
TABLE I
Ceftizoxime Cefotaxime Cefsulodin Ceftazidime Moxalactam Cefmenoxime Cefotetan Ceftriaxone Cefoperazone
Relationship of Molecular Weight, Serum Protein Binding, and Presence of Nmethylthiotetrazole Group Biliary Excretion
Molecular Weight
Protein Binding
Methylthlotetrazole Group
minimal minimal minimal minimal low moderate moderate high high
383 455 545 547 521 511 574 554 646
31 37 26 17 53 77 85 95 90
No No No No Yes Yes Yes No Yes
July 31, 1984
The American Journal of Medicine
27
CHANGING PATIERNS OF HOSPITAL INFECTIONS SYMPOSIUM-CRAIG and VOGELMAN
fold longer in patients with cirrhosis and biliary obstruction [16-19]. However, these half-life values are still shorter in duration than the usual 12-hour dosing interval; thus, significant accumulation does not occur with multiple dosing. Dosage modification of these drugs with significant dual elimination from the body would only be required when there is concomitant renal and hepatic dysfunction. In summary, ingenious chemical modifications have improved the stability of the third-generation cephalosporins to various beta lactamases, extended their spectrum of activity, and altered their protein binding, handling by
the kidney, and biliary excretion. These drugs have demonstrated that protein binding can have beneficial effects on the pharmacokinetics of tissue penetration, providing that the drug is excreted by glomerular filtration. Large, widely-spaced doses will provide high initial tissue concentrations, and the dual excretion by both kidney and liver make dosage modification unnecessary in patients with either renal or hepatic dysfunction. These pharmacokinetic changes have been achieved without loss of antimicrobial properties of these agents, thus allowing for the best possible in vivo efficacy.
REFERENCES 1.
2. 3.
4.
5.
6.
7.
8.
9.
10.
28
Craig WA, Suh B: Protein binding and the antimicrobial effects: Methods for the determination of protein binding. In: Lorian V, ed. Antibiotics in laboratory medicine, ed 2. Baltimore: Williams & Wilkins Co, 1984; in press. Bergan T: Pharmacokinetics of tissue penetration of antibiotics. Rev Infect Dis 1981; 3: 45-66. Wise R, Gillet AP, Cadge B, et al: The influence of protein binding on the tissue levels of six !3-lactams. J Infect Dis 1980; 142: 77-82. Tan JS, Trott A, Phair JP, et al: A method for measurement of antibiotics in human interstitial fluid. J Infect Dis 1972; 126: 492-497. Tan JS, Salstrom SJ: Levels of carbenicillin, ticarcillin, cephalothin, cefazolin, cefamandole, gentamicin, tobramycin, and amikacin in human serum and interstitial fluid. Antimicrob Agents Chemother 1977; 11: 698-700. Hoffstedt B, Walder M: Influence of serum protein binding and mode of administration on penetration of five cephalosporins into subcutaneous tissue fluid in humans. Antimicrob Agent Chemother 1981; 20: 783-786. Craig WA, Welling PG: Protein binding of antimicrobials: clinical pharmacokinetic and therapeutic implications. Clin Pharmacokinet 1977; 2: 252-268. Vogelman B, Craig WA: The pharmacokinetics of the third-generation cephalosporins in man: a review. In: Klastersky J, ed. Nosocomial infections: current problems and role of the new cephalosporins 1983 update. Brussels: lmprimerie Europrint, 1983; 53-70. Van Etta LL, Peterson LR, Fasching CE, et al: Effect of the ratio of surface area to volume on the penetration of antibiotics into extravascular spaces in an in vitro model. J Infect Dis 1982; 146: 423-428. Ryan OM, Mason U, Harding SM: The penetration of eel-
July 31, 1984
The American Journal of Medicine
11. 12.
13. 14.
15.
16.
17.
18.
19.
tazidime into extravascular fluid. J Antimicrob Chemother 1981; 8(suppl B): 283-288. Barza M: Principles of tissue penetration of antibiotics. J Antimicrob Chemother 1981; 8(suppl C): 7-28. Van Etta LL, Peterson LR, Fasching CE, et al: Effect of method of administration on extravascular penetration of four antibiotics. Antimicrob Agents Chemother 1982; 21 : 873-880. Turnidge JD, Craig WA: !3-lactam pharmacology in liver disease. J Antimicrob Chemother 1983; 11 : 493-501. Komatsu T, Nishiya H, Iwata K, et al: Binding of antibiotics to human liver ligandin. In: Petrii P, Grassi GG, eds. Current chemotherapy and immunotherapy, Proceedings of the 12th International Congress of Chemotherapy. Washington, D.C.: American Society for Microbiology, 1981; 150-151. Wright W, Line VD: Biliary excretion of cephalosporins in rats: influence of molecular weight. Antimicrob Agents Chemother 1980; 17: 842-846. Greenfield RA, Gerber AU, Craig WA: Pharmacokinetics of cefoperazone in patients with normal and impaired hepatic and renal function. Rev Infect Dis 1983; 5: S127-S136. Boscia L, Korzeniowski 0, Snepar R, et al: Pharmacokinetics of cefoperazone in normals, cirrhotics, and patients with hepatosplenic schistosomiasis. In: Program and Abstracts of 22nd lnterscience Conference on Antimicrobial Agents and Chemotherapy. Washington, D.C.: American Society for Microbiology, 1982; 116. Cochet B, Belareff J, Allaz AF, et al: Serum levels and urinary excretion of cefoperazone in patients with hepatic insufficiency. Infection 1981; 9(suppl1): S37-S39. Marti MC, Varquet C, Fabre J, et al: Pharmacokinetic and biliary excretion of cefoperazone in patients with bile duct drainage. Infection 1981; 9(suppl1): S34-S36.
WILLIAM A. CRAIG, M.D. BENNETT VOGELMAN, M.D. Madison, Wisconsin
From the William S. Middleton Memorial Veterans Hospital and Department of Medicine, University of Wisconsin, Madison Wisconsin. Requests for reprints should be addressed to Dr. William A. Craig, William S. Middleton Memorial Veterans Hospital, 2500 Overlook Terrace, Madison, Wisconsin 53705.
24
July 31, 1984
Antibiotics with high protein binding have a lower percentage of free drug available for tissue penetration than antibiotics with lower protein binding. High protein binding, however, may have a beneficial effect on drug distribution. The smaller volume of distribution and reduced glomerular filtration of highly bound agents result in higher serum levels that are sustained longer. Although intermittent and continuous dosing regimens produce similar areas under the concentration-versus-time curves for serum and tissue, intermittent dosing produces higher peak and potentially earlier effective antibiotic levels at the site of infection. The excretion of certain antibiotic agents in the bile may be related to hepatic protein binding, high molecular weight, or unique structural features. Biliary excretion is important not only for bile concentrations but also for dosage modification. Antibiotics with dual elimination by the kidney and biliary tract require minimal dosage modification unless there is concomitant hepatic and renal dysfunction. The third-generation cephalosporin& provide good examples of how protein binding, tissue penetration, and excretory mechanisms can be used to alter pharmacokinetics advantageously. The prime objective of antimicrobial chemotherapy is to help eradicate invading microorganisms by delivering an adequate amount of drug to the focus of infection. The choice of antimicrobial agent not only depends on a drug's activity against the suspected pathogens, but also its ability to reach and maintain effective concentrations at sites of infection. Pharmacokinetics deals with the movement of drugs through the body-their absorption, distribution, metabolism, and excretion. Some changing concepts on the pharmacokinetics of antibiotics and their penetration into tissues and inflammatory fluids are reviewed here. Serum protein binding appears to be one of the important determinants of drug distribution in the body [1]. Only the free, unbound drug molecules can readily pass through capillary pores into tissue fluids. One would anticipate, therefore, that highly bound drugs would tend to remain within the intravascular compartment, giving rise to high serum concentrations but low tissue concentrations. A variety of clinical studies, using peripheral lymph, skin-blister fluid, skin-window fluid, and wound exudates document this inhibitory effect of protein binding [2-7]. The effect of protein binding on the area under the concentration-versus-time curve in both serum and blister fluid for six beta lactam antibiotics is illustrated in Figure 1 [3]. In serum, increasing protein binding was associated with an increase in the area under the curve of total drug but a decrease in the area under the curve of free drug. However, in blister fluid, increasing
The American Journal of Medicine
CHANGING PATTERNS OF HOSPITAL INFECTIONS SYMPOSIUM-CRAIG and VOGELMAN
200
.
E
~ 100 0
~
"' ::>, 60 40 ;:::
~
.,
·---
·~·
•
o~ •
:::>
u
....,
0
20
~~
•
"0
c: ::>
10
0
Figure 1. Relation between the area under the concentration curves and the percentage of drug bound in serum. Data from Wise et at [3] corrected tor the small differences in half-life among six antibiotics (flucloxacillin, amoxicillin, cetoxitin, cefuroxime, cetsulodin, Bay k 4999).
!!!
<[
6
0
4
...:::>
I
SERUM
BLISTER FLUID
' !D
2 20
protein binding reduced the area under the curve of both total and free drug, with the major decrease occurring at binding values over 80 percent. The inhibitory effect of serum protein binding on tissue penetration does not automatically imply that a drug that is only partly bound will be superior to one that is bound extensively. It is well known that protein binding reduces the rate of drug elimination by glomerular filtration. Thus, binding can have a positive effect on drug distribution into tissues by slowing elimination and producing higher and more sustained serum concentrations. One of the characteristics of many of the new third-generation cephalosporins, such as cefoperazone, moxalactam, ceftriaxone, cefsulodin, ceftazidime, and cefotetan, is that they are excreted by the kidney, primarily by glomerular filtration [8]. The lack of a probenecid effect with these drugs supports the relative absence of tubular secretion. As shown in Figure 2, the renal clearance of these drugs decreases as the protein binding increases [8]. With the exception of cefoperazone, which is eliminated primarily in the bile, the half-life of these drugs is directly related to the degree of protein binding. This varies from one and a half to two hours for cefsulodin and ceftazidime (low binding) to eight hours for ceftriaxone (high binding). If ceftriaxone had low binding, its half-life would be similar to that of cefsulodin or ceftazidime as its filtration clearance would increase. Without high binding, the half-life of cefoperazone would be shorter and similar to that of the low-bound cefotaxime. Figure 3 illustrates the total and free drug concentrations following a 1 g intravenous dose of cefoperazone, cefotaxime, ceftriaxone, or ceftazidime. By comparing free serum concentrations of these drugs, one can see that high protein binding is compensated for entirely by the higher, more sustained serum levels. For example, the total serum levels of cefotaxime and
40
60
80
100 20
40
60
80
100
Serum Protein Binding (%)
cefoperazone at four hours after a 1 g bolus dose are approximately 2 and 15 mg/L, respectively. As the fraction of free drug for cefotaxime is 0.7 whereas that for cefoperazone is only 0.1 , the actual free drug concentrations in serum are virtually identical (1.4 versus 1.5 mg/L). Similar calculations show that the free-drug concentrations of ceftriaxone, despite its very high binding of 95 percent, are similar to, but more sustained than, those of the low-bound ceftazidime. Thus, it is not surprising that
200 0
160
.Ec:
0
......
!.,
0
120
()
c: ~
., u 80 c
.,c:c
0:
40
0 0
20
40
60
80
Serum Protein Binding (%)
100
Figure 2. Relation between renal clearance and the percentage of drug bound in serum. From left to right (low to high binding) the filtered drugs are ceftazidime, cetsulodin, moxalactam, cefotetan, cetoperazone, and ceftriaxone; the secreted drugs are ceftizoxime, cefotaxime, and cetmenoxime. Data from [8].
July 31, 1984
The American Journal of Medicine
25
CHANGING PATTERNS OF HOSPITAL INFECTIONS SYMPOSIUM- CRAIG and VOGELMAN
Total Drug
Total Drug
:J ....._ o>
E
10
z 0
CTZ
~ 1 a:: I- 100 z w 50 u z 25
Free
Free
Drug
Drug
0
u
10 ::::> 5 a:: w 2.5 (j} ~
1 0.5 0.25 0
2
4
6
8 10 12
0
2
4
TIME (hours)
these highly protein-bound drugs are very effective in the therapy of a variety of infections, even when administered at widely spaced intervals. In contrast to glomerular filtration, tubular secretion is largely independent of protein binding. It is even possible that protein binding, by raising the total serum concentration, could increase drug elimination by the kidney. This is suggested in Figure 2 for the three third-generation cephalosporins excreted in significant quantities by tubular secretion [8]. The actual antibiotic concentrations obtained in various tissues and body fluids depend on the ratio of diffusion surface area to the volume of fluid [9]. For example, in most tissues, the ratio of the surface area of capillaries to the small volume of interstitial fluid is very large. Antibiotic levels in these fluids show little lag and closely approximate free drug levels in serum. The use of implanted subcutaneous threads have demonstrated this concept in man [6, 10]. On the other hand, there is a smaller ratio of diffusion surface area to fluid volume in inflammation producing excess pleural, peritoneal, or synovial fluids. There is also a significant lag between changes in concentration in serum and in these fluids; peak levels are generally smaller and occur later than those in serum. In addition, the rate of drug elimination from these fluids is slower than that from serum. This results in higher concentrations in
26
July 31, 1984
The American Journal of Medicine
6
8 10 12
Figure 3. Total- and free-drug serum concentrations following 1 g intravenous doses of cefoperazone (CPZ), cefotaxime (CTX), ceftriaxone (CTRX), and ceftazidime (CTZ).
the fluids than in the serum toward the end of a dosing interval. Penetration of antibiotics into human skin-blister fluid is a modef of this type of pharmacokinetics [2,3]. Many investigators have used these and other models to determine whether intermittent or continuous dosing of antibiotics produces higher tissue concentrations. Several conclusions can be drawn from these studies [11]. Firstly, different methods of drug administration will produce different concentration-versus-time curves in tissues as well as in serum. Secondly, the areas under the concentrationversus-time curves at tissue sites are similar, despite different methods of dosing with fixed amounts of drug. Thirdly, effective concentrations in tissues are obtained more rapidly when a drug is administered intermittently. Thus, intermittent or pulse dosing of antimicrobial agents produces higher peak concentrations and, potentially, earlier effective concentrations at the site of infection. For example, Van Etta et al [12] studied the effect of mode of administration on antimicrobial penetration into subcutaneous dialysis sacs in rabbits. As illustrated in Figure 4 for ampicillin and oxacillin, steady-rate concentrations occurred earlier with intermittent dosing than with continuous infusion. In addition, with continuous dosing, the more protein-bound oxacillin takes longer to reach steady-state concentrations than does the low-bound ampicillin. For this reason, and because protein binding is lower at high
CHANGING PATTERNS OF HOSPITAL INFECTIONS SYMPOSIUM-CRAIG and VOGELMAN
10
E
......
Cl'
3 z
0
~ 0::
Ampict71in
OxacH!in
5
2.5
f-
z
LLJ
1.)
z
0
1.) 1.)
0.5
o- Continuous •- Intermittent
i=
0
1!! f-
0.25
z <( Figure 4. The effect of continuous and intermittent administration on extravascular penetration of penicillins in the rabbit. Adapted from [12].
0.1
0
concentrations, intermittent dosing enhances tissue penetration of the highly protein-bound, third-generation cephalosporins [12]. The biliary excretion of several of the new third-generation cephalosporins is quite high. As much as 75 percent of cefoperazone, 45 percent of ceftriaxone, and 20 percent of cefotetan is eliminated in the bile [8]. The molecular basis of the high biliary excretion of these drugs is not known. Investigators have suggested that the ability of a drug to bind to intracellular hepatic proteins such as ligandin may be an important determinant of uptake and transport of antibiotics by the liver [13]. Although cefotetan and cefazolin bind similarly to serum proteins, cefotetan binding to human liver proteins is much greater [14]. Its biliary excretion is also much greater than that of cefazolin. In rats, biliary excretion of the cephalosporins is directly related to molecular weight above a threshold value of 450 [15]. Molecular weight also appears to be a determinant of biliary excretion in man (Table 1). Of the new third-generation cephalosporins, cefotaxime and ceftizoxime have molecular weights below 500 and have low biliary excretion, whereas cefoperazone has the highest molecular weight (646) and is eliminated predominantly in the bile. For those drugs with intermediate molecular weights in the 500s, biliary excretion is higher for those with high protein binding (ceftriaxone, cefotetan, and cefmenoxime) and decreases progressively as binding becomes intermediate (moxalactam), or low (ceftazidime and cefsulodin). Chemical structure also appears to be
2
4
6
8 10 12
..L
I
I
2 .4 24 TIME (hours)
I
6
I
I
I
8 10 12
..L
24
important for biliary excretion, as four of the five cephalosporins with significant biliary excretion have the same methylthiotetrazole group at the 3-position of the dihydrothiazine ring (Table 1). Enhanced biliary excretion provides not only high bile concentrations, even in the presence of severe hepatic dysfunction [16], but, in addition, an alternative route for drug elimination in patients with renal disease. Thus, little if any dosage modification is required for these drugs in patients with renal impairment. Several investigators have shown that the half-life of cefoperazone-the cephalosporin with the highest biliary excretion-is two- to four-
TABLE I
Ceftizoxime Cefotaxime Cefsulodin Ceftazidime Moxalactam Cefmenoxime Cefotetan Ceftriaxone Cefoperazone
Relationship of Molecular Weight, Serum Protein Binding, and Presence of Nmethylthiotetrazole Group Biliary Excretion
Molecular Weight
Protein Binding
Methylthlotetrazole Group
minimal minimal minimal minimal low moderate moderate high high
383 455 545 547 521 511 574 554 646
31 37 26 17 53 77 85 95 90
No No No No Yes Yes Yes No Yes
July 31, 1984
The American Journal of Medicine
27
CHANGING PATIERNS OF HOSPITAL INFECTIONS SYMPOSIUM-CRAIG and VOGELMAN
fold longer in patients with cirrhosis and biliary obstruction [16-19]. However, these half-life values are still shorter in duration than the usual 12-hour dosing interval; thus, significant accumulation does not occur with multiple dosing. Dosage modification of these drugs with significant dual elimination from the body would only be required when there is concomitant renal and hepatic dysfunction. In summary, ingenious chemical modifications have improved the stability of the third-generation cephalosporins to various beta lactamases, extended their spectrum of activity, and altered their protein binding, handling by
the kidney, and biliary excretion. These drugs have demonstrated that protein binding can have beneficial effects on the pharmacokinetics of tissue penetration, providing that the drug is excreted by glomerular filtration. Large, widely-spaced doses will provide high initial tissue concentrations, and the dual excretion by both kidney and liver make dosage modification unnecessary in patients with either renal or hepatic dysfunction. These pharmacokinetic changes have been achieved without loss of antimicrobial properties of these agents, thus allowing for the best possible in vivo efficacy.
REFERENCES 1.
2. 3.
4.
5.
6.
7.
8.
9.
10.
28
Craig WA, Suh B: Protein binding and the antimicrobial effects: Methods for the determination of protein binding. In: Lorian V, ed. Antibiotics in laboratory medicine, ed 2. Baltimore: Williams & Wilkins Co, 1984; in press. Bergan T: Pharmacokinetics of tissue penetration of antibiotics. Rev Infect Dis 1981; 3: 45-66. Wise R, Gillet AP, Cadge B, et al: The influence of protein binding on the tissue levels of six !3-lactams. J Infect Dis 1980; 142: 77-82. Tan JS, Trott A, Phair JP, et al: A method for measurement of antibiotics in human interstitial fluid. J Infect Dis 1972; 126: 492-497. Tan JS, Salstrom SJ: Levels of carbenicillin, ticarcillin, cephalothin, cefazolin, cefamandole, gentamicin, tobramycin, and amikacin in human serum and interstitial fluid. Antimicrob Agents Chemother 1977; 11: 698-700. Hoffstedt B, Walder M: Influence of serum protein binding and mode of administration on penetration of five cephalosporins into subcutaneous tissue fluid in humans. Antimicrob Agent Chemother 1981; 20: 783-786. Craig WA, Welling PG: Protein binding of antimicrobials: clinical pharmacokinetic and therapeutic implications. Clin Pharmacokinet 1977; 2: 252-268. Vogelman B, Craig WA: The pharmacokinetics of the third-generation cephalosporins in man: a review. In: Klastersky J, ed. Nosocomial infections: current problems and role of the new cephalosporins 1983 update. Brussels: lmprimerie Europrint, 1983; 53-70. Van Etta LL, Peterson LR, Fasching CE, et al: Effect of the ratio of surface area to volume on the penetration of antibiotics into extravascular spaces in an in vitro model. J Infect Dis 1982; 146: 423-428. Ryan OM, Mason U, Harding SM: The penetration of eel-
July 31, 1984
The American Journal of Medicine
11. 12.
13. 14.
15.
16.
17.
18.
19.
tazidime into extravascular fluid. J Antimicrob Chemother 1981; 8(suppl B): 283-288. Barza M: Principles of tissue penetration of antibiotics. J Antimicrob Chemother 1981; 8(suppl C): 7-28. Van Etta LL, Peterson LR, Fasching CE, et al: Effect of method of administration on extravascular penetration of four antibiotics. Antimicrob Agents Chemother 1982; 21 : 873-880. Turnidge JD, Craig WA: !3-lactam pharmacology in liver disease. J Antimicrob Chemother 1983; 11 : 493-501. Komatsu T, Nishiya H, Iwata K, et al: Binding of antibiotics to human liver ligandin. In: Petrii P, Grassi GG, eds. Current chemotherapy and immunotherapy, Proceedings of the 12th International Congress of Chemotherapy. Washington, D.C.: American Society for Microbiology, 1981; 150-151. Wright W, Line VD: Biliary excretion of cephalosporins in rats: influence of molecular weight. Antimicrob Agents Chemother 1980; 17: 842-846. Greenfield RA, Gerber AU, Craig WA: Pharmacokinetics of cefoperazone in patients with normal and impaired hepatic and renal function. Rev Infect Dis 1983; 5: S127-S136. Boscia L, Korzeniowski 0, Snepar R, et al: Pharmacokinetics of cefoperazone in normals, cirrhotics, and patients with hepatosplenic schistosomiasis. In: Program and Abstracts of 22nd lnterscience Conference on Antimicrobial Agents and Chemotherapy. Washington, D.C.: American Society for Microbiology, 1982; 116. Cochet B, Belareff J, Allaz AF, et al: Serum levels and urinary excretion of cefoperazone in patients with hepatic insufficiency. Infection 1981; 9(suppl1): S37-S39. Marti MC, Varquet C, Fabre J, et al: Pharmacokinetic and biliary excretion of cefoperazone in patients with bile duct drainage. Infection 1981; 9(suppl1): S34-S36.