PHARMACOKINETIC-PHARMACODYNAMIC RELATIONSHIPS Clin Pharmacokinet 2001; 40 (9): 685-694

0312-5963/01/0009-0685/$22.00/0

© Adis International Limited. All rights reserved.

Clinical Use of Ceftriaxone

A Pharmacokinetic-Pharmacodynamic Perspective on
the Impact of Minimum Inhibitory Concentration and
Serum Protein Binding
Tracy R. Perry and Jerome J. Schentag
School of Pharmacy, University at Buffalo, Buffalo, New York, USA

Contents
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 685
1. Pharmacokinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 686
2. Pharmacodynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 687
3. Healthy Volunteers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 688
4. Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 691
5. The Elderly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 692
6. Patients with Renal and Hepatic Impairment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 692
7. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 693

Abstract Ceftriaxone is a third-generation cephalosporin that is used for a variety of

infections such as meningitis, gonorrhoea and community-acquired pneumonia.
The most important aspects of its pharmacokinetics include a long half-life, ex-
cellent tissue penetration and saturable (dose-dependent) serum protein binding
of the drug. A pharmacodynamic analysis [total area under the concentration-time
curve (AUC)/minimum inhibitory concentration (MIC)] was performed in several
populations (healthy volunteers, children, the elderly, and patients with renal and
hepatic impairment) against various bacterial species (Streptococcus pneu-
moniae, the Enterobacteriacieae, methicillin-susceptible Staphylococcus aureus,
and Pseudomonas aeruginosa). AUC/MIC [area under the inhibitory time curve
(AUIC)] was chosen as the pharmacodynamic parameter for this analysis since
ceftriaxone is a time-dependent killer and high peak concentrations are not
needed. In addition, there is a significant correlation between AUIC, time when
concentration exceeds the MIC (t > MIC) and time to eradication. Total and free
AUICs (assuming a free fraction = 10%) were calculated since it is highly protein
bound. It was postulated that a free AUIC of at least 125 would be required to
achieve efficacy. From our analysis of these various populations, we were able
to conclude that the free AUIC values support the use of 1g daily in infections
where MIC values are below 2 mg/L. In addition, consistent with its reported
good activity against CSF organisms with MICs ≤1.0 mg/L and marginal activity
against organisms with MICs ≥2.0 mg/L, we also recommend the target free AUIC
values of at least 125 for patients with severe infections such as meningitis.
Patients with mild infections may recover with values below 125 but they may
remain at risk of the development of resistant organisms. Furthermore, it is
686 Perry & Schentag

essential to further validate these findings in patients who have received treat-
ment, calculate AUICs and correlate these parameters with both clinical and
microbiological outcomes.

Ceftriaxone is a third-generation cephalosporin that include the long half-life, excellent penetration into
has been commercially available since 1985.[1] It mani- the fluids surrounding cells, and saturable (dose-
fests a broad spectrum of in vitro activity, including dependent) serum protein binding of the drug. Con-
both Gram-positive and Gram-negative organisms sensus pharmacokinetic parameters of ceftriaxone
as well as a few anaerobic bacteria. It is very active are provided in table I.
against most of the members of the Enterobacter- For most protein-bound antibacterials, the per-
iaceae [minimum inhibitory concentration against centage of protein binding remains relatively con-
90% of isolates (MIC90) <1.0 mg/L], Streptococcus stant throughout the dose range. For ceftriaxone,
pneumoniae (MIC90 <0.5 mg/L), S. pyogenes, S. there can be saturation of all available plasma pro-
agalactiae and S. viridans. It is also very active in tein binding sites within the normal dose range,
low concentrations against β-lactamase-positive with the expected result being an increase in the
and -negative isolates of Haemophilus influenzae, excreted amount. A study by Stoeckel[8] confirms
Neissseria gonorrhoeae and N. meningitidis (MIC90 this phenomenon. Intravenous doses of 150, 500
<0.025 mg/L). Ceftriaxone has moderate activity and 1500mg resulted in disproportionate increases
against penicillin-susceptible and -resistant strains in the total area under the concentration-time curve
of Staphylococcus aureus (MIC90 2 to 8 mg/L) but (AUC) and total concentrations of ceftriaxone. In
it should not be used to treat oxacillin-resistant fact, the free fraction increased from 4 to 17% as
strains of the species.[2,3] Organisms such as Entero- the serum concentrations varied between 0.5 and
coccus and Pseudomonas aeruginosa are usually 300 mg/L. This is the range of plasma values after
resistant to ceftriaxone.[2,4] In addition, ceftriaxone doses of 1 to 2g. The resulting higher concentrations
distributes only into the extracellular space and does of unbound drug in the plasma allow more rapid
not enter cells. Intracellular pathogens such as Myco- excretion and may also increase the free concentra-
plasma, Rickettsia, Chlamydia and Legionella are tions in the tissues. This occurs because the high
usually resistant in vivo, even if they appear to be extent of protein binding protects the drug from
susceptible in vitro.[3] renal excretion, producing a longer plasma half-
The past 15 years of clinical experience with life.[3,8] The effect is theoretically greater if higher
this agent provide evidence that it is reliably active doses are given once daily as opposed to smaller
against organisms for which MIC is ≤1.0 mg/L, but
doses given twice daily. Although the effect of an
fails with increasing frequency where MIC values
increase in free ceftriaxone on excretion is rela-
are ≥2.0 mg/L. We decided to examine whether
tively easy to measure, it is much more difficult to
pharmacokinetic-pharmacodynamic calculations
show a change in antimicrobial activity that results
might enhance our understanding of these clinical
from an increase in the unbound drug concentra-
observations. Consequently, the intent of this manu-
tion, since antimicrobial testing is performed in
script is to establish pharmacokinetic-pharmaco-
protein-free broth media. It is also true that adding
dynamic understanding of this antibacterial. It is
albumin to broth interferes with the action of
not intended to provide an exhaustive review of the
ceftriaxone, particularly against Gram-positive
pharmacokinetics of ceftriaxone, as this has been
organisms such as S. aureus.[9,10] Proving the exist-
done previously.[3,5-7]
ence of these differences in vivo has not yet been
attempted.
1. Pharmacokinetics
Another parameter affected by the increase in
The ‘highlights’ of ceftriaxone pharmacokinetics free fraction is tissue fluid penetration. Ceftriaxone

© Adis International Limited. All rights reserved. Clin Pharmacokinet 2001; 40 (9)
Clinical Use of Ceftriaxone 687

is highly bound to albumin, and unbound drug is 125 (where AUC24 is the AUC to 24 hours). Further
the active form. It should be pointed out that more increases in concentration above these values do
than 50% of the body’s total albumin concentration not kill bacteria more rapidly.
is in the extracellular fluid space, not circulating in For antimicrobials such as the aminoglycosides
the bloodstream. Ceftriaxone displays saturable and fluoroquinolones, concentration-dependent
protein binding, so once the available binding sites killing mechanisms cause more rapid killing above
on albumin in the bloodstream are saturated, the an AUC/MIC ratio of 125, up to a maximum killing
free concentrations increase disproportionately and rate as the AUC/MIC ratio exceeds 250. For con-
more free ceftriaxone can equilibrate by diffusing centration-dependent killers, higher peaks trans-
into tissue fluids, where binding to interstitial fluid late into higher AUC/MIC ratios and longer times
proteins in organs and tissues can occur. But when concentration exceeds the MIC (t > MIC),
ceftriaxone has a higher affinity for binding to the both of which are correlates of increased rates of
penicillin-binding proteins of bacteria than it has killing.[11]
to albumin, so the drug molecule can be stripped Pharmacodynamic parameters used to assess
off the albumin binding site by contact with a bacte- efficacy in β-lactam antibacterials are t > MIC and
rial pathogen. Once this occurs, a free ceftriaxone area under the inhibitory time curve (AUIC),
molecule in the bloodstream moves to the tissue which is also known as AUC/MIC.[12] In a study
space. Perhaps the bound fraction of this drug conducted by our group, there was a significant
should be considered a reservoir, with the release correlation between AUIC, t > MIC and time to
of a constant supply of active drug for up to 24 eradication.[13] Further analysis of the data re-
hours.[3,6] vealed that an AUIC >125 also correlated with mi-
crobiological response. For long half-life antimi-
2. Pharmacodynamics crobials such as vancomycin and ceftriaxone, there
From concentrations greater than zero up to is a particularly close concordance between AUIC
twice the MIC, all antibacterials kill bacteria more >125 and t > MIC of 80 to 100% of the dose ad-
rapidly as concentrations increase. After 2 to 4 ministration interval.[14] A study by Highet et al.[15]
times MIC, the mechanisms of killing diverge. β- confirmed earlier findings. They found that pa-
lactam agents, vancomycin, clindamycin and the tients with culture-documented Gram-negative
macrolides kill bacteria in a time-dependent fash- lower respiratory tract infections treated with β-
ion. The aminoglycosides, fluoroquinolones and lactam regimens had a better response to antimi-
metronidazole are concentration-dependent kill- crobial therapy if the AUIC was at least 125.[15]
ers. Although no clinical studies of ceftriaxone have
Time-dependent killing is characterised by maxi- measured its AUIC or t > MIC values, or peak con-
mum efficacy of an antimicrobial at 2 to 4 times centration : MIC ratio, it is highly likely that find-
the MIC, an exposure profile achieved when 80 to ings will be similar. This is especially true since the
100% of the concentrations are above the MIC. long half-life of ceftriaxone would allow for its
This also coincides with an AUC24/MIC ratio of concentrations to exceed the MIC for the desired

Table I. Pharmacokinetic parameters of ceftriaxone[1,6]

Parameter Healthy adults Children with meningitis
Elimination half-life (h) 5.8-8.7 4.3-4.6
Volume of distribution 5.78-13.5L 0.338-0.373 L/kg
Plasma clearance 0.58-1.45 L/h 0.049-0.060 L/h/kg
Fraction excreted unchanged in the urine (%) 33-67
Protein binding Concentration dependent

© Adis International Limited. All rights reserved. Clin Pharmacokinet 2001; 40 (9)
688 Perry & Schentag

Table II. Pharmacokinetic-pharmacodynamic parameters for ceftriaxone based on total concentrations. All AUCs are adjusted to 24 hours
Population and dose AUC AUIC
(mg
h/L) Streptococcus Enterobacteriaciae MSSA Pseudomonas
pneumoniae (MIC <1.0 mg/L) (MIC = 2-8 mg/L) aeruginosa
(MIC = 0.5-1 mg/L) (MIC >64 mg/L)

Healthy adults
500mg IV od[7] 610 610-1220 610 76.3-305 9.5
500mg IV q12h[17] 998 998-1996 998 124.8-499 15.6
500mg IM q12h[18] 823.8 823.8-1647.6 823.8 103-412 12.9
1g IV od[7] 1006 1006-2012 1006 125-1006 15.7
1g IM od[20] 901.44 901-1803 901 112.7-450.7 14.1
1g IV q12h[8,17] 863 863-1726 863 107.9-431.5 13.5
1g IM q12h[18] 1647 1647-3294 1647 205.9-823.5 51.5
2g IV od[1] 3084.0 3084-6168 3084 385.5-1542 48.2
2g IV q12h[8,17] 1441 1441-2882 1441 180.1-720.5 22.5

Children[22]
Infants (average dose 462mg) 1223.9 1223.9-2447.8 1223.9 153-612 19.1
Young children (average dose 699mg) 1225.1 1225.1-2450.2 1225.1 153.1-612.6 19.1

Elderly[23]
1g IV od 1289 1289-2578 1289 161-644.5 20.1
[24]
Renal impairment (1g IV od)
Dialysis (CLCR <15 ml/min) 1858.7 1858.7-3717.4 1858.7 232.3-929.4 29
Severe impairment (CLCR 5-15 ml/min) 1666.7 1666.7-3333 1666.7 208.3-833.4 26
Moderate impairment (CLCR 16-30 1890.4 1890.4-3780.4 1890.4 236.3-945.2 29.5
ml/min)
Mild impairment (CLCR 31-60 ml/min) 1418.4 1418.4-2836.8 1418.4 177.3-709.2 22.2
[25]
Hepatic impairment (1g IV od)
Alcoholic fatty liver 1234.6 1234.6-2469.2 1234.2 154.3-617.3 19.3
Cirrhosis without ascites 1048.2 1048.2-2096.4 1048.2 131-524.1 16.4
Cirrhosis with ascites 715.3 715.3-1430.6 715.3 89.4-357.7 11.2
AUC = area under the plasma concentration-time curve; AUIC = area under the inhibitory time curve; CLCR = creatinine clearance; IM
= intramuscular; IV = intravenous; MIC = minimum inhibitory concentration; MSSA = methicillin-susceptible S. aureus; od = once daily;
q12h = every 12 hours.

period and an AUIC value of >125 is roughly this drug may be somewhere between these 2 ex-
equivalent to antimicrobial concentrations being tremes.
above the MIC of an organism for approximately
80% of the dosage interval.[14] In addition, since
3. Healthy Volunteers
ceftriaxone is a time-dependent killer, it is highly
unlikely that peak : MIC is an adequate pharmaco- By using the data from pharmacokinetic studies
dynamic parameter to correlate with efficacy, since in healthy adult volunteers, one can derive AUC
high peaks are not needed.[11,16] values and then calculate both free and total
The unknown effect on all of these calculations AUC.[2-8,17-21] The use of AUC is particularly ap-
is that of saturable protein binding. Theoretically, propriate because it provides an integrated measure
this effect would minimise the differential impact of drug exposure and removes the need to rely on
of free versus total drug, and so we will evaluate a single time point along a long half-life disposition
this with consideration of free and total concentra- curve to base conclusions. From AUC, both free
tions, realising that the final antimicrobial effect of and total, one can calculate predicted AUIC values

© Adis International Limited. All rights reserved. Clin Pharmacokinet 2001; 40 (9)
Clinical Use of Ceftriaxone 689

for target organisms. This exercise was performed trations. This is the usual laboratory breakpoint for
for both total (table II) and free (table III) concen- this drug in many countries. The figure also illus-
trations of ceftriaxone. Calculations were also per- trates why ceftriaxone should not be used in the
formed for ceftriaxone blood concentrations re- treatment of serious pseudomonal infections, since
ported from other populations, including children, its AUIC is well below 125 against this organism,
the elderly and patients with renal or hepatic im- even using total concentrations, and total concen-
pairment (see sections 4, 5 and 6). trations are below the MIC for most of the dosage
Table II shows the AUIC values (total drug) for interval.
different doses of ceftriaxone against 4 different The AUIC data for free drug in table III may be
pathogens arranged on the basis of ascending MICs. considered more important by some investigators,
Ceftriaxone was highly active, using the AUIC > especially in the light of the high extent of protein
125 criterion, against S. pneumoniae and the En- binding of ceftriaxone. Free drug is responsible for
terobacteriaciae at all dosages, even at 500mg in- activity in vitro, since in vitro testing systems are
travenously daily. This is of great importance, free of albumin. The other attributes attached to free
since the MICs used for S. pneumoniae are 0.5 concentrations are renal excretion and penetration
mg/L (susceptible) and 1.0 mg/L (intermediate). into fluids surrounding infected tissues, such as the
So ceftriaxone can be used for intermediate strains lung in the case of community-acquired pneumonia.
even at lower dosages. The conclusion is that 1g of Table III shows the AUIC values for free drug cal-
ceftriaxone either intramuscularly or intravenously culated from the available pharmacokinetic studies.
can be safely used in the treatment of pneumococ- ‘Free’ AUIC was subsequently calculated by as-
cal pneumonia to cover even an intermediately sus- suming a free fraction of approximately 10%. The
ceptible strain, at least if total concentrations are problem with analysing these data on the basis of
the marker for efficacy. Table II also confirms the AUIC or t > MIC is that there are no studies suggest-
current belief that ceftriaxone has little to no useful ing what a target free AUIC would be to achieve
activity against Pseudomonas, since the MICs are clinical success. Therefore, to continue the exercise,
much too high and the resulting AUICs are too low. we shall postulate that a free AUIC of 125 is the
The activity of ceftriaxone against methicillin-sus- target, since most of the drugs from which this
ceptible S. aureus is marginal in some cases, even value has been derived display insignificant plasma
using total AUIC, and outcomes against staphylo-
cocci are likely to be dose- and MIC-dependent.
Ceftriaxone serum concentration (mg/L)

Figure 1 depicts a simulated concentration-time 250

curve for ceftriaxone 1g given intravenously. This
figure shows both total serum concentrations and
MIC values and establishes that the serum concen-
trations of this drug are above the MICs of 1.0 MIC = 64 AUIC = 16

mg/L (Escherichia coli and S. pneumoniae) and 8.0

mg/L (methicillin-susceptible S. aureus) for most MIC = 8 AUIC = 125
of the dosage interval, based on total drug concen-
MIC = 1 AUIC = 1006
trations. Again, AUIC correlates very well with t >
MIC for total concentrations, as would be expected 0 12 24
for a compound with a long half-life. At this dose Time (h)
of ceftriaxone, total AUICs ≥ 125 are achieved Fig. 1. Simulated serum total concentration-time curve for intra-
against MICs of 1.0 and 8.0 mg/L, and one could venous ceftriaxone 1 g/day showing effect of different MIC values
on total AUIC. Area under the total concentration-time curve was
justify a laboratory breakpoint of 8.0 mg/L for this assumed to be 1006 mg
h/L.[13] AUIC = area under the inhibitory
drug if outcome were based on total drug concen- time curve; MIC = minimum inhibitory concentration.

© Adis International Limited. All rights reserved. Clin Pharmacokinet 2001; 40 (9)
690 Perry & Schentag

Table III. Pharmacokinetic-pharmacodynamic parameters for ceftriaxone based on free concentrations. All AUCs are adjusted to 24 hours
Population and dose AUC AUIC
(mg
h/L) Streptococcus Enterobacteriaciae MSSA Pseudomonas
pneumoniae (MIC <1.0 mg/L) (MIC = 2-8 mg/L) aeruginosa
(MIC = 0.5-1 mg/L) (MIC >64 mg/L)

Healthy adults
500mg IV od[10] 610 61-122 61 7.63-30.5 0.95
500mg IV q12h[11] 998 99.8-199.6 99.8 12.48-49.9 1.56
500mg IM q12h[12] 823.8 82.38-164.8 82.4 10.3-41.2 1.29
1g IV od[13] 1006 100.6-201 100.6 12.6-100.6 1.6
1g IM od[14] 901.44 90.1-180.3 90.1 11.3-45.1 1.41
1g IV q12h[15] 863 86.3-172.6 86.3 10.8-43.2 1.35
1g IM q12h[12] 1647 164.7-329.4 164.7 20.59-82.4 5.15
2g IV od[11] 3084.0 308.4-616.8 308.4 38.55-154.2 4.82
2g IV q12h[15] 1441 144.1-288.2 144.1 18.0-72.1 2.25

Children[22]
Infants (average dose 462mg) 1223.9 122.4-244.8 122.4 15.3-61.2 1.9
Young children (average dose 699mg) 1225.1 122.5-245.0 122.5 15.3-61.3 1.9
[23]
Elderly
1g IV od 1289 193.4-386.7 193.4 24.2-96.7 3.0
[24]
Renal impairment (1g IV od)
Dialysis (CLCR <15 ml/min) 1858.7 185.9-371.2 185.9 23.3-92.9 2.9
Severe impairment (CLCR 5-15 ml/min) 1666.7 166.7-333.3 166.7 20.8-83.3 2.6
Moderate impairment (CLCR 16-30 1890.4 189.0-378.0 189.0 23.6-94.5 2.95
ml/min)
Mild impairment (CLCR 31-60 ml/min) 1418.4 141.8-283.7 141.8 17.7-70.9 2.22
[25]
Hepatic impairment (1g IV od)
Alcoholic fatty liver 1234.6 123.5-247 123.4 15.4-61.7 1.93
Cirrhosis without ascites 1048.2 104.8-209.6 104.8 13.1-52.4 1.64
Cirrhosis with ascites 715.3 71.5-143.1 71.5 8.94-35.8 1.12
AUC = area under the plasma concentration-time curve; AUIC = area under the inhibitory time curve; CLCR = creatinine clearance; IM
= intramuscular; IV = intravenous; MIC = minimum inhibitory concentration; MSSA = methicillin-susceptible S. aureus; od = once daily; q12h
= every 12 hours.

protein binding. Furthermore, at free AUICs of tion of approximately 10%, a 1g dose produces a free
125, free concentrations would be above the MIC AUIC of approximately 100, and this would validate
for approximately 80% of the time.[26,27] We recog- the usually good outcome of a dosage regimen of
nise that concentration-dependent protein binding 1g daily against pneumococcal pneumonia and
could result in higher free AUIC than predicted even meningitis. Unfortunately, MICs for ceftriaxone
from a multiplication factor applied to total AUC. are rising against pneumococci, and there may
Patients with diseases may experience variable pro- soon be some failures, for example with MICs at
tein binding, a phenomenon that we have described 2.0 mg/L or higher. The remedy for this problem can
previously.[28] be higher doses or more frequent administration.
The breakpoint of AUIC > 125 is a logical target If AUIC > 125 is the target exposure, a clinician
for free concentrations of ceftriaxone. There is ex- could make an argument against using ceftriaxone
tensive evidence to support the use of ceftriaxone 1g to treat a serious infection caused by methicillin-
either intramuscularly or intravenously daily to treat susceptible S. aureus, especially when their MICs
pneumococcal pneumonia.[1-3] Assuming a free frac- exceed 2.0 mg/L. Figure 2 shows the free drug

© Adis International Limited. All rights reserved. Clin Pharmacokinet 2001; 40 (9)
Clinical Use of Ceftriaxone 691

concentration-time curve for a 1g dose of ceftriaxone of this regimen (assuming approximately 10% free
given intravenously, and calculates the free AUIC fraction) would be approximately 125, illustrating
values for the target MICs. If the target free AUIC that an MIC of 1.0 mg/L is actually borderline if
value is to be approximately 100, a 1g dose is ob- the free concentrations are predictive of the outcome.
viously too low. On further inspection of figure 2, An MIC of 2.0 mg/L would produce an AUIC well
one can also conclude that the free drug concentra- below 100, with potential for resistance development
tion is above an MIC of 1.0 mg/L for most of the and even clinical failure in serious infections. On
dosage interval, whereas at an MIC of 8.0 mg/L the this basis, it is reasonable to designate organisms
concentration clearly does not exceed the MIC for with MIC values of 2.0 mg/L as intermediate. One
most of the dosage interval. This may be increas- potential solution to this borderline exposure pro-
ingly important, as resistance selection is a grow- file is to double the dosage to 2g daily.
ing problem with third generation cephalosporins, The free AUIC may indeed be most important,
and low AUICs predict the selection of resistant especially since only free drug can pass the blood-
microbes.[29] Clearly, clinical trials are justified to brain barrier. The free fraction for both groups of
establish a target value for free AUICs. Neverthe- patients in this study was approximately 16%, al-
less, there is already general clinical support for the lowing us to calculate a free AUIC value of 184
interpretation of ceftriaxone outcomes using free against an intermediately susceptible strain of
AUICs, and these calculations may be useful along pneumococcus with an MIC of 1.0 mg/L. This
with locally measured MICs in the practice of anti- would be too low for the treatment of an organism
bacterial selection. Alternatively, the data in tables with an MIC of 2.0 mg/L. Even more importantly,
II and III suggest that ceftriaxone may best be con- it should be noted that ceftriaxone penetration into
sidered effective in relation to total drug rather than CSF results in concentrations that are between 5
free drug concentrations, since the total drug con- and 15% of total serum concentrations in the CSF.
centrations and AUICs appear high enough to Assuming that the dose is 1g, the free AUIC data
achieve efficacy against marginal organisms such would also support the use of once daily adminis-
as S. aureus. As many clinicians realise, they should tration in paediatric meningitis caused by organisms
not select a low dose of ceftriaxone for treatment with MIC values ≤1.0 mg/L.[22]
of a severe methicillin-susceptible S. aureus infec-
tion in preference to oxacillin, even though the bac-
terium is reported as being susceptible to ceftriaxone MIC = 64 Free AUIC = 1.6
Ceftriaxone serum concentration (mg/L)

at an MIC of 8.0 mg/L.[17-21]

4. Children

Schaad and Stoeckel[22] studied single dose

pharmacokinetics of ceftriaxone in 5 infants and 5
young children. On clinical grounds, they concluded MIC = 8 Free AUIC = 12.6
that 50 mg/kg as a single daily dose was sufficient
to treat childhood meningitis caused by S. pneu-
moniae, H. influenzae and N. meningitidis. This MIC = 1 Free AUIC = 100.6
dose produces an AUC (total drug) of 1225 (table
II). The total AUIC value would be 1225 against Time (h)
an organism with an MIC of 1.0 mg/L, a very high Fig. 2. Simulated serum free concentration-time curve for intra-
total AUIC value and one consistent with 100% t venous ceftriaxone 1 g/day showing effect of different MIC values
on free AUIC. Area under the free concentration-time curve was
> MIC against S. pneumoniae, which has the high- assumed to be 100.6 mg
h/L.[13] AUIC = area under the inhibitory
est MIC values of the 3 organisms. The free AUIC time curve; MIC = minimum inhibitory concentration.

© Adis International Limited. All rights reserved. Clin Pharmacokinet 2001; 40 (9)
692 Perry & Schentag

15% penetration CSF AUIC = 184 basis of half-life, they concluded that no dosage ad-
Ceftriaxone CSF concentration (mg/L)

10% penetration CSF AUIC = 123

5% penetration CSF AUIC = 61 justments are needed in the geriatric population,
except in patients who are debilitated or malnour-
ished, or who have marked renal impairment. Anal-
ysis of their AUC data reveals that no substantial
alterations to the conclusions derived in section 3
for younger adults are necessary for geriatric patients,
except for the free AUIC being higher in these pa-
tients because of a modest increase in the free frac-
MIC = 1 tion of about 15% as compared with their younger
Time (h)
counterparts (about 11%). The free AUCs in the
geriatric patients were disproportionately higher
Fig. 3. Simulated cerebrospinal fluid (CSF) concentration-time
than in the younger individuals because of the ef-
curves for intravenous ceftriaxone 50 mg/kg/day in children,
showing effect of different extents of CSF penetration on AUIC fects of aging on plasma protein binding and also
in CSF. Area under the total serum concentration-time curve was the effect of reduced creatinine clearance in the el-
assumed to be 1225 mg
h/L.[22] AUIC = area under the inhibitory derly, lowering the ability to excrete free drug.
time curve; MIC = minimum inhibitory concentration.
As overall AUC was higher (not lower) in the
geriatric patients, the effect of lower renal function
These conclusions are supported by the AUIC appears to be more important than the effect of
analysis of figure 3. A 5% penetration of ceftriaxone aging on the amount of free ceftriaxone. If the free
into CSF results in AUIC values of approximately fraction and protein binding effects were predom-
61 against an intermediately susceptible strain of inant, the drug could be cleared more rapidly and
pneumococci with an MIC value of 1.0 mg/L. As- have a shorter half-life than otherwise expected,
suming 10 or 15% penetration into CSF, the result- and the overall AUC of this drug would be lower
ing AUICs would be 123 or 184, respectively, in the geriatric patients, because drug displaced
against this same organism. These CSF AUICs are would be cleared more rapidly. The data in tables
(realistically) free AUICs. The efficacy of this drug II and III argue that this was not the case.
in paediatric meningitis can be supported by AUIC As might be expected from the AUIC values,
values, given that the MIC value is 1.0 mg/L or ceftriaxone remains highly active against pneumo-
lower. t > MIC would be in agreement with AUICs, cocci and the Enterobacteriaceae in elderly patients
but also show high dependence on the actual MIC at MICs ≤1.0 mg/L, even without reliance on in-
of the treated pathogen. A pneumococcal organism creased concentrations associated with decreased
with an MIC value of ≥2.0 mg/L would necessitate renal function. Unreliable ceftriaxone activity is
either a higher dose or possibly the addition of a the rule against Pseudomonas, and variable activity
second active agent such as vancomycin. The total is possible against methicillin-susceptible S. aureus,
AUIC value for a combination of ceftriaxone plus as in younger patients, because the somewhat higher
vancomycin would be the sum of the individual AUC in the elderly does not seem likely to force
AUICs of each drug against the respective pathogen, the AUIC high enough to improve the chances of
as AUIC values in vitro for these 2 agents show eradicating these marginal organisms.[30]
additivity.[12]
6. Patients with Renal and
5. The Elderly Hepatic Impairment

Luderer et al.[23] derived the single dose pharmaco- Patel et al.[24] reported ceftriaxone pharmaco-
kinetic parameters of ceftriaxone in 8 healthy elderly kinetic parameters in patients with varying degrees
individuals by giving 1g doses intravenously. On the of renal dysfunction. The AUC of ceftriaxone in-

© Adis International Limited. All rights reserved. Clin Pharmacokinet 2001; 40 (9)
Clinical Use of Ceftriaxone 693

creased as renal function decreased (tables II and in treated patients and calculating AUICs, correlating
III). This is because the free drug is cleared renally these pharmacokinetic-pharmacodynamic measure-
by glomerular filtration. The increases in AUC are ments with both clinical and microbiological out-
more modest than would be seen with most third comes. Doing so would undoubtedly shed further
generation cephalosporins, because the clearance light on the phenomenon of lower AUC and AUIC
of ceftriaxone also occurs via the biliary route into in patients with normal renal function and low al-
the gastrointestinal tract, and there is no enterohep- bumin levels (see section 6). In the case of older
atic cycling because the drug is not absorbed in the patients treated with ceftriaxone, the potential for
gastrointestinal tract. The authors used these data lower protein binding to increase excretion and
to conclude that dosage adjustments were not lower AUC is largely offset by the tendency of their
needed as long as the daily dose was less than 2g. decreasing renal function to raise the AUC, and
As with most of the third generation cephalosporins, thus few patients risk concentrations more than 3
ceftriaxone is not appreciably removed by dialysis, times normal. This extends the ‘accidental spectrum
and a supplemental dose following haemodialysis of action’ of this antibacterial to organisms having
was not recommended.[24] MIC values as high as 2.0 mg/L.
Stoeckel and Koup [25] reported the effects of According to the calculated AUICs, ceftriaxone
hepatic disease on ceftriaxone pharmacokinetics. is an active antimicrobial for infections caused by
Cirrhotic patients without ascites had free AUCs most S. pneumoniae, N. meningitidis, N. gonor-
that were about the same as those of normal vol- rhoeae, H. influenzae and Enterobacteriaceae. It is
unteers, but the AUCs were one-third lower than not reliably active against P. aeruginosa. It has
those of normal volunteers if the patient had cir- variable activity against methicillin-susceptible S.
rhosis. This is presumed to be the effect of lower
aureus, which is dependent on the actual MIC of
serum protein concentrations causing an increased
the infecting organism and the dose used. Ceftri-
free fraction, which is cleared renally. The authors
axone may soon begin to display variable reliabil-
concluded that it was not appropriate to reduce the
ity against penicillin-resistant S. pneumoniae, par-
dosage in hepatically impaired patients with normal
ticularly as the MIC values of these organisms rise
renal function. In these patients, there was more
above 2.0 mg/L and approach 4.0 or even 8.0 mg/L.
free drug, which increased the amount excreted
Anecdotal reports indicate that this is occurring,
renally. The renal elimination pathway was clearly
and the calculated AUICs for the free drug (fig. 3)
more important for this drug.
explain why. These infections include community-
Clearly, anephric patients with major liver im-
pairment would require dosage adjustments, be- acquired pneumonia and bacterial meningitis.
cause they are unable to excrete the drug renally, The AUIC values calculated here support the
offsetting the effect of a higher free fraction and use of ceftriaxone 1g daily in infections where
the consequent ability to clear drug more rapidly. MIC values are below 2.0 mg/L, because in these
In addition, the authors concluded that the dosage conditions the free AUIC will be ≥125. This anal-
of ceftriaxone should not exceed 2 g/day in patients ysis leads us to conclude that the clinical experi-
with either hepatic or renal impairment, because of ences with ceftriaxone are in good concordance
decreased plasma protein binding and higher free with its pharmacokinetics and pharmacodynamics,
AUC values; such a restriction of dosage would be both when total and free concentrations are consid-
cost saving.[25] ered. Consistent with its reported good activity
against CSF organisms with MIC ≤1.0 mg/L, and
7. Conclusion marginal activity against organisms with MIC values
≥2.0 mg/L,[22] we recommend the perspective of free
It is essential to further validate these findings AUICs when prescribing this drug, with the target
by measuring total and free ceftriaxone concentrations for free AUIC being 125. This is the mathematical

© Adis International Limited. All rights reserved. Clin Pharmacokinet 2001; 40 (9)
694 Perry & Schentag

equivalent of the free concentration being above an 15. Highet VS, Forrest A, Ballow CH, et al. Antibiotic dosing issues
in lower respiratory tract infection: population-derived area
MIC of 1.0 mg/L for 80 to 100% of the dosage under inhibitory curve is predictive of efficacy. J Antimicrob
interval, and we advocate that this target be used Chemother 1999; 43 Suppl. A: 55-63
for patients with serious infections. Patients with 16. Craig WA. Choosing an antibiotic on the basis of pharmacody-
namics [discussion 11-2]. Ear Nose Throat J 1998; 77 (6
mild infections may recover with free AUICs be- Suppl.): 7-11
low 125, but they may remain at increased risk of 17. Pollock AA, Tee PE, Patel IH, et al. Pharmacokinetic charac-
teristics of intravenous ceftriaxone in normal adults. Anti-
acquiring or spreading resistant organisms.[29] microb Agents Chemother 1982; 22 (5): 816-23
18. Holazo AA, Patel IH, Weinfeld RE, et al. Ceftriaxone pharma-
cokinetics following multiple intramuscular dosing. Eur J
References Clin Pharmacol 1986; 30 (1): 109-12
1. Rocephin (ceftriaxone sodium) product information. Nutley 19. Patel IH, Kaplan SA. Pharmacokinetic profile of ceftriaxone in
(NJ): Roche Laboratories Inc., 2000. Available from: URL: man. Am J Med 1984; 77 (4C): 17-25
http://www.rocheusa.com/products/rocephin/pi.html
20. Fraschini F, Braga PC, Scarpazza G, et al. Human pharmacoki-
[Acessed 2001 Aug 1]
netics and distribution in various tissues of ceftriaxone. Che-
2. Brogden RN, Ward A. Ceftriaxone: a reappraisal of its antibac-
motherapy 1986; 32 (3): 192-9
terial activity and pharmacokinetic properties, and an update
on its therapeutic use with particular reference to once-daily 21. Patel IH, Miller K, Weinfeld R, et al. Multiple intravenous dose
pharmacokinetics of ceftriaxone in man. Chemotherapy 1981;
administration. Drugs 1988; 35 (6): 604-45
27 Suppl. 1: 47-56
3. Yuk JH, Nightingale CH, Quintiliani R. Clinical pharmaco-
kinetics of ceftriaxone. Clin Pharmacokinet 1989; 17 (4): 22. Schaad UB, Stoeckel K. Single-dose pharmacokinetics of
223-35 ceftriaxone in infants and young children. Antimicrob Agents
4. Thornsberry C, Burton PH, Vanderhoof BH. Activity of peni- Chemother 1982; 21 (2): 248-53
cillin and three third-generation cephalosporins against US 23. Luderer JR, Patel IH, Durkin J, et al. Age and ceftriaxone kinet-
isolates of Streptococcus pneumoniae: a 1995 surveillance ics. Clin Pharmacol Ther 1984; 35 (1): 19-25
study. Diagn Microbiol Infect Dis 1996; 25 (2): 89-95 24. Patel IH, Sugihara JG, Weinfeld RE, et al. Ceftriaxone pharma-
5. Scully BE, Fu KP, Neu HC. Pharmacokinetics of ceftriaxone cokinetics in patients with various degrees of renal impair-
after intravenous infusion and intramuscular injection. Am J ment. Antimicrob Agents Chemother 1984; 25 (4): 438-42
Med 1984; 77 (4C): 112-6 25. Stoeckel K, Koup JR. Pharmacokinetics of ceftriaxone in pa-
6. Blumer J. Pharmacokinetics of ceftriaxone [discussion 52-4]. tients with renal and liver insufficiency and correlations with
Hosp Pract (Off Ed) 1991; 26 Suppl. 5: 7-13 a physiologic nonlinear protein binding model. Am J Med
7. Patel IH, Chen S, Parsonnet M, et al. Pharmacokinetics of 1984; 77 (4C): 26-32
ceftriaxone in humans. Antimicrob Agents Chemother 1981; 26. Schentag JJ, Nix DE, Adelman MH. Mathematical examination
20 (5): 634-41 of dual individualization principles (I): relationships between
8. Stoeckel K. Pharmacokinetics of Rocephin, a highly active new AUC above MIC and area under the inhibitory curve for
cephalosporin with an exceptionally long biological half-life. cefmenoxime, ciprofloxacin, and tobramycin. DICP 1991; 25
Chemotherapy 1981; 27 Suppl. 1: 42-6 (10): 1050-7
9. Van der Auwera P, Klastersky J. Study of the influence of pro- 27. Schentag JJ, Nix DE, Forrest A, et al. AUIC: the universal
tein binding on serum bactericidal titres and killing rates in parameter within the constraint of a reasonable dosing in-
volunteers receiving ceftazidime, cefotaxime and ceftriaxone. terval [editorial; comment]. Ann Pharmacother 1996; 30
J Hosp Infect 1990; 15 Suppl. A: 23-34 (9): 1029-31
10. Leggett JE, Craig WA. Enhancing effect of serum ultrafiltrate 28. Reitberg DP, Cumbo TJ, Smith IL, et al. Effect of protein bind-
on the activity of cephalosporins against gram-negative ba- ing on cefmenoxime steady-state kinetics in critical patients.
cilli. Antimicrob Agents Chemother 1989; 33 (1): 35-40 Clin Pharmacol Ther 1984; 35 (1): 64-73
11. Craig WA. Pharmacokinetic/pharmacodynamic parameters: ra- 29. Thomas JK, Forrest A, Bhavnani SM, et al. Pharmacodynamic
tionale for antibacterial dosing of mice and men [quiz 11-2]. evaluation of factors associated with the development of bac-
Clin Infect Dis 1998; 26 (1): 1-10 terial resistance in acutely ill patients during therapy. Anti-
12. Schentag JJ, Strenkoski-Nix LC, Nix DE, et al. Pharmacody- microb Agents Chemother 1998; 42 (3): 521-7
namic interactions of antibiotics alone and in combination. 30. Hayton WL, Stoeckel K. Age-associated changes in ceftriaxone
Clin Infect Dis 1998; 27 (1): 40-6 pharmacokinetics. Clin Pharmacokinet 1986; 11 (1): 76-86
13. Schentag JJ, Swanson DJ, Smith IL. Dual individualization:
antibiotic dosage calculation from the integration of in-vitro
pharmacodynamics and in-vivo pharmacokinetics. J Anti-
microb Chemother 1985; 15 Suppl. A: 47-57
Correspondence and offprints: Dr Jerome J. Schentag, Uni-
14. Hyatt JM, McKinnon PS, Zimmer GS, et al. The importance of
pharmacokinetic/pharmacodynamic surrogate markers to out- versity at Buffalo School of Pharmacy and Pharmaceutical
come: focus on antibacterial agents. Clin Pharmacokinet Sciences, Cooke Hall 556, Amherst Campus, Buffalo, NY
1995; 28 (2): 143-60 14260, USA.

© Adis International Limited. All rights reserved. Clin Pharmacokinet 2001; 40 (9)