ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Mar. 1996, p. 819–821 Vol. 40, No. 3
Copyright q 1996, American Society for Microbiology
Inﬂuence of Experimental Rat Model of Multiple Organ Dysfunction
on Cefepime and Amikacin Pharmacokinetics
* A. JACOLOT,
AND O. PETITJEAN
Service d’Anesthe´sie-Re´animation Chirurgicale,
and Service d’Anatomo-Pathologie,
94270 Le Kremlin-Biceˆtre, and Cre´pit 93, Centre de Recherche en Pathologie Infectieuse
and De´partement de Pharmacologie-Toxicologie Hospitalie`re,
Hoˆpital Avicenne, 93009 Bobigny, France
Received 8 August 1995/Returned for modiﬁcation 28 September 1995/Accepted 6 December 1995
We adapted an experimental model of multiple organ dysfunction to study the alterations it induces in the
pharmacology of cefepime and amikacin. The half-lives of both antibiotics were signiﬁcantly prolonged because
of nonsigniﬁcant enhancement of the volume of distribution and reduced renal elimination. In the presence of
multiple organ dysfunction, the concentration of each antibiotic in the lungs, compared with that in the lungs
of healthy controls, was signiﬁcantly decreased, despite similar concentrations in plasma, indicating that the
application of a standard antibiotic concentration in plasma could lead to underdosage in tissues during the
initial days of therapy.
Previous studies have shown that the pharmacokinetic pa-
rameters of aminoglycosides are profoundly modiﬁed in inten-
sive care unit patients, but little is known about other classes of
antibiotics, such as b-lactams (16). In this study, we adapted an
experimental model described by Goris et al. (5) by injecting
smaller doses of zymosan to evaluate the multiple organ dys-
function-induced pharmacokinetic modiﬁcations of a new b-
lactam, cefepime, in plasma and lungs. Amikacin was also
studied to compare the pharmacokinetic modiﬁcations in-
duced by our model with those observed in humans and thus to
determine whether it could constitute a valuable model.
Experimental multiple organ dysfunction. Twenty Wistar
rats were randomly assigned to a zymosan group or a healthy
group (10 per cohort). Each rat in the former group received
an intraperitoneal (i.p.) injection of 0.5 mg of zymosan (Sigma)
per g suspended in liquid parafﬁn to induce a chronic release of
inﬂammatory mediators, which resulted in maximal multiple
organ dysfunction 12 days later. Previous studies showed that
all of the surviving animals 12 days after zymosan challenge
improved until complete recovery. Because zymosan induced
acute peritonitis during the ﬁrst 3 days after its administration,
0.1 mg of imipenem-cilastatin (Merck Sharp and Dohme-Chi-
bret) per g was administered simultaneously to diminish the
initial mortality due to overwhelming infection. No i.p. injec-
tion was given to rats in the healthy group which served as a
control. To simulate further the human situation, all rats re-
ceived i.p. injections of 0.9 M saline equal to 0.025 ml/g on days
0, 1, 2, 9, 10, and 11 following randomization. Clinical condi-
tions were recorded daily, and on day 12, all surviving rats were
anesthetized and bled by aortic puncture to determine biolog-
ical parameters, and the lungs, kidneys, spleen, liver, and a part
of the digestive tract were removed and weighted. Relative
organ weights were calculated by using the formula (organ
weight/body weight) 3 100. After ﬁxation with 4% formalde-
hyde, microscopic sections of these organs were prepared for
staining with hematoxylin-eosin-saffron and examined (n 5 six
per group). Studies of cefepime and amikacin (Bristol-Myers
Squibb) pharmacokinetics were performed when multiple or-
gan dysfunction was well established, i.e., 12 days after i.p.
injection of zymosan, and the results were compared with those
of similar studies with healthy rats.
Pharmacokinetic parameters. Each rat (ﬁve or six per
group) received a single 1-ml i.p. injection of either 50 mg of
cefepime per kg or 18 mg of amikacin per kg. Multiple blood
samples (300 ml) were collected 30, 60, 90, 120, 150, 180, 240,
300, and 360 min after antibiotic injection. Saline (600 ml) was
injected intraarterially after collection of each blood sample to
restore blood volume. Individual cefepime and amikacin phar-
macokinetic parameters were determined. The peak concen-
tration in plasma and the time it took to reach that peak were
the experimental values. A noncompartmental model was used
to ﬁt the data by weighted least squares with the weighting
to determine half-life, volume of distribution in
the b phase total body clearance, and the area under plasma
concentration-time curve extrapolated to inﬁnity for cefepime
and amikacin in healthy rats and rats with multiple organ
dysfunction, taking the fraction of the cefepime or amikacin
dose absorbed to be equal to 1 (Siphar).
Antibiotic concentrations in lung tissue. Six to 10 other rats
per group were sacriﬁced 1, 2, and 3 h after being given a single
1-ml i.p. injection of 50 mg of cefepime per kg and 18 mg of
amikacin per kg. Blood was collected to determine the antibi-
otic concentration in plasma. The lungs were homogenized in
1 ml of saline, and the resulting suspension was centrifuged for
10 min at approximately 1,000 3 g and 38C. The supernatant
was collected, and antibiotic levels were assayed. Antibiotic
measurements were not corrected for blood contamination
because blood represents less than 6% of lung weight and
neither antibiotic accumulated in erythrocytes; therefore, the
presence of blood does not lead to substantial errors in the
evaluation of amikacin or cefepime levels (11).
Measurement of antibiotic concentrations. The cefepime
concentration was determined by using a modiﬁed version of
the high-performance liquid chromatography assay described
by Barbhaiya et al. (2). The quantiﬁcation limit of the assay was
1 mg/liter, and the between-days coefﬁcient of variation ranged
from 7% at 1 mg/liter to 6% at 50 mg/liter. The amikacin
concentration was determined by using an immunoenzyme as-
* Corresponding author. Phone: (33) 1 45213441. Fax: (33) 1
say (EMIT assay). The limit of quantiﬁcation was 1 mg/liter,
and the coefﬁcients of variation were below 8% over the entire
range of measurement. The biological matrices used for high-
performance liquid chromatography and EMIT standards were
bovine plasma for plasma samples and water for tissue sam-
Statistical analysis. Results are expressed as medians with
ranges. Data from healthy rats and rats with multiple organ
dysfunction were compared by using the Mann-Whitney U
test. A P value below 0.05 was considered to be signiﬁcant.
From days 1 to 4, all rats in the zymosan group showed
symptoms of illness. They became lethargic and anorexic, hy-
perventilated, and had rufﬂed fur and diarrhea. Epistaxis and
bleeding conjunctivae were also observed. From days 4 to 8,
the rats improved, becoming very active, gaining weight, and
no longer hemorrhaging or having diarrhea. Between days 9
and 12, their general condition deteriorated again, with in-
creasing hyperventilation, tachycardia, and loss of weight. All
control rats remained healthy throughout the experimental
period and progressively gained weight. One zymosan group
rat died on day 10. The remaining 19 rats were sacriﬁced on
day 12 for biological and histological studies. Premortem ex-
amination showed that healthy rats gained signiﬁcantly more
weight during the study period than did rats with multiple
organ dysfunction (7.8 versus 1.5%; P , 0.002). Postmortem
examination of rats with multiple organ dysfunction showed
signs of extensive acute peritonitis, edematous lungs, pale liver,
and large spleen, whereas postmortem examination results of
healthy rats were normal. As shown in Table 1, in the presence
of multiple organ dysfunction, the relative weights of each
organ was signiﬁcantly increased, indicating increased water
content. The results of blood analyses are summarized in Table
2. Experimentally induced multiple organ dysfunction gener-
ated several microscopic modiﬁcations in different organs.
Lung specimens showed thickening of alveolar walls, edema,
and inﬁltrates of mononuclear cells. Kidney specimens exhib-
ited focal necrosis of proximal tubules and edema. Spleens
were depleted of lymphoid tissue, while the numbers of mega-
karyocytes and erythroblasts were increased. Liver specimens
showed engorged veins, inﬂammatory lymphocyte inﬁltrates,
and hyperplasia of Kupffer cells. The wall of the small intestine
specimen was inﬁltrated by mononuclear cells.
Cefepime and amikacin pharmacokinetic parameters in
healthy animals were similar to those previously reported (Ta-
ble 3). For both antibiotics, the presence of multiple organ
dysfunction signiﬁcantly increased the antibiotic elimination
half-lives because of nonsigniﬁcant enhancement of the vol-
ume of distribution and reduced renal elimination (Table 3).
Despite similar concentrations of the two antibiotics in plasma
in both study groups, their concentrations in the lungs were
signiﬁcantly lower in rats with multiple organ dysfunction 1 and
2 h after administration (Table 4). No antibiotic was detected
in the lungs 3 h after injection.
Experimental models of multiple organ dysfunction in small
laboratory animals generally utilize a bacterial or endotoxin
challenge (7, 12, 17). The mortality induced by these models is
high. Recently, Goris et al. (5) developed an experimental
animal model of multiple organ dysfunction that we adapted by
using a lower dose of zymosan to stimulate the human clinical
syndrome by its prolonged activation of the alternative com-
plement pathway and macrophages induced by an i.p. injection
of zymosan into the rats (5). Twelve days after zymosan injec-
tion, the clinical, biochemical, and microscopic alterations in
our model closely resembled those reported in critically ill
septic patients and in other animal models of bacterium or
endotoxin-induced multiple organ dysfunction (5). Impor-
tantly, mortality in our model was relatively low in comparison
with that observed in previously reported models. Modiﬁca-
TABLE 1. Comparison of relative organ weights between healthy
rats and those with multiple organ dysfunction
Median relative organ wt (range)
(n 5 10)
Rats with multiple
(n 5 9)
Lung 0.4 (0.3–0.4) 0.7 (0.4–0.9) 0.002
Liver 3.3 (2.8–3.7) 4.0 (3.6–4.7) 0.002
Kidney 0.6 (0.5–0.6) 0.7 (0.6–0.8) 0.009
Spleen 0.2 (0.2–0.4) 0.7 (0.5–1.1) 0.001
TABLE 2. Comparison of blood analysis results of healthy
rats and rats with multiple organ dysfunction
(n 5 10)
Rats with multiple
(n 5 9)
(mm Hg) 92 (82–103) 77 (65–90) 0.01
Glycemia (mmol/liter) 9.1 (7.2–11.3) 7.0 (4.3–9.1) 0.003
Creatinine (mmol/liter) 54 (38–58) 51 (43–55) 0.5
92 (69–114) 127 (85–150) 0.003
Bilirubin (U/liter) 7 (4–19) 12 (4–14) 0.2
/liter) 3.0 (1.6–8.2) 5.4 (3.7–8.2) 0.02
37 (16–61) 60 (55–72) 0.004
Lymphocytes (%) 59 (36–78) 35 (22–42) 0.003
Hematocrit (%) 41 (39–46) 37 (35–38) 0.001
/liter) 710 (532–852) 1,263 (1,086–1,486) 0.001
TABLE 3. Pharmacokinetic parameters measured in plasma after
i.p. injection of 50 mg of cefepime per kg or 18 mg of amikacin per
kg into healthy rats or rats with multiple organ dysfunction
Rats with multiple
(mg/liter) 54 (39–83) 78 (60–94) 0.17
(min) 30 (30–60) 30 (30–30) 0.21
(liters/kg) 0.4 (0.4–0.5) 0.5 (0.4–0.8) 0.27
(ml/min z kg) 7.2 (6.3–7.3) 6.7 (4.3–9.6) 0.29
(h) 0.7 (0.6–0.8) 0.9 (0.7–1.0) 0.05
AUC (h z mg/liter) 116 (113–131) 123 (96–195) 0.37
(mg/liter) 18 (15–41) 24 (22–27) 0.34
(min) 30 (30–30) 30 (30–30) 1
(liters/kg) 0.6 (0.3–0.8) 0.7 (0.5–0.9) 0.37
(ml/min z kg) 8.7 (4.3–9.8) 5.3 (3.5–7.2) 0.15
(h) 0.9 (0.9–1) 1.2 (1–2.4) 0.01
AUC (h z mg/liter) 35 (31–70) 57 (42–86) 0.15
, peak concentration in plasma; T
, time to peak concentration in
, volume of distribution in the b phase; CL
, total body clearance;
; half-life during the b phase; AUC, area under the plasma concentration-
There were seven healthy rats in the cefepime group and six in the amikacin
There were six rats each with multiple organ dysfunction in the cefepime and
820 NOTES ANTIMICROB.AGENTS CHEMOTHER.
tions of cefepime and amikacin pharmacokinetics induced by
our experimental multiple organ dysfunction were comparable,
with signiﬁcantly increased half-lives and nonsigniﬁcant en-
hancement of the volume of distribution. Expanded volumes of
distribution and prolonged half-lives of aminoglycosides and
some b-lactams (including cefepime) have already been re-
ported for critically ill subjects (1, 3, 6, 9, 10, 15, 16). The
increase in the volume of distribution is largely explained by
the enhancement of extracellular ﬂuid following endothelial
damage and tissue edema (4, 13). The slight increase in the
volume of distribution in our model compared with that ob-
served in humans could be explained by the lack of various
factors affecting pharmacokinetic parameters, i.e., massive vol-
ume expansion, mechanical ventilation, positive end-expiratory
pressure, fever, old age, etc., which could not be reproduced in
our model (8, 14). The addition of an experimental infection,
which could modify per se the pharmacokinetic parameters of
both antibiotics and emphasize the modiﬁcations observed (es-
pecially enhancement of the volume of distribution could be
expected), needs to be addressed. The variability of the phar-
macokinetic parameters observed in animals with multiple or-
gan dysfunction was not greater than in healthy rats, which is
different from reported ﬁndings on humans. However, inten-
sive care unit patients do not form a homogeneous population
and the gravity of the illness varies from one patient to an-
other. By contrast, with the same dose of zymosan the intensity
of inﬂammatory system stimulation was similar among all of
the animals, and this can explain the lack of great variability
among pharmacokinetic parameters in rats with multiple organ
Concentrations of both antibiotics in the lungs were signif-
icantly lower in rats with multiple organ dysfunction than
in healthy rats, despite similar concentrations measured in
plasma. Even if we were unable to conﬁrm that antibiotic
concentrations in the extracellular ﬂuid proportionally de-
crease with regard to antibiotic concentrations in whole tissue,
in light of our results we can assume it. Multiple organ dys-
function induces an increase in the water content of organs
(Table 1) that is located primarily in the extracellular space (4,
13). Since b-lactams and aminoglycosides are distributed in the
total extracellular water, the proportion of ﬂuid containing
antibiotics was increased and the apparent antibiotic level in
the whole tissue should have increased, but the opposite was
observed, suggesting that antibiotic concentrations in the ex-
tracellular space were decreased in animals with multiple or-
gan dysfunction. Decreased blood ﬂow to the tissues, as ob-
served in humans and other experimental models (12, 17),
could explain, at least in part, the results observed.
Our ﬁnding that in the presence of multiple organ dysfunc-
tion, concentrations of both antibiotics in the lungs were sig-
niﬁcantly decreased, despite their similar concentrations in
plasma, suggests that the recommendation of a standard anti-
biotic concentration in plasma could lead to suboptimal con-
centrations in tissues during the initial days of therapy. These
pharmacokinetic alterations need to be conﬁrmed, and their
effects in an experimental model of infection must be evalu-
We thank R. J. A. Goris and Monique Jansen for help with the
adaptation of the zymosan model and P. Norwood for rereading the
We thank Bristol-Myers Squibb (France) for supporting this work.
1. Albanese, J., C. Martin, P. Saux, M. N. Mallet, B. Bruguerolles, X. Viviand,
P. de Micco, and F. Gouin. 1989. Pharmacocine´tique d’un aminoside: la
tobramycine, et de deux b-lactamines: l’azlocilline et la ce´fme´noxime chez
des patients en ventilation me´canique. Med. Mal. Infect. 19:372–376.
2. Barbhaiya, R. H., S. T. Forgue, W. C. Shyu, E. A. Papp, and K. A. Pittman.
1987. High-pressure liquid chromatographic analysis of BMY-28142 in plas-
ma and urine. Antimicrob. Agents Chemother. 31:51–59.
3. Beckhouse, M. J., I. M. Whyte, P. L. Byth, J. C. Napier, and A. J. Smith.
1988. Altered aminoglycoside pharmacokinetics in the critically ill. Anaesth.
Intensive Care 16:418–422.
4. Ellman, H. 1984. Capillary permeability in septic patients. Crit. Care Med.
5. Goris, R. J. A., W. K. F. Boekholtz, I. P. T. van Bebber, J. K. S. Nuytinck, and
P. H. M. Schillings. 1988. Multiple organ failure and sepsis without bacteria.
An experimental model. Arch. Surg. 121:897–901.
6. Kieft, H., A. I. M. Hoepelman, C. A. Knupp, A. van Dijk, J. M. Branger, A.
Struyvenberg, and J. Verhoef. 1993. Pharmacokinetics of cefepime in pa-
tients with the sepsis syndrome. J. Antimicrob. Chemother. 32(Suppl. B):
7. Lang, C. H., G. J. Bagby, G. H. Bornside, L. J. Vial, and J. J. Spitzer. 1983.
Sustained hypermetabolic sepsis in rats: characterization of the model. J.
Surg. Res. 35:201–210.
8. Mackowiak, P. A. 1989. Inﬂuence of fever on pharmacokinetics. Rev. Infect.
9. Mann, H. J., D. W. Fuchs, R. Awang, F. A. Ndemo, and F. B. Cerra. 1987.
Altered aminoglycoside pharmacokinetics in critically ill patients with sepsis.
Clin. Pharm. 6:148–153.
10. Marik, P. E. 1993. Aminoglycoside volume of distribution and illness severity
in critically ill septic patients. Anaesth. Intensive Care 21:172–173.
11. Nix, D. E., D. Goodwin, C. A. Peloquin, D. L. Rotella, and J. Schentag. 1991.
Antibiotic tissue penetration and its relevance: models of tissue penetration
and their meaning. Antimicrob. Agents Chemother. 35:1947–1952.
12. Oh, G. R., L. M. Mela-Riker, R. E. Bryant, and D. K. Lowe. 1984. A new
experimental model of chronic high-output sepsis. Circ. Shock 13:99–104.
13. Tanaka, H., T. E. Dahms, E. Bell, K. S. Naunheim, and L. J. Baudendistel.
1993. Effect of hydroxyethyl starch on alveolar ﬂooding in acute lung injury
in dogs. Am. Rev. Respir. Dis. 148:852–859.
14. Triginer, C., R. Fernandez, I. Izquiero, and S. Benito. 1989. Gentamicin
pharmacokinetic changes related to mechanical ventilation. Drug Intell.
Clin. Pharm. 138:499–504.
15. Triginer, C., I. Izquiero, R. Fernandez, J. Rello, J. Torrent, S. Benito, and A.
Net. 1990. Gentamicin volume of distribution in critically ill septic patients.
Intensive Care Med. 16:303–306.
16. Van Dalen, R., and T. B. Vree. 1990. Pharmacokinetics of antibiotics in
critically ill patients. Intensive Care Med. 16(Suppl. 3):235–238.
17. Wichterman, K. A., A. E. Baue, and I. H. Chaudry. 1980. Sepsis and septic
shock: a review of laboratory models and a proposal. J. Surg. Res. 29:
TABLE 4. Concentrations of cefepime or amikacin in plasma and
lungs measured 1, 2, and 3 h after antibiotic administration in
healthy rats and rats with multiple organ dysfunction
No. of rats, median cefepime
No. of rats, median amikacin
Rats with mul-
Rats with mul-
Plasma 10, 57 (21–82) 9, 76 (40–97) 10, 26 (20–39) 9, 22 (16–36)
Lungs 10, 23 (8–28) 9, 13 (7–15)
10, 4 (3–6) 9, 2 (2–2)
Plasma 10, 35 (26–38) 8, 29 (20–38) 10, 9 (7–16) 8, 13 (10–16)
Lungs 10, 9 (3–13) 8, 3 (2–6)
Plasma 6, 19 (9–28) 6, 13 (9–18) 6, 5 (3–7) 6, 7 (5–9)
Lungs 6, ,16,,16,,16,,1
Concentrations in plasma are in milligrams per liter. Concentrations in lung
tissue are in micrograms per gram.
P , 0.05 for rats with multiple organ dysfunction versus healthy rats for the
VOL. 40, 1996 NOTES 821