You are on page 1of 5

Journal of Antimicrobial Chemotherapy (2004) 53, 848–852

DOI: 10.1093/jac/dkh158
Advance Access publication 31 March 2004

Concentrations of fosfomycin in the cerebrospinal fluid of


neurointensive care patients with ventriculostomy-associated
ventriculitis

Bettina Pfausler1, Heinrich Spiss1, Peter Dittrich2, Markus Zeitlinger3, Erich Schmutzhard1 and
Christian Joukhadar3,4*
1Department
of Neurology, University Hospital, Innsbruck; 2Department of Pharmacology and Toxicology,
Karl-Franzens-University, Graz; 3Department of Clinical Pharmacology, Division of Clinical Pharmacokinetics,
University of Vienna Medical School, Allgemeines Krankenhaus, Waehringer Guertel 18-20, A-1090 Vienna;
4Institute of Pharmacology, University of Vienna, Vienna, Austria

Received 1 December 2003, returned 23 December 2003; revised 26 January 2004; accepted 2 February 2004

Objective: The present study was performed to test the ability of fosfomycin to penetrate into the CSF of
neurointensive care patients with ventriculostomy-associated ventriculitis.
Patients and methods: Six patients requiring neurointensive care monitoring, including extraventricular
drainage due to secondary obstructive hydrocephalus, were enrolled into the study. All patients received 8 g
of fosfomycin intravenously three times a day over a period of at least 5 days. Concentrations of fosfomycin
in the CSF and plasma were measured after single-dose administration and at steady state.
Results: Mean values of the fosfomycin area under the time–concentration curves for the dosing interval of
8 h (AUC8) were 929 ± 280 and 225 ± 131 mg·h/L for plasma and CSF after single-dose administration, respect-
ively (P < 0.03). The ratios of the AUC8 for CSF to the AUC8 for plasma were 0.23 ± 0.07 after a single dose and
0.27 ± 0.08 following multiple doses (P > 0.05, not significant). Additional in vitro experiments have shown
that fosfomycin exerts non-concentration-dependent microbial growth inhibition. At steady state, the time
above MIC (t > MIC) values were 98%, 92% and 61% for pathogens with MIC values of 8, 16 and 32 mg/L,
respectively.
Conclusion: The present pharmacokinetic study indicates that 8 g of fosfomycin three times per day should
provide sufficient antimicrobial concentrations in the CSF for the overall treatment period. Thus, the
co-administration of fosfomycin could be useful for the treatment of ventriculitis caused by susceptible
pathogens.

Keywords: CSF, intensive care, pharmacokinetics, humans

Introduction Fosfomycin is a bactericidal antimicrobial agent with high in vitro


activity against a range of Gram-positive and distinct Gram-negative
Severe brain infections require immediate administration of anti- bacteria, such as Pseudomonas aeruginosa,2 Escherichia coli3 and
biotics that effectively combat prevalent pathogens. However, in other Enterobacteriaceae.4 Tissue concentrations of fosfomycin were
some cases empirical antibiotic therapy fails to kill the causative identical to plasma levels in healthy volunteers,5 and differed only
pathogen. This can be attributed to an inadequate choice of anti- moderately in critically ill patients and diabetics.6,7 Thus, although
microbial agent, impaired penetration of the drug into tissues, or to there was almost complete plasma-to-tissue equilibration for tissues
the development of bacterial resistance during antimicrobial therapy. in these patients, we had no knowledge of concentrations of fosfo-
In particular, impaired penetration of antibiotics into brain tissue and mycin in the CSF of patients with ventriculitis. Previous pharmaco-
CSF has been observed in patients with extraventricular drainage kinetic (PK)/pharmacodynamic (PD) simulations have raised the
(EVD) and presenting with ventriculitis.1 speculation that optimal microbial killing by fosfomycin in tissues is
..................................................................................................................................................................................................................................................................

*Corresponding author. Tel: +43-1-40400-2981; Fax: +43-1-40400-2998; E-mail: christian.joukhadar@univie.ac.at


...................................................................................................................................................................................................................................................................

848
JAC vol.53 no.5 © The British Society for Antimicrobial Chemotherapy 2004; all rights reserved.
CSF kinetics of fosfomycin

not necessarily linked to high Cmax values, as fosfomycin exerts was extended to 4 h. Sampling of CSF and plasma was continued for up to
non-concentration-dependent microbial growth inhibition.5,6,8 These 180 h after first drug administration. All samples were stored at –80°C
PK/PD characteristics of fosfomycin would, theoretically, compen- until analysis.
sate for impaired penetration of fosfomycin into brain tissues as
long as concentrations in brain tissue and CSF exceed MIC values for PK calculations
causative pathogens for a distinct period.
PK analysis was performed using the commercially available computer
Thus, the present study addressed the ability of fosfomycin to
software Kinetica 3.0 (Innaphase Sarl, Paris, France). All parameters
penetrate into the CSF of patients with EVD-associated ventriculitis were calculated using non-compartmental analysis. Area under the con-
after single-dose administration and at steady state. In addition, we centration time–curve (AUC) values for plasma and CSF were calculated
aimed to show in vitro that fosfomycin exerts concentration- from non-fitted data by employing the linear trapezoidal rule. The
independent growth inhibition of a select model strain of methicillin- volume of drug distribution (V) and clearance (CL) were calculated for
susceptible Staphylococcus aureus. plasma by use of standard formulae as follows: V = dose/AUCtotal × kel,
CL = kel × V; where kel represents the elimination rate constant. The half-
Materials and methods life calculated for the terminal slope (t1/2β) was calculated by the equation
t1/2β = ln(2)/kel.
The study protocol was approved by the Ethics Committee of the University The following PK parameters were determined: AUC, maximum
of Innsbruck Medical School and was performed in accordance with the concentration (Cmax) and time to maximum concentration (Tmax). Main
Declaration of Helsinki (1964) in the revised version of 1996 (Somerset- PK data are summarized in Table 3.
West), the Guidelines of the International Conference of Harmonization,
the Good Clinical Practice Guidelines and the Austrian drug law. The In vitro experiments
study met all criteria set by the local ethics committee for patients who
were unable to give written consent because of incapacity or being Bacterial strain and antibiotic. S. aureus was obtained from the American
comatose. Type Culture Collection (ATCC 29213). Between experiments,
S. aureus was stored in liquid nitrogen until use. Fosfomycin was
Patients purchased from Biochemie (Austria) and was stored, handled and
prepared according to the manufacturer’s guidelines. In all experiments,
Six Caucasian patients (two women and four men) with severe haemor- glucose-6-phosphate (Boehringer Mannheim, Germany) was added at a
rhage infarction (n = 1) or acute subarachnoid haemorrhage (n = 5) concentration of 25 mg/L according to the NCCLS guidelines.10
requiring EVD due to obstructive hydrocephalus were treated with
broad-spectrum intravenous antibiotic therapy. Patients’ demographic Susceptibility tests. The MIC was determined by a two-fold serial Muel-
and laboratory data are shown in Tables 1 and 2. Bacterial ventriculitis ler–Hinton microdilution method, according to NCCLS criteria.10 S.
was diagnosed based on the isolation of pathogens in the CSF, on inflam- aureus was pre-cultured overnight on a Columbia agar plate (Columbia +
matory parameters in plasma, and elevated counts of leucocytes and pro- 5% sheep blood; bioMérieux, France) and then introduced into Mueller–
tein levels in the CSF. Methicillin-susceptible Staphylococcus aureus Hinton broth (MHB; Mikrobiologie Mueller-Hinton Bouillon, Merck,
(MSSA) was identified in two patients, and methicillin-susceptible Germany) at an initial inoculum of ∼5 × 105 cfu/mL. The lowest concen-
Staphylococcus epidermidis (MSSE) was found in two other patients. No tration of fosfomycin, which inhibited visible bacterial growth after 20 h
pathogen was isolated in the CSF of two patients (Table 2). In these two of incubation at 37°C, was defined as the MIC value.
patients, ventriculitis was diagnosed based on a significant increase in
granulocytes in the liquor, a decrease in the ratio of glucose in liquor to
serum, an increase in protein concentration in liquor and positive MSSE Time–kill curves. Bacterial growth inhibition curves were obtained by
blood cultures. Fosfomycin was administered at a dosage of 8 g three inoculating MHB with ∼5 × 105 cfu/mL of S. aureus ATCC 29213 at
times a day. No interaction of fosfomycin with concomitantly adminis- defined fosfomycin concentrations. Samples were drawn and bacteria
tered drugs occurred. Each patient underwent common diagnostic and counted at defined time-points over a period of 8 h. The bacterial growth
therapeutic intensive care measures. inhibition period of 8 h was chosen because of the three-times daily
dosing regimen of fosfomycin with a dosing interval of 8 h. After vortex-
ing the culture tubes, two 50 µL samples were drawn and serially diluted
Measurement of fosfomycin concentrations with 0.9% sodium chloride. Twenty microlitres of each fraction was
Total fosfomycin levels in plasma and CSF were determined by a plated onto Columbia agar plates, which were incubated for 24 h at 37°C.
previously published but modified gas chromatography method.9 The Afterwards the colonies were counted and back-extrapolated to the
coefficients of variation were 7.0% at 6 mg/L, 6.6% at 240 mg/L and original volume to determine the number of cfu/mL. Controls present
2.2% at 480 mg/L (n = 4). The limit of detection was 1 mg/L. Intra-day bacterial growth in MHB when no antibiotic was added.
and inter-day coefficients of variations were <7%.
Statistical calculations
Study protocol
For statistical comparison of PK parameters, paired Wilcoxon tests were
Patients received 8.0 g of fosfomycin three times a day over a period of at employed. All PK data are presented as means ± S.D. A two-sided P value
least 5 days. Eight grams of fosfomycin dry powder (Fosfomycin <0.05 was considered significant.
‘Biochemie’, 8.0 g, Trockenstechampulle, Biochemie, Vienna, Austria)
was reconstituted with 200 mL of sterile water just before dosing and
administered to patients. Study drug was administered over a period of
Results
∼30 min. The sampling interval was 1–2 h over a period of 8 h after the Patients
first drug administration of fosfomycin. Thereafter, the blood and CSF
samples were collected at 2 h intervals. After 24 h of treatment, steady- Six patients were enrolled in the study. Patient diagnosis on admis-
state conditions of fosfomycin were reached and the sampling interval sion identified pathogens for EVD ventriculitis; the demographic and

849
B. Pfausler et al.

Table 1. Demographic data are presented as means ± S.D. concentrations following single- and multiple-dose administration
(Table 3 and Figure 1; P < 0.03). Steady-state conditions of fosfo-
Parameters Pre-treatment Post-treatment mycin were reached in plasma and CSF after administration of the
third intravenous dose. The mean ratios of the AUC8 for CSF to the
Sex (female/male) 2/4 2/4 AUC8 for plasma were 0.23 ± 0.07 and 0.27 ± 0.08 after single and
Age (years) 53 ± 8 53 ± 8 multiple doses, respectively (Table 3; P = 0.17, not significant).
Weight (kg) 71 ± 9 ND Mean Cmax values were 260 ± 85 mg/L and 43 ± 20 mg/L for plasma
Body mass index (kg/m2) 24.0 ± 1.6 ND and CSF, respectively, following single dose administration of 8 g of
Leucocyte count (109/L) 12.5 ± 2.4 9.1 ± 1.5 fosfomycin (P < 0.03). After multiple doses of fosfomycin, mean
Fibrinogen (mg/dL) 616 ± 72 470 ± 46 Cmax values were 307 ± 101 mg/L for plasma and 62 ± 38 mg/L for
Serum creatinine (mg/dL) 0.82 ± 0.24 0.82 ± 0.23 CSF (P < 0.03). Main PK parameters of fosfomycin for plasma and
C-reactive protein (mg/dL) 13.2 ± 4.5 1.9 ± 1.0 CSF are presented in Table 3.
Cell count in liquor/mm3 757 ± 421 26 ± 8
Protein in liquor (g/L) 146 ± 35 64 ± 27
Glucose liquor/serum ratio 0.39 ± 0.08 0.77 ± 0.07 Safety and tolerability
Study drug administration was well tolerated by all patients. No study
ND, not determined.
drug-related severe adverse events were observed during the
patients’ stay at the neurointensive care unit.

laboratory data are presented in Tables 1 and 2. All patients com- Bacterial growth inhibition experiment
pleted the study and were included in safety analysis.
Bacterial growth inhibition of S. aureus ATCC 29213 by fosfomycin
at different concentrations is presented in Figure 2. Initial inhibition
In vivo experiment
of growth of S. aureus was observed for all fosfomycin concen-
Specimens were collected at defined time-points. However, in some trations used. Similar and effective bacterial growth inhibition was
individuals, sampling of specimens was incomplete or deviated detected for all fosfomycin concentrations that were equal to or
somewhat from the time schedule. The ‘sawtooth’ levels of fosfo- higher than the strain’s MIC value (2 mg/L), which indicates that
mycin in the CSF (Figure 1) resulted from the flushing of the CSF fosfomycin exerts non-concentration-dependent antimicrobial effects.
drain every 2 h with ∼3 mL of physiological saline solution. This was At steady state, the calculated time above MIC (t > MIC) values
carried out to avoid blocking of the drainage system by blood cells. for CSF were 98%, 92% and 61% for pathogens with MIC values of
For this reason, 3 mL of CSF was drawn every 2 h and the drain was 8, 16 and 32 mg/L, respectively.
subsequently rinsed with saline solution. This procedure, however, is
not routine in all intensive care units and concentrations of fosfo- Discussion
mycin are expected to be ∼15% higher in liquor if the EVD is not
periodically flushed with saline solution. One of the most important mainstays in the therapy of severe brain
The time–concentration courses of fosfomycin for plasma as well infections is the administration of antimicrobial agents that are
as for CSF were not significantly different after single and multiple known to reach effective concentrations in the CSF and brain tissue.
doses (Figure 1; P > 0.12; not significant). Concentrations of fosfo- This issue becomes particularly important when looking at recent
mycin in CSF were consistently lower than corresponding plasma reports that have linked the development of bacterial resistance to

Table 2. Patients’ characteristics and concomitant therapya

Duration of Duration of
Age Antibiotic Haemodynamically fosfomycin therapy ventriculitis
Patient no. Sex (years) Underlying disease Pathogen co-treatment active drugs (days) (days)

1 F 47 subarachnoid MSSA amikacin dopamine, dobutamine, 9 5


haemorrhage ceftazidime phenylephrine
2 M 61 subarachnoid MSSE penicillin G dopamine, dobutamine, 9 7
haemorrhage clindamycin norepinehrine
3 M 50 subarachnoid MSSE amikacin dopamine, norepinephrine 10 6
haemorrhage ceftazidime
4 F 63 subarachnoid none netilmicin dopamine, phenylephrine 8 4
haemorrhage ceftazidime
5 M 43 cerebral infarction none penicillin G none 10 5
clindamycin
6 M 56 subarachnoid MSSA clindamycin none 10 4
haemorrhage ceftazidime

aSix
patients were studied.
MSSA, methicillin-susceptible Staphylococcus aureus.
MSSE, methicillin-susceptible Staphylococcus epidermidis.

850
CSF kinetics of fosfomycin

with single-dose administration, although the limited number of


patients and the high inter-individual variability did not allow for sta-
tistical confirmation of this trend (P not significant).
This important finding confirms previous data derived from
patients with brain infections.14,15 Thus, there is evidence that pene-
tration of fosfomycin into the CSF is restricted. This, most probably,
is due to the hydrophilic character of fosfomycin, the partially intact
blood–brain barrier and the presence of transport pumps, such as
P-glycoprotein that actively act against plasma-to-tissue equilibration.
Our observation of incomplete fosfomycin plasma-to-CSF equili-
bration might be of clinical relevance because β-lactams and fosfo-
mycin exert optimal bacterial killing when the time above the MIC
(t > MIC) of the respective target is optimized (Figure 2). Exceeding
Figure 1. Time–concentration profiles of fosfomycin for plasma (open circles) the MIC over a dosing interval of 60%–70% (t > MIC) is considered
and CSF (filled squares) after single and multiple intravenous doses of 8 g over 30 the lower threshold value for favourable clinical and bacteriological
min in neurointensive care patients (n = 6). Each arrow indicates intravenous
outcome.16,17 Several recent clinical studies suggested that t > MIC
administration of fosfomycin. The solid horizontal lines represent MIC values for
pathogens. Data are shown as means ± S.D.
should almost cover 100% of the dosing interval for ‘difficult-to-
treat’ pathogens, although studies in animal infection models have
found that t > MIC does not necessarily need to reach this value to
subinhibitory antibiotic concentrations11–13 at the target site. Against exert a significant antimicrobial effect,18–21 providing that unbound
this background, we set out the present study and addressed the pene- drug levels are used for these calculations. Thus, at steady state,
tration properties of fosfomycin into the CSF in neurointensive care fosfomycin appears to be highly effective when considering calcu-
patients with EVD-associated ventriculitis. lated t > MIC values of unbound fosfomycin in the CSF as values of
In the present study, we found that fosfomycin exerts non-concen- 98%, 92% and 61% will be reached for bacteria with MIC values of
tration-dependent bacterial growth inhibition of S. aureus (Figure 2). 8, 16 and 32 mg/L, respectively. These MIC values do not include
The antimicrobial effect of fosfomycin on Gram-negative bacteria MSSE, but cover frequently isolated bacteria such as methicillin-
has not been explored in the present study, but previous experiments resistant S. aureus (MRSA),22 MSSA,22 Streptococcus species6,22,23
from our group have shown that fosfomycin kills strains of Entero- and Enterobacteriaceae.15 Nevertheless, it is recommended to
bacter cloacae and Serratia marcescens comparable with S. aureus.5 administer fosfomycin in combination with other classes of anti-
Thus, we expect that the antimicrobial effect of fosfomycin on microbial agents to improve antimicrobial activity and to avoid the
susceptible Gram-negative bacteria is comparable with Gram-posi- development of resistance.24 Cephalosporins as a class have been
tive pathogens. Steady-state conditions were observed after the third suggested for combined use with fosfomycin2,25 because of their
intravenous dose of fosfomycin, but the plasma-to-CSF equilibration similar tissue and plasma PK profiles in healthy volunteers5 and
was not complete (Figure 1). Concentrations of fosfomycin in the critically ill patients.6 The antimicrobial efficacy of the simultaneous
CSF were ∼four-fold lower in comparison with plasma (P < 0.03). administration of cephalosporins with fosfomycin versus mono-
This is indicated by the ratio of the AUC8 for CSF to the AUC8 for therapy with fosfomycin to intensive care patients has currently been
plasma of 0.23 ± 0.07 and 0.27 ± 0.08 after single and multiple dose demonstrated in vitro.2
administration, respectively. Peak concentrations of fosfomycin One might expect the ability of hydrophilic compounds such as
increased by ∼18%–40% in plasma and CSF at steady state compared fosfomycin—and the class of cephalosporins—to penetrate the CSF

Table 3. Main PK parameters calculated for the study population (n = 6)

Single dose Multiple doses

Parameters plasma CSF plasma CSF

AUC8 CSF/AUC8 plasma 0.23 ± 0.07 0.27 ± 0.08


AUC8 (mg·h/L) 929 ± 280a 225 ± 131 1035 ± 383b 295 ± 179
Cmax (mg/L) 260 ± 85a 43 ± 20 307 ± 101b 62 ± 38
Tmax (h) 1.2 ± 0.4a 3.8 ± 1.8 1.5 ± 1.2 4.5 ± 2.7
t1/2β (h) 3.0 ± 1.0 ND 4.0 ± 0.5 ND
CL (L/h) 7.4 ± 2.3 ND 5.0 ± 2.0 ND
V (L) 30.8 ± 10.2 ND 26.3 ± 9.7 ND

Data are presented as means ± S.D.


aP < 0.05 compared with CSF single dose.
bP< 0.05 compared with CSF multiple dose.

Cmax, the maximum concentration of fosfomycin; Tmax, the time to reach Cmax; t1/2β, the termi-
nal elimination half-life; AUC, area under the concentration–time curve; CL, total body clear-
ance; V, apparent volume of distribution; ND, not determined.

851
B. Pfausler et al.

7. Legat, F. J., Maier, A., Dittrich, P. et al. (2003). Penetration of


fosfomycin into inflammatory lesions in patients with cellulitis or diabetic
foot syndrome. Antimicrobial Agents and Chemotherapy 47, 371–4.
8. Grif, K., Dierich, M. P., Pfaller, K. et al. (2001). In vitro activity of
fosfomycin in combination with various antistaphylococcal substances.
Journal of Antimicrobial Chemotherapy 48, 209–17.
9. Dios-Vieitez, M. C., Goni, M. M., Renedo, M. J. et al. (1996).
Determination of fosfomycin in human urine by capillary gas chromato-
graphy: application to clinical pharmacokinetic studies. Chromatographia
43, 293–5.
10. National Committee for Clinical Laboratory Standards. (2002).
Performance Standards for Antimicrobial Susceptibility Testing—Twelfth
Informational Supplement M100-S12. NCCLS, Wayne, PA, USA.
11. Gould, I. M. & MacKenzie, F. M. (2002). Antibiotic exposure as a
Figure 2. Growth inhibition versus time curves of a select S. aureus ATCC 29213 risk factor for emergence of resistance: the influence of concentration.
strain, which was exposed to different fosfomycin concentrations over a period of Journal of Applied Microbiology 92, Suppl., 78S–84S.
8 h. 12. Huovinen, P. (2002). Macrolide-resistant group A Streptococcus—
now in the United States. New England Journal of Medicine 346, 1243–5.
13. Dagan, R., Klugman, K. P., Craig, W. A. et al. (2001). Evidence to
and brain tissue, to decrease in the course of treatment because of the support the rationale that bacterial eradication in respiratory tract infection
reconstitution of the blood–brain barrier. As a consequence, the is an important aim of antimicrobial therapy. Journal of Antimicrobial
combination of fosfomycin with lipophilic antimicrobial agents, Chemotherapy 47, 129–40.
such as the fluoroquinolones,26 appears to be a particularly important 14. Brunner, M., Reinprecht, A., Illievich, U. et al. (2002). Penetration
approach to assure that at least one antimicrobial agent sufficiently of fosfomycin into the parenchyma of human brain: a case study in three
patients. British Journal of Clinical Pharmacology 54, 548–50.
penetrates the CSF, even if the blood–brain barrier has recovered.
15. Kuhnen, E., Pfeifer, G. & Frenkel, C. (1987). Penetration of fosfo-
However, we have demonstrated that the ability of the hydrophilic mycin into cerebrospinal fluid across non-inflamed and inflamed menin-
agent fosfomycin to penetrate into the CSF remains unaffected over a ges. Infection 15, 422–4.
period of more than 5 days, despite progressive healing and the 16. Hyatt, J. M., McKinnon, P. S., Zimmer, G. S. et al. (1995). The
improvement of ventriculitis (Table 3 and Figure 1). importance of pharmacokinetic/pharmacodynamic surrogate markers to
In conclusion, the present study has shown that, with the exception outcome. Focus on antibacterial agents. Clinical Pharmacokinetics 28,
143–60.
of MSSE, effective concentrations of fosfomycin can be demon-
17. Scaglione, F. (2002). Can PK/PD be used in everyday clinical
strated in the CSF, covering a range of clinically relevant bacteria, practice? International Journal of Antimicrobial Agents 19, 349–53.
including MRSA, MSSA, Streptococcus species and Enterobac- 18. Craig, W. A. (1995). Interrelationship between pharmacokinetics
teriaceae. Thus, CSF pharmacokinetics of fosfomycin indicate that and pharmacodynamics in determining dosage regimens for broad-
fosfomycin might qualify for the management of severe brain spectrum cephalosporins. Diagnostic Microbiology and Infectious Disease
infections in neurointensive care patients with ventriculostomy- 22, 89–96.
associated ventriculitis. Nevertheless, fosfomycin should be admin- 19. Vogelman, B., Gudmundsson, S., Leggett, J. et al. (1988). Corre-
lation of antimicrobial pharmacokinetic parameters with therapeutic
istered in combination with other classes of antimicrobial agents to
efficacy in an animal model. Journal of Infectious Diseases 158, 831–47.
avoid the development of bacterial resistance.
20. Leggett, J. E., Fantin, B., Ebert, S. et al. (1989). Comparative anti-
biotic dose-effect relations at several dosing intervals in murine pneumo-
References nitis and thigh-infection models. Journal of Infectious Diseases 159, 281–
92.
1. Pfausler, B., Spiss, H., Beer, R. et al. (2003). Treatment of 21. Roosendaal, R., Bakker-Woudenberg, I. A., van den Berg, J. C.
staphylococcal ventriculitis associated with external cerebrospinal fluid et al. (1985). Therapeutic efficacy of continuous versus intermittent
drains: a prospective randomized trial of intravenous compared with administration of ceftazidime in an experimental Klebsiella pneumoniae
intraventricular vancomycin therapy. Journal of Neurosurgery 98, 1040–4. pneumonia in rats. Journal of Infectious Diseases 152, 373–8.
2. Zeitlinger, M. A., Marsik, C., Georgopoulos, A. et al. (2003). Target 22. Van der Auwera, P., Godard, C., Denis, C. et al. (1990). In vitro
site bacterial killing of cefpirome and fosfomycin in critically ill patients. activities of new antimicrobial agents against multiresistant Staphylo-
International Journal of Antimicrobial Agents 21, 562–7. coccus aureus isolated from septicemic patients during a Belgian
national survey from 1983 to 1985. Antimicrobial Agents and Chemo-
3. Ungheri, D., Albini, E. & Belluco, G. (2002). In-vitro susceptibility of
therapy 34, 2260–2.
quinolone-resistant clinical isolates of Escherichia coli to fosfomycin
23. Takahashi, A., Yomoda, S., Tanimoto, K. et al. (1998). Strepto-
trometamol. Journal of Chemotherapy 14, 237–40.
coccus pyogenes hospital-acquired infection within a dermatological
4. Garcia-Rodriguez, J. A., Trujillano, M. I., Baquero, F. et al. (1997).
ward. Journal of Hospital Infection 40, 135–40.
In vitro activity of fosfomycin trometamol against pathogens from urinary
24. Pardo, Z. A., Gimeno, B. L. & Andonegui, O. J. (1977). Evaluation
tract infections: a Spanish multicenter study. Journal of Chemotherapy 9, of fosfomycin in an intensive care unit. Chemotherapy 23, Suppl. 1, 411–5.
394–402. 25. Borowski, J. & Linda, H. (1977). Combined action of fosfomycin
5. Frossard, M., Joukhadar, C., Erovic, B. M. et al. (2000). Distribution with β-lactam and aminoglycoside antibiotics. Chemotherapy 23, Suppl.
and antimicrobial activity of fosfomycin in the interstitial fluid of human 1, 82–5.
soft tissues. Antimicrobial Agents and Chemotherapy 44, 2728–32. 26. Pea, F., Pavan, F., Nascimben, E., et al. (2003). Levofloxacin dis-
6. Joukhadar, C., Klein, N., Dittrich, P. et al. (2003). Target site penetra- position in cerebrospinal fluid in patients with external ventriculostomy.
tion of fosfomycin in critically ill patients. Journal of Antimicrobial Chemo- Antimicrobial Agents and Chemotherapy 47, 3104–8.
therapy 51, 1247–52. –

852