Journal of Antimicrobial Chemotherapy (2002) 49, 331-335
© 2002 The British Society for Antimicrobial Chemotherapy
ß-Lactam modification of the bacteraemic profile and its relationship with mortality in a pneumococcal mouse sepsis model
a Centro Nacional de Microbiología, Instituto de Salud Carlos III, Ctra. Majadahonda-Pozuelo, Km. 2, 28220 Majadahonda, Madrid; b Medical Department, GlaxoSmithkline, Tres Cantos, Madrid, Spain
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResults and discussionAcknowledgementsReferences |
|---|
A sepsis BALB/c mice model was used to investigate the relationship between mortality and the bacteraemic profile produced by a serotype 6B Streptococcus pneumoniae clinical isolate (MIC/MBC of amoxicillin 4/4 mg/L and of cefotaxime 2/4 mg/L). Animals were treated subcutaneously with doses of amoxicillin or cefotaxime ranging from 6.25 to 50 mg/kg tds for 48 h, starting 1 h after intraperitoneal inoculation (2 x 107 cfu/mouse). Blood cultures were carried out daily over 15 days. A survival rate of 100% was obtained with amoxicillin 25 mg/kg and of 60% with cefotaxime 50 mg/kg. A statistically significant (P = 0.012) relationship was found between the maximum cfu/mL in blood and mortality. A maximum log cfu/mL of 6.5 was associated with an 84% probability of death.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionAcknowledgementsReferences |
|---|
Two parameters are commonly used for the evaluation of antibiotic efficacy in animal models: bacterial counts in tissue fluids and survival rate.1 The second parameter is easy to assess and seems more relevant from the clinical perspective; a curative dose is used as a reliable end-point for showing subtle differences in antibiotic behaviour.2 Reduction in bacterial counts provides information on bactericidal activity in vivo;1 nevertheless, in vitro inocula and an inocula reduction rate equivalent to those in vivo have never been established3 and reduction in bacterial count may not be related to animal survival.2
Bacterial virulence is the minimal bacterial mass capable of producing injury to a given host4 and can be defined as the competence of an infectious agent to produce pathological effects, as indicated by case fatality rates and/or the ability to invade the host.5 This study explores the modification by ß-lactam therapy of the bacteraemic profile (representative of ability to invade the host) produced by a bacterial mass (representative of virulence) and its relationship with mortality rates (representative of competence to produce pathological effects). A serotype 6 penicillin-resistant isolate of Streptococcus pneumoniae was used as infecting strain, as this is one of the most frequently isolated serotypes in bacteraemia and respiratory-tract infections in Spain.6,7
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionAcknowledgementsReferences |
|---|
Infecting strain
A serotype 6B S. pneumoniae strain isolated from a blood culture [MIC and minimum bactericidal concentration (MBC) of penicillin 2 and 4 mg/L, respectively] was used. The microorganism was grown until an absorbance of 0.3 (UV-VIS spectrophotometer, Shiimadzu UV-1203, Japan) was obtained in Todd–Hewit broth supplemented with 0.5% yeast extract (THYB) (Difco, Detroit, MI, USA), and aliquots were stored at -70°C in 15% glycerol.
In vitro studies
MICs and MBCs of amoxicillin and cefotaxime were determined by broth dilution following NCCLS procedures.8 Modal values of five separate determinations were considered.
Animals
Eight- to 12-week-old female BALB/c mice weighing 19–22 g were used. The study was approved by the Spanish Central Laboratory of Public Health—Instituto de Salud Carlos III.
Determination of minimal lethal dose
Groups of 10 mice per dilution were injected intraperitoneally with 0.2 mL of different inocula, 102, 104, 106 and 108 cfu/mL (spectrometrically measured), to determine the minimal dose that produced a 100% mortality over a 15 day follow-up period (minimal lethal dose, MLD). Bacteria in a logarithmic phase of growth in THYB were centrifuged and the pellets washed three times and resuspended in PBS pH 7.2 to reach the desired turbidity. The inoculum was confirmed by the culture of serial dilutions onto blood Mueller–Hinton agar incubated at 37°C in 5% CO2 air. Mouse mortality was recorded daily. The MLD was determined from the results obtained in three independent experiments.
Dose-ranging treatment
Survival and the bacteraemic profile of inoculated animals over a 15 day follow-up period were determined in a dose-ranging study with amoxicillin and cefotaxime doses ranging from 6.25 to 50 mg/kg. Animals were inoculated intraperitoneally with 200 µL of the MLD, and antibiotic treatment was initiated 1 h after bacterial inoculation. Groups of five animals per dose were treated with 100 µL subcutaneously tds for 48 h. The five animals of the control group received placebo (apyrogen sterile distilled water). Animals were observed and deaths were recorded for 15 days.
Blood samples were obtained daily (except on day 1 when they were collected at 2, 6 and 24 h) over the 15 day follow-up period, from five animals per antibiotic dose, to study the bacteraemic profile. Tails were disinfected and anaesthetized, and the end portion of the tail was amputated with scissors. Using a calibrated loop, 0.008 mL of blood were taken and resuspended in Todd–Hewitt broth containing 10% trisodium citrate and plated onto blood agar for colony counting to obtain the first sample. Plates were incubated at 37°C in 5% CO2 air for 24 h. To obtain the subsequent blood samples, the crust was removed and the anaesthetized tail was pressed to collect the 0.008 mL. The lower limit of detection was 1.25 x 102 cfu/mL.
Determination of antibiotic concentrations in serum
Amoxicillin and cefotaxime concentrations in serum were determined in healthy animals after a single subcutaneous dose of the dose that obtained 100% survival or of the maximum dose tested if 100% survival was not achieved. Blood samples were collected 5 min, 15 min, 30 min, 1 h, 2 h, 4 h, 6 h and 8 h after dose administration from groups of five animals per dose and antibiotic. Concentrations were measured by bioassay using Micrococcus luteus ATCC 4698 for amoxicillin and Escherichia coli ATCC 25922 for cefotaxime as reference organisms in 9 cm diameter plates with 14 mL of antibiotic agar no. 2 (Difco) for amoxicillin and Mueller–Hinton agar for cefotaxime, containing a final inoculum of 8 x 108 cfu/mL. Aliquots (30 mL) of each sample were deposited into 6 mm diameter wells in inoculated plates that were incubated at 36.5°C for 18 h. Standards containing 0.012–1.6 and 0.4–50 mg/L were prepared in apyrogen distilled water for amoxicillin and cefotaxime, respectively, to determine the assay regression line (standard curve) and to extrapolate the antibiotic concentrations from the corresponding inhibition zone diameters.
Pharmacokinetic study
Concentration–time curves for each antibiotic were analysed by a non-compartmental approach using the Win-Nonlin program (Pharsight, Mountainview, CA, USA). The areas under the serum concentration–time curves 0 h to 
(AUC0–) were calculated from the equation AUC0– = AUC0–480min + AUC480min–. The values for AUC0–480min were calculated from plots of serum concentrations versus time by using the trapezoidal rule. The values for AUC480min– were calculated from the expression AUC480min– = C480min/ß, where ß is the slope obtained from least-square regression of the terminal elimination phase. The value of ß was calculated for each antibiotic using at least the last three sample time values of serum concentrations. The theoretical concentration at time 0 (obtained by back-extrapolation to the origin of the elimination regression line) was considered the maximum concentration in serum (Cmax). Time above MIC (
T > MIC) was calculated graphically from the semi-logarithmic plot representing the concentration–time data.
Statistical analysis
Survival curves were obtained by the Kaplan–Meier method. A dose-adjusted Cox regression analysis was used to compare survival with each antibiotic. A Probit regression analysis was carried out to calculate the 50% efficacy dose (EC50) in the theoretical model using the dose as covariable and the antibiotic as factor. A binary logistic regression analysis was used to study the relationship between the maximum cfu/mL and outcome (death/survival) per individual.
|
Results and discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionAcknowledgementsReferences |
|---|
Modal MIC/MBC values of amoxicillin and cefotaxime were 4/4 and 2/4 mg/L, respectively, for the infecting strain. Considering the MIC90 of both drugs determined in previous studies in Spain,6,7 the susceptibility of the strain used in the present study represents the resistance pattern in this country. The MLD was 108 cfu/mL (i.e. 2 x 107 cfu/mouse). Since 108 cfu/mL was the minimal bacterial mass producing 100% mortality, this value could be considered representative of the virulence of this type 6B strain. When evaluating the competence of this MLD to produce pathological effects through the ability to invade the host, by studying the bacteraemic profile, it was observed that the mean blood colony counts in untreated animals were
107 cfu/mL at 2, 6 and 24 h sample times (Figures 1 and 2
), with all untreated animals having bacterial counts >106 cfu/mL until death.
|
, inoculum; 
, 12.5 mg/kg (death); *, 12.5 mg/kg (survival); 
, 25 mg/kg; •, 50 mg/kg.
|
The Table
shows percentages of survival with the different antibiotic regimens over the first 144 h of the follow-up period. Survival percentages at 360 h were identical to those at 144 h. The mean survival time obtained for amoxicillin was greater than that obtained for cefotaxime at each dose (mean days for amoxicillin versus cefotaxime: 2.1 vs 1.7 for 6.25 mg/kg, 8.7 vs 2.9 for 12.5 mg/kg, >15 vs 2.9 for 25 mg/kg and >15 vs 11 for 50 mg/kg). Statistically significant (P = 0.014) differences were found between both antibiotics in survival curves. The dose obtaining EC50 in the theoretical model was 12.6 mg/kg for amoxicillin and 47.5 mg/kg for cefotaxime; the median relative efficacy between antibiotics was 0.2657 favouring amoxicillin.
|
The Cmax and the AUC were, respectively, 270.6 mg/L and 7657.7 mg/min/L for amoxicillin 25 mg/kg and 621.4 mg/L and 16483.8 mg/min/L for cefotaxime 50 mg/kg. The time that serum concentrations were higher than the MIC (
T > MIC) was 123 min for amoxicillin 25 mg/kg and 156 min for cefotaxime 50 mg/kg, representing 26% and 32% of the dosing interval, respectively. Previous studies have shown that T > MIC of 40% of the dosing interval is needed for ß-lactams to obtain therapeutic efficacy.9 Percentages are lower for penicillins than for cephalosporins (reflecting the fastest killing with penicillins);9 this fact is corroborated in the present study, since 100% survival was obtained with T > MIC of 26% for amoxicillin 25 mg/kg but not with T > MIC of 32% for cefotaxime 50 mg/kg.
Figures 1 and 2
show the mean bacteraemic profile (blood culture cfu/mL over 360 h) with the different amoxicillin and cefotaxime doses, with animals divided in each study group by outcome: death and survival. As can be seen, mean blood colony counts in surviving mice were 
106 cfu/mL over the 360 h. In contrast, maintained mean bacteraemic colony counts >106 cfu/mL over the first 144 h were found in those animals that died. This can be clearly seen with the amoxicillin 12.5 mg/kg and cefotaxime 50 mg/kg regimens, where some animals survived and some died, and where both bacteraemic profiles were present. A statistically significant (P = 0.012) relationship was found between the maximum cfu/mL in blood and mortality, with ß0 = -15.21 and ß1 = 2.5944 as logistic regression coefficients. In the theoretical model, a maximum log cfu/mL of 6.5 was associated with an 84% probability of death.
In contrast to previous studies, where pneumococcal counts in blood were poorly related to the outcome of infection in penicillin-treated animals,10 in the present study, using a penicillin-resistant strain, bacterial counts were good indicators of mouse survival and death. The two ß-lactams (amoxicillin to a higher degree) reduced fatality rates of the serotype 6B pneumococcal strain by decreasing colony counts of the bacteraemic profile.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionAcknowledgementsReferences |
|---|
We thank J. Prieto and M. L. Gómez-Lus (Universidad Complutense, Madrid, Spain) for their technical advice and critical review of the manuscript, A. Carcas (Universidad Autónoma, Madrid) for the pharmacokinetic analysis, A. Pedromingo (www.e-biometria.com) for the statistical analysis and F. Molero (Centro Nacional de Microbiología. Instituto de Salud Carlos III) for her technical assistance. This study was supported by European Funds for Regional Development and by funds from the Spanish National R+D Program (Project 2FD 97-0554).
|
Notes |
|---|
* Corresponding author. Tel: +34-91-509-7975; Fax: +34-91-509-7966; E-mail: jcasal{at}isciii.es
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionAcknowledgementsReferences |
|---|
1 . Frimodt-Moller, N., Knudsen, J. D. & Espersen, F. (1999). The mouse peritonitis/sepsis model. In Handbook of Animal Models of Infection, (Zak, O. & Sande, M. A., Eds), pp. 127–36. Academic Press, London.
2 . Frimodt-Moller, N. (1993). The mouse peritonitis model: present and future use. Journal of Antimicrobial Chemotherapy 31 Suppl. D, 55–60.
3 . Amsterdam, D. (1990). Assessing cidal activity of antimicrobial agents: problems and pitfalls. The Antimicrobic Newsletter 7, 49–56.
4 . Lorian, V. & Gemmell C. G. (1991). Effect of low antibiotic concentrations on bacteria: effects on ultrastructure, virulence, and susceptibility to immunodefenses. In Antibiotics in Laboratory Medicine, 3rd edn, (Lorian, V., Ed.), pp. 493–555. Williams & Wilkins, Baltimore, MD.
5 . Gemmell, C. G. & Lorian, V. (1996). Effects of low concentrations of antibiotics on bacterial ultrastructure, virulence, and susceptibility to immunodefenses: clinical significance. In Antibiotics in Laboratory Medicine, 4th edn, (Lorian, V., Ed.), pp. 397–452. Williams & Wilkins, Baltimore, MD.
6
.
Fenoll, A., Jado, I., Vicioso, D., Pérez, A. & Casal, J. (1998). Evolution of Streptococcus pneumoniae serotypes and antibiotic resistance in Spain. An update: 1990–1996. Journal of Clinical Microbiology 36, 3447–54.
7
.
Marco, F., Bouza, E., García-de-Lomas, J., Aguilar, L. & the Spanish Surveillance Group for Respiratory Pathogens. (2000). Streptococcus pneumoniae in community-acquired respiratory tract infections in Spain: the impact of serotypes, geographical, seasonal and clinical factors on its susceptibility to the most commonly prescribed antibiotics. Journal of Antimicrobial Chemotherapy 46, 557–64.
8 . National Committee for Clinical Laboratory Standards. (2001). Performance Standards for Antimicrobial Susceptibility Testing—Eleventh Informational Supplement: Document M100-S11. NCCLS, Wayne, PA.
9 . Craig, W. A. (1998). Pharmacokinetic/pharmacodynamic parameters: rationale for antibacterial dosing of mice and men. Clinical Infectious Diseases 26, 1–10.[Web of Science][Medline]
10 . Frimodt-Moller, N. & Frolund Thomsen, V. (1986). The pneumococcus and the mouse protection test: inoculum, dosage and timing. Acta Pathologica Microbiologica et Immunologica Scandinavica Sect. B 94, 33–7.
Received 15 June 2001; returned 25 September 2001; revised 17 October 2001; accepted 12 November 2001
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||