Experimental model for treatment of extended spectrum betalactamase producing-Klebsiella pneumoniae.

BACKGROUND: Animal models are useful to evaluate the efficacy of antimicrobials in experimental sepsis. AIM: To elucidate the steps of producing an experimental model for the treatment of extended-spectrum beta-lactamase (ESBL)-producing Klebsiella pneumoniae sepsis
METHODS: Several ESBL inoculums ranging from 1.5x109 colony-forming units per milliliter (CFU/mL) to 2.0x1010 CFU/mL were administered by peritoneal injection in adults Wistar rats. Outcomes and microbiological data of quantitative peritoneal and blood cultures were observed in untreated animals. Animals which received 2.0x1010 CFU/mL inoculums were treated with single meropenem dose (30mg/kg) after one hour and those which received 1.0x1010 CFU/mL inoculums were treated immediately with three doses of meropenem 50 mg/kg. Outcomes were observed for 24 hours after inoculation.
RESULTS: Solutions with 1.5 x109 and 6.0x109 CFU/mL were not lethal within 24 hours. Inoculums of 1.0x1010 CFU/mL were lethal in 80% and solutions with 2.0x1010 CFU/mL were lethal in 100% of animals. ESBL lethal sepsis (1.0x1010CFU/mL) was treated immediately with 50 mg/kg of meropenem every eight hours for 24 hours and presented 40% mortality compared with 80% mortality of the control group (p=0.033). Quantitative cultures of peritoneal fluid presented 104 CFU/mL or less for treated animals compared to more than 105 for untreated animals (p=0.001).
CONCLUSION: Inoculums of 1.0x1010CFU/mL achieved the best results to study a model of lethal sepsis and this model of treatment of carbapenem-susceptible Enterobacteriaceae can serve as control to further evaluation of treatment of carbapenemase-producing Enterobacteriaceae models.

INTRODUCTION

The incidence of carbapenemase-producing Enterobacteriaceae has increased and the ideal treatment has not been established. Some retrospective studies suggest an association of different drugs to improve outcomes[10,11,12]. Antibiotic therapy to specific bacteria can be evaluated based on global mortality, time to death and rate of microbiological cure. Considering these findings, an experimental model might be helpful to evaluate the combination of different drugs to treat carbapenemase-producing Enterobacteriaceae until clinical studies confirm the benefits of this approach. To study animal models on carbapenemase-producing Enterobacteriaceae, a treatment control of extended-spectrum betalactamase (ESBL)-producing Klebsiella pneumonia must be validated.

Several animal models of peritonitis, pneumonia and thigh infection after imunossupression using Enterobacteriaceae were reviewed, but none defines a peritonitis model of ESBL-producing Klebsiella pneumoniae treatment with meropenem. Models of peritonitis in rats evaluated inoculums ranging from 105 to 1010 colony-forming unit per milliliter (CFU/mL) of E. coli. Lethal sepsis was observed at higher inoculum concentrations (109to 1010 CFU/mL)[2,3,5]. Non-lethal models were done with 105 to 108 CFU/mL inoculums[1,13]. Klebsiella pneumoniae was evaluated in peritonitis of neutropenic mice (3x105 CFU/mL)[4], thigh infection in neutropenic rats (106 to 108 CFU/mL)[6,8] and pneumonia models in rats (106 to 1010 CFU/mL)[7]. Enterobacter spp. was also evaluated in a pneumonia model of 1010 CFU/mL[9].

Klebsiella pneumoniae inoculum concentrations must be standardized to determine a sepsis model that might be able to evaluate the efficacy of antimicrobial therapy in preventing lethality serving as a control for treatment of carbapenem-resistant Klebsiella pneumoniae.

This study aims to describe the more adequate inoculum concentration to induce lethal but treatable sepsis. Timing and dose of antimicrobial therapy for ESBL peritoneal sepsis induced in non-neutropenic rats were evaluated.

METHODS

Animals

The experiment was performed with adult (20-24 week old) male and female Wistar rats weighting 200-340 g. Animals were maintained under artificial day-night cycles, adequate temperature (22-24 ºC) and humidity. The rats received a standard diet and water ad libitum. Animals were allowed to adapt to laboratory conditions for two days. The animal research ethics committee of the Universidade Estadual de Ponta Grossa approved the study. Fifty rats were included in the phases of this experiment.

Bacterial strain, inoculum production and sepsis induction

ESBL-producing strain (ATCC 700603) was inoculated into Mueller-Hinton broth and incubated at 37° C for 24 h. Colonies were suspended in sterile isotonic saline solution to form the inoculums.

To accurately measure the inoculum a densimeter (Densimat Biomerieux®) capable of measuring densities of 0.5 to 7.5 McFarland was used to evaluate the inoculums of 1.5x109 CFU/mL which was obtained at 5 McFarland. To accurately measure more concentrated inoculums, spectrophotometry (Lambda 25 UV/Vis Spectrophotometer Perkin Elmer®) was performed at optic density of 625 nm. Inoculums with 1.5x1010 and 2.0x1010 CFU/mL corresponded to solutions of barium chloride and sulfuric acid of 50 and 67 McFarland standards and the absorbencies of these solutions were 2.343 and 2.764 respectively. According to Beer-Lambert law, absorbencies over 0.890 are not accurate for measuring microorganism counts. After 1:20 dilution, inoculums with 1.5x1010 and 2.0x1010 CFU/mL presented absorbencies of 0.543 and 0.633. Inoculums of 6.0x109 CFU/mL and 1x1010 were obtained by injection of 0.4 mL and 0.6 mL of 1.5x1010CFU/mL solution.

All inoculums were incubated at Mueller-Hinton and CFU were counted eight hours latter to confirm the concentration before animal injection.

Sepsis was induced by intra-peritoneal injection of the inoculum using a 26 gauge needle in the lower right abdomen. All the procedure was performed under aseptic conditions.

Inoculum lethality was defined by injection of 1.5x109 CFU/mL solution in six animals, 6.0x109 CFU/mL in five, 1.0x1010 CFU/mL in ten animals and 2.0x1010 CFU/mL in ten animals.

Antimicrobial therapy

Two groups of ESBL lethal sepsis were treated with meropenem (Astra-Zeneca®). Twelve rats were inoculated with 2.0x1010CFU/mL and six of them were treated with one dose of meropenem 30 mg/kg after one hour of inoculation. Twenty animals were inoculated with 1.0x1010CFU/mL and ten were treated immediately with 50 mg/kg of meropenem every eight hours for 24 hours. Homogeneous distribution of animals by weight and sex were done in treated and untreated groups.

Outcome evaluation

The rate of lethality, length of survival, blood cultures positivity and quantitative peritoneal fluid and peritoneal tissue cultures were evaluated. Cultures were obtained aseptically.

Animals not presenting lethal sepsis after 24 h suffered euthanasia with lethal doses of xylazine and quetamine.

Blood cultures (0.5-1.0 mL) were collected through cardiac puncture after death or euthanasia and incubated in brain heart infusion broth.

Peritoneal fluid was obtained after laparotomy and injection of 5 mL of isotonic saline and aspiration. One microliter of this fluid was cultured in McConkey agar. Quantitative cultures were performed after 1:100 dilutions of the peritoneal solution in isotonic saline and incubation of 1μL in McConkey agar.

Statistical analysis

Continuous data were expressed as mean±standard deviation (SD), frequencies were expressed as percentages. Dichotomous variables were compared using Mann-Whitney test. Kruskal-Wallis test was used to evaluated hours of survival of the four untreated groups. Significance level was set at 0.05. All data were stored using the software Excel (Microsoft, New York, USA) and statistical analysis was performed using the software SPSS 16 (SPSS, Chicago, USA). Graphics and statistical analysis by Mann-Whitney were performed with GraphPad Prism 5.0 (GraphPad, La Jolla, USA).

RESULTS

ESBL solutions ranging from 1.5x109 to 2.0x1010CFU/mL were evaluated. Solutions with 1.5x109 CFU/mL were not lethal in 100% of animals. Inoculums of 6.0x109 and 1.0x1010CFU/mL were lethal in 80% rats. Solutions with 2.0x1010 CFU/mL were lethal in 100% of animals (Figure 1). ESBL lethal sepsis (2.0x1010CFU/mL) was treated with meropenem one dose of 30 mg/kg after one hour of inoculation with no improvement on mortality. Other group of ESBL lethal sepsis (1.0x1010CFU/mL) was treated immediately with 50 mg/kg of meropenem every eight hours for 24 hours presented 40% mortality, significantly lower than 80% mortality of the control group (p=0.042, Figure 2). Quantitative cultures of peritoneal fluid presented 104CFU/mL or less for treated animals compared to more than 105 for untreated animals (p=0.001, Figure 3)

FIGURE 1

Survival time in hours of untreated animals according to inoculum concentrations (CFU/ml)

FIGURE 2

Survival after 1.0x1010 CFU/mL inocullum in untreated animals and treated imediatelly with higher meropenem dose

FIGURE 3

Quantitative peritoneal cultures after treatment of peritonitis. 2.0x1010 CFU/mL inocullum treated with lower meropenem single dose vs 1.0x1010 CFU/mL inocullum treated imediatelly with higher meropenem dose

DISCUSSION

Previous studies described models of peritoneal inoculums of E. coli between 105 and 1010 CFU/mL to achieve lethal and non-lethal sepsis[2,5,13]. Recent models of neutropenic rats with thigh infection are performed with lower concentrated inoculums and usually do not evaluate mortality, only microbiologic efficacy[8]. Models with Klebsiella spp. are less frequent and must be validated. Here is described the standardization of a lethal model of peritonitis by ESBL-producing K. pneumoniae passible of treatment in non immunosuppressed rats.

Solutions of 108 and 109 UFC/mL cause non-lethal sepsis in immunocompetent rats, which are useful to stratify antimicrobial dosing and compare antimicrobial efficacy on microbiological results, but are not ideal to compare antimicrobial efficacy on clinical outcomes. Was observed that a single antimicrobial dose after inoculation might not be adequate to differentiate treated and untreated animals. Furthermore, was also observed that inoculation of 2.0x1010 CFU/mL with no immediate treatment, cause lethal sepsis that may not be adequate to evaluate antimicrobial efficacy on survival, since most animals may die in spite of treatment.

Inoculums of more than 1.0x1010 and less than 2.0x1010 colony-forming units per milliliter, accurately measured by spectrophotometry, produce lethal sepsis. Immediate treatment after inoculation, administered for 24 hours permits to compared outcomes and microbiological samples of treated and untreated animals. The immediate antimicrobial infusion was based on previous studies[2,5]. Was thought that immediate infusion of antibiotic could reduce the bacterial burden, but both groups had positive cultures in the end of the experiment.

This study validates an animal model of sepsis which induced lethal peritonitis in the control group between six and 24 hours and the treated group had cultures with significantly fewer microorganisms. Data from quantitative cultures, length of survival and mortality can serve as a control to evaluate the treatment of carbapenemase-producing Enterobacteriaceae models.

CONCLUSION

Inoculums of 1.0x1010CFU/mL achieved the best results to study a model of lethal sepsis and this model of treatment of carbapenem-susceptible Enterobacteriaceae can serve as control to further evaluation of treatment of carbapenemase-producing Enterobacteriaceae models.