Relationship between the inoculum dose of Streptococcus pneumoniae and pneumonia onset in a rabbit model

Animals and study design

The study was reviewed and approved by the US Army Institute of Surgical Research (Fort Sam Houston, San Antonio, TX, USA) Research Council and Animal Care and Use Committee, the manuscript was reviewed for compliance prior to submission for publication, and guidelines for the care and use of laboratory animals were followed 11.

Male specific pathogen-free New Zealand White rabbits (mean±sd 4.75±0.25 kg (n = 27)) were obtained from Myrtle's Rabbitry, Inc. (Thompson Station, TN, USA). After arrival, all animals were placed in quarantine for 7 days in individual cages and fed ad libitum with water and commercial feed according to current recommendations. The health status of each rabbit was confirmed by the US Army Institute of Surgical Research veterinary staff before commencing each experiment.

Pneumococcal pneumonia was induced as previously described 7, 8. Briefly, on the day of the experiment (day 1), rabbits were anaesthetised intramuscularly with a ketamine–xylazine mixture and orally intubated with a 3.5-mm cuffed endotracheal tube. For inoculation of bacteria, a sterile silicone catheter (diameter 1 mm) was introduced aseptically through the endotracheal tube into the trachea and gently inserted until it reached the bronchi, and 1 mL of freshly prepared pneumococcal inoculum (see below) was flushed slowly through the catheter. Animals were placed in the upright position for 15 s in order to facilitate distal alveolar migration by gravity. Every effort was made to minimise the potential adverse effects of anaesthesia and intubation. After awakening and extubation, the animals were placed back in their individual cages separate from other rabbits and given free access to food and water. The inoculum doses used ranged 3.27–7.56 log₁₀ cfu·mL^-1. Higher doses were not used as they always result in pneumonia onset in rabbits 7. In total, 20 different inoculum doses were used in 24 rabbits. A negative control group included three rabbits that were intubated but not inoculated with S. pneumoniae.

Serotype 3 S. pneumoniae (American Type Culture Collection (ATCC®) 6303; ATCC, Manassas, VA, USA) was used for preparation of the inoculum in the present study. This serotype is frequently seen in patients with community-acquired pneumonia 9, and is widely used in experimental medicine, including in animal pneumonia models 10, 12. A working stock of the organism was made by growing the organism overnight in nutrient broth and then freezing 1-mL aliquots in evaporated milk at -80°C in accordance with standard operating procedures at the US Army Institute of Surgical Research.

One day prior to inoculation, a vial of S. pneumonia ATCC 6303 was removed from the freezer and thawed at room temperature, and 1 mL of the stock pipetted into 25 mL of brain–heart infusion broth. The culture was incubated for 18 h at 37°C in 5% CO₂. On the day of inoculation, the overnight culture was serially diluted in physiological saline to the desired concentration, again according to standard operating procedures of the US Army Institute of Surgical Research.

In order to confirm that live bacteria were inoculated into rabbits, a viable plate count was also performed at this time, for the determination of inoculum concentration. Plate counts were performed in duplicate, using the standard spread plate method on tryptic soy agar. All plates were incubated for 18–24 h at 37°C in 5% CO₂. Colonies were then counted and the number of cfu per mL calculated. Only one strain was used in all experiments.

Rabbits were observed for pneumonia onset over the 96 h after inoculation (days 2–5). The diagnostic criteria for pneumonia were based on the Dennesen et al. 13 approach and on the Clinical Pulmonary Infection Score 14, modified slightly to account for rabbit physiology (higher normal body temperature, etc.) along with tissue microbiological and morphological findings. In the present study, a diagnosis of pneumonia was accepted when at least four of six (i.e. >50%) diagnostic criteria group results were positive (table 1⇓). The accuracy of this diagnostic approach was verified by comparing the results obtained with the histological data. In previous publications, histological results were considered the gold standard for pneumonia verification 15, 16.

Procedures

Computed tomographic (CT) imaging (Aquilion 16; Toshiba America, Inc., New York, NY, USA) of the rabbit thorax was performed under intravenous ketamine–xylazine anaesthesia and mechanical ventilation for ∼10 min at 24, 48, 72 and 96 h after inoculation. The width of the slices was 0.5 mm. CT scans were performed during breath-hold (inspiratory phase) and always before bronchoalveolar lavage (BAL) in order to avoid the possible appearance of artefacts caused by any lavage fluid remaining.

BAL was performed according to previously published recommendations for paediatricians 19. Briefly, BAL was performed immediately after each CT scan under ketamine–xylazine anaesthesia. Sterile technique was utilised during all manipulations. BAL fluid (BALF) was collected by using up to five separate 8-mL aliquots of 0.9% sodium chloride (38°C). BAL was considered complete when 5 mL of the fluid was obtained back. Acquired BALF specimens were transferred to the laboratory within 5 min for microbiological and cellular evaluation. Cells in BALF were separated using a Millipore filter (Millipore Corp., Bedford, MA, USA) and cytocentrifugation. Cell counts and distributions in BALF were determined using a ABX Pentra 120 Blood Analyzer (HORIBA ABX, Montpellier, France).

For microbiological evaluation of BALF, specimens were serially diluted in physiological saline and then spread on tryptic soy agar plates, using 100 μL per plate and two plates per dilution 20. The lavage sample was also streaked on trypticase soy agar plates containing 5% sheep blood and MacConkey II agar plates for culture isolation. All plates were incubated overnight at 37°C in 5% CO₂. All isolates were identified using standard microbiological methods 21. S. pneumoniae identification was based on the presence of alpha haemolysis on trypticase soy agar plates containing 5% sheep blood and the production of a zone of inhibition of ≥14 mm against a 5-μg ethylhydrocupreine disc (optochin test; BBL; Becton Dickinson, Sparks, MD, USA).

For determination of total blood cell count and arterial blood gas levels, 1 mL of blood was obtained from the arterial catheter during initial catheterisation and 24, 48, 72 and 96 h after inoculation. Total cell count and arterial blood gas analyses were performed immediately after sample collection using the ABX Pentra 120 Blood Analyzer and the AVL Omni Modular System (AVL Medical Instruments; Roche Omni Systems, Indianapolis, IN, USA), respectively. All blood samples were collected and analysed according to current clinical standards 24.

Lung tissue specimens were collected from each animal at the time of death or euthanasia. Euthanasia was performed via intravenous administration of a veterinary euthanasia solution (Fatal Plus; Vortech Pharmaceuticals, Dearborn, MI, USA). Autopsy was performed such that the lungs were aseptically removed. For each pulmonary lobe, the macroscopic aspect was noted using a scoring grid 7 based upon human morphological findings (table 2⇓). An overall macroscopic score was calculated as the sum of all lobar macroscopic scores, plus 2 points in the case of pleural effusion (range 0–29). This method has already been successfully used in a rabbit model of pneumonia 7, 8. Macroscopic diagnosis of pneumonia was defined as positive when the score in one lobe was >2.

Two portions of each lobe with the maximum macroscopic lesions were harvested aseptically (one for quantitative cultures and one for histological study) in order to compare histological and bacteriological findings. For pathological studies, specimens were fixed in buffered formalin according to standard methods. A veterinary pathologist, without knowledge of the animal and bacterial data, evaluated the specimens. The lesions were graded into five categories: oedema, alveolitis, bronchiolitis/bronchopneumonia, necrosis, and fibrin vasculitis. Only bronchiolitis/bronchopneumonia and necrosis were accepted as histological confirmation of pneumonia consistent with previous studies 18.

For bacteriological processing of specimens, tissue samples, aseptically taken at necropsy, were flamed quickly in order to remove any surface contaminants and then placed in pre-weighed Stomacher® 80 bags (Seward, London, UK) and weighed to determine tissue weight. Physiological saline was added to the Stomacher bags and the tissue blended in a Stomacher 80 laboratory blender for 60 s at normal speed. The resulting solution was serially diluted and quantitative plate counts and culture isolation were performed as previously described 20, 21. All isolates were identified using standard microbiological methods. S. pneumoniae identification was based on the presence of alpha haemolysis on trypticase soy agar plates containing 5% sheep blood and the production of a zone of inhibition of ≥14 mm against a 5-μg ethylhydrocupreine disc (optochin test).

Data analysis

The independent variable was microbe dose and the dependent variable morbidity. For estimation of the inoculum size corresponding to ED₅₀ in rabbits, the Reed–Muench method was used 23. Logistical regression and probit were also used to evaluate the relationship between S. pneumoniae dose and morbidity in the rabbit pneumonia model and to confirm ED₅₀ estimation using the Reed–Muench method. The sensitivity and specificity of the diagnostic criteria for pneumonia, as well as prevalence of pneumonia, were calculated according to the method of Sackett et al. 24. Fisher's exact test was used to compare categorical variables. Correlation was assessed using the Spearman rank test. A p-value of <0.05 was considered significant. Where appropriate, data are presented as mean±sd.