Source of preterm pigs and initial care. The Institutional Animal Care and Use Committees of the University of Tennessee Health Science Center (location of caesarean section) and the University of Memphis (site of ventilation and critical care) approved the protocols for the harvest, care, and sampling of preterm pigs (Sus scrofa). Antenatal steroids were not provided, and preterm pigs were delivered via caesarean section on gestational day 102 (89% of 115-day term) from two specific pathogen-free, artificially inseminated sows obtained from a production herd with genetics derived from crossing multiple pig strains. After suctioning and clearing the airway, the pigs were placed in a 38-39°C incubator with supplemental oxygen. After spontaneous breathing was established, the pigs were placed in a warmed transport carrier with supplemental oxygen provided by masks that fit over the snout and transported to a neonatal intensive care unit developed for the care of preterm pigs (pNICU).
Instrumentation, processing, and intensive care of preterm pigs.Pigs were weighed in the pNICU. An umbilical catheter (UAC; 3.5F Argyle TM, Covidien, MA) was inserted via one of the two umbilical arteries. The UAC was advanced to the descending aorta, and the position was confirmed by radiography (Duoview high Resolution Digital Radiography System, Revo2, Kennesaw, GA). The UAC was used to provide parenteral nutrition (PN), sample arterial blood, and administer Cefazolin (50 mg/kg/dose) as a prophylactic antibiotic. Maternal plasma (5 ml/kg) was also administered via the UAC to provide passive immunity and compensate for the absence of colostrum.
There were 6 pigs in each group from 2 separate litters included in study. From each litter, 6 pigs (spontaneously breathing and of similar body weights) were randomly allocated to 3 controls and 3 treatments. They were intubated with red rubber 2.5 French endotracheal (ET) tubes (Jorgensen Laboratories, Loveland, CO) that minimize leaks during mechanical ventilation (<10%). The position of the ET tubes was confirmed using digital radiography and repositioned, if necessary. The pigs were connected to Dräger VN500 ventilators (Dräger Medical, Incorporated, Dräger, Telford, PA) with initial assist control volume guarantee (AC+VG) settings of a tidal volume of 5 cc/kg, a respiration rate of 40 breaths per min, positive-end expiratory pressure (PEEP) of 5 cm H2O, iTime 0.35 seconds, and FiO2 of 40%. Surfactant was not administered as it would be difficult to distinguish between endogenous and exogenous sources. The peak inspiratory pressure (PIP) and ventilation rate were adjusted based on the blood gases to maintain normal gas exchange values.
Within 1 hr. after ventilation was established, the pigs were randomized to the KGF treatment (3 per litter) or sham/control group (3 per litter). rhKGF (ProSpec Protein Specialists, Ness Ziona, Israel) is produced by E. coli as a single, non-glycosylated polypeptide chain and reconstituted in normal saline. Based on the mouse homolog, this gene is required for embryonic epidermal morphogenesis, including brain development and lung morphogenesis. This gene may also be a primary factor in wound healing. A single dose of rhKGF (20 µg/kg) was diluted and mixed with normal saline to prepare a volume of 1 ml that was divided into two equal aliquots. These aliquots were administered via the ET tube to each side of the lung. rhKGF was then hand bagged (PIP: 15 cm H2O; PEEP: 5 cm H2O) to enhance adequate distribution throughout the lungs. The control pigs received a similar volume of normal saline. The heart rate, oxygen saturation, and perfusion index were monitored continuously (Masimo Radical 7, Masimo, Irvine, CA). Arterial blood gas measurements (iSTAT®, Abbott, Abbott Park, IL) were performed after placement of the UAC and every 3 hr. or after adjustment of ventilator settings to maintain pulse oximetry saturation of > 90-95%, pH 7.25 to 7.4, pCO2 40 to 55 mmHg, and pO2 > 60. Ventilators automatically recorded the mean airway pressure, PIPs, tidal volumes, and oxygen requirements every 5 min. The pigs were repositioned each hour from one side to the other to avoid dependent edema. All study pigs could spontaneously breathe during the study and were not paralyzed. Pigs that became excessively active during the 24 hr. of mechanical ventilation were sedated using ketamine (Bioniche Teoranta, Galway, Ireland) via the UAC. After 24 hr., the pigs were extubated, and supplemental oxygen was provided by nasal cannula for 12 hr. The pigs were then followed for survival. At the end of the 36-hr study period, all surviving pigs were euthanized (Euthasol; Virbac AH, Inc., Fort Worth, TX, 1 ml/kg; IV).
PN was provided continuously at a rate of 4 ml/kg-hr, beginning immediately after placement of the UAC. For the first 4 to 6 hr, the pigs received a low potassium (2 mmol) PN solution that provided (per L) 116 g dextrose, 60.5 g amino acids (Travasol), and 31.3 g lipid (Intralipid 30%) with electrolytes, vitamins, and trace elements. Thereafter, a PN solution with normal potassium (5 mmol) was provided. Supplemental fluid was provided via the UAC as needed to maintain tissue perfusion using lactated Ringers and averaged 3-4 ml/kg-hr, with the same relative volume (by weight~ 100 ml/kg/day) administered to all pigs to avoid possible differences caused by variable fluid volumes. The volume of fluid administered was insufficient to cause significant pulmonary edema. Metabolic acidosis was corrected with a normal saline bolus (10 ml/kg) as indicated by a base deficit on arterial blood gas.
Radiography.A chest x-ray image was obtained after placement of the ET tube and UAC to confirm proper positioning and to assess initial lung volume recruitment. An additional chest x-ray was obtained at the end of the study to assess lung volume recruitment.
Necropsy. The lungs were collected from pigs that died prior to 36 h and from pigs that were euthanized after 24 h of mechanical ventilation, followed by 12 h of oxygen provided by nasal cannulation. The lungs were removed en bloc and inflated using the ET tube and a NeoPuff ™ (Fisher & Paykel Healthcare, Irvine, CA) to a PIP of 20 cm H2O and PEEP of 5 cm H2O pressure. The trachea was immediately clamped, and the right lower lobe was tied off, excised, and submerged in formalin for routine histology and immunohistochemistry (IHC).
Histologic analysis.Formalin-fixed tissues were processed in paraffin, embedded in paraffin, and sectioned (4 μm). For routine histology, the sections were stained with hematoxylin and eosin. A pediatric pathologist (JZ) who was blinded to the study protocol semi-quantitatively graded inflammation, hemorrhage, edema, necrosis, and atelectasis of each lung. Each parameter was individually scored using a Likert scale from 0 (no injury), 1 (25% injury), 2 (50% injury), 3 (75% injury), and 4 (100% injury) (15).
Immunohistochemistry. For IHC, the sections were deparaffinized, rehydrated with graded ethanol and treated with methanol and hydrogen peroxide to remove any endogenous peroxidase. The sections were treated with guanidinium hydrochloride followed by trypsin to enhance antigen detection. Then, the sections were incubated for 20 min in PBS containing 3% goat serum (Gibco, Thermo Fisher Scientific, Waltham, MA) to block nonspecific binding sites. Following manufacturer instructions, the slides were incubated overnight with primary antibodies for surfactant protein B (SP-B rabbit polyclonal, 20 µg/ml, Hycult Biotech, Wayne, PA) and transforming growth factor β1 (TGF-β1, 2 ng/ml, EMD Millipore, Billerica, MA). Slides were incubated for 1 hr with primary antibodies for E cadherin (0.5 µg/ml, Novus Biologicals, Centennial, CO), Vimentin (1:300), Ki-67 (1:100) and β-catenin (1:50) (Dako North America Inc., Carpinteria, CA). After washing, secondary antibodies (anti-rabbit or anti-mouse biotinylated horseradish peroxidase) were applied for 30 min according to the primary antibody. Color was developed by 3,3'-diaminobenzidine (DAB), and the slides were counterstained with hematoxylin.
Aperio© Image Analysis Algorithm (version 9, Leica Biosystems, Wetzlar, Germany) was used to quantify IHC-stained cells. The algorithm was optimized for fetal pig lung sections stained for β-catenin, E-cadherin, Ki-67, vimentin, prosurfactant, SP-B, and TGF-β. The algorithm classified nuclei as 0, 1+, 2+, and 3+ based on staining intensity. The percentage of positively stained nuclei, average staining intensity of positive nuclei, and percentage of nuclei in each classification were exported as Excel spreadsheets. The spreadsheets were combined into a single master file for each animal.
Categorical data were compared using non-parametric tests. Continuous variables were analyzed using ANOVA for physiologic parameters. Post hoc Tukey’s tests were performed on continuous data. Quantitative immunohistochemistry and histology data were analyzed using a Mann-Whitney U test after testing for normality. Data are presented as the means ± SD. The selected level of significance was p< 0.05.