Animals
In total 28 pigs (Sus scrofa domesticus) were used, 20 for combined experiment 1 and 2, and 8 for experiment 3. The pigs weighed 50–70 kilos and were housed per pairs in a pen with water and standard food ad libitum. A standard 12-hour light/dark cycle was maintained. All pigs had a standard acclimatization period of one week before surgery. The pens were enriched with toys. All animals were checked at least twice a day and controls consisted of body weight, temperature, apathetic behavior and poor food intake. Humane endpoints were defined according to the local animal welfare body guidelines. In the event of a deteriorated condition, such as being unable to get up or eating or drinking independently, the veterinarian was contacted to discuss further policy (treatment or withdrawal from the experiment). In between surgeries, animals were housed on a farm after the first postoperative week.
Hemostatic patches
Experiment 1
Two experimental patches: 1) GFC-NHS-POx (GATT Technologies, Nijmegen, The Netherlands): this hemostatic patch consists of a Gelatin Fibrous Carrier (GFC) with NHS-ester functionalized polyoxazoline polymer P(EtOx-c-OH-c-NHS) 60-20-20 (NHS-POx). 2) ORC-PLGA-NHS-POx which is an Oxidized Regenerated Cellulose and PolyLactic-co-Glycolic Acid (ORC-PLGA Ethicon, New Jersey, USA, 10 x 5 cm) carrier with NHS-POx combined on a 1:1 molar ratio with P(EtOx-NH2) 80 − 20 (referred to as nucleophilically activated polyoxazoline, NU-POx).
One negative control: GFC blank: Gelatin Fibrous Carrier (GFC) (Gelita Rapid, origin: porcine, 5 × 7.5 cm, Gelita Medical GmbH, Eberbach, Germany). This patch is a control for the polymer effect.
Two positive controls: 1) TachoSil® (Takeda Austria GmbH, Linz, Austria) is a collagen sponge coated with the human coagulation factors fibrinogen and thrombin (18). 2) Veriset™ (Covidien Inc., Mansfield, MA, USA) consists of oxidized regenerated cellulose impregnated with buffer salts, triglycine and a reactive polyethylene glycol (17).
Experiment 2
Three experimental patches: 1) GFC-NHS-POx 1.5: GFC with NHS-POx 1.5 is made in the same way as GFC-NHS-POx, but contains 1.5 times more Gelatin and 1.5 times more NHS-POx. 2) GFC-NHS-POx and 3) ORC-PLGA-NHS-POx (see experiment 1)
Two positive controls: TachoSil® and Veriset™ (see experiment 1)
Experiment 3
One Experimental patch: GFC-NHS-POx:NU-POx (referred to as ‘GATT-Patch’): GFC with NHS-POx and added 20% as nucleophilically activated polyoxazoline (NU-POx).
One Negative control: GFC blank (see experiment 1)
Two positive controls: TachoSil® and Veriset™ (see experiment 1)
Study design
The study was approved by the Dutch Animal Ethics Committee (CCD, numberAVD10300 2015 348) and performed in the Central Animal Laboratory of Radboud University. The experiments were performed according to ARRIVE guidelines (22).
The study was divided into three experiments (Fig. 1). In experiment 1, twenty pigs underwent a laparotomy, and were randomly divided into five groups of four animals surviving 1, 2, 3, 4, or 6 weeks. Each pig received 15 standardized punch bleedings, ten on the liver and five on the spleen. Five products (two prototype patches, two positive controls and one negative control) were applied three times (two on liver, one on spleen) per pig for evaluation of hemostasis of standardized moderate bleedings (23). At relaparotomy after 1, 2, 3, 4 or 6 weeks, the sites of the former patch application were inspected for signs of tissue and foreign body responses and biopsied for histological analyses. After this inspection, pigs were used for a second experiment. During the relaparotomy repeated partial liver lobe and spleen resections were performed on previously uninjured parts under normal and heparinized conditions (Fig. 2). The severe bleedings were treated with five products (three prototypes and two positive controls) and hemostasis was assessed.
Based on results of both experiments, the prototype GFC-NHS-POx patch was optimized by adding a cross-linker to reduce stickiness to surgical gauze (‘GATT-Patch’), and evaluated in a third experiment. In this experiment, eight pigs underwent a laparotomy to inflict standardized liver punch bleedings and perform partial liver and spleen resections for assessment of rapid, within 10 seconds, hemostasis with the GATT-Patch and two positive and one negative controls.
Surgical Procedures
Pigs were anesthetized according to the standard protocol of the animal research facility (see appendix). After disinfection of the abdomen, sterile drapes were placed and a midline incision was performed from xyphoid to the umbilicus. During all surgeries, blood pressure, heart rate, CO2, O2 saturation, temperature and breathing frequencies were recorded after every five applications of patches. At the end of the surgery in experiment 1, the abdominal wall was closed in two layers: the fascia with a continuous suture (PDS VIO LOOP 150cm M4, Ethicon, Sommerville, NJ), the skin with intracutaneous continuous suture (Vicryl 2.0 Ethicon, Sommerville, NJ), after which anesthesia was ended, the pig decannulated and brought to the pen to recover. At the end of the surgeries in experiment 2 and 3, animals were euthanized by an overdose of pentobarbital.
For the heparinized condition, heparin was given by intravenous bolus injection to increase activated clotting time (ACT) to 2–4 times the baseline value. ACT was measured with an Istat® portable clinical analyzer (Abbott Point of Care Inc. Princeton, NJ USA) with ACT cartridges before the start of surgery for baseline, at least 5 minutes after heparin was administered, and after each block of 5 treatments. Heparin was titrated according to ACT values to stay between 2–4 times baseline.
Experiment 1: Surgical punch (moderate) bleedings
Bleeding was inflicted with an 8 mm biopsy punch with a depth of 3 mm: n = 5 on the left lateral lobe of the liver, n = 5 on the left medial lobe of the liver, and n = 5 on the spleen (Fig. 2). Blood flow was determined by collecting outflowing blood in a pre-weighed gauze for 30 seconds, measuring the difference in weight, corresponding to the amount of blood loss.
Each of three locations received five different patches; patches were allocated to the lesions using computer generated block randomization. Each patch was applied with digital pressure for one minute, except for TachoSil®, which needs a pressing time of three minutes according to the instructions for use (18). Pressure was gently applied using a standard woven, wetted surgical gauze. Upon removal of the gauze, hemostasis was assessed by visual inspection by two observers at 10 seconds (secondary efficacy endpoint) and 3 minutes (primary efficacy endpoint) after application. When the patch failed to stop the bleeding within 3 minutes after gauze removal, rescue measures (e.g. additional digital pressure) were taken to stop the bleeding with as final rescue measure the application of Veriset™.
At 1, 2, 3, 4 or 6 weeks pigs underwent a relaparotomy, at which the abdominal cavity was macroscopically inspected for abnormalities. Adhesions to the liver and spleen at patch locations were categorized as: 0, no adhesions; 1, filmy adhesions: easy to separate by blunt dissection; 2, stronger adhesions: blunt dissection possible but partly sharp dissection necessary (beginning of vascularization); 3, strong adhesions: lysis possible but sharp dissection only; clear vascularization; 4, very strong adhesions: lysis possible by sharp dissection only (organ strongly attached with severe adhesions and damage of organs hardly preventable) (24). Macroscopic signs of patch remnants were evaluated as present or absent. Liver and spleen were biopsied at patch locations (10 x 5 mm and 5 mm deep) and stored in formaldehyde solution for histological analyses.
Experiment 2: liver and spleen resection (severe) bleedings with and without heparin
At relaparotomy (experiment 1), after macroscopic inspection of the abdominal cavity, adhesion scoring and evaluation of patches’ remnants, the 20 pigs were re-used for the second experiment. A series of four partial liver resections of 5cm in length and 1.5cm in depth was performed with surgical scissors (Fig. 2 provides an overview of the location of the resections) to generate a severe bleeding. The patch was folded around the resected area and digitally pressed with a wet gauze for two minutes (3-minute for TachoSil® per instructions for use). Upon removal of the gauze, hemostasis was assessed at 10 seconds (secondary endpoint) and 3 minutes (primary endpoint), with rescue measures as described previously. Subsequently, four partial spleen resections ensuring the splenic artery was cut were made with surgical scissors to create an arterial, pulsating, severe bleeding. The patch was applied by folding it around the resected area and pressed as described for the liver resection. Upon removal of the gauze, hemostasis was again assessed at 10 seconds and 3 minutes with rescue measures as previously described. After hemostasis, a new resection and bleeding was created ~ 3 cm more proximally on the spleen.
During experiment 2, each product was tested on severe bleedings four times on the liver and four times on spleen. These were distributed over two non-heparinized and two heparinized pigs.
Experiment 3
After opening the abdomen and surgical exposure, liver punch bleedings and liver and spleen resections were performed followed by patch application, as described above, in eight pigs. Four patches’ types were evaluated, each applied on 6 liver punches, 6 liver resections, and 2 spleen resections divided over the eight pigs. In contrast to the former experiments, pressure was applied for only 10 seconds for all patches, including TachoSil®. After removal of the gauze, hemostatic efficacy was assessed at 10 seconds (primary endpoint) and at 3 minutes (secondary endpoint). When after 10 seconds no hemostasis was achieved, an additional 30 seconds of pressure time was administered and repeated as needed until the 3-minute secondary endpoint.
Histology
Liver and spleen biopsies (experiment 1) were fixated using a 4% formaldehyde solution. Tissues were infiltrated and embedded in paraffin, sectioned (4 µm) using a microtome, pasted on superfrost-plus slides (Thermo-Scientific, Waltham, MA, USA), and dried overnight in the stove at 56°C. Three different stains were done: HE (Hematoxylin and eosin), MSB (Martius Scarlet Blue) and SR (Sirius Red Staining). All slides were digitalized using a Panoramic p250 scanner (3DHISTECH, Budapest, Hungary). For scoring, the slides were randomly mixed (for sequence). Two researchers independently scored the thickness of the patch, inflammatory cells, and wound healing using a digital scoring form. The thickness of the patch was digitally measured (in mm) using Pannoramic viewer (3DHISTECH, Budapest, Hungary). Degradation of the patches in time was initially quantitatively scored by the number of times the patch was still visible by researchers during the surgical inspection of the patches and subsequently qualitatively by the microscopic thickness of the patch on histology. Inflammatory cells in the patch, new tissue, and in the liver under the wound as well as the amount of fibroblasts in the new tissue were scored using the adjusted Ehrlich and Hunt scale (25) (scoring: 0: no evidence. 1: occasional evidence. 2: light scattering. 3: abundant presence. 4: confluent cells or fibers). Wound healing was scored by general wound healing, according to Shafer (26) (scoring: 1: very light healing. 2: moderate healing. 3:advanced healing. 4: well organized). The scores of the two researchers were compared; differences in scoring were discussed until consensus.
Statistical analysis
Continuous variables are presented as mean (SD); discrete data are presented as frequencies. Data on the primary and secondary endpoints of hemostatic efficacy in experiment 1 are presented as percentages of hemostatic success (pass/fail) at three minutes and Linear Mixed Models were used for comparing efficacy between groups with post-hoc LSD test; we tested for the dependency of the bleeding rate as a modifying factor confirming no correlation between bleedings within subjects. Hemostatic efficacy during experiment 2 and 3 were descriptively evaluated and percentages of hemostatic success (pass/fail) were depicted. Furthermore, histological outcomes were descriptive evaluated without further analyses. All analyses were performed using SPSS (version 25, IBM corporation, Armonk, NY) and the significance level was set at a p-value of < 0.05.