Small intestinal segments were collected from 15 client-owned horses euthanized for reasons unrelated to disease of the gastrointestinal system. Horses were free of any signs of colic within 24 hours before euthanasia and owner consent was obtained for donation for research. Intestinal segments were harvested immediately after euthanasia and anastomosis, bursting pressure test and lumen reduction measurements were all performed within the following 4 hours.
Intestinal specimens
A segment of 1.5 to 2 meters of jejunum was harvested. Three to five centimeters of mesentery was kept on the mesenteric border. The segment was lavaged with tap water to remove ingesta and stored in a saline (0.9% NaCl2) solution at room temperature throughout the study except during anastomosis and testing procedures. Three samples of 30 to 40 cm were obtained from each harvested segment to perform a two-layer hand-sewn anastomosis (2L-CT group), a one-layer hand-sewn anastomosis sealed with an UV-PMA layer (1L-UV-PMA group) and a control segment (Control group).
Anastomosis techniques
A two-layer anastomosis and a one-layer anastomosis with application of methacrylate glue were realized on each sampled horse by the same ECVS Resident. All intestinal samples for the suture and glue groups were transected at mid-distance from each end at about 60° to the mesenteric attachments before performing the anastomosis.
2-layer Anastomosis (2L-CT Group)
Polyglycolic acid USP 2-0 on a taper cutting needle3 was used in a hemicircumferential simple continuous full thickness pattern to appose all layers of the intestinal wall, which was interrupted at mesenteric and anti-mesenteric borders to avoid a string-purse effect. Suture bites were taken approximately 5 mm apart and 3 mm from the incised edge. The same suture material was then used in a hemicircumferential Cushing pattern, started at the middle of the first layer sutures (3 and 9 o’clock) in order to avoid knot superposition between the two layers. Bites were placed 3 mm from the first layer and 5 mm apart (Figure 1).
Construction time from the first bite of the second layer to the final knot was recorded.
1-layer Anastomosis and UV-PMA(1L-UV-PMA Group)
The same technique as in the 2L-CT group was used for the first layer. After this the intestinal sample was dried with a gauze swab, before application of the glue.
The adhesive consists of 2 separate layers and is liquid in its initial form. It solidifies with the aid of UV-light at a wavelength of 395 nm.
The first layer of the surgical glue was applied directly on the suture, on a width of about 5 mm on each side of the suture and was polymerized by the action of the UV LED curing lamp for 30 seconds. Then the second layer was applied over the first one and polymerized by the action of the UV LED curing lamp for 30 seconds. The intestine was rolled between the surgeon’s fingers while the UV lamp was applied to polymerize the glue on the entire circumference of the bowel.
Construction time for glue application was recorded from the end of the first suture to completion of the second 30-sec UV-light period.
Evaluation of Luminal Diameter Reduction (LDR)
Before mechanical testing, each segment was distended to an intraluminal pressure of 20 mmHg, as previously described (5). Ultrasonography was used to determine luminal diameter reduction for 2L-CT and 1L-UV-PMA groups.
Ultrasonography was performed using a 6-15 MHz linear transducer.4 For each group, longitudinal images on the anastomosis were taken. Care was taken to obtain the largest diameter possible. The samples were held in a linear position by an assistant. Luminal diameter measurements were determined from these images using a linear function of the ultrasound unit, on the anastomosis site (Measurement A) and at 2 cm proximal and distal to the anastomosis (Measurements B and C, Figure 2).
Mean value of measurements B and C served as reference for normal luminal diameter. LDR at the anastomosis site was calculated by dividing the luminal diameter calculated at the anastomosis site (Measurement A) by the mean normal diameter:

Bursting Strength Pressure (BSP) Testing
After completion of the anastomosis, the intestinal segment was placed in a water tank for bursting strength pressure testing. The same method as previously described was used (3,5,6,31–35). Briefly, each intestinal segment was submerged in 30 L saline solution within a water tank at room temperature. Infusion sets were inserted at each end of the intestine, and a knot was made with a polypropylene string around the intestine to provide a watertight seal between the intestinal wall and the infusion set (Figure 3). One infusion set was connected to a roller pump5 for fluid delivery. The other infusion set was connected to a T-connector attached to a pressure transducer.6 The pressure transducer was then connected to a computer to assess real-time pressure within the bowel on a dedicated software.7 A balanced electrolyte solution (Hartmann’s solution) tainted with methylene blue (2mL/L, 0,2%) was infused in the bowel at constant rate (700 mL/min, pump’s maximum) until failure occurred. Bursting Strength Pressure (BSP) was determined as the maximal pressure obtained before failure. Failure was first detected as apparition of blue tainted flow inside the saline bath. With time visual rupture of the intestine could be observed. Every trial was video recorded and re-assessed to describe the mode of failure.
Modes of failure were recorded as: “Extremities” when the knots around the infusion sets ruptured before the intestine; “Mesenteric Mucosa” when the mucosa and muscularis ruptured and the intestine ballooned without rupture of the serosa on the mesenteric border of the intestine at a location more than 2 cm away from the suture; “Mesenteric Suture” when the suture line ruptured at the mesenteric border of the intestine, “Non-mesenteric Suture” when the suture ruptured at a location different from the mesenteric border (Figure 4).
The same mechanical testing was performed for the control group.
Macroscopic evaluation of the anastomosed segments was performed before and after testing. The anastomosis before and after BSP testing was observed for abnormalities and shredding of the glue, and the mode of failure was confirmed after BSP test by visual assessment.
Statistical Analysis
Descriptive statistics were reported as mean [95% CI]. Data was tested for normality with a Shapiro-Wilk test.
A paired t-test was realized to evaluate a statistical difference in construction time, BSP and LDR.
Mode of failure was assessed using a chi²-test between suture-related failures (Mesenteric and Non-mesenteric Suture types of failure) and non-suture-related failures (Mesenteric Mucosa types of failure).
Statistical analysis was performed using excel software8 and for all statistical tests a P<0.05 was considered significant.