All adult Labrador dogs and Bama miniature pigs (provided by Taizhou Meifengli Animal Experimental Center, Jiangsu Province) weighed 20–30 kg. There were 3 male and 3 female dogs, and 2 male pigs and one female pig. All animal experiments were performed in accordance with the Guidelines for Ethical Review of Laboratory Welfare and followed the Regulations on the Administration of Laboratory Animals of the Ministry of Science and Technology, PRC. All animals were housed in single rooms at a temperature of 16–26 ℃ and fed standardized food with free access to food and water.

All animals received an intramuscular injection of Zoletil (7–25 mg/kg) for sedation and anaesthesia, which were maintained by 1–6% of isoflurane via endotracheal intubation during the operation. Blood pressure, heart rate, electrocardiogram, and oxygen saturation were monitored during surgery.

For this model, we established a stenosis model with thrombin, forming an immediate and solid thrombus in the left iliac vein. Animals were placed in a supine position after anaesthesia in a hybrid operating room. Conventional skin sterilization was performed in abdominal surgery, and a median abdominal incision was made from the first to the third nipple. Then, we incised the skin and subcutaneous tissues to the rectus abdominis muscle and separated the lateral peritoneum from the left abdominal wall to the posterior peritoneum. We suspended the left iliac vein and artery with an absorbable thread after exposing the abdominal aorta, inferior vena cava, and left common iliac artery and vein (Fig. 1A). The perimeter of the left common iliac vein was measured, and the diameter of the common iliac vein was calculated (Fig. 1B, C). The diameter of the ligating glass rod was calculated according to the degree of 80% stenosis (Fig. 1D). The left common iliac vein near the bifurcation of the inferior vena cava was tied with a rod of the calculated diameter, and then the rod was pulled out to show the constriction of the iliac vein at the ligation site (Fig. 1E). The diameter of the iliac vein after ligation was remeasured with silk thread, and the percentage by which the vein diameter was reduced was calculated (Fig. 1F, G). Then, the first vascular clip was placed at the ligation site, and the second clip was used at the distal end from the ligation site, taking care to avoid the branch veins. The filling and dilatation of the common iliac vein were observed, and we measured the distance between the two clips to determine the thrombus length (Fig. 1H). Immediately after that, we sequentially injected 1 ml thrombin (500 IU/mL, Harbin HanBang Medical Science and Technology Co.) and 1 ml fibrinogen (60 mg/mL, Harbin HanBang Medical Science and Technology Co.) with a 5 ml syringe under direct vision, and the blood flow was blocked with clips for 10 min (Fig. 1I). Streptomycin 300 mg and ceftiofur sodium 300 mg were given postoperatively for 7 days. Studies have shown that rivaroxaban, at a dosage of 1–2 mg/kg per day, has a significant anticoagulant effect in both pigs and dogs [12,13,14,15]. To maintain the thrombus until the chronic phase, each animal was given oral treatment with rivaroxaban tablets at 20 mg per day, starting 2 weeks after thrombosis model establishment.

Induced venous thrombosis in the left iliac veins of canines. A Separated bilateral iliac veins. B, C Measured iliac vein circumference. D Diameter of the ligation rod. E Ligation of the rod and left iliac vein. F, G Remeasured iliac vein circumference. H Thrombosis length. I Injection with thrombin and fibrinogen

Venous blood samples were collected from the right jugular vein before the induction surgery for venous thrombosis and on Days 7, 30, 60, and 90 after the surgery. Blood samples were taken from dogs and pigs in the fasting state for the following tests: prothrombin time (PT), international normalized ratio (INR), white blood cell count, neutrophil count, lymphocyte count, monocyte count, total protein, albumin, and creatinine. Testing was performed at Taizhou Meifengli Animal Experimental Center.

Left iliac vein blood flow was observed by anterograde angiography at 7, 30, 60, and 90 days after thrombus model establishment. Under anaesthesia as described previously, the left inguinal area was disinfected and sheeted. The femoral vein was located by ultrasound, a 5F puncture sheath was placed into the femoral vein, and 5 ml heparin water (8 U/mL, Shanghai ShangPharma First Biochemical Company) was injected through the catheter to prevent thrombosis from occluding the contrast catheter. Iodixanol contrast agent (100 mL, General Electric Pharmaceutical Co.) and heparin water (10 ml, 1:1) were added for imaging. The left iliac vein should be visualized, and the contrast agent should be returned to the inferior vena cava directly through the partially recanalized veins or the compensatory collateral vessels until the contrast agent dissipates in the visual field.

At 90 days after venous thrombosis induction, the bilateral iliac veins of the surviving animals were isolated and fixed in 4% paraformaldehyde for at least 24 h after dissection, followed by dehydration in ethanol with a gradient of concentrations. The dehydrated samples were embedded in paraffin and serially sectioned at 4 μm thickness. These sections were subjected to haematoxylin–eosin (HE) and Masson staining and photographed by light microscopy.

We calculated the number (%) for categorical variables and the mean ± standard deviation for normally distributed data. The results of thrombosis modelling in pigs and dogs were tested by independent sample T test (continuous variables) or Chi-square test (categorical variables). A two-sided p value < 0.05 represented statistically significant differences. We analyzed all data using the SPSS 26.0 statistical package and Prism 8.4.2 software.