Cranial vena cava syndrome refers to the constellation of clinical signs resulting from the partial or complete obstruction of the cranial vena cava (2). There are numerous causes of cranial vena cava syndrome, such as thrombus, neoplasia, and granuloma (2). These factors pose a high risk of obstruction in the cranial vena cava, resulting in edema of the head, neck, and forelimb, and, occasionally, pleural effusion.

Among several causes, thrombus formation can be described by what is called Virchow’s triad. Virchow’s triad comprises three broad categories of mechanisms underlying thrombosis (1). These are endothelial damage, abnormal blood flow, and hypercoagulability (10). They may be independent or interact with each other to form a thrombus. The causes of endothelial damage include sepsis, intravenous catheterisation, disseminated intravascular coagulation and heartworm injury. The causes of abnormal blood flow include local stasis, reduced blood flow, cardiac disease, hypovolemia, and heartworm injury. The causes of hypercoagulability include DIC, pancreatitis, sepsis, and infection (6).

This report reviews the case of cranial vena cava syndrome associated with the administration of CPN through the central line and assesses the risk of cranial vena cava syndrome when applying CPN.

A 5-year-old castrated male Golden Retriever dog weighing 15 kg was presented at Jeonbuk Animal Medical Center (Jeonbuk National University, College of Veterinary Medicine, Jeollabukdo, Korea) with evidence of intestinal intussusception. He had a history of heartworm infection and was severely cachectic and dehydrated before his referral to our department. Further physical examination and blood work revealed lethargy, acidemia, and hypoproteinemia.

Ultrasonography of the abdomen (Aplio 300, Canon Medical Systems; Tokyo, Japan) with a linear-array probe (12 MHz) confirmed ascites and a target sign located at the jejunum (suspected), which is indicative of intestinal intussusception (Fig. 1). Shortly after diagnosis, emergency surgery was performed using the enteroplication method. After successful surgery, the patient was hospitalised for postoperative care. Initial treatment was aimed at the reversion of dehydration and the provision of adequate nutrition. Fluid therapy (crystalloid 12 mL/kg/hr, colloid 1 mL/kg/hr, butorphanol-Lidocaine-Ketamine 2 mL/kg/hr) and central parenteral nutrition (60 mL/h, 1116 mOsmol/L) were administered via peripheral and central venous catheters, respectively.

Figure 1. Cross-sectional ultrasonographic image of the jejunoileal segment. This retriever dog showed a target sign consisting of multiple hyperechoic and hypoechoic concentric rings. The diameter of the target sign ranged from 2.0 cm to 2.8 cm. Arrows indicate the border of the intussusception.

On the second postoperative day, the serum albumin and total protein concentrations were low, with a serum albumin concentration of 1.6 g/dL. Low perioperative serum albumin is known to be a poor prognostic factor for intestinal healing (7); therefore, albumin was supplemented with 20% human albumin (SK human albumin, SK plasma; Seongnam, Korea) diluted to 10%. On the following day, the serum albumin concentration immediately increased to 2.9 g/dL. From the second day, however, serum albumin decreased to 1.9 g/dL and remained constant at low concentrations.

Ten days postoperatively, swelling and edema were observed in the head and neck. Ultrasonographic Doppler scan of the jugular region showed an absence of blood flow in both jugular veins (Fig. 2). An additional CT scan (Alexion TSX-034A, Canon Medical Systems; Tokyo, Japan) and angiography using a non-ionic iodine contrast medium (iohexol, Omnipaque 300, GE healthcare; Illinois Chicago, USA) were performed, and a complete blockage of the cranial vena cava was diagnosed by confirming the absence of contrast media cranial to the heart (Fig. 3). Anticoagulant (Dalteparin sodium, Fragmin, 2500 IU/mL, Pfizer, New York, USA) was administered (100 U/kg, SC, tid) and thrombolytic (recombinant tissue plasminogen activator, Actilyse, 20 mg; Boehringer Ingelheim, Ingelheim, Germany) treatments (0.4 mg/kg) were immediately administered, but the patient had completely lost all of its vital functions at daybreak on postoperative day 11.

Figure 2. Sagittal ultrasonographic images of the right (A) and left (B) external jugular vein. In both jugular veins, a complex internal architecture, indicated by a dotted black circle with no venous flow, was observed.

Figure 3. Sagittal (A) and transverse (B) images related to the obstruction of the cranial vena cava. (A) angiography showing a hyper-attenuated caudal vena cava and a relatively hypo-attenuated cranial vena cava (dotted white circle). This indicates a lack of contrast media in cranial vena cava and, therefore, an obstruction. (B) Dotted white circle indicates mixed texture of gas and soft tissue, instead of contrast media in the cranial vena cava.

Necropsy was performed the following day, and the thrombus was confirmed in the jugular vein, including a thickened endothelial wall and changes associated with endothelial damage. (Fig. 4) Histological samples obtained from necropsy showed a loss of endothelial cells and fibrosis secondary to inflammation. Evidence of recanalization was also observed, which proved the existence of vascular occlusion (Fig. 5). With a history of heartworm infection, we confirmed multiple heartworms through both necropsy and histological examination (Fig. 6). This finding supports the existence of endothelial damage.

Figure 4. Necropsy image of an external jugular vein. Thrombus, thickened vascular wall, and changes associated with endothelial damage were observed.

Figure 5. Histologic images of the thickened jugular vein. These images show loss of endothelial cells and fibrosis secondary to inflammation. Dotted black circle indicates recanalization, suggesting there had been a vascular occlusion. H&E. (A) Magnification = ×40, bar = 500 µm. (B) Magnification = ×100, bar = 200 µm.

Figure 6. Histologic images of encapsulated heartworm within pulmonary parenchyma. Chronic heartworm infection is confirmed through these images. H&E. (A) Magnification = ×12.5, bar = 1600 µm. (B) Magnification = ×40, bar = 500 µm.

In human medicine, the reported rate of central line complications associated with thrombus formation ranges from 14% to 18% (10). To the author’s knowledge, however, there are only a few reported cases in veterinary medicine. In veterinary medicine, various studies (2,5,9) have previously discussed issues related to cranial vena cava syndrome, but all three of their reports involved pacemaker implantation and no CPN application. Likewise, only a few studies in veterinary medicine have directly linked thrombus formation to a complication of CPN application.

Virchow’s triad refers to three major categories of mechanisms that interact to facilitate thrombus formation, and these are endothelial damage, abnormal blood flow, and hypercoagulability (1). Although several known causes fall within these three categories, it is unlikely that the alteration of any single component of the triad would be sufficient to induce thrombosis (3). Several factors have to interact to facilitate thrombus formation. Likewise, the patient in this report had severe cachexia, hypovolemic shock, and heartworm infection, which all fell within the criteria of Virchow’s triad and were considered to have contributed simultaneously to thrombus formation. Furthermore, all of these clinical states that have a negative influence on both prognosis and mortality led to the need for CPN administration and central line placement. All of these numerous negative factors, the characteristics of CPN, and the long-term use of CPN contributed to the consequent cranial vena cava syndrome.

Protein-losing enteropathy is one of the factors influencing Virchow’s triad. Although not the sole cause of protein-losing enteropathy the loss of antithrombin contributes to hypercoagulability (4). Protein-losing enteropathy was not definitively diagnosed in this study, but clinical signs and laboratory findings supported its existence and, therefore, is a potential contraindication for central line placement. However, parenteral feeding via a central venous catheter was necessary because the patient had anorexia and cachexia. The condition of the patient presented a clinical dilemma in deciding the placement of the central line, which can be a further consideration for clinicians confronted with similar situations.

In a retrospective study by Reuter et al. (8), the overall mortality rate of patients who had received CPN was 48.8%. Among the 209 dogs included in this study, 5% had jugular thrombosis and a mortality rate of 50%. Furthermore, in two studies previously mentioned in this report, all patients were euthanized because of their debilitating state and complications. From the data obtained from these reports, clinicians should acknowledge that the status of a patient requiring CPN application is associated with mortality rate, and close monitoring is important.

Limitation of this case report is that its number of case is not enough to provide any solid opinion. The patient showed several signs associated with thrombus formation caused by CPN application and, thus, cranial vena cava syndrome; however, it is insufficient to say that CPN application was the sole cause. Additional studies with larger sample sizes are required to verify their association. Furthermore, coagulation profile was not performed during hospitalization. Anticoagulant and thrombolytic treatments were given only after diagnosis of clinical sign and ultrasound findings. Therefore it is hard to say that treatment strategy was optimal.

In conclusion, venous thrombus formation secondary to CPN application via the central line is a rare but possible complication. Veterinarians who are concerned about taking care of patients receiving CPN through the central line should keep the possibility of venous thrombus formation in mind. Patients with concurrent hypercoagulability disease or protein-related disease are at greater risk of thrombus formation.

This research was financially supported by the Ministry of Small and Medium-sized Enterprises (SMEs) and Startups (MSS), Korea, under the “Regional Specialized Industry Development Plus Program (R&D, S3244754)” supervised by the Korea Technology and Information Promotion Agency (TIPA).

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No.2020R1F1A1075219).

The authors have no conflicting interests.