- Top of page
- Materials and Methods
Background: Splenic venous thrombosis (SVT) is usually considered an incidental finding on abdominal ultrasound examination but can indicate the presence of underlying disease. Concurrent disease processes and conditions in dogs with SVT have not been identified previously.
Objectives: To identify concurrent diseases and conditions in dogs with SVT.
Animals: Eighty dogs with SVT.
Methods: Retrospective review. Medical records from 1994 through 2008 were searched for dogs with SVT identified by ultrasound examination. These records were then reviewed for signalment, medical history, clinicopathologic testing, diagnostic imaging, and clinical diagnosis.
Results: The most common concurrent conditions were neoplasia (54%), exogenous corticosteroid administration (43%), systemic inflammatory response syndrome (26%), disseminated intravascular coagulation (20%), pancreatitis (18%), and immune-mediated disease (16%). The most common neoplastic disease was lymphoma, and the most common immune-mediated disease was immune-mediated hemolytic anemia. Protein-losing nephropathy and naturally occurring hyperadrenocorticism were identified in <10% of the dogs. Concurrent splenic infarcts were identified in 33% of dogs, and concurrent portal vein thrombi were found in 18% of dogs.
Conclusions: SVT is a sonographic finding of clinical importance, and dogs with SVT can have 1 or more coexisting diseases.
The splenic vein is part of the portal system and joins with the cranial mesenteric vein, caudal mesenteric vein, and gastroduodenal vein to form the portal vein in the dog. Thrombosis of the splenic vein has seldom been reported in animals and is usually asymptomatic. At our institution, splenic venous thrombosis (SVT) is occasionally identified via abdominal ultrasound examination. Thrombi typically cause increased echogenicity in the vessel lumen and disruption of blood flow identified with color Doppler (Figs 1 and 2).1 Although SVT is usually an incidental finding, it can suggest the presence of underlying disease. Concurrent disease processes and conditions in dogs with SVT have not been identified previously.
Figure 2. Abdominal ultrasound image with color Doppler demonstrating lack of blood flow within one of the splenic veins.
Download figure to PowerPoint
For vascular thrombosis to occur, the normal regulatory mechanisms of coagulation must be disturbed. Disease states that alter 1 or more of these 3 factors can predispose to venous or arterial thrombus formation. Diseases previously reported to be associated with venous thrombosis in dogs include immune-mediated hemolytic anemia (IMHA), neoplasia, corticosteroid therapy, protein-losing nephropathy (PLN), and inflammatory disease conditions.2–7
A complete understanding of the etiology of isolated SVT in people is yet to be elucidated. Although it is commonly grouped with portal venous thrombosis in the literature, the causes and clinical significance of isolated SVT are thought to differ.8 In human medicine, the most common cause of isolated SVT is chronic pancreatitis. Other reported causes in people include pancreatic neoplasia, pancreatic abscess or pseudocyst, trauma, nonpancreatic neoplasia, gastric surgeries, splenectomy, splenic artery aneurysms, myeloproliferative diseases, renal disease, and inflammatory disorders.8–13 There are very few reports of SVT in animals. Isolated SVT is reported in a dog with nephrotic syndrome as well as in dogs with splenic torsion.14,15 One case report describes a dog with megakaryoblastic leukemia that developed systemic thrombosis with splenic vein involvement.4 The purpose of this retrospective study was to identify concurrent diseases and conditions associated with SVT in the dog.
- Top of page
- Materials and Methods
Between 1994 and 2008, 86 dogs were identified with a diagnosis of splenic vein thrombosis (SVT) made via ultrasound examination. Three dogs with splenic torsion and 3 dogs with insufficient diagnostic or clinical information were excluded from this study. Eighty dogs met the inclusion criteria, and all were diagnosed with diseases in 1 or more categories (Table 1). These remaining dogs had an approximate median age of 9 years (range 2–19 years) with 39/80 (49%) female spayed, 30/80 (38%) male castrated, 7/80 (9%) male, and 4/80 (5%) female. A variety of breeds were represented, but most were medium and large breed dogs with a median weight of 28.9 kg (range 2.1–73 kg).
Table 1. Diseases and conditions seen in 80 dogs with splenic venous thrombosis in order of frequency.
| ||Number of Cases||% of Total (80)|
|Renal failure without proteinuria||11||14|
A total of 47 neoplasms were identified in 43/80 dogs (54%). The most common neoplastic category was round cell neoplasia, which was identified in 20/80 dogs (25%) and included lymphoma (9 dogs), leukemia (4), mast cell disease (5), histiocytic disease (2), and melanoma (1). One dog had both chronic lymphocytic leukemia and mast cell neoplasia. Nine of 20 dogs (45%) with round cell neoplasia had recent or concurrent corticosteroid therapy. The remaining neoplastic categories included epithelial (12/80 dogs, 15%), mesenchymal (4/80 dogs, 5%), and suspected neoplasia (9/80 dogs, 11%). Epithelial neoplastic diseases included a variety of underlying cell types such as anal sac gland adenocarcinoma, hepatic epithelial carcinoma, schirrous mammary adenocarcinoma, bronchoalveolar carcinoma, schirrous gastric carcinoma, and renal carcinoma. Three of the dogs with epithelial neoplasia had confirmed metastatic disease to the liver, spleen, and lymph nodes.
Mesenchymal neoplastic disease (4/80 dogs) included right atrial hemangiosarcoma (2), metastatic rib osteosarcoma (1) and concurrent osteosarcoma and gastric leiomyosarcoma (1). Both dogs with hemangiosarcoma had secondary DIC and SIRS and qualified for 2 categories. Two dogs had neoplastic disease in more than 1 category (mammary carcinoma/auricular hemangiosarcoma and lymphoma/apocrine gland adenocarcinoma) at the time of thrombus diagnosis. Two dogs had more than 1 neoplasia within the same category (osteosarcoma/gastric leiomyosarcoma and chronic lymphocytic leukemia/mast cell neoplasia).
Twenty-one dogs (26%) had SIRS, and 12 of these 21 dogs had concurrent DIC. Four dogs had DIC without evidence of SIRS during their hospitalized period. Concurrent diseases identified with dogs that had DIC without SIRS included corticosteroid administration and immune-mediated disease; corticosteroid administration and epithelial neoplasia; round cell neoplasia and PLN; and hepatopathy. Three dogs with SIRS and 1 dog with both SIRS and DIC did not have a known underlying cause or fall under any other category.
Fourteen dogs had pancreatitis (18%). Six of these 14 dogs (43%) had concurrent immune-mediated disease; all were treated previously with corticosteroids. Five of the 14 dogs with pancreatitis also had suspected or confirmed underlying neoplasia. The 3 remaining dogs with pancreatitis had other concurrent disease including diabetes mellitus with renal failure, gastrointestinal hemorrhage of unknown etiology, and DIC with SIRS and concurrent renal failure. Nine dogs had an underlying hepatopathy (9/80 dogs, 11%), which included hepatic insufficiency as determined by biochemical analysis (3 dogs) and a histopathologic diagnosis of cirrhosis (2), chronic active hepatitis (1), acute severe massive necrosis (1), common bile duct obstruction (1), and severe multifocal degenerative hepatopathy (1).
Immune-mediated disease was identified in 13/80 dogs (16%). Twelve of the dogs with immune-mediated disease (92%) were treated with corticosteroids before identification of a splenic vein thrombus. One dog diagnosed with vasculitis and a hepatopathy did not receive corticosteroids. The most common immune-mediated disease was IMHA (8 dogs), followed by IMHA with IMTP (3 dogs) and IMPA (2 dogs). Pemphigus and SLE was diagnosed in 1 dog each.
Eleven of 80 dogs (14%) were diagnosed with renal failure without proteinuria. Ten of the 11 dogs qualified for additional categories including corticosteroid administration. The single remaining dog with renal failure had a decreased AT level of 64% (reference range: 80–120%) of unknown etiology.
Seven of 80 dogs (9%) were diagnosed with a PLN with a median UPC of 5 (range 3.0–21.8). Five of these dogs had a median AT level of 58% (range 38–104%). One dog was diagnosed with a PLE with concurrent SIRS, DIC, and corticosteroid therapy.
Thirty-four of 80 dogs (43%) had been treated with corticosteroids at the time of diagnosis of the splenic thrombus. Two of these dogs were also diagnosed with PDH. An additional dog had a cortisol-secreting tumor without a history of corticosteroid administration, but had suspected hepatic metastatic neoplasia because of the presence of multiple liver masses on ultrasound examination.
Three dogs had adrenal masses. One dog had a cortisol-secreting tumor (described above) and two did not. None of the adrenal masses identified had vascular invasion evident with ultrasound.
One dog was healthy previously, had been hit by a car and was diagnosed with concurrent pulmonary contusions, pleural effusion, suspected renal hematoma, and ureteral trauma, in addition to SVT and splenic infarctions.
Concurrent splenic infarcts were identified on ultrasound examination in 26/80 dogs (33%). Concurrent incomplete portal vein thrombi were found in 14/80 dogs (18%). Three of the dogs with portal venous thrombi had multiple acquired extrahepatic portosystemic shunts identified with abdominal ultrasound or exploratory laparotomy, and 9/14 dogs had peritoneal effusion. Concurrent diseases and conditions in dogs with portal venous thrombosis included corticosteroid administration (7 dogs), pancreatitis (5), immune-mediated disease (5), SIRS (5), neoplasia (5), DIC (4), renal failure (2), and hepatopathy (2).
- Top of page
- Materials and Methods
In the present study, SVT was detected with a variety of coexisting disease processes, and in many incidences dogs had more than 1 predisposing factor. Neoplasia and corticosteroid exposure were the most common concurrent conditions in dogs with SVT in this study.
Over half of the dogs in the present study had at least 1 neoplastic process. The most common neoplastic cell type was round cell, particularly lymphoma. Neoplasia is a well-recognized risk factor for the development of venous thromboembolism in human patients and has been associated with a 4-fold increase in risk of thrombosis.22,23 Furthermore, thromboembolic disease is the second leading cause of death in human cancer patients. The mechanism of cancer-associated thrombosis is complex and includes neoplastic cell expression of procoagulant molecules including tissue factor, abnormal endothelial cell growth, neoplastic cell-derived microparticles, platelet activation, thrombophilia, increased von Willebrand's factor levels and in some circumstances blood flow stasis or turbulence.24,25 Chemotherapy, in particular corticosteroids and l-asparaginase, further increase the risk of venous thrombosis.26–29 The cancer types most frequently associated with thrombosis in human medicine, in descending order, are pancreatic, brain, acute myelogenous leukemia, gastric, esophageal, renal, pulmonary, ovarian, hepatic, and lymphoma.24 Cancer-associated venous thrombosis has been reported previously in animals, including a case of acute megakaryoblastic leukemia in a dog causing systemic thrombosis with involvement of the splenic vasculature.4,6,30 Both malignancy and chemotherapy can represent independent but additive risk factors for thrombus formation.
A large number (44%) of dogs in this study had excess glucocorticoid exposure. All but one of these animals had received recent exogenous corticosteroid therapy; 3 dogs had naturally occurring HAC, 2 of which also received exogenous corticosteroids. Both exogenous and endogenous corticosteroids have prothrombotic effects, including increased coagulation factor levels such as factor VIII and von Willebrand's factor levels, in addition to decreased AT levels.a27,28,31,32 Cushing's syndrome in people is associated with an increased risk of thromboembolism of 4 times that of the general population.31 In veterinary medicine, evidence of hypercoagulability has been identified in dogs with Cushing's disease and dogs receiving exogenous glucocorticoids, including several reports of pulmonary thromboembolism associated with HAC.22,26,27 In a report of 11 dogs with portal vein thrombosis, 10 had received exogenous corticosteroid therapy.5 Animals receiving corticosteroid therapy commonly have 1 or more underlying disease processes that could further increase the risk for venous thrombosis. In the present study, the 34 dogs receiving exogenous corticosteroids also had concurrent neoplasia (15 dogs), immune-mediated disease (12), and pancreatitis (8). It is not possible to determine the exact role of corticosteroids in the development of SVT from the results of this study. When choosing to administer corticosteroids, their significant effects on the coagulation system should be considered.
Both SIRS (26%) and DIC (20%) were common clinical findings in dogs with SVT in the present study, and 12 dogs had both SIRS and DIC. Systemic inflammation is known to cause a prothrombotic state via increasing procoagulant processes such as the expression of tissue factor, decreased levels of anticoagulants such as AT and protein C in addition to inhibition of fibrinolytic pathways. DIC describes the prothrombotic state associated with systemic inflammation. Clinically, DIC is commonly recognized in the late bleeding stage after platelet and coagulation factor consumption, but the underlying coagulation abnormality remains prothrombotic in nature.33–35 A majority of the dogs in this study had SIRS or DIC in conjunction with 1 or more other coexisting disease processes, such as neoplasia, that could further contribute to thrombus formation. Stimulation of systemic inflammation and activation of the coagulation system is likely to be a common contributor to the formation of venous thrombosis in many disease states.
Pancreatitis was evident in 18% of dogs in the current study, and no dogs were diagnosed with pancreatic neoplasia. In people, isolated SVT is associated with pancreatic disease, particularly pancreatitis. The pathogenesis is related to the anatomic location of the splenic vein just posterior to the pancreas and adjacent to peri-pancreatic tissues and lymph nodes. Chronic pancreatitis is believed to cause perivascular inflammation of the splenic vein and subsequent thrombosis, with a reported incidence of up to 45% in human patients.36 As the left limb of the pancreas in the dog is anatomically located adjacent to the splenic vein, similar to humans, with the vascular supply to this portion of the pancreas arising from branches of the splenic artery and vein, it is possible that pancreatitis can increase the risk of splenic thrombosis in dogs in a similar manner to that reported in human patients.37 In addition, the majority of the dogs with pancreatitis in this study had other potential factors for venous thrombosis including neoplasia and immune-mediated disease with concurrent corticosteroid therapy.
Protein-losing disease processes are well-reported causes of thrombosis. A relatively small number of dogs in this study were diagnosed with PLN (7 dogs) or PLE (1 dog). In an effort to exclude dogs with preglomerular causes of proteinuria, our inclusion criteria for PLN arbitrarily required a UPC of ≥3.0. Therefore, it is possible that these results could have underestimated the number of dogs with SVT and concurrent PLN. The cause of the hypercoagulable state of dogs with glomerular disease is multifactorial and is influenced by low AT activity, increased platelet activation, and hyperfibrinogenemia.38–40 The reported incidence of thromboembolism in a study of 137 dogs with protein-losing glomerular disease was 13% and was associated with a high mortality.38 In that study, the pulmonary vasculature was the most common site of thromboembolism. Although it is uncommon, SVT appears to be a possible manifestation of thromboembolic disease in this patient population. Renal disease in the absence of glomerular disease can cause endothelial dysfunction, coagulation factor changes, and inflammation, which may also contribute to a hypercoagulable state.40,41 Although there were 11 dogs in the present study that were classified as having renal failure, all but 1 had other predisposing conditions for thrombosis in addition to renal disease. The role of renal failure in the development of SVT in these animals is unclear at this time.
Immune-mediated disease is a well-recognized risk factor for venous and arterial thrombosis in human patients. The most common mechanism of thrombosis is the development of a hypercoagulable state secondary to antiphospholipid antibody production.42,43 Although antiphospholipid syndrome is known to increase the risk of thrombosis, the exact mechanism by which this occurs has not been elucidated.44 In this study, the most common immune-mediated disease represented was IMHA, which has a high risk of thromboembolic complications in dogs.2,7,45,46 Mechanisms by which IMHA can increase the risk of thrombosis include cytokine-mediated up-regulation of tissue factor expression, direct membrane activation, erythrocyte microparticle release, increased platelet aggregation secondary to plasma hemoglobin binding local nitric oxide, hyperbilirubinemia, and antiphospholipid antibody production.47,48 Additionally, most dogs with immune-mediated disease receive exogenous corticosteroid administration; as a consequence, it is impossible to separate the hypercoagulable effects of immune-mediated disease from corticosteroid effects in this patient population. Mechanisms of hypercoagulability have been poorly investigated in canine IMHA, but the presence of antiphospholipid antibodies is reported in dogs with IMHA.43,49
Concurrent portal venous thrombosis was evident in 18% of dogs with SVT in the present study. Unlike isolated SVT, portal venous thrombosis is likely to cause portal hypertension, which can lead to the formation of acquired extrahepatic portosystemic shunt vessels and clinical signs such as ascites. In human medicine, the etiology and pathogenesis of isolated SVT is thought to be different than SVT occurring in association with portal venous thrombosis.8 Pancreatitis is the most common cause of isolated SVT in human patients, and hepatic cirrhosis and neoplasia are the most common diseases associated with portal venous thrombosis in people.8 Of 11 dogs identified on necropsy as having portal venous thrombosis, 4 were reported to have pancreatic necrosis, 3 had neoplasia, 2 had peritonitis, and 10 of the dogs had received corticosteroid therapy.5 No conclusions regarding the etiology of isolated SVT compared with portal venous thrombosis in dogs can be made from the results of the present study, and would be a subject worthy of future investigations.
Adrenal masses of >2 cm in diameter are more likely to be neoplastic in nature and can predispose to SVT by similar mechanisms as other neoplastic processes described above.50 Additionally adrenal masses that are functional cortisol-secreting tumors could further contribute to a hypercoagulable state. Vascular invasion and extensive local thrombosis can also occur with adrenal masses. In this study, only 3 dogs with adrenal masses were identified with SVT. Of these dogs, 1 had a functional cortisol-secreting tumor, and the nature of the other 2 masses was not determined. None of these adrenal masses had local vascular invasion evident at the time of ultrasound.
One dog in this study developed an SVT after abdominal trauma as a result of being hit by a car. Although trauma is a recognized risk factor for the development of venous thromboembolism in human patients, the pathogenesis is not yet fully understood.51,52 Local vessel and endothelial injury could be the initiating event leading to platelet adhesion and thrombin production.52 Inflammation and cytokine production can further alter hemostatic balance and contribute to a prothrombotic state in trauma patients. Thrombosis secondary to trauma has been seldom reported in the veterinary literature. As abdominal ultrasound is not routinely performed on all trauma patients, it is possible that SVT occurs more frequently in these animals than is currently recognized.
The incidence of splenic infarction in dogs is unknown. In a study of 1,480 cases of splenic disease, infarction was reported in <2% of cases.53 However, this study evaluated splenic biopsy specimen and splenectomy samples only, and likely underrepresented a disease process that is thought to be clinically silent. Of the 80 dogs in our study identified with SVT, one third of them had splenic infarcts evident on abdominal ultrasound examination. This would suggest that similar diseases are associated with both SVT and splenic infarction in dogs. There is little discussion of splenic infarction in the veterinary literature; 1 retrospective study of 16 dogs reported neoplasia and corticosteroid therapy to be the most common concurrent diseases.54 The identification of splenic infarction in dogs can mark the presence of underlying disease processes in a manner similar to SVT.
The retrospective design of this study is an obvious limitation that reduced the accuracy of categorization of the concurrent disease processes and led to the exclusion of some dogs because of incomplete clinical or diagnostic information. By searching ultrasound reports and medical records, the dogs were identified from a selective population. Because most dogs with SVT are asymptomatic, many will go undiagnosed, particularly if the underlying medical condition does not warrant ultrasound examination. Hence the results of this study cannot be used to estimate the incidence of SVT in dogs. In this retrospective review, progression or resolution of SVT also could not be determined, nor could prognosis or outcome be evaluated. A prospective clinical study would be ideal in order to answer some of these questions.
To date there is little evidence that isolated SVT leads to clinical illness in dogs. In human patients, isolated SVT is commonly asymptomatic and requires no specific therapy.8 In 4–55% of people, SVT leads to gastric variceal bleeding, an abnormality that is rarely reported in the dog.8,36 When SVT is considered clinically important in human patients, a splenectomy is performed. When performing an abdominal ultrasound examination in the dog, it is recommended to include a complete evaluation of the splenic vasculature for thrombosis and abnormal blood flow. Evidence of SVT should prompt the clinician to evaluate for coexisting disease processes or drug therapy that could predispose to this abnormality. The results of this study suggest that neoplasia and corticosteroid exposure should be of particular concern.