**2. Epidemiology**

The frequency of deep arm vein thrombosis relative to all deep thromboses has been reported to be between 1 and 14% (Hill & Berry, 1990; Joffe et al*.*, 2004; Spencer et al*.*, 2007). Recently, the prospective RIETE registry and the population based Malmö thrombophilia study reported both very similar rates of upper extremity deep vein thrombosis (4.4% and 5% of all thrombosis, respectively (Munoz et al*.*, 2008; Isma et al*.*, 2010). Therefore, it can be assumed that about 5% of all thrombosis will involve the deep arm veins, which corresponds to an annual incidence of approximately 3 per 100.000 patients per year (Bernardi et al*.*, 2006). Less than 50% of these arm vein thromboses can be expected to extend into the internal jugular vein (Gbaguidi et al*.*, 2011). About one

Deep Vein Thrombosis of the Arms 131

Compared to deep vein thrombosis of the legs, local factors play a dominant role in deep arm vein thrombosis. By far the highest risk for thrombosis in this region is caused by foreign material in the lumen of the arm veins, most importantly indwelling central venous catheters and pacemaker leads. The odds ratio for arm vein thrombosis of patients carrying these intravascular devices compared to patients that do not has been reported to be as high as 10 to more than 1000 (Joffe et al*.*, 2004; Blom et al*.*, 2005), Table 1. This large variation in risk may in part be explained by specific features of the central venous catheter, as e.g. catheter type and material, site, technique and level of insertion as reviewed by Van Rooden et al*.*, 2005. Additional factors that have an impact on the risk of thrombosis in patients with central venous catheters include the number of punctures during catheter insertion, the duration of catheterization, the fluid administered, and catheter related infections (Koksoy et al*.*, 1995; Hernandez et al*.*, 1998; Martin et al*.*, 1999). In addition, wrong placement of the catheter tip in the upper half of the superior vena cava, subclavian or innominate veins results in a higher risk of thrombosis (Luciani et al*.*, 2001; Verso et al*.*, 2008). Implanted port a cath systems and pacemaker leads significantly increase the risk of arm vein thrombosis as well (Van Rooden et al*.*, 2004; Goltz et al*.*, 2010). The major pathogenetic mechanism appears to be the thrombogenicity of the foreign material itself. Other possible factors include injury of the vascular wall and disturbances of venous blood flow (Beathard, 2001). Different types of thrombi associated with central venous catheters have been described, ranging from fibrin sleeves that may be embolized following catheter removal, nonocclusive mural thrombi and complete venous obstruction (Brismar et al., 1981; Martin et al. 1999; Beathard, 2001). The second major risk factor for arm vein thrombosis is the presence of active malignant disease. Since chemotherapeutic agents are frequently delivered via central venous catheters, both major risk factors are often present in cancer patients. However, malignant disease carries a significant risk also in the absence of foreign material in the arm veins. The mechanisms by which malignant tumors promote thrombosis in various venous segments include local arrosion or invasion of blood vessels, hypercoagubility of the blood by the expression of tumor antigens, and stasis by tumor compression of venous segments proximal to the site of thrombosis (Sood, 2009; Martinelli et al*.*, 2010). It is the experience of clinicians treating patients with thrombosis that this disease is most aggressive and difficult to treat in tumor patients. Hospitalisation has also been cited as a strong risk factor, which may be explained by the increasing frequency of complex therapeutic regimens requiring central venous lines for various indications (Joffe et al*.*, 2004; Mai & Hunt, 2011). Other major risk factors are listed in Table 1 and include local factors (arm surgery, arm injury and immobilisation of the upper extremities by plaster casts), unusual strenuous arm exercise ("thrombosis par effort"), a family history of venous thromboembolism and inherited forms of thrombophilia, and the use of estrogen containing contraceptive drugs (Martinelli et al*.*, 2004; Joffe et al*.*, 2004; Blom et al*.*, 2005). Although cited frequently, a thoracic outlet syndrome is diagnosed in comparably few cases (Blom et al*.*, 2005). Several unusual risk factors for arm vein thrombosis have been reported in case reports, including backpacking (Schoen et al., 2007), portable computer games (Phipps & Joo, 2008), ambulatory blood pressure monitoring (Marschang et al*.*, 2008), intravenous calcium guconate injection (Chen et al*.*, 2009), and infraclaviculary lipoma (Palamari et al*.*, 2010). Interestingly, two known risk factors for lower extremity deep vein thrombosis, namely age and obesity, do not appear to confer additional risk for upper deep vein thrombosis (Joffe et al*.*, 2004; Mai & Hunt, 2011). Isolated thrombosis of the internal jugular vein may be observed in the context of two

**4. Risk factors for arm vein thrombosis** 

third of patients with deep arm vein thrombosis will have primary thrombosis, i.e. idiopathic and effort-related thrombosis (Paget-von Schroetter syndrome). The remaining two thirds of patients will have secondary upper extremity thrombosis with exogenous (e.g. central venous catheters) or endogenous (e.g. cancer) risk factors. In intensive care patients as well as in patients suffering from malignant disease with central venous catheters, rates of asymptomatic thrombosis as high as 30% to over 60% have been reported (Timsit et al*.*, 1998; Van Rooden et al*.*, 2005). There appears to be an increase in upper extremity deep vein thrombosis in the last decades, which may reflect the increasing use of central venous catheters (S. Mustafa et al*.*, 2003; Czihal & Hoffmann, 2011), improved diagnostic methods, or both.
