Patients with UEDVT most often present with unilateral swelling and discomfort or localized pain [4, 8, 12–14]. Other symptoms and signs that have been described are weakness, paresthesia, heaviness, low-grade fever, visible collateral veins, erythema, a palpable cord, cyanosis, and warmth (Table 27.1) [8, 12, 16–19]. The majority of UEDVT associated with a CVC or pacemaker remains subclinical, as most cases are discovered during the work-up of a dysfunctional catheter or PE [20–22]. Concomitant symptomatic PE is present in 3–12% of all patients with UEDVT [6, 7, 23–28], which is less than in patients with lower extremity DVT, in which prevalences of around 30% have been reported [23, 28].

UEDVT is subdivided into primary and secondary UEDVT, based on the pathogenesis. Primary UEDVT represents 20–50% of all cases and includes effort-related thrombosis (also known as the Paget-Schroetter syndrome) in combination with the thoracic outlet syndrome (TOS) and idiopathic thrombosis. The majority of UEDVT is secondary to a predisposing risk factor [3, 9, 29–32]. The risk factors most strongly associated with UEDVT are cancer and the presence of a CVC. Other risk factors include pacemakers, previous venous thromboembolism (VTE), a positive family history of VTE, arm surgery or trauma, immobilization, the use of estrogens, and thrombophilia (Table 27.2).

An accurate diagnosis of UEDVT is important, as appropriate treatment can reduce the clinical burden and prevent complications in the acute phase, such as PE. The prevalence of UEDVT in patients with a clinical suspicion of UEDVT varies from 10 to 45% in several cohort studies, which might be explained by differences in study design and the proportions of cancer patients, CVCs, and the number of inpatients (Table 27.4) [7, 12, 49, 50]. In patients with a CVC, the prevalence of UEDVT was 53% in one study [7], compared to only 18% in patients without a CVC (p < 0.01). These figures were 31 and 23% for cancer and non-cancer patients, respectively (p = 0.07, manuscript under revision).

Venography is the gold standard to diagnose UEDVT, as it visualizes the entire deep vein system of the upper extremity, but it is invasive, expensive, and involves the use of contrast, which may cause complications including renal failure and allergic reactions. Due to these disadvantages, venography has been largely replaced in clinical practice by compression ultrasonography, which is noninvasive, relatively cheap, and easy to perform [19]. In a systematic review, identifying nine studies on the role of compression ultrasonography in the diagnosis of UEDVT, the overall sensitivity was 97% (95% CI 90–100%), with a specificity of 96% (95% CI 87–100%) [51]. The presence of the clavicle may hinder evaluation of the middle part of the subclavian vein, and in case of indeterminate compression ultrasonography results, venography may provide a definitive answer. Other diagnostic options include computed tomography (CT) angiography and magnetic resonance angiography (MRA), which are both noninvasive. However, both have only been evaluated in studies with very few patients with a clinical suspicion of UEDVT, and the diagnostic performance of both modalities is therefore unclear [52, 53].

Several attempts have been made to improve the diagnostic process in patients with a clinical suspicion of UEDVT. Constans and colleagues developed a clinical decision rule, incorporating four items (Table 27.5) [12]. If the total score is one or less, UEDVT is deemed unlikely, whereas if the total score is two or higher, the diagnosis is likely. The prediction of UEDVT based on this score was consistent in three study samples, with prevalences of 64–70% in patients with a total score indicating “UEDVT likely” and 9–13% in those with a total score indicating “UEDVT unlikely,” suggesting that this score can be a valuable tool in a diagnostic algorithm [12].

The diagnostic value of D-dimer has been tested in 2 studies, 1 including 52 patients of whom 15 (29%) had UEDVT, and the other including 239 patients of whom 24 (10%) were diagnosed with UEDVT [49, 54]. Both studies applied a cutoff value of 500 ng/mL. The sensitivity was high in both studies with 100% (95% CI 78–100%) and 92% (95% CI 73–99%), respectively, whereas the specificity was low (14%, 95% CI 4–29% and 60%, 95% CI 52–67%, respectively). These figures were similar for cancer patients and patients with a CVC [49, 54].

Recently, a multicenter, international, prospective diagnostic management study evaluated an algorithm consisting of the Constans score, D-dimer testing, and compression ultrasonography in consecutive patients with a clinical suspicion of UEDVT [7]. In total, 406 patients were included, and the algorithm was feasible in 390 (96%). UEDVT was confirmed in 103 patients (25%). In 87 patients (21%; 95% CI 17–25%), ultrasonography could be withheld. One patient, in which UEDVT was initially excluded, developed a UEDVT during 3-month follow-up, for an overall failure rate of the algorithm of 0.4% (95% CI 0–2.2%). In another study, 483 patients with a clinical suspicion of UEDVT all underwent immediate compression ultrasonography and were followed for 3 months prospectively. The failure rate, defined as the rate of recurrent VTE, was 0.6% (95% CI 0.2–2.2%) for single ultrasonography and 0.2% (95% CI 0.1–1.7%) for serial ultrasonography. Of note, the prevalence of UEDVT was relatively low in this cohort (13%) [50].

While there have been important improvements in the field, the best diagnostic strategy in patients with a clinical suspicion of UEDVT remains to be determined. Hence, at present, objective imaging remains the cornerstone of UEDVT diagnosis. D-dimer testing may help to reduce the number of patients who require imaging, although the efficiency of the test appears moderate in this population with high prevalences of cancer and CVCs. The use of an algorithm has been shown to be efficient and safe but needs to be validated prospectively before it can be implemented in clinical practice . Furthermore, improvement of the algorithm appears to be desirable, for example, by applying age-adjusted D-dimer cutoff values (van Es, in press). In patients with a CVC and a suspicion of UEDVT, direct imaging seems justified, as only two examinations must be performed to detect one UEDVT.

In the acute phase of UEDVT, the goal is to relieve acute symptoms and prevent complications, such as the loss of venous access or development of PE. The long-term goals of treatment are mainly the prevention of recurrent VTE, including fatal PE, and the development of PTS. Treatment of UEDVT is based on anticoagulation predominantly with selective use of thrombolytic therapy, mechanical catheter interventions, first rib resection, and vena cava filter (VCF) placement. No randomized controlled trials have evaluated any of these therapies in patients with UEDVT. Therefore, treatment recommendations by the major guidelines are largely extrapolated from studies on DVT of the leg and are only based on small observational studies in UEDVT patients [55].