Although much of the focus of venous disorders concentrates on the lower extremities, pathologic conditions such as deep venous thrombosis (DVT) also affect the upper extremities with significant frequency. Certain venous disorders, such as Paget-Schroetter syndrome (PSS), arise specifically because of anatomic relationships seen only in the upper limbs. The increasing usage of central venous catheters is also linked to the increased prevalence of upper extremity venous pathology. Advancements in noninvasive diagnosis and treatment have expanded the therapeutic options for upper extremity venous disorders.

The similarities in arterial development of the upper and lower extremities are also present in the venous system. The upper limb has both a superficial and deep system of veins. These both drain into a single outflow tract, the axillary vein. The dorsal aspect of the upper extremity receives venous drainage primarily via the cephalic vein laterally and the basilic vein medially. The median cubital vein joins both of these veins at the antecubital fossa and is a frequent site of blood draws. The cephalic vein continues along the surface of the biceps muscle in the upper arm and then pierces the clavipectoral fascia to join the axillary vein. The basilic vein travels along the medial aspect of the arm and pierces the fascia of the upper arm to join the deep brachial vein to become the axillary vein. The deep veins of the upper extremity begin with the palmar interosseous veins and continue as paired communicating veins traveling alongside the radial and ulnar arteries. These radial and ulnar veins join with a third group of deep veins, the forearm interosseous veins. These then form the paired brachial veins, which continue as the axillary, subclavian, and innominate (brachiocephalic) veins centrally.

The upper central venous system consists of both internal jugular veins and subclavian veins. These unite behind the sternoclavicular joints to form the right and left innominate veins. Both innominate veins then join to form the superior vena cava (SVC). The right and left innominate veins are asymmetric, the left being 6 cm in length compared with 2.5 cm on the right. The SVC descends along the right side of the ascending aorta to enter the right atrium at the level of the third costal cartilage. The azygous vein joins the SVC posteriorly just before it enters the pericardium.

Variations in the venous anatomy of the upper extremity are frequent but are rarely of clinical significance. Unlike the lower extremities, venous drainage of the upper limb is largely dependent on cardiac function. This is secondary to the lack of a muscle pump, lessening the importance of venous valves. In addition, blood flow in the arm veins may increase with respiration and decrease with expiration because there is no high-pressure area corresponding to the abdomen between the thorax and the arm. Because of the human erect posture, pathologic venous conditions of the upper extremities are less common than with the lower extremities. Only a few patent veins are required for adequate arm drainage, and collateral compensation is robust.

Upper extremity DVT usually refers to thrombosis of the axillary or subclavian veins and can be divided into primary and secondary types. Primary upper extremity DVT is most commonly caused by external venous trauma at the thoracic inlet and is commonly referred to as effort thrombosis or PSS. Secondary upper extremity DVTs are far more common and are related to iatrogenic trauma from central venous catheters.

Among all patients with DVT, 89% had lower extremity DVT and 11% had upper extremity DVT, according to a U.S. multicenter DVT registry.¹ The number of upper extremity DVTs has increased significantly over the past two decades. Originally thought to be self-limited, recent studies have demonstrated upper extremity DVT to be associated with complications such as pulmonary embolism (PE), loss of vascular access, SVC syndrome (SVCS), and postthrombotic insufficiency manifested as chronic arm and hand aching and swelling.^2,3,4,5,6

Catheter-Related Deep Venous Thrombosis

The mechanism of catheter-related venous thrombosis may be secondary to irritation of the vessel wall, malignancy, or infusion of sclerosing chemotherapeutic agents. Other causes include vessel wall trauma associated with insertion, antibiotics, and total parenteral nutrition if not adequately diluted.^7,8,9 Additionally, the simple presence of the catheter within the vein impacts laminar blood flow around it, resulting in an area of stasis (Figure 18-1A). Clinical manifestations include overt signs and symptoms such as pain, fever, tenderness, jugular venous distension, warmth, swelling or edema, bluish discoloration, or visible collateral circulation.¹⁰ Subclinical thrombosis is defined as thrombosis in the absence of signs and symptoms and is demonstrated by screening diagnostic imaging. Risk factors specifically associated with increased incidence of central venous catheter- related thrombosis, including increasing age, presence of cancer, active cancer treatment, high platelet count at time of central venous catheter insertion, prior thromboembolism, major surgery within 30 days, or history of immobilization within 30 days. Increased central venous catheter thrombosis has also been demonstrated with specific catheter types. Central venous catheters composed of silicone or polyurethane are less often associated with local thrombosis than central venous catheters made of polyethylene.^11,12,13

Duplex ultrasound imaging has revolutionized vascular imaging because of its noninvasive nature, low cost, lack of ionizing radiation, and ability to be performed at bedside. It is the initial imaging test of choice for diagnosing upper extremity DVT because of its high sensitivity and specificity in evaluation of the jugular, distal subclavian, and axillary veins. Limitations are encountered because acoustic shadowing from the clavicle and sternum hinders visualization of the proximal central veins.^1,14,15 Contrast venography remains the gold standard for evaluation of venous anatomy; however, drawbacks include difficulty with accessing the vein with an edematous upper extremity; need for contrast administration, which may lead to nephrotoxicity or allergic reactions; and radiation exposure. Magnetic resonance angiography (MRA) is gaining acceptance as a modality for venous thrombosis imaging. It is noninvasive, correlates well with venography, and provides more complete evaluation of central collaterals. The availability of high-quality MRA is still limited and it is not suitable in claustrophobic patients or those with implanted metal.¹⁶

Multiple treatment options exist for the management of patients with catheter-related upper extremity thrombosis, ranging from anticoagulation to thrombolysis to surgery. When maintenance of the central catheter is not mandatory, removal of the offending central line is recommended (Figures 18-1B and C). Caution must be exercised with catheter removal because pulmonary embolus can be triggered upon removal. This occurs when the pericatheter fibrin sheath peels off the catheter, dislodges from the vessel wall, and embolizes. Anticoagulation remains the cornerstone of thrombotic therapy. It serves to maintain the patency of collaterals as well as reducing the potential for clot propagation. Current recommendations include either fractionated or unfractionated heparin as a bridge to warfarin therapy with a minimal of 3 months of therapy and a goal International Normalized Ratio (INR) of 2.0 to 3.0. A minimum of 6 months of anticoagulation has been proposed for patients detected as having hypercoagulable conditions.¹⁶

The role of thrombolytics in catheter-related DVT has not been definitively established, although the best candidates for thrombolytic therapy are young, otherwise healthy patients with primary upper extremity DVTs. Certain patients with recent catheter-related DVT and symptomatic SVCS or the requirement for preservation of a central venous access may be targeted for thrombolysis.^16,17 Postthrombotic syndrome of the upper extremity is not as well recognized. It has been demonstrated that patients with conservatively treated (standard anticoagulation) upper extremity DVTs have reduced venous outflow, residual thrombus, and arm swelling.¹⁸ This provides additional support to a more aggressive stance regarding thrombolytic therapy. Standard precautions with thrombolytic usage pertain in upper extremity DVT treatment.

Prophylaxis for patients with central venous catheters has been demonstrated to be effective. Boraks et al¹⁹ performed a study to look at “minidose” warfarin to prevent catheter thrombosis in patients with malignancies. A total of 108 patients with hematologic malignancies were started on 1 mg/d of warfarin at the time of catheter insertion. This group was then compared with a historic group of 115 consecutive patients who did not receive warfarin. Results showed that five (5%) patients of the 108 patients who received warfarin developed thrombosis at a median of 72 days after line placement compared with 15 (13%) of the 115 patients who did not receive warfarin at a median of 16 days. These results suggest that low-dose warfarin may reduce central venous catheter–related thrombosis in patients with hematologic malignancies.

Surgical reconstruction of the central veins or SVC should be reserved for patients with severe symptoms who have failed endovascular therapy and those with extensive DVT who are not anatomically suitable for endovascular treatment. Patients with malignant disease are best served with a combination of endovascular treatment, radiation, and chemotherapy for palliation. Benign central vein occlusion treated with surgical reconstruction has had good long-term results. In a series reported by Kalra et al,²⁰ the primary and secondary 5-year patency rates of upper central venous bypasses for SVCS were 53% and 80%, respectively. Conduits used for bypass include spiral-saphenous vein grafts, femoral vein grafts, externally supported ePTFE (expanded polytetrafluoroethylene), and cryopreserved homografts.

Superior Vena Cava Filters

The reported incidence of PE related to upper extremity DVT varies from 4% to 28%.²¹ The true risk of PE likely falls somewhere in the middle of this broad range and depends on many variables. Although the risk may be slightly lower than with lower extremity DVTs, it remains significant.^3,22,23 Indications for SVC filter placement are similar to those for inferior vena cava (IVC) filters. These include contraindication to anticoagulation, complications of anticoagulation, and failure of anticoagulation. From a technical standpoint, SVC filter insertion is more demanding than IVC filter insertion because of the shorter length (7 cm) of the SVC and proximity to cardiac pulsations.^24,25 A superior venacavogram is essential to identify the SVC and junction of the innominate veins. Filter placement can be performed using either a femoral approach or jugular approach, preferably on the right. Before SVC filter placement, existing central venous catheters need to be removed or retracted. Caution should be exercised when performing central line placement in patients with SVC filters. Straight guidewires with fluoroscopy should be used instead of the usual J-wire and blind approach.²⁴ Complications of SVC filter placement include SVC perforation, malplacement in the innominate vein, filter migration, guidewire entrapment during line placement, and SVC occlusion. In general, the patient population in whom SVC filters are placed have severe comorbid conditions such as multiorgan system dysfunction, sepsis, and metastatic carcinoma. Ascher et al.²¹ reported a survival rate of only 48% at 6 months in this patient group, with a high proportion of those dying within the same hospitalization from causes unrelated to the SVC filter or recurrent thromboembolism. Although their analysis failed to identify a subset of patients with limited life expectancy in whom SVC filter placement would have little benefit, they did attempt to limit the use of SVC filters to patients with a life expectancy longer than 1 month.