by Julie Hollberg, MD
A 45-year-old female with a history of cellulitis requiring peripheral inserted central catheter (PICC) line placement for intravenous antibiotics presents two weeks after line removal with persistent, dull, aching pain in her right shoulder and difficulty removing the rings on her right hand. The pain worsens with exercise and is relieved with rest. The physical exam reveals nonpitting edema of her hand. The ultrasound shows subclavian vein thrombosis. What is the best approach to treating her upper extremity deep venous thrombosis (UEDVT)?
DVT and pulmonary embolism (PE) have been subject to increased publicity recently, and both conditions are recognized as serious entities with life-threatening consequences. In fact, more people die annually from blood clots than breast cancer and AIDS combined.1,2 Still, the increased DVT and PE awareness is primarily focused on lower extremity DVT (LEDVT)), while UEDVT is thought of as a more benign entity. However, current data suggest that UEDVT is associated with equally significant morbidity and mortality.
UEDVT prevalence has increased in step with the increased use of central venous catheters (CVCs) and pacemakers. Although most patients present with pain, swelling, parathesias, and prominent veins throughout the arm or shoulder, many patients will not display any local DVT symptoms. For example, Kabani et al recently presented data for 1,275 patients admitted to the surgical ICU over a 12-month period. They found the incidence of UEDVT was higher than that of LEDVT (17% vs. 11%; P=0.11). They also determined that scanning all four extremities diagnosed more DVT than two-extremity scans (33% vs. 7%; P<0.001).3
While current medical literature has pushed for increased UEDVT attention, there is no consensus on its treatment. Recent American College of Chest Physicians (ACCP) guidelines addressed UEDVT treatment specifically and recommended analogous treatment to LEDVT with heparin and warfarin.4 This follows prospective studies that have shown patients with UEDVT and LEDVT have similar three-month clinical outcomes. The ACCP guidelines do not specifically recommend different treatment courses based on whether the UEDVT is catheter-related or not. Furthermore, while one might assume that removal of an associated catheter might reduce the treatment duration, there is limited data to support shorter courses in this scenario.
Incidence: UEDVT is becoming more common secondary to increased interventions in the upper extremity (CVC, pacemaker), and is more easily recognized due to improvement in noninvasive ultrasound technology. UEDVT accounts for up to 10% of all DVT, with an incidence of approximately three per 100,000 persons in the general population.5-8 Because UEDVT can also be asymptomatic, it is believed that the incidence likely is higher than previously reported, but prospective data are lacking.
Risk factors: UEDVT is further categorized as either primary or secondary, depending upon the cause. First described in the late 1800s, spontaneous primary thrombosis of the upper extremity, or Paget-Schroetter syndrome, accounts for approximately 20% of UEDVT.9 Primary UEDVT includes both idiopathic and “effort-related” thrombosis. Effort-related thrombosis usually develops among young people after strenuous or repetitive exercise, such as pitching a baseball. Some hypothesize that effort-related thrombosis is related to a hypercoaguable state or anatomic abnormalities, although a specific cause, such as thoracic outlet syndrome, is found in only 5% of these cases.10,11
Secondary UEDVT characterizes thrombosis in which an endogenous or exogenous risk factor is present. Endogenous risk factors include coagulation abnormalities, such as antithrombin, protein C and protein S deficiencies; factor V Leiden gene mutation; hyperhomocysteinemia; and antiphospholipid antibody syndrome. Exogenous risk factors include CVC pacemakers, intracardiac defibrillators, malignancy, previous or concurrent LEDVT, oral contraceptives, some artificial reproductive technologies (women can develop ovarian hyperstimulation syndrome, which is associated with increased hypercoaguability), trauma, and IV drug use (especially cocaine).5,12-14
Clinical presentation and diagnosis: Swelling (80% of patients) and pain (40% of patients) are the most common UEDVT symptoms at presentation.2 Other clinical features include new, prominent veins throughout the shoulder girdle, erythema, increased warmth, functional impairment, parathesias, and non-specific feelings of arm heaviness or discomfort. Symptoms typically worsen with arm use and improve with rest and elevation.15 Patients with UEDVT related to CVC are more likely to be asymptomatic and may present only with PE.16 The differential diagnosis includes superficial phlebitis, lymphatic edema, hematoma, contusions, venous compression, and muscle tears.17
Contrast venography is the gold standard for the UEDVT diagnosis. However, it is more expensive and invasive than ultrasound, and thus serial compression ultrasound is now the standard test in UEDVT evaluation. Then again, contrast venography remains the test of choice in patients with high pre-test probability and negative ultrasound results.18,19
Prevention: Nearly 70% of secondary UEDVT is associated with a CVC.5 Further, CVC use is the most powerful predictor of UEDVT (adjusted odds ratio (OR), 9.7; 95% CI, 7.8 to 12.2).2 Despite the association between CVCs and UEDVT, anticoagulant prophylaxis is not recommended. Studies evaluating the results of 1-mg warfarin conflict and include small populations. Warfarin’s potential interaction with antibiotics and dosing variance based on nutritional intake logically prompted studies on the potential benefit of low-weight molecular heparain (LWMH); however, these studies have failed to show benefit.20,21
Treatment: Recent ACCP guidelines recommend treating UEDVT patients with unfractionated heparin (UFH) or LMWH and warfarin, with an INR goal of 2 to 3 for at least three months depending upon the overall clinical scenario. Two small studies evaluating catheter-related thrombosis (15 patients in each trial) reported no subsequent embolic phenomenon.22,23 Some authors interpreted this data to mean UEDVT was not as morbid as LEDVT and, subsequently, that catheter-related UEDVTs require only one month of therapy. Since the small studies were published, the increasing incidence and relevance of UEDVT have become more widely recognized, and most authors are recommending three months of treatment.
Still, it’s important to note that there aren’t any published data directly comparing the one-month and the three-month anticoagulation therapies. The RIETE registry, which is the largest ongoing published registry of patients with confirmed DVT or PE, reports similar three-month clinical outcomes between those with UEDVT and LEDVT.
Small, single-center trials have shown that active intervention, such as thrombolysis, surgery, or multi-staged approaches are associated with increased vein patency and decreased rates of post-thrombotic syndrome.24,25 However, ACCP has withheld general recommendations for these interventions based on a lack of sufficient data to comment on their overall safety and efficacy, as well as comparable rates of post-thrombotic syndrome (15% to 50%) in studies that directly compared surgical and medical intervention. In fact, the ACCP recommends against interventional treatments unless the patient has failed anticoagulation therapy, has severe symptoms, and expertise is available.4
Superior vena cava filters are available at some centers for patients in whom anticoagulation is contraindicated, but efficacy data is limited. While the data for filter use in UEDVT is limited, its use should be considered in patients who have a contraindication to anticoagulation and remain high risk for UEDVT (e.g., prolonged central line placement).
Complications: Post-thrombotic syndrome (PTS) is the most significant local complication of UEDVT. PTS characteristics are edema, pain, venous ulcers, and skin pigmentation changes, and it is the result of chronic venous insufficiency due to the clot. A meta-analysis of clinical studies on UEDVT noted that PTS occurs in 7% to 46% (mean 15%) of patients.26 One hypothesis for the wide range in frequency is the lack of clear diagnostic criteria for PTS.27 No clear beneficial treatment or prevention for PTS exists, but many recommend graduated compression stockings for the arm.
Residual and recurrent thrombosis are associated with increased PTS risk, which emphasizes the need for further study of interventional treatment because preliminary work has shown increased rates of vein patency in comparison to anticoagulants alone. Recurrent venous thromboembolism (VTE), another local complication, appears to occur less often than it does in patients with LEDVTs, but reaches 8% after five years of followup.28
PE is less common on presentation among patients with UEDVT when compared to patients with LEDVT, but when PE occurs, the three-month outcome is similar.3 PE appears to be more frequent in patients who have a CVC, with an incidence as high as 36% of DVT patients.4,13,21,29
Increased mortality: The mortality among UEDVT patients has been described as 10% to 50% in the 12 months after diagnosis, which is much higher than the ratio in LEDVT patients.21,30 This in part is due to sicker cohorts getting UEDVT. For example, patients with distant metastasis are more likely to develop UEDVT than those with confined malignancy (adjusted OR 11.5; 95% CI, 1.6 to 80.2).31
Occult malignancy, most commonly lung cancer or lymphoma, has been found in as many as 24% of UEDVT patients.32 The high rate of mortality associated with UEDVT appears to be related more with the patient's overall poor clinical condition rather than directly related to complications from the DVT. However, its presence should alert hospitalists to the patient's potentially poorer prognosis and prompt evaluation for occult malignancy if no risk factor is present.
This patient should be started on either UFH or LMWH while simultaneously beginning warfarin. She should continue warfarin treatment for at least three months, with a goal INR of 2.0 to 3.0, similar to treatment for LEDVT. The ultimate treatment duration with warfarin follows the same guidelines as treatment with a LEDVT. Although prophylaxis is not routinely recommended, dosing 1 mg of warfarin beginning three days before subsequent CVC placement should be considered if this patient requires a future CVC. Additionally, an evaluation for occult malignancy should be considered in this patient.
Upper extremity DVT is not a benign condition, and is associated with a general increase in mortality. It should be treated similarly to LEDVT in order to decrease PTS, recurrent DVT, and PE. TH
Dr. Hollberg is a clinical instructor in the section of hospital medicine at Emory University Hospital in Atlanta.
The Hospitalist newsmagazine reports on issues and trends in hospital medicine. The Hospitalist reaches more than 25,000 hospitalists, physician assistants, nurse practitioners, residents, and medical administrators interested in the practice and business of hospital medicine.