Back to Annual Meeting Program

Introduction:
The reported incidence of pulmonary embolism (PE) associated with isolated upper extremity deep venous thrombosis (UEDVT) is variable (1-13%). As per Chest 2008 guidelines, many clinicians treat UEDVT with anticoagulation (AC) to ameliorate symptoms, and decrease the risk of subsequent PE. Following the analysis of a smaller cohort of UEDVT patients, we had previously reported a low risk of PE attributable to UEDVT. We here expand our analysis to 500 UEDVT patients and compare the observed outcomes among both anticoagulated and non-anticoagulated (NAC’d) patients in terms of PE and hemorrhagic risk.
Methods:
Between April 2005 and July 2010, 500 consecutive UEDVT patients were identified from a prospectively maintained Registry. A retrospective analysis was then performed among AC’d and NAC’d patients, comparing patient demographics and outcomes, as well as UEDVT characteristics. The decision to AC patients reflected the judgment of the treating physicians, and was not randomized.
Results:
UEDVTs were identified in the distal innominate (n=113), internal jugular (n= 239), subclavian (n= 261), axillary (n= 190), and brachial veins (n=164), with many patients having multiple named venous segments involved. Most UEDVTs had sonographically acute components (86%).
Among the 500 patients, 297 were AC’d (59%), while 203 patients were NAC’d. There were 403 patients (81%) with catheter-associated UEDVT. Most patients were symptomatic, and patients who were anticoagulated had more UEDVT segments involved (See Table). There were no statistically significant differences in the treatment groups related to trauma or post-operative status, malignancy, renal failure, diabetes, coronary artery disease or obesity. AC’d patients were younger than NAC’d patients, with less associated co-morbidities. The NAC’d group demonstrated decreased survival, perhaps owing to older patient age and more co-morbidities, but not due to increased PE rates.

Patient/UEDVT Characteristics	NAC (n = 203)	AC (n = 297)	P-Value
Patient Age	54 +/- 17	50 +/- 18	0.02*
Mortality at 3 months	57/203 (28%)	44/297 (15%)	<0.0001**
Symptomatic	174/203 (86%)	268/297 (90%)	0.12**
UEDVT associated PE	2/203 (1.0%)	13/297 (4%)	0.022**
Malignancy	77/203 (38%)	96/297 (32%)	0.12**
No of named DVT segments	1.7 +/- 1.0	2.1 +/- 1.1	<0.0001*
*t-test **Chi-square

The overall rate of UEDVT-associated PE was 3%, with no associated mortality. Among the 15 patients suffering PE in association with their UEDVT, 8 patients presented with simultaneous PE and UEDVT, and were anticoagulated. Only 7 patients suffered PE subsequent to a primary diagnosis of UEDVT, of which most (5 of 7) were anticoagulated.
After hospital discharge, four of the UEDVT patients treated with AC developed fatal intracranial bleeds. Seven additional patients had their AC discontinued early due to hemorrhagic complications requiring hospital readmission and transfusion.
Conclusions:
We continue to observe that clinicians choose to AC UEDVT patients that are younger, with more extensive thrombus burdens. The risk of PE subsequent to UEDVT diagnosis remains small however, and we observed no evidence that AC dampens this risk. In light of the persistent risk of hemorrhagic complications demonstrated among this fragile patient group, our patient outcomes do not support routine anticoagulation among UEDVT patients.