Pulmonary embolism without deep venous thrombosis

Abstract

Background: To identify patients with pulmonary embolism (PE) without deep venous thrombosis (DVT), and to compare them with those with an identifiable source on upper (UED) and lower-extremity venous duplex scans (LED).

Methods: We performed a retrospective review of 2700 computed tomography angiograms of the chest between January 2008 and September 2010 and identified 230 patients with PE. We then evaluated the results of UED and LED and divided the patients into four groups based on the results of their duplex studies. We compared patients with PE and DVT with those with PE and no DVT in terms of age, gender, size and location of PE, critical illness, malignancy, and in-hospital mortality.

Results: We identified 152 women and 78 men (mean age, 68 years) with PE. One hundred thirty-one patients had a documented source of PE (group 1). Fifty-three patients had negative LED results, but did not undergo UED (group 2). Thirty-one patients did not undergo either LED or UED (group 3). Seven men and eight women had no documented source of PE on UED and LED (group 4). Ten of 15 patients in group 4 had a documented malignancy listed as one of their diagnoses. Because patients in groups 2 and 3 did not undergo complete duplex studies, we excluded them from our analysis. We then reviewed the discharge summaries of patients in groups 1 and 4. There was no statistically significant difference in age and gender distribution, size and location of PE, critical illness, smoking status, cardiovascular disease, trauma, and in-hospital mortality between patients in group 1 and 4. Patients in group 4 had a statistically significant increased prevalence of malignancy (67% vs. 40%, P = 0.046). Patients in group 4 also had a higher percentage of active cancer than those in group 1 (47% vs. 24%, P = 0.084), although not statistically significant. We defined active cancer as either a metastatic disease or a malignancy diagnosed shortly before or after the diagnosis of PE. Patients who were undergoing treatment for cancer at the time of diagnosis of PE were also considered to have active cancer.

Conclusion: We demonstrated a statistically significant increased prevalence of malignancy in patients with PE without DVT. However, pathophysiology and clinical significance are the aspects that remain to be understood after accrual of more patients and further research. Possibilities such as de novo thrombosis of pulmonary arteries, complete dislodgement of thrombi from peripheral veins, or false-negative venous duplex need to be explored.