The association between malignancy and venous thromboembolism (VTE) is well known. Non-Hodgkin lymphoma (NHL) increases VTE risk, though most studies do not differentiate between low- and high-grade NHL. In order to better understand the natural history of VTE in NHL, we examined the incidence and factors associated with VTE in patients with diffuse large B-cell lymphoma (DLBCL) and follicular lymphoma (FL).


Patients diagnosed with DLBCL or FL between 1998 and 2008 and treated within the VHA system with CHOP or CHOP-like regimens (+/- rituximab) were identified in the VA Central Cancer Registry. Those with VTE prior to the lymphoma diagnosis or history of atrial fibrillation were excluded. Data on sex, race, stage, lactate dehydrogenase (LDH), disease histology, B-symptoms, body mass index, co-morbidities, medications, and date of VTE event were obtained. Incident cases of VTE were determined through the use of ICD-9 codes for VTE combined with either pharmacy records of anti-coagulation therapy, placement of a vena-cava filter, or death within 30 days of an ICD-9 code for VTE. Logistic regression analysis was used to identify baseline factors associated with VTE. VTE incidence was then analyzed to evaluate differences before, during, and after treatment.


2606 NHL patients (1838 DBLCL, 768 FL) were identified, of whom 179 developed VTE (119 DLBCL, 33 FL) within 2 years of treatment initiation. The mean age at NHL diagnosis was 64 years. A majority of patients were male (97%), Caucasian (87%), and had a diagnosis of DLBCL (70%). The mean Charlson co-morbidity score in the cohort was 2.2. Annualized VTE incidence rates were significantly different between DLBCL and FL patients during the time from diagnosis to treatment initiation (18.7% vs. 6.1%; p<0.001) and in the 6 months after treatment initiation (9.3% and 3.8%; p=0.001). The annualized incidence dropped to 1% or less in both DLBCL and FL patients during the period 6 months to 2 years following treatment initiation. On multivariate logistic regression analysis, factors associated with increased odds of developing VTE within 6 months of treatment initiation included: DLBCL histology (OR 2.04; 95% CI 1.27 – 3.31), BMI ≥ 30 (OR 2.58; 95% CI 1.61 – 4.17), and stage III/IV disease (OR 1.7; 95% CI 1.14 – 2.55). There was a protective association observed for each point increase in the Charlson co-morbidity score (OR= 0.85; 95% CI 0.78 – 0.96).


VTE risk was greatest in time between diagnosis and 6 months after treatment initiation in all NHL patients. This is potentially due to greater disease burden during the time between diagnosis and treatment initiation and the pro-thrombotic effects of chemotherapy in the six months after treatment initiation. After controlling for potential confounding factors, the odds of VTE were twice as high in DLBCL compared to FL patients. While prophylactic anticoagulation in cancer patients remains controversial, future efforts in NHL should be focused on patients with higher grade histology during the timeframe between diagnosis and shortly after treatment cessation.


Carson:Spectrum Pharmaceuticals: Honoraria, Research Funding, Speakers Bureau.

Author notes


Asterisk with author names denotes non-ASH members.

This icon denotes a clinically relevant abstract