Frequency, Risk Factors and Mortality Effect of Venous Thromboembolism in Adult Patients with Central Nervous System Lymphoma

Background: Venous thromboembolism (VTE) is an established complication of cancer, with a high incidence in patients with both lymphoma (5-15%) and gliomas (15-20%). There is minimal data describing the incidence of VTE in patients with CNS lymphoma. Known VTE risk factors including obesity, immobility, and hospitalization are present in those with central nervous system (CNS) lymphoma; however, concern for bleeding complications limits the use of pharmacologic prophylaxis. The aim of this study is to characterize the frequency of VTE in adult patients with CNS lymphoma treated at a large academic medical center.

Methods: The University of California, San Francisco (UCSF) Cancer Registry was queried for all cases of adult CNS lymphoma diagnosed from 2008-2014. Chart review was performed and patients were excluded if the primary treatment occurred elsewhere. Demographics, histology, treatment and mortality data were examined. The presence of VTE was defined by radiographic evidence of a pulmonary embolism (PE) or deep vein thrombosis (DVT).

Results: 134 cases of adult CNS lymphoma were reported at UCSF between 2008-2014; 18 patients were excluded. The average follow up time was 3.8 years from date of diagnosis. The mean age at diagnosis was 63. Of the 116 patient included in the study, 77 (66%) were identified as having Primary CNS Lymphoma and 39 (34%) had Secondary CNS Lymphoma including the following subtypes: Diffuse Large B-cell, Burkitt's, Mantle cell, Marginal zone, and NK cell lymphoma.

There were 34 cases of VTE (29.3%): 12 were PE and 22 were DVT, 32 (94%) patients were hospitalized at the time of VTE diagnosis. Twenty eight (82%) were symptomatic. Ten (29%) were line-associated VTE. Fifty percent of VTE cases occurred during cycle 1 (18%) or cycle 2 (32%) of chemotherapy and 27 patients (79%) received systemic steroids within 30 days of VTE diagnosis. The median time from diagnosis of CNS lymphoma to VTE was 70 days.

Confounding variables including age at diagnosis, race, sex, smoking history and histology were studied and no difference was found between the VTE and non-VTE groups. Notably, body mass index was significantly higher in the VTE group (29.85, 95% CI: 26.5-33.2) as compared to the non-VTE group (25.66, 95% CI: 24.76-26.55, p=0.019). There was a trend toward shorter time to death after the diagnosis of CNS lymphoma in the VTE group (382 days, 95% CI: 106-657) vs. the non-VTE group (699 days, 95% CI: 348-1049, p= 0.18). There was no difference in overall survival (log rank p=0.09).

Conclusions: Here we report a remarkably high frequency of VTE (29%) in CNS Lymphoma patients treated at a large academic center. While there was no difference in age, sex, race, smoking history or histology, those who developed VTE did have a significantly greater BMI. Because many patients receive eight cycles of high dose methotrexate as treatment for CNS Lymphoma, our determination that half of the VTE episodes occurred during the first two cycles may be useful in defining the optimal timing of VTE prophylaxis in this high risk population. Most (82%) of the VTE reported in this study were symptomatic, which suggests VTE may have a significant impact on quality of life. There was a trend toward shorter time to death after diagnosis of CNS lymphoma in the VTE group, suggesting VTE may affect mortality. Larger, prospective studies are needed to characterize the timing and incidence of VTE in patients with CNS Lymphoma, but this study reports one of the largest single institution experiences in the literature.