Incidence of Unprovoked Venous Thromboembolic Events in Patients with Chronic Lymphocytic Leukemia (CLL)

While lymphoma is a known risk factor for venous thromboembolism (VTE), the thrombogenicity of chronic lymphocytic leukemia (CLL) has not been investigated. One study reported a ten-fold increased risk in VTE compared to the general population, but this included both provoked and unprovoked events, making it difficult to discern the independent contribution of CLL to the thrombogenic predisposition (Whittle A et al. Leuk Res 2011; 35:419-21). Since we found no studies examining only unprovoked VTE in CLL, we sought to estimate the risk of spontaneous VTE occurring in patients with CLL in the absence of known thrombogenic risk factors. The study was approved by the institutional human subjects committee. We examined patients diagnosed with CLL between 1997-2014 by the persistent presence of ≥ 5,000 circulating small, mature, monoclonal, CD5+, CD23+ B lymphocytes (per diagnostic criteria in: Halleck M et al. Blood 2008;111:5446-56) and listed in the Minneapolis VA Tumor Registry. Clinical and laboratory data were obtained from the electronic medical record. Each patient's chart was reviewed individually for VTE, either pulmonary embolism (PE) or deep vein thrombosis (DVT), and the report of a confirmatory imaging study was examined to confirm the diagnosis.

We followed patients (n = 308, 99% males, mean age 70 years) starting at the date of CLL diagnosis or the date after CLL diagnosis when the patient began follow-up at the VA and ending at last patient contact or death (122/308 [39.6%] patients died during the observation period). Each patient made at least annual contact with the VA between the start and end dates, with a mean VA follow-up time of 4.6 +/- 2.9 years; 90% patients were followed continuously at the VA since after the diagnosis of CLL. Patients were excluded if there was a preceding period of immobility, serious illness requiring hospitalization, diagnosed hypercoagulable state, or recent chemotherapy known to be associated with increased thrombotic risk. The total follow-up time was 1408 patient-years, during which 10 unprovoked VTE occurred, for an estimated incidence rate of unprovoked VTE of 0.71 per 100 patient-years, 95% CI [0.38, 1.32]. Eight VTE were originally diagnosed at our institution. Confirmatory imaging was available for the 2 VTE diagnosed elsewhere. All 10 patients were males between the ages of 55 to 95 years at the time of CLL diagnosis. Six had DVT, three had PE, and one had both DVT and bilateral PE. Nine patients had symptoms that prompted diagnostic imaging, while one PE was diagnosed incidentally on a chest CT scan. Nine patients had bulky lymphadenopathy at the time of VTE, but these did not compress the relevant vein in any patient. Small, stable monoclonal paraproteinemia was identified in 2/5 VTE patients tested (one patient at 0.39 and 0.35 g/dL, one patient at 0.32 and 0.27 g/dL). At time of development of VTE, 4, 1, 1 and 3 patients had Rai stage I, II, III or IV CLL, respectively. Thus, while there were relatively few events, there did not appear to be any clear relationship of VTE with advanced CLL stage. Recurrence of unprovoked VTE occurred in only 1 of 10 (10%) patients.

In a study of the general population in our region of the US (Olmsted County, MN), the overall incidence of VTE in 70-74 year old males was 0.65 per 100 patient-years (Silverstein M et al. Arch Internal Med 1998;158:585-93). Since 17-47% of all VTEs were unprovoked in various population series (Cushman M et al. Thromb Haemost 2001;86[suppl 1]:OC2349, Heit JA et al. Arch Intern Med 2002;162:1245-8, Spencer FA et al. J Thromb Thrombolysis 2009;28:401-9, White RH Circulation 2003;107:I-4-8), the incidence rate of unprovoked VTEs observed in CLL patients in the current series (0.71; 95% CI: 0.38, 1.32) approximates the expected incidence of unprovoked VTEs in the general population. Further, the development of unprovoked VTE would have been expected to be skewed towards patients with advanced stage (Rai stages III/IV), and/or have been associated with more frequent recurrences, if CLL was an independently thrombogenic condition.

Conclusion: This is the first long-term study of unprovoked VTE in patients with CLL, with >1400 patient-years of follow-up. The low incidence of unprovoked VTEs observed in our study suggests that primary thromboprophylaxis may not be required in patients with CLL who do not have additional risk factors for thromboembolism.