Increased Incidence of Thrombosis with Lenalidomide in Chronic Lymphocytic Leukemia: Effect of Lenalidomide On Inflammation, Endothelial Cell Damage and Coagulation.

Abstract 455

There is an increased risk of deep venous thrombosis (DVT) in patients (patients) undergoing lenalidomide (L) therapy. Previous results from our group indicated significant upregulation of tumor necrosis factor alpha (TNFα) in CLL patients on L therapy. Here, we hypothesized that L may increase the risk of DVTs in CLL due to endothelial cell damage resulting from high TNFα serum levels. We studied 27 patients with relapsed CLL enrolled on an investigator initiated phase II single agent L trial (ClinicalTrials.gov Identifier: NCT00465127, 20mg for patients 1-10, 10mg for patients 11-27). The pre-specified accrual goal has been reached. L was given for 4 cycles of L daily for 21 days followed by 21 days off drug. Patients with at least partial response could receive treatment for up to 8 cycles. One patient did not receive therapy (patient's choice). Venous doppler studies (VDS) were obtained at baseline, cycle 4, cycle 8 and when DVT was suspected. Aspirin prophylaxis was not mandated. Serum samples were evaluated for TNFα, soluble vascular endothelial adhesion molecule 1 (sVCAM) and C-reactive protein (CRP). Plasma samples were evaluated for quantitative D-dimer and soluble thrombomodulin (TM). 26 patients received on average 4 cycles of L (total 107, range 1 to 8 cycles). The median age was 64 years (36-78). Results of hypercoagulabel work-up prior to enrollment were: protein C deficient 0%, mild protein S deficient 12%, mild antithrombin III deficient 8%. Factor V Leiden heterozygosity was found in 1 out of 19 patients, prothrombin gene 20210 polymorphism was absent in all 19 patients tested. Five patients developed six incidents of DVT: hypercoagulability screen was negative in all; 2 patients were fully anticoagulated at the time due to previous DVT. Time to DVT was 134 days (56-259). DVTs occurred during cycle 2, 3, 4, 7 and 8. No PE was detected. Three out of 6 DVTs were treated with tPA thrombolysis. All 5 patients were maintained on full dose anticoagulation for 6 months and 4 continued L. One patient had a second DVT on routine VDS at the end of cycle 8. We have previously shown that L induces a prominent inflammatory state in CLL especially in the first cycle. TNFα was significantly upregulated on day 8 of cycle 1 (increase from mean 32 to 122 pg/ml, P = 0.01) and was persistently upregulated to day 22 (78 pg/ml, P <0.01). C-reactive protein as a global marker of inflammation was also significantly increased on day 8 of cycle 1 (from 0.5 to 4.1 mg/dl, P<0.01). Notably, TNFα levels on day 8 of cycle 1 were significantly higher in patients who subsequently developed DVTs than in patients who did not (269 pg/ml vs 87 pg/ml, respectively P=0.02). Next, we assessed effects of L on the coagulation system: D-Dimers increased on day 8 (from 0.5 to 1.1ug/ml, P <0.01) indicating increased fibrin turnover. TM, an endothelial cell surface marker, increased from 93 ug/ml at baseline to 130 and 129 ug/mL on day 8 and day 22, respectively (P <0.01). Soluble VCAM an endothelial cell surface marker independent of TM increased from 2.6 ug/ml at baseline to 3.8 and 3.4 ug/ml on day 8 and day 22, respectively (P<0.01) and patients with subsequent DVTs had higher VCAM levels on day 8 than patients without DVTs (4.2 vs 3.8 ug/ml, P=0.04). Consistent with a link between TNFα and endothelial damage we found a strong correlation between TNFα and TM serum levels (r=0.61, P<0.01) and between TNFα and VCAM serum levels (r=0.74, P<0.01). In summary, in CLL patients treated with L we find a 19% incidence of DVTs that occur in the context of a systemic inflammatory syndrome induced by L, whereby increased release of TNFα correlates with endothelial cell damage. Strikingly, high levels of TNFα in the first cycle identify patients who are more likely to develop DVTs on L. Our results not only identify at risk patients, but also suggest that effective strategies for DVT prevention in L treated patients may have to focus on reducing the inflammatory syndrome and in particular TNFα release. This could be accomplished with the use of anti-inflammatory drugs, and TNFα antagonists may be particularly effective.