Thalidomide causes platelet activation, which can be abrogated by aspirin

Thalidomide is an important therapeutic option for patients with myeloma; however, its usefulness is hindered by an increased incidence of venous thrombosis. This is particularly seen in patients receiving combination therapy with doxorubicin and high-dose dexamethasone, where the incidence is as high as 30% and 26% respectively [1]. Even with the traditional oral regimen of melphalan and prednisone, the addition of thalidomide has been reported to increase the 2-year thrombosis rate from 1.5% to 18.5% [2].

Fortunately, antithrombotic prophylaxis with low- molecular-weight heparin or warfarin can abrogate the thrombotic tendency [3]. Aspirin has also been shown to significantly reduce the frequency of venous thromboembolism (VTE) even in high-risk patients treated with both doxorubicin and high-dose dexamethasone [4]. In a recent publication in the Journal of Thrombosis and Haemostasis, the Italian Multiple Myeloma Network reported that the incidence of VTE in patients treated with melphalan and prednisone was not elevated by the addition of lenalidomide (a thalidomide analog) if aspirin prophylaxis was used [2]. Interestingly, the incidence of thrombosis was lower in this group compared with patients receiving thalidomide who received prophylactic enoxaparin, albeit not in a direct comparison.

The mechanism of thrombogenesis is uncertain. Endothelial cell activation has been documented as has elevated von Willebrand factor, which may be a surrogate marker of endothelial cell injury [4,5]. The role of platelet activation has not been thoroughly examined, although it is an attractive hypothesis given the documented efficacy of aspirin. In a study by Streetly et al. [5], an increase in soluble P-selectin was seen after patients were exposed to the thalidomide analog actimid. In addition, Baz et al. [4] found increased platelet aggregation to platelet agonists in patients on thalidomide who later developed VTE[A].

We explored platelet activation because of thalidomide with and without aspirin. We examined five patients (Table 1) who received thalidomide for progressive disease as second-line therapy [four patients received thalidomide alone, and one patient (patient 2) received thalidomide in combination with dexamethasone at a dose of 40 mg day^–1 for 4 days every fortnight]. There was no history of prior VTE in any patient. Patient 4 had Type 1 diabetes mellitus and poorly controlled hypertension. Samples were collected at baseline, at 1 month (average 32 days, range 28–42 days) after treatment with thalidomide, and later again after at least 2 weeks of aspirin prophylaxis (average 16 days, range 14–19 days). Thalidomide was commenced at 100 mg day^–1 for one week and then escalated to 200 mg day^–1. Aspirin was administered at a dose of 150 mg day^–1. No thrombotic event occurred in the follow-up period (average 11 months, range 6–18 months).

Platelet surface P-selectin expression and platelet–leukocyte aggregates were examined by flow cytometry using previously published methods [6]. Briefly, 5 μL of whole blood (3.2% citrate sample) was incubated with 5 μL CD42b-PE (Immunotech, Marseille, France), 10 μL CD62p-PE (Becton Dickinson, San Jose, CA, USA), or corresponding isotypic control (Immunotech and Becton Dickinson, respectively), together with 40 μL (35 μL for CD62p) of HEPES-buffered saline (HBS), for 10 min at room temperature and immediately analyzed. Analysis regions were verified to be platelets by expression of anti-CD42b. An analysis marker placed to the right of the negative control fluorescence histogram was set so that 0.5% events were positive, and a total of 10 000 events was analyzed. For platelet–leukocyte aggregates, 25 μL of whole blood was incubated with 5 μL CD42b, 5 μL CD45-PECy5 (DAKO, Buckinghamshire, UK) and 65 μL HBS for 10 min at room temperature. The sample was then fixed, lyzed and analyzed, isolating monocytes by their characteristic CD45-side-scatter plot. Platelet-specific (CD42b) events occurring within this region were measured and counted as monocyte–platelet complexes. Data are presented as percentage positive cells. The flow cytometer used was a Becton Dickinson FACSCalibur. The aggregate data from each group were compared using Student’s t-test. A P-value < 0.05 was considered statistically significant.

There was increased platelet P-selectin expression (P = 0.005; Table 1) and platelet–monocyte complexes (P = 0.01) after patients received thalidomide, suggesting platelet activation. Following the prophylactic administration of aspirin, both P-selectin levels (P = 0.01) and platelet–monocyte complexes (P = 0.01) decreased, returning to levels of platelet activation seen at baseline (P = 0.9 for both). Patient 4 had evidence of significant platelet activation prior to exposure to thalidomide; however, from our previous experience these levels are not uncommon in patients with poorly controlled Type 1 diabetes mellitus. Patient 5 showed no change in platelet activation in response to thalidomide.

These results support the hypothesis that thalidomide can cause platelet activation, which can be abrogated by aspirin. This may help to explain the clinical benefit of aspirin over anticoagulants. The mechanism of platelet activation needs further investigation as it may be directly induced by thalidomide or indirectly via thalidomide-induced endothelial cell activation. Indeed, aspirin has been shown to reduce endothelial cell activation in response to cytokine stimulation [7,8]. A further study involving markers of both endothelial cell and platelet activation in patients receiving combination chemotherapy and thalidomide is underway.

Disclosure of Conflict of Interests

The authors state that they have no conflicts of interest.