BACKGROUND: CD36 is a widely distributed transmembrane glycoprotein, called GpIV on platelets, involved in various physiologic processes. It has been implicated in thrombogenic pathologies such as diabetes, atherosclerosis and anti-phospholipid syndrome (APS). It is also an adhesion molecule capable of interacting with collagens and is considered a platelet collagen receptor. We investigated autoantibodies (Ab) against CD36 in TTP [Schultz et al, BrJH 103:849, 1998]. The present study was undertaken to explore relationships of anti-platelet Ab (aPlt-Ab), antiphospholipid Ab (APLA), and circulating cell-derived microparticles (MP) in patients with thrombosis.

METHODS: Forty-five patients with documented thrombosis (TB) referred for hypercoagulable workups were recruited consecutively. Of them, 18 suffered from recurrent thrombosis (rTB) while 27 had a single event (sTB) within the past 5 years. We measured APLA, lupus anticoagulant (LA), antiplatelet antibody (aPlt-Ab) and MP. The APLA were IgG and IgM against cardiolipin (CL) and β2-glycoprotein-I (β2GP1) measured by ELISA. The aPlt-Abs were IgG and IgM reacting to GpIIb/IIIa (CD41b), GpIb/IX (CD42b), and GpIV (CD36), assayed by the PAICA method of Macchi et al [Thromb Haemost 76:1020, 1996]. The following types of MP were measured by flow cytometry: endothelial MP defined by CD31+/CD42− (EMP31) or by CD62E+ (EMP62); platelet MP defined by CD31+/CD42+ (PMP); leukocyte MP defined by CD45+ (LMP); and red blood cell MP by glycophorin+ (RMP). Lupus anticoagulant (LA) was measured in the hospital clinical laboratory and was considered positive if 2 of 3 different test methods were positive. Platelet activation was measured by CD62p expression in flow cytometry.


  1. Although both groups had increased prevalence of APLA and LA, there were no significant differences between the sTB and rTB groups. About 70% of patients in both groups were positive for at least one APLA.

  2. Of the aPlt-Ab, only aGpIV (aCD36) statistically discriminated between the groups, being more frequent in rTB for both IgG (p<0.02) and IgM (p<0.02).

  3. Among the MPs assayed, PMP were elevated in all patients in both groups. EMP62 discriminated best between rTB and sTB (p=0.003). LMP and RMP were also significantly higher in rTB than sTB, p<0.025. When we compared patients CD36+ vs CD36− we found that EMP62 was more frequently elevated in CD36+ patients (7/12, 58%) than in CD36− patients (2/32, 6%), p=0.001.

  4. Platelet activation marker CD62p was more often increased in rTB than sTB, p=0.022.

CONCLUSION: GpIV (CD36) antibodies are the most prominent risk factor for rTB to emerge from this study. This confirms a previous report [Pelegri et al, Clin Exp Rheum 21:221, 2003]. EMP, LMP, RMP and platelet activation (CD62p) are also increased in the recurrent TB group vs single TB. EMP62 were associated with aCD36. Since CD36 is found also on endothelial cells, this suggests that aCD36 could be responsible for the high EMP62 in the rTB group. The present study demonstrates an association between high EMP62 and aCD36. We suggest that persisting autoantibodies against GpIV activate EC to predispose to recurrent thrombosis.

Author notes

Disclosure: No relevant conflicts of interest to declare.