Differences of soluble CD40L in sera and plasma: Implications on CD40L assay as a marker of thrombotic risk

Introduction: Soluble CD40L (sCD40L) ELISA has emerged as a promising predictor of poor outcomes in acute coronary syndrome. Yet many blood processing techniques have been used with little consideration of their effect on the results. Methods: We measured sCD40L by ELISA in 10 patients with thrombocytopenia and 12 with normal or high platelet counts and 8 healthy controls using three sampling techniques: serum clotted on ice (serum-I) or at room temperature (serum-RT) and platelet poor plasma (PPP). Results: Serum-RT samples, compared to serum-I, gave significantly higher CD40L values (p = 0.003), demonstrating that ex vivo sCD40L release by activated platelets is inhibited by cold temperature. Although serum-I and PPP were comparable in patients with normal platelet counts, serum-I gave significantly higher values than PPP in the thrombocytosis group (p = 0.01), suggesting that cold inhibition is insufficient in the latter group. To estimate the fraction of sCD40L that was microparticle-bound CD40L (mp-CD40L), 16 samples underwent 0.1-mum filtration. 50.6% of sCD40L was mp-CD40L in serum-RT, whereas 21.3% and 29.9% were observed in serum-I and PPP, respectively. Lastly, plasma sCD40L was assayed in 46 patients with and 35 without thrombosis. Plasma sCD40L did not correlate with platelet count in non-thrombotic, non-inflammatory patients but did (p < 0.01) in those with thrombosis. Conclusions: Sample processing and temperature profoundly affect sCD40L assay. Serum-I and PPP minimize the release of sCD40L ex vivo and better represent sCD40L in vivo. However, PPP may be preferable particularly in patients with thrombocytosis. The existence of mp-CD40L highlights the importance of centrifuge conditions. (C) 2004 Elsevier Ltd. All rights reserved.