Diagnosis of atrial fibrillation as the source of cryptogenic retrieved clots by a novel thrombin secretion assay.

Background: Secondary prevention of acute ischemic stroke depends on identifying the source of cryptogenic clots. We previously reported that secreted thrombin activity from endovascularly retrieved clots is significantly different in atrial fibrillation (AF) vs atherosclerosis (AS), probably due to the in vivo biology of the clots.

Objectives: The purpose of this study was to validate and optimize thrombin secretion for clot source diagnosis.

Methods: Strokes were defined as either AS, AF, or cryptogenic, with the latter group defined as AF or non-AF by long-term clinical follow-up. Artificial venous blood clots were either prewashed or unwashed to model AS and AF sources. Thrombin secretion from the clot was measured fluorometrically by manually transferring the clots between wells or by washing them in an agarose bead column, collecting samples manually, or by using the fraction collector. The secretion pattern was represented as the ratio of maximal initial values to minimal baseline.

Results: The new column-based assay was faster and more informative than the manually transferring approach. Significant differences in thrombin secretion ratios were observed between prewashed and unwashed artificial clots, and between AF and AS clots, with a cutoff ratio of 0.52 showing 78.9% specificity and 90.0% sensitivity. This cutoff identified 84% of cryptogenic patients who developed definite AF, and lifetable analysis found significantly more definite AF confirmation by cardiac monitoring in this group than in patients with a probable non-AF ratio (P = .021 by log-rank [Mantel-Cox] test).

Conclusion: We developed a rapid and sensitive method for determining cryptogenic clot sources that correlates well with long-term clinical outcomes.