Platelet Reactivity and CD39 Ectonucleotidase Activities in Cryptogenic Stroke

In the US, 795,000 patients are diagnosed with stroke annually. Of the 90% of strokes that are ischemic, a substantial number are of unknown etiology. In younger adults, the proportion of such cryptogenic strokes exceeds 40%. Cryptogenic stroke patients have an increased prevalence of interatrial shunts (IAS), which can be found in as many as 60% of cases as compared to 26% for the general population. Yet, percutaneous closure of such defects has not been found to reduce the recurrence rate as compared with medical therapy. Hypercoagulable disorders have also been implicated in the pathogenesis of premature ischemic strokes, but their occurrence in cryptogenic stroke series is modest, leaving many such strokes unexplained. We hypothesized that cryptogenic stroke may in part represent a disorder of enhanced platelet reactivity. An understanding of the responsiveness of platelets from young cryptogenic stroke patients to a broad panel of agonists is required because platelet reactivity to a single agonist or a thromboxane measurement does not provide sufficient information about platelet reactivity. CD39/NTPDase1, the ecto-enzyme on endothelial cells and leukocytes, is responsible for control of adenine nucleotide levels in the local vascular microenvironment, and, therefore, blood fluidity. CD39 rapidly metabolizes ATP and ADP released from activated platelets and other cells, thereby abolishing the prothrombotic cascade of platelet activation and recruitment.

We sought to determine whether platelet activation and recruitment, markers of platelet activation, and CD39 enzymatic activity are increased in younger patients with unexplained stroke. To this end, we evaluated such platelet-related measures in a subsample (n=167 subjects) participating in the THrombophilia In Cryptogenic stroKe (THICK) study, a prospective case-control study evaluating the prothrombotic determinants of unexplained stroke. Cases consisted of patients, ages 18–64, referred to Weill Cornell Medical Center for evaluation of cryptogenic stroke, whereas controls were stroke-free volunteers within the same age range recruited at our center and the surrounding area. Cryptogenic stroke was defined according to modified TOAST criteria.

Cases (n=99) were similar in age (median 45 years old), sex (51% women), and race-ethnic composition (8% African American) as compared with controls (n=68). They also had similar prevalences of hypertension (18%), diabetes (7%), dyslipidemia (40%), and current smoking (6%). Cases had over twice the prevalence of IAS (56%) as controls (24%). Extensive testing for humoral prothrombotic disorders revealed a potential explanation for stroke in only half of the patients. We developed a Broad Range Aggregometry System to study profiles of platelet reactivity to multiple agonists, and classified patient responsiveness based on their reactivity to these agonists. We have also been able to make a laboratory diagnosis of aspirin non-responsiveness in these patients and controls. In addition, in stroke subjects, trends to higher total ADPase activity were observed on lymphocytes (p=0.09) and PMN (p=0.09), while ATPase activities were similar (p=0.81 and 0.68, respectively), as determined by our radio-TLC nucleotidase assay. Also, the ADPase to ATPase activity ratio was greater in the stroke patients than controls (p=0.003 for lymphocytes; p=0.13 for PMN). Cryptogenic stroke patients also displayed increased levels of platelet-monocyte aggregates (p=0.034) and the platelet activation marker CD63 (p=0.048), but not CD154 (p=0.183), as compared to controls (determined by FACS analyses).

These findings support a potential role for enhanced platelet reactivity and a low threshold for platelet responsiveness in cryptogenic stroke patients.

No relevant conflicts of interest to declare.