First long-term evaluation of the risk of recurrence and prognosis of cryptogenic stroke in patients with and without patent foramen ovale

Abstract

Background: Between 30% and 40% of ischemic strokes remain unclassified after an exhaustive etiologic study, being defined as cryptogenic strokes (CS). The presence of a patent foramen ovale (PFO) has been strongly associated with CS, especially in younger patients. The best treatment for PFO-related stroke has not been fully stablished. 6 randomized clinical trials (RCT) comparing PFO closure with medical treatment have been published but the obtained results were inconclusive, with short follow-up times (<6 years) and a narrow cost-benefit balance with high NNT. Furthermore, when PFO is considered as causal, patients have a better short-term prognosis (modified Rankin Scale) than in CS in general what increases even more the doubts about their management. No studies analyzing long-term PFO-related stroke recurrence risk and prognosis have been conducted. The cohort of the CODICIA study offers the opportunity to make a first long-term evaluation of this group of patients. Objective: The aim of this study is to make an extension of the CODICIA study (2000-2005) (1), a prospective cohort that analyzed the short-term risk of CS recurrence in patients with and without PFO treated with medical therapy. Our objective is to analyze retrospectively the same sample with a mean follow-up time of approximately 15 years, obtaining long-term data about the risk of stroke recurrence and prognosis of these patients. Design: The study will be a multicenter observational retrospective cohort. 16 hospitals will participate with the University Hospital Dr. Josep Trueta (HJT) as the coordinator center. Participants: Patients previously included in the CODICIA study. People older than 18 years who suffered a CS (cerebral infarct or TIA) more than 10 years ago treated with medical therapy (antiplatelet/anticoagulation or both). All the analysis will be repeated in the group of patients younger than 55 years. Methods: We will record any vascular event in case report forms (CRF). The data will be obtained mainly by phone contact and/or medical histories if available. Some data are already registered in the database of the original CODICIA study. For the statistical analysis we will use χ2 test and Exact Fisher test for qualitative variables and t-Student or Mann-Whitney for quantitative variables. A confidence interval of 95% will be assumed and a p<0,05 will be considered statistically significant. The association between the independent and dependent variables will be adjusted by means of logistic-regression analysis in order to avoid possible confounding factors