22 The importance of the multidisciplinary team (MDT) in guiding patent foramen ovale (PFO) device closure (DC) in patients with cryptogenic stroke (CS): a clinical follow-up study

Background There is a high prevalence of PFO in patients with CS. Paradoxical embolisation (PE) may only be implicated in some of these patients. Careful assessment by an MDT is essential for the appropriate selection of patients for percutaneous PFO DC. Methods Consecutive patients with CS and PFO, referred to a single tertiary cardiac centre between June 2005 and November 2019, were retrospectively studied. Clinical MDT work-up to ascertain cause for the clinical event and risk of recurrent events included clinical consultation, expert review of cardiac imaging (TOE/bubble contrast TTE), brain imaging (CT, CT angiography, DWI MRI), thrombophilia and arrhythmia investigations. PASCAL was classified retrospectively. MDT conclusions on final diagnosis and treatment recommendations were obtained from meeting minutes, electronic and paper records while clinical event follow up was assessed through routine electronic health records. Results 168 patients (Mean age: 43.41 years, Female 50.6%) with a suspected PFO were referred with an index stroke (140 patients), TIA (23 patients) or peripheral emboli (5 patients). The median MDT meeting number per patient was 1 (IQR: 1–2). Cardiac imaging confirmed PFO presence in 87%, ASD in 5% and both in 4%: the intra-atrial septum was intact in 4%. Proximal source of thromboembolism was adjudged in 60.1% (101/168) of patients with atrial arrhythmia identified in 5 (table 1). Among the remaining 92 patients with possible PE, 64 were adjudged to be high risk of recurrence and PFO DC (closure group - CG) was recommended. In the remaining 28, PE was speculated, with a low risk of recurrence and DC was therefore not recommended (non-closure group – NCG). There was a higher proportion of cortical infarctions (79.4% vs 53.6%, p=0.024) in CG than NCG. High-risk TOE characteristics such as mobile septum (28.1% vs 17.9%, p=0.561), atrial septum aneurysms (28.1% vs 21.4%, p=0.759) and spontaneous shunt (53.1% vs 28.6%, p=0.084) were numerically higher in CG compared to NCG although not statistically significant. PASCAL classification showed higher proportions of "probable" causal associations (59.4% vs 10.7%, p<0.001) in CG compared to NCG. In a median follow-up of 6.37 years in all 168 patients, neurological events occurred in 1 patient in CG, 0 in NCG group and 2 in the group in whom PE was not considered likely. Arterial or venous thrombotic events (including DVTs, PEs and limb ischaemia) occurred in 1 patient in CG, 0 in the NCG and 2 in the non-PE group. Significant bleeding (defined as BARC3a and above) occurred in 1 patient in CG, 3 in NCG (HR 0.143, 95%CI 0.015–1.371, p=0.092) and 0 in non-PE group. Conclusion In patients referred with CS and PFO, possible PE was implicated in only 54.8% and DC recommended in the minority (38.1%). Recurrent neurological or embolic event rates were low, and this corroborates the validity of the decision-making and importance of an MDT process in selecting patients for DC. Conflict of Interest Nil