Factors associated with shunt dynamic in patients with cryptogenic stroke and patent foramen ovale: an observational cohort study.

Background: As previously reported there is evidence for a reduction in right to left shunt (RLS) in stroke patients with patent foramen ovale (PFO). This occurs predominantly in patients with cryptogenic stroke (CS). We therefore analysed factors associated with a shunt reduction on follow-up in stroke patients suffering of CS.

Methods: On index event PFO and RLS were proven by transesophageal echocardiography and contrast-enhanced transcranial Doppler-sonography (ce-TCD). Silent PE was proved by ventilation perfusion scintigraphy (V/Q) within the stroke work-up on index event; all scans were re-evaluated in a blinded manner by two experts. The RLS was re-assessed on follow-up by ce-TCD. A reduction in shunt volume was defined as a difference of ≥20 microembolic signals (MES) or the lack of evidence of RLS on follow-up. For subsequent analyses patients with CS were considered; parameters such as deep vein thrombosis (DVT) and silent pulmonary embolism (PE) were analysed.

Results: In 39 PFO patients suffering of a CS the RLS was re-assessed on follow-up. In all patients (n = 39) with CS a V/Q was performed; the median age was 40 years, 24 (61.5%) patients were female. In 27 patients a reduction in RLS was evident. Silent PE was evident in 18/39 patients (46.2%). Factors such as atrial septum aneurysm, DVT or even silent PE were not associated with RLS dynamics. A greater time delay from index event to follow-up assessment was associated with a decrease in shunt volume (median 12 vs. 6 months, p = 0.013).

Conclusions: In patients with CS a reduction in RLS is not associated with the presence of a venous embolic event such as DVT or silent PE. A greater time delay between the initial and the follow-up investigation increases the likelihood for the detection of a reduction in RLS.