Transthoracic (TTE) and transesophageal (TEE) echocardiography with saline contrast injection (with and without Valsalva manoeuvre), have been considered the most sensitive methods to detect PFO. Although TTE may identify patients with RLS, TEE with saline contrast injection is more sensitive by allowing visualization of microbubbles in the left atrium (LA) that would otherwise be filtered by the lung capillary [12]. Using agitated saline contrast, the sensitivities of traditional TTE for RLS detection are at most half those of TEE: in large active laboratories, PFO detection by TTE was 10–18 % in comparison to 18–33 % incidence using TEE [5]. The incidence of PFO using TEE is quite similar to the incidence of PFO in pooled autopsy data [13] and this finding prompt the evidence that a properly performed TEE is the clinical gold standard for the detection of PFO during life.

From an echocardiographic point of view PFO could be defined as flap-like opening in the atrial septum secundum (thick, muscular), with the septum primum (thin, fibrous) serving as a one-way valve allowing for permanent or transient RLS. As PFO represents a tunnel-like communication, TEE may help to define this complex anatomical structure by measuring both the size of the tunnel (the largest separation between the primum and secundum septa) and the length of the tunnel itself (Fig. 5.1).

TEE also provides additional information regarding the association with atrial septal aneurysm (ASA) (Fig. 5.2), Chiari’s network or Eustachian valve (Fig. 5.3), lipomatous hypertrophy of the atrial septum, the presence or absence of intratrial thrombus or masses; it allows as well to rule out associated congenital heart defects requiring surgical correction.

A PFO is judged to be present if any microbubble is seen in the left-sided cardiac chambers within three cardiac cycles from the maximal right atrial opacification [14]. The degree of interatrial shunting across the PFO is determined by the count of the maximum number of microbubbles seen in the LA in any single frame after intravenous contrast injection both in basal condition and during Valsalva manoeuvre; “significant” shunts were categorized as >20 bubbles in the LA. However, the criteria used to define the different degrees of shunt by echocardiography are based upon different and quite arbitrary criteria. Braun et al. [15] considered the RLS as “small”, “moderate” and “large” when 3–10 bubbles, 10–20 bubbles and >20 bubbles respectively were detected in the LA. Windecker et al. [16] graded spontaneous or provoked RLS semi-quantitatively according to the amount of bubbles detected in the LA after crossing the interatrial septum on a still frame: grade 1 = minimal (1–5 bubbles), grade 2 = moderate (6–20 bubbles), and grade 3 = severe (>20 bubbles). Otherwise, Serena et al. [9] graded RLS as “moderate” when uncountable microbubbles were less echogenic in the LA than in the right atrium, and “severe” when the same microbubbles echogenicity was documented in both the atria. A methodological study using TEE demonstrated that contrast injection via the femoral vein approach is superior to the antecubital route for PFO detection [17]. A false-negative TEE may result from inadequate visualization within the oesophagus, elevated LA pressures preventing right-to-left passage of contrast [18] and difficulty to perform a correct Valsalva maneuver. Use of harmonic imaging with TTE and coughing during contrast injection may increase the sensitivity of PFO detection [14, 19, 20]. Even though less sensitive than TEE, TTE has been used to quantify the microbubbles amount by mean of Doppler signal across the mitral valve [21]. Because of its semi-invasive nature TEE has some limitations, especially in patients with acute strokes. Although rare, aspiration, airway obstruction, oesophageal perforation, and vocal cord dysfunction have also been reported [22, 23].

More recently, three-dimensional echocardiography (3D Echo) has been employed to better describe the three-dimensional changes of PFO shape during the cardiac cycle [24] and to provide an useful insight on the relationship between PFO or interatrial communications and the neighbour cardiac structures. 3D Echo also allows a better definition of the rim tissues and more accurate detection of additional defects or multiple fenestrations of the septum primum. Therefore, it might become the main imaging modality for monitoring and guiding PFO interventional closure procedures [25, 26].

Transcranial Doppler (TCD) is a non-invasive method of assessing the state of the intracranial circulation. TCD is an ultrasonic technique measuring blood flow velocity and direction in the proximal portions of large intracranial arteries (Fig. 5.4). The velocity of flow can be measured in the ophthalmic, in the internal carotid arteries and in the middle, anterior and posterior cerebral arteries (MCA, ACA and PCA) (Fig. 5.5). TCD’s principal use is in the evaluation and management of patients with cerebrovascular disease. A 2-MHz pulsed transcranial Doppler device with 1 or 2 insonating probes are generally used. The small ultrasound probe is generally placed on the temple just above the ear: in this position ipsilateral MCA, ACA and PCA could be insonated (Figs. 5.4 and 5.5). c-TCD of the MCA during contrast injection is a non invasive method to detect the presence and the amount of RLS; therefore, it has been proposed as an alternative to contrast echocardiography for detecting the presence of PFO (Tables 5.2 and 5.3). However, it must be underlined that c-TCD does not identify the site of RLS; actually, the detection of microbubbles in the cerebral circulation may depend not necessarily on the presence of a PFO, but may also be the consequence of intrapulmonary shunts or pulmonary arteriovenous malformations (Table 5.3). For the interventional cardiologist, the most important clinical applications of c-TCD are the detection of cerebral microemboli and the quantification of right-to-left cardiac shunts. A number of recent reports have emphasized the amount of RLS as the crucial factor underlying the likelihood of paradoxical brain embolism in stroke patients [9, 10, 27–29]. c-TCD should be generally performed by an experienced neurosonologist following the guidelines of the Consensus Conference of Venice [30] with the patient in the supine position (Fig. 5.6a): the right MCA is insonated with a 2-MHz hand-held probe at the depth where the MCA and the ACA are both visible. The amount of contrast agent per injection should be 10 ml of air mixed saline solution (using 1 ml air and 9 ml saline). A three-way stopcock is connected to a 10-ml syringe I (containing 9 ml of 0.9 % saline) and the syringe II (containing 1 ml air). 1 ml air (syringe I) and 9 ml saline (syringe II) should be rapidly and energetically exchanged between the syringes at least ten times (Fig. 5.6b). The contrast agent is injected (bolus) in the antecubital vein while simultaneously recording the Doppler signal from the right MCA during normal breathing and after a Valsalva manoeuvre (Fig. 5.6c, d). In case of RLS, air microbubbles are detected on the spectral display of the insonated artery and may be counted, allowing a quantitative assessment of the amount of shunt [30]. Using agitated saline the microbubbles do not survive the first pulmonary passage. The number of microbubbles detected using unilateral MCA monitoring allows the RLS categorization (Figs. 5.7, 5.8, 5.9 and 5.10).