Comparative analysis of the diagnostic value of several methods for the diagnosis of patent foramen ovale

The aim of this research was to compare the sensitivity and positive predictive value of contrast transcranial Doppler (c‐TCD), contrast transthoracic echocardiography (c‐TTE), and contrast transesophageal echocardiography (c‐TEE), to determine the best method for diagnosing patent foramen ovale (PFO) and to provide a reference for the further improvement of clinical practice. We investigated 161 patients who suffered from migraines, cryptogenic stroke, TIA, and cerebral infarction of unknown cause. All patients underwent transcatheter examination, and the results of the right heart catheterization (RHC) were considered the gold standard for PFO diagnosis. The present study revealed that c‐TTE with the Valsalva maneuver had a higher sensitivity in detecting PFO related right‐to‐left shunt (PFO‐RLS), c‐TCD performed similarly to c‐TEE but maybe produce more false positives. Moreover, when we observed color shunt from the slit‐like channel between the septum primum and the septum secundum on TEE, the positive predictive value was the highest. Patients with suspected PFO should be examined with c‐TTE and c‐TEE for confirmation. When there were a large number of bubbles in the left heart, especially in the presence of color shunt, the positive predictive value was the highest. The positive results of c‐TCD only point out the presence of right‐to‐left shunt and cannot exclude extracardiac shunt, so c‐TCD should not be used as a screen for PFO, additional measures such as c‐TTE and c‐TEE should be used.

blockages in the venous system can flow to the left atrium. Patent foramen ovale is the cause of 95% of paradoxical embolisms. 5 There are three methods for diagnosing PFO: contrast transcranial Doppler(c-TCD), transthoracic echocardiography (TTE), and transesophageal echocardiography (TEE). TEE has a major role in the diagnosis of PFO. The diagnostic sensitivity of PFO related right-to-left shunt (PFO-RLS) with contrast transcranial Doppler (c-TCD) is similar to that with TEE. However, c-TCD has a limited ability to differentiate cardiac from pulmonary RLS. The use of contrast echocardiography was first reported by Gramiak. 6 Currently, contrast transthoracic echocardiography (c-TTE) with the Valsalva maneuver is extensively used for the detection and semiquantitative assessment of PFO-RLS. 7 However, the diagnostic criteria are not unified. Thus, the aim of this research was to compare the sensitivity and specificity of each examination method to determine which is the best method for diagnosing PFO and to provide a reference for the further improvement of clinical practice. The study was approved by the local ethics committee, and all patients or their relatives provided written informed consent to participate in this study prior to the examination.

| Saline contrast preparation
The right anterior elbow vein was selected for the placement of an indwelling needle. The contrast agent was prepared by mixing it with 3 ml of air, 30 ml of saline solution, and 3-5 mL of the patient's blood and intensively mixing it back and forth 20 times between two 30-ml syringes connected by a T-branch pipe.

| c-TCD Examination
TCD was performed using the Multi-DopX4 Transcranial Doppler (DWL Electronic Systems, Sipplingen, Germany) with a 2 MHz probe.
Middle cerebral artery flow was monitored through the temporal bone window at a depth of 50-65 mm. The study was performed by a neurologist who specialized in this technique and was blinded to the results of the echocardiographic study. After value of the middle cerebral artery blood flow velocity was obtained, the contrast agent was quickly injected via the established route of the anterior elbow vein. The contrast agent was prepared, and right-to-left shunt was diagnosed when TCD detected microsignals in the middle cerebral artery, in both the resting state and during performance of the Valsalva maneuver. The severity of the shunt was quantified as negative (no microsignals), mild (1-20 microsignals, 1-10 per side), moderate (≥20 microsignals, >10 per side), or extensive (>30 microsignals or rain curtain-like signals) ( Figure 1).

| c-TTE and c-TEE Examination
Contrast-TTE was conducted using a GE Vivid E9 or E95 platform equipped with a 3.7-5 MHz M5S transducer (Horten, Norway) or a Philips EPIQ7 platform equipped with a 1-5 MHz S5-1 transducer  Figure 2). If microbubbles appeared in the left atrium within five cardiac cycles after they are released, RLS was primarily considered to be derived from a PFO. If microbubbles appeared in the left atrium after more than five cardiac cycles, RLS was assumed to originate from a pulmonary arteriovenous malformation.
Contrast TEE was performed using the same system fitted with a 2-7 MHz multi-frequency transesophageal probe. To improve the tolerance of the test, pharyngeal topical anesthesia was administered using tetracaine hydrochloride gel 15 minutes before the examination. The probe was pushed in 30-40 cm, rotated within 30°-90° to clearly display ascending aorta root, the septum primum and the septum secundum, and to observe whether an opened PFO existed in the two-dimensional view, measured PFO width and length at the same time. Color Doppler was used to observe whether there was a shunt across the PFO. The shunt flow signal must locate between the septum primum and the septum secundum, parallel to the septum primum, which was bundle-like and persistent. 8 Then, with the patient at rest and performing the Valsalva maneuver, prepared saline contrast agent was injected according to the procedure described above. To ensure maximal diagnostic yield, a standard apical four-chamber view was performed with the administration of contrast agents. The severity of the microbubbles was quantified as the same way as for c-TTE ( Figure 3).

| Right heart catheterization and transcatheter closure
All patients for whom RLS was indicated underwent transcatheter examination, and the results of the right heart catheterization (RHC) were regarded as the gold standard for PFO diagnosis in present study. If the catheter could pass though the foramen ovale to reach the left atrium, then PFO could be diagnosed. If not, then pulmonary arteriovenous fistula (PAVF) was excluded by pulmonary angiography ( Figure 4).

| Statistical analysis
Qualitative variables are presented as percentages and continuous data and expressed as the mean ±standard deviation. The chi-square test was used to compare the sensitivity and specificity of the three methods. A P value of <.05 indicated statistical significance. All data were analyzed using SPSS software (version 18.0.1, SPSS Inc).

| Diagnosis of PFO
Using the results of the right heart catheterization (RHC) as the gold standard for PFO diagnosis, 141 of the 161 patients studied were diagnosed with PFO, 18 patients could not be diagnosed with PFO, 2 patients were diagnosed with pulmonary arteriovenous malformation, and the morbidity of PFO was 88.17% (Table 1).  (Table 2).

| Diagnostic value of the different methods
In total, the sensitivity of c-TCD at baseline was 70.28% (97/138): In the comparison among groups, there were no significant difference between the sensitivity of total c-TCD, c-TTE, and c-TEE at baseline. When pairs of methods were compared, c-TCD was significantly more sensitive than c-TEE (P = .004). With the Valsalva maneuver, the sensitivity of the total c-TTE was significantly higher than that of c-TEE (P = .041). For the within-group comparisons, when bubbles were extensive in c-TTE examination at baseline, c-TTE had significantly higher sensitivity than c-TCD (P = .011) and c-TEE (P = .009).

| D ISCUSS I ON
Foramen ovale is a persistent fetal communication between the right and left atrium caused by the incomplete closure of the atrial septum. After birth, the foramen ovale normally closes, but the morbidity of PFO was about 30% in adults. 9 PFO normally has no obvious symptoms, it is difficult to hear an associated heart murmur, and there are no abnormalities on electrocardiogram or chest X-ray, therefore, PFO is easy to overlook. It was recently proposed that for patients who suffer from migraine headaches, cryptogenic PFO is a three-dimensional structure that changes shape with the cardiac cycle. Therefore, it is difficult to fully display a PFO using TEE's two-dimensional technology, and in our results, there were some false positive cases when we used 2D TEE images (5 cases).
The use of three-dimensional (3D) technology can solve the onesidedness problem, but the quality of 3D images can be poor compared to that of 2D images; sometimes, 3D images were only used for reference. 10 In terms of sensitivity, total c-TTE after the Valsalva maneuver had the highest sensitivity, especially for extensive cases, we consider that sensitivity increased with increasing quantities of The results sometimes were different between the resting state and the Valsalva maneuver, and the positive rate was higher with the Valsalva maneuver. In the resting state, RLS could not be observed in all the patients. In some PFO patients, the foramen ovale is closed in resting state, only under some particular situation, for example, cough, cry, constipation, Valsalva maneuver and so on, that's also the causes of nervous system symptoms on PFO patients. 12 Only under those situations, the right atrial pressure can higher than the left atrial transiently, the foramen ovale then opened, so we could observe RLS. Therefore, when no bubbles were present in the left heart, the Valsalva maneuver was necessary, but the results were positive with or without the Valsalva maneuver when bubbles were present in left heart.
Pulmonary arteriovenous fistula (PAVF) is an abnormal pulmonary vascular structure that connects a pulmonary artery to a pulmonary vein, bypassing the normal capillary bed and resulting in an intrapulmonary right-to-left shunt. 13   In conclusion, the present study reveals that c-TTE with the Valsalva maneuver yields a higher sensitivity in detecting PFO-RLS.
When there were a large number of bubbles in the left heart, especially in the presence of color shunt from the slit-like channel, the positive predictive value was the highest. The results of c-TCD can only point out the presence of right-to-left shunt and cannot exclude extracardiac shunt, so TCD should not be used as a screen for PFO, only can be used as an supplement. The examination results of PVAF were similar to PFO, we should pay more attention to the differential diagnosis in clinical work. ALL those procedures may be helpful for the diagnosis of patent foramen ovale in practice.

CO N FLI C T O F I NTE R E S T
We declare that we do not have any commercial or associative interest that represents a conflict of interest in connection with the work submitted.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data included in this study are available upon request by contact with the corresponding author.