The Valsalva maneuver, the most sensitive test for patent foramen ovale (PFO) detection, is difficult during transesophageal echocardiography (TEE), especially after sedation. The aim of this study was to compare PFO detection effectiveness between inferior vena cava (IVC) compression and the Valsalva maneuver.
A total of 293 patients with paroxysmal atrial fibrillation undergoing TEE before initial atrial fibrillation ablation were prospectively enrolled. Agitated saline was injected in 290 patients under three conditions: Valsalva maneuver under conscious sedation, at rest, and IVC compression under deep sedation. Three patients with newly diagnosed atrial septal defects on TEE were excluded. The IVC compression maneuver consisted of manual compression 5 cm to the right of the epigastric region and depressed the abdominal wall by 5 cm for 30 sec and compression release immediately before right atrial opacification with microbubbles by agitated intravenous saline.
The overall PFO detection rate was better with IVC compression (57 PFOs [19.7%]) than at rest (15 patients [5.2%]) ( P < .0001) or with the Valsalva maneuver (37 patients [12.8%]) ( P = .024). There were no significant differences in PFO detection between IVC compression and the Valsalva maneuver (IVC compression, 43 patients [22.5%]; Valsalva maneuver, 35 patients [18.3%]; P = .31), even in patients who could perform the Valsalva maneuver effectively ( n = 191).
IVC compression is feasible and effective for detecting PFO and is not inferior to the Valsalva maneuver. In particular, IVC compression could be an alternative diagnostic method for PFO using TEE when the Valsalva maneuver cannot be performed under deep sedation.
The authors propose IVC compression as new provocation maneuver to detect PFO.
IVC compression is feasible and not inferior to the Valsalva maneuver.
IVC compression should be performed when the Valsalva maneuver is not effective.
Patent foramen ovale (PFO) is associated with a variety of disorders, including cryptogenic stroke, migraine, platypnea-orthodeoxia syndrome, and decompression illness. Moreover, PFO is associated with a substantially increased risk for embolic stroke, even in patients with endocardial leads as implantable cardioverter-defibrillators or pacemakers. Recently, as-treated analysis of the Randomized Evaluation of Recurrent Stroke Comparing PFO Closure to Established Current Standard of Care Treatment trial showed that PFO closure by a transcatheter occluder is superior to conservative treatment. Therefore, transesophageal echocardiography (TEE) plays a more important role in the evaluation of PFO because of its improved imaging of cardiac structures and easier identification of PFO with the use of intravenous agitated saline contrast compared with transthoracic echocardiography (TTE). Although an adequate Valsalva maneuver is crucial for diagnosing PFO, sedation status and/or the presence of a transesophageal echocardiographic probe inside the esophagus and laryngopharynx often decreases patient effort during this maneuver. Thus, a more feasible and reproducible maneuver to detect PFO, irrespective of sedation status and/or the presence of a probe, is needed. As a feasible provocation test, inferior vena cava (IVC) compression maneuver, which involves manual compression of the abdomen to produce partial IVC collapse and increased IVC flow upon release, has been developed. Accordingly, we aimed to compare the effectiveness of IVC compression and the Valsalva maneuver to diagnose PFO in a rigorous manner.
Between May 2014 and June 2015, we prospectively enrolled 293 patients with paroxysmal atrial fibrillation (AF) who underwent TEE before left atrial (LA) catheter ablation at the Gunma Prefectural Cardiovascular Center. Exclusion criteria were AF rhythm at the time of TEE, known atrial septal defects, moderate or severe valvular heart disease, prior cardiac surgery or repeated AF ablation, and inability to perform the Valsalva maneuver because of cognitive or coordination impairment. The present study was approved by the ethical committee of our hospital. All patients provided written consent for the study.
Before TEE, all patients were instructed to perform the Valsalva maneuver. The examiner placed a hand on the patient’s abdomen to check for abdominal muscular contraction and confirmed the efficacy of the Valsalva maneuver in all patients. Complete TEE using a 5-MHz multiplane probe (iE33 with an X7-2t probe; Koninklijke Philips, Amsterdam, the Netherlands) was performed. Following oropharyngeal anesthesia, 1% propofol was used for conscious sedation, in an intravenous bolus of 0.2 to 0.4 mg/kg (10–30 mg) before and during TEE. Blood pressure, pulse oximetry, and the electrocardiogram were monitored. A search for other cardiac sources of emboli (assessment of the aorta and LA appendage) was systematically undertaken. Images of the interatrial septum (IAS) were obtained from the best imaging plane for septal membrane visualization, typically 50° to 75°. At that time, atrial septal defects were newly documented in three patients, who were subsequently excluded from the present study. The Valsalva maneuver was then attempted in the remaining 290 patients (mean age, 65 ± 10 years; 189 men) during TEE, with simultaneous abdominal strain assessment. In case the Valsalva maneuver was ineffective, coughing and conventional abdominal compression ( Figure 1 B) were used to complement the examination. At least two saline contrast injections were administered during TEE, and additional contrast injections were administered with images obtained from other transesophageal echocardiographic planes if the site of microbubble passage in the septum was not clear. Contrast injections consisted of 1 mL air and 9 mL saline agitated by a three-way stopcock with two connected syringes, one of which was filled with saline; the injections were administered intravenously from an antecubital vein. Contrast was injected during the strain phase of the Valsalva maneuver, and normal respiration resumed as the first bubbles appeared in the right atrium. The Valsalva maneuver was considered effective if we could observe the leftward deviation of the atrial septum following the maneuver. PFO was judged as present when microbubbles were seen in the left chambers within three heartbeats, after full opacification of the right chambers. Quantification of LA opacification was regarded as grade I (mild; three to 10 microbubbles), grade II (moderate; 11–30 microbubbles), or grade III (severe; >30 microbubbles). Following the Valsalva maneuver, additional intravenous propofol (0.5–1.0 mg/kg, 30–80 mg) was administered to relieve patient discomfort. A level of deep sedation was targeted in this study according to the statement on the use of non-anesthesiologist-administered propofol sedation. Subsequently, contrast was injected at least twice at rest and under IVC compression. The IVC compression maneuver was performed as follows ( Figures 1 and 2 ): We firmly performed manual compression 5 cm to the right of the epigastric region and depressed the abdominal wall of the patients by 5 cm for 30 sec ( Figure 1 A). IVC flow was interrupted during the compression, which resulted in decreased venous return and LA pressure. IVC compression was released immediately before right atrial (RA) opacification with microbubbles by agitated intravenous saline. The augmented IVC stream increased RA pressure and the RA-LA pressure gradient, which provoked right-to-left shunting in patients with PFOs. Figure 2 A shows the concept of the IVC compression maneuver, and Figure 2 B and Video 1 (available at www.onlinejase.com ) shows a representative case. Before the actual maneuver, we performed a pretest using color Doppler imaging to confirm that the compression effectively indirectly interrupted the IVC flow ( Figure 2 C and Video 2 available at www.onlinejase.com ). Color Doppler flow (Nyquist limit, 30–40 cm/sec) from the IVC was assessed following release of the compression for 5 sec, as a trial. In 260 patients (89.7%), increased IVC flow to the right atrium by color Doppler imaging could be observed by initial compression. We could obtain increased IVC flow by adjusting compression more firmly in the remaining 30 patients. Effectiveness of IVC compression was evaluated by leftward atrial septal motion as with the Valsalva maneuver. A right-to-left shunt was diagnosed as described previously. Moreover, to assess the change in preload as a result of IVC compression, pulse Doppler transmitral early diastolic (E) velocity during IVC compression was also measured from a midesophageal long-axis view in the first 21 patients after the study protocol.
Demographic Data Collection
Clinical data, including age, sex, body surface area, body mass index (BMI), brachial blood pressure, heart rate, documented diagnosis of hypertension, diabetes mellitus, history of heart failure or thromboembolism, CHADS 2 score, and comprehensive transthoracic echocardiographic data according to guidelines, were collected at the time of TEE ( Table 1 ).
|Age (y)||65 ± 10|
|Body surface area (m 2 )||1.68 ± 0.19|
|BMI (kg/m 2 )||23.5 ± 3.4|
|Diabetes mellitus||36 (12)|
|History of heart failure||12 (4)|
|Previous thromboembolism||17 (6)|
|CHADS 2 score (0/1/2/≥3)||98/122/48/22|
|Left ventricular diastolic dimension (mm)||45 ± 5|
|Left ventricular systolic dimension (mm)||28 ± 5|
|Left ventricular diastolic volume index (mL/m 2 )||56 ± 13|
|Left ventricular systolic volume index (mL/m 2 )||19 ± 8|
|Left ventricular ejection fraction (%)||67 ± 8|
|LA anteroposterior dimension (mm)||39 ± 6|
|LA volume index (mL/m 2 )||33 ± 14|
The agreement of PFO quantification between the IVC compression and Valsalva maneuvers was evaluated. In addition, the agreement of the presence or absence of PFO between the first and second IVC compression maneuvers was assessed. A second observer who was blinded to the results of IVC compression independently evaluated all studies for the presence or absence of a right-to-left shunt to assess interobserver variability.
Results are expressed as mean ± SD or percentage unless otherwise specified. Data were compared between patients developing predefined events and those without events using Student’s t test, the χ 2 test, or the Fisher exact test as appropriate. Sensitivities of PFO detection at rest and by the Valsalva and IVC compression maneuvers were assessed using all PFO cases diagnosed by any maneuvers as the gold standard. Agreement between two procedures was tested using κ statistics. P value < .05 were considered to indicate statistical significance. All statistical analyses were performed using commercial statistical software (JMP version 8; SAS Institute, Cary, NC).
Comparisons of PFO Detection Rate and Sensitivity
All patients were in normal sinus rhythm during TEE in this study. Of the 290 patients, 59 patients (20.3%) were diagnosed with PFOs by the Valsalva and/or IVC compression maneuver. Overall, the IVC compression maneuver diagnosed 57 PFOs (19.7%); the PFO detection rate with IVC compression was superior to that at rest (15 patients [5.2%], P < .0001) or with the Valsalva maneuver (37 patients [12.8%], P = .001 vs at rest, P = .024 vs.IVC compression maneuver) ( Figure 3 ).
Although 191 patients (65.9%) could perform the Valsalva maneuver effectively, the IVC compression maneuver provoked an effective leftward shift of the IAS in 266 patients (91.7%, P < .05 vs Valsalva maneuver). In the 191 patients who could effectively perform the Valsalva maneuver, even though the Valsalva maneuver showed a higher PFO detection rate, the PFO detection rate with IVC compression was equivalent (IVC compression, 43 patients [22.5%]; Valsalva maneuver, 35 patients [18.3%]; P = .31). In the remaining 99 patients whose Valsalva maneuvers were suboptimal, the PFO detection rate with the Valsalva maneuver was only 2%, whereas that with IVC compression was 14.1% ( P < .01).
The sensitivities of PFO detection at rest, using the Valsalva and IVC compression maneuvers, were 25.4%, 62.7%, and 96.6% overall and 20.0%, 77.8%, and 95.6% in patients in whom the Valsalva maneuver was performed effectively, respectively.
Preload Reduction Effect of IVC Compression
To confirm the preload reduction effect of IVC compression, transmitral E velocity was assessed in the first 21 consecutive patients who were enrolled in the present study. IVC compression for 30 sec decreased the E velocity (from 61 ± 19 to 47 ± 17 cm/sec, P < .0001), and the reduction in E velocity was recovered following release of IVC compression (61 ± 22 cm/sec, P < .0001 vs IVC compression for 30 sec) ( Figure 4 ).
Overall, the IVC compression and Valsalva maneuvers showed good agreement of quantification of right-to-left shunts (κ = 0.651; P < .0001; agreement, 90.0%) ( Table 2 ). Among patients who could effectively perform the Valsalva maneuver, better agreement was demonstrated between the two maneuvers (κ = 0.758; P < .0001; agreement, 91.6%) ( Table 2 ). There was excellent agreement between the first and second IVC compression maneuvers with respect to the presence or absence of a right-to-left shunt (κ = 0.978; P < .0001; agreement, 99.3%). There was 100% agreement between the two observers with respect to the presence or absence of a right-to-left shunt by the IVC compression maneuver.
|Valsalva maneuver||IVC compression maneuver||Total|
|<3 bubbles||3–10 bubbles||11–30 bubbles||>30 bubbles|
|Overall patients ( n = 290) ∗|
|<3 bubbles||231 (79.7%)||14 (4.8%)||6 (2.1%)||2 (0.7%)||253 (87.2%)|
|3–10 bubbles||2 (0.7%)||22 (7.6%)||1 (0.3%)||1 (0.3%)||26 (9.3%)|
|11–30 bubbles||0 (0%)||1 (0.3%)||3 (1.0%)||0 (0%)||4 (1.4%)|
|>30 bubbles||0 (0%)||0 (0%)||2 (0.7%)||5 (1.7%)||7 (2.4%)|
|Total||233 (80.3%)||37 (12.8%)||12 (4.1%)||8 (2.8%)||290 (100%)|
|Patients who could perform an “effective” Valsalva maneuver ( n = 191) †|
|<3 bubbles||146 (76.4%)||7 (3.7%)||2 (1.1%)||1 (0.5%)||156 (81.7%)|
|3–10 bubbles||2 (1.1%)||21 (11.0%)||1 (0.5%)||0 (0%)||24 (12.6%)|
|11–30 bubbles||0 (0%)||1 (0.5%)||3 (1.6%)||0 (0%)||4 (2.1%)|
|>30 bubbles||0 (0%)||0 (0%)||2 (1.1%)||5 (2.6%)||7 (3.7%)|
|Total||148 (77.5%)||29 (15.2%)||8 (4.2%)||6 (3.1%)||191 (100%)|