Right-to-left shunting via a patent foramen ovale (PFO) has a recognized association with embolic events in younger patients. The use of agitated saline contrast injection (ASCi) for detecting atrial shunting is well documented, but the optimal technique is not well described. The purpose of this study was to assess the efficacy and safety of transthoracic echocardiographic (TTE) ASCi for the assessment of right-to-left atrial communication in a large cohort of patients.
A retrospective review was undertaken of 1,162 consecutive patients who underwent TTE ASCi, of whom 195 had also undergone clinically indicated transesophageal echocardiography. ASCi shunt results were compared with color flow imaging, and the role of provocative maneuvers (PM) was assessed.
Four hundred three TTE studies (35%) had paradoxical shunting seen during ASCi. Of these, 48% were positive with PM only. There was strong agreement between TTE ASCi and reported transesophageal echocardiographic findings (99% sensitivity, 85% specificity), with six false-positive and two false-negative results. In hindsight, the latter were likely due to suboptimal right atrial opacification and the former to transpulmonary shunting. TTE color flow imaging was found to be insensitive (22%) for the detection of a PFO compared with TTE ASCi.
TTE color flow imaging is too insensitive for PFO screening. TTE ASCi, however, is simple and highly accurate for the detection of right-to-left atrial communication, on the proviso that a dedicated protocol, including correctly implemented PM, is followed. It is recommended that TTE ASCi with PM be considered the primary diagnostic tool for the detection of PFO in clinical practice.
The foramen ovale is an important component of the fetal circulation, allowing oxygenated blood to pass from the right atrium to the left atrium, thus ensuring that oxygen-rich blood reaches the fetal brain. At birth, aeration of the neonatal lungs markedly reduces pulmonary vascular resistance, so that right atrial (RA) pressure falls below left atrial (LA) pressure, effectively closing the foramen ovale. Subsequently, any transient increase in RA pressure above LA pressure will result in right-to-left shunting. Permanent fusion of the foramen ovale occurs by 2 years of age in approximately 75% of individuals, with patency remaining in the other 25%.
The presence of a patent foramen ovale (PFO) has been linked to many illnesses, including cryptogenic stroke, transient ischemic attack, migraine, platypnea-orthodeoxia syndrome, and even decompression sickness in scuba divers. The presumed mechanism for cryptogenic stroke is the migration of a thrombus, air bubble, or fat embolus from the venous system into the left atrium through a PFO, with subsequent systemic embolization. Implication of the PFO as the cause for paradoxical embolization is strongest when the “PFO triad” is present. This triad consists of venous source of thrombosis, temporal association of transiently raised RA pressures with the neurologic event, and the presence of a PFO.
The detection of PFO on transthoracic echocardiographic (TTE) imaging is greatly improved by using agitated saline contrast injection (ASCi). Intravenous injection of saline after agitation between two syringes enhances the backscatter of the ultrasound beam, thus highlighting venous blood flow. The “bubbles” generated by using agitated saline are too large to cross the pulmonary capillary bed, so visualization of contrast in the left-heart chambers indicates either intracardiac or transpulmonary shunting. Provocative maneuvers such as the Valsalva maneuver (VM), sniff, and cough transiently increase RA pressure above LA pressure. These maneuvers further enhance the sensitivity of ASCi for the detection of atrial shunting ( Figure 1 ) and improve diagnostic confidence by making evidence of right-to-left shunting more obvious ( Figure 2 ).
Transesophageal echocardiographic (TEE) imaging is the gold standard for visualization of atrial septal anatomy. Color flow imaging (CFI) during TEE examination is very effective at demonstrating left-to-right shunting when present. However, right-to-left shunting through a PFO is often absent at rest, and provocative maneuvers are required. There are technical challenges in eliciting an adequate increase in RA pressure during TEE imaging. The presence of the TEE probe prevents closure of the glottis, which is required for an effective VM. This can be further complicated by varying degree of patient sedation. The fasting state of patients undergoing TEE imaging, along with sedation-induced hypotension, can result in lowered RA pressure, leading to a reduced detection rate. These factors are eliminated with TTE imaging, which if properly performed can be a superior technique for demonstrating right-to-left shunting with ASCi.
The purpose of this study was to assess the efficacy and safety of ASCi via TTE imaging for the assessment of right-to-left atrial communication in a large cohort of patients.
Study Population and Patient Selection
More than 18,000 consecutive, digitally stored TTE studies from the Hearts 1st echocardiography database (Greenslopes Private Hospital, Greenslopes, Australia) performed between December 1, 2003, and September 7, 2009, were available for analysis. Before commencement of the study, approval was sought and granted from the Queensland University of Technology University Human Research Ethics Committee, which is registered and accredited with the National Health and Medical Research Council. A total of 1,162 patients (559 men, 603 women; mean age, 51 ± 16 years) were retrospectively identified as having undergone digitally stored comprehensive TTE studies with ASCi.
A subgroup of 195 patients were identified to have also undergone TEE studies for further delineation of atrial septal anatomy. All studies were reported by experienced echocardiologists. Independent blinded review was performed by a study investigator (A.F.) for all CFI studies and a subset ( n = 50) of the ASCi studies (K.M.) to ensure consistency of reporting.
TTE examinations were carried out by experienced sonographers using either an Acuson Sequoia C512 or an Acuson Aspen (Siemens Medical Solutions USA, Inc., Mountain View, CA). Standard echocardiographic protocols were followed on the basis of the recommendations of the American Society of Echocardiography, including Doppler-derived assessment of LA pressure.
CFI assessment for left-to-right shunting across the atrial septum was performed in parasternal short-axis, apical four-chamber, and subcostal four-chamber and short-axis views, and findings were reported as positive, negative, indeterminate, or technically inadequate. During assessment of the atrial septum, respiration was paused appropriately for image optimization. Provocative maneuvers were not specifically performed during CFI. Findings on CFI were reported as positive if a discrete color jet was visualized passing through the atrial septum from at least one view. A study was deemed indeterminate if color flow could not be differentiated from caval flow or possible artifact. Studies that had inadequate imaging due to suboptimal acoustic windows or challenging color optimization were also coded as such.
To ensure maximal diagnostic yield from ASCi, a standard imaging protocol for all ASCi studies was used, as described below. ASCi studies were undertaken with 18-gauge or 22-gauge cannulas in an antecubital vein (rarely in a dorsal hand vein) using 8.5 mL normal saline, 0.5 mL air, and 1 mL blood aspirated from the cannula. Blood was used in the injection because of its enhanced contrast appearance. This solution was then agitated forcefully between two syringes connected to extension tubing with a three-way stopcock and rapidly injected once a suitable TTE image was obtained. Most commonly, an apical four-chamber view was used, but on occasion, the parasternal short-axis view at the level of the atrial septum (aortic valve level) or a subcostal four-chamber view was used. On first visualization of contrast entering the right atrium, a 4-sec or 6-sec digital loop was acquired. Dense opacification of the right atrium was imperative for accuracy in the performance of ASCi. Competitive flow arising from the inferior vena cava can result in localized loss of contrast along the RA side of the atrial septum, thereby resulting in a false-negative. In these cases, external manual compression of the liver was used as a tool to reduce inferior vena cava flow and help differentiate between competitive inferior vena cava flow and negative contrast from an atrial septal defect.
In addition to normal respiration, images were also obtained during provocative maneuvers to elicit right-to-left shunting; these maneuvers included the VM, sniff, and/or cough. Satisfactory performance of a provocative maneuver was defined as complete opacification of the right atrium adjacent to the atrial septum at the time of leftward bowing of the atrial septum. Failure to demonstrate at least transient leftward bowing of contrast-opacified atrial septum on two-dimensional imaging indicated insufficient performance of this maneuver, and the attempt was repeated. When this was not achieved despite repeated efforts, the ASCi study was deemed indeterminate. A minimum of five injections with a combination of normal respiration, sniff, and the VM were performed in all studies, as previously recognized by Johansson et al. to achieve maximum diagnostic yield. When both leftward bulging of interatrial septum and dense contrast filling of the right atrium were present, the sensitivity for PFO detection after a single injection has been reported to be as high as 95%.
The results of an ASCi study were coded as negative if there was no left-heart contrast seen within five beats of the entrance of contrast into the right atrium after provocative maneuvers or at rest. Microbubbles visualized in the left-heart chambers after five cardiac cycles were presumed to represent transpulmonary transit and were thus considered negative. Study results were coded as positive if there was any contrast seen in either of the left-heart chambers within five beats of the entrance of contrast into the right atrium. In addition, which maneuver was associated with contrast visualization was noted. Studies were considered indeterminate if the image quality was suboptimal for visualization of contrast, when adequate maneuvers could not be performed, or when opacification of the RA septum could not be achieved.
Of the 1,162 patients who underwent TTE imaging with ASCi, 195 also underwent TEE imaging, considered the gold standard for anatomic assessment of the atrial septum. All examinations were performed under sedation (fentanyl/midazolam, with or without propofol) or general anesthetic. All examinations were performed by echocardiologists using the standard imaging protocol in our practice and an Acuson Sequoia C256 or C512 ultrasound system (Siemens Medical Solutions USA, Inc.) with a TE-V5M multiplane transesophageal probe. A comparison of detection of right-to-left atrial communication by TTE imaging with ASCi and TEE imaging was performed in this subgroup.
Statistical analysis was performed using commercially available statistical software (SPSS version 15.0; SPSS, Inc., Chicago, IL). Levine’s test for homogeneity as a quantitative analysis and a bell curve for qualitative demonstration of equal distribution of variance demonstrated normal distribution of variables. Continuous and categorical variables are expressed as mean ± SD and percentages, respectively. One-way analysis of variance was used for comparison of means for continuous variables between the groups of patients. Post hoc analysis was performed using the Scheffé method when the analysis of variance demonstrated a significant difference. Significance between categorical variables was assessed using a χ 2 test for goodness of fit. P values < .05 were considered statistically significant. Sensitivity and specificity tests were used to compare TTE ASCi, TTE CFI, and the TEE subgroup.
Clinical Indications for ASCi
The clinical indications for performing ASCi in all patients are demonstrated in Table 1 . The most common indications for ASCi were transient ischemic attack or cerebrovascular accident (25.6%), assessment of right ventricular function (17.6%), and episodes of visual loss (12.7%). Some form of neurologic symptom or sign was the basis for referral in 70.8% of the indications. The clinical indication for performing ASCi was not documented on the report or referral in 5.2% of studies.
|Assess RV function||205 (17.6)|
|Episode of visual loss||147 (12.7)|
|Limb weakness||90 (7.7)|
|Exclude shunt||74 (6.4)|
|Other CNS symptoms||68 (5.8)|
|Brain changes||50 (4.3)|
|Source of embolus||29 (2.5)|
|Loss of consciousness||2 (0.2)|
Patient Demographics and Clinical Characteristics
Patient demographics and clinical characteristics are summarized in Table 2 . There were small but significant differences ( P < .01) in the mean age, heart rate, and both systolic and diastolic blood pressures of the negative group compared with the positive group. There were also significant differences ( P < .01) in both systolic and diastolic blood pressures of the indeterminate group compared with the positive group.
( n = 1162)
( n = 403)
( n = 715)
( n = 44)
|Age (y)||51.1 ± 16.1||49.0 ± 16.2||52.1 ± 15.9 ∗||54.0 ± 15.8|
|Height (cm)||170.3 ± 9.6||170.8 ± 9.56||170.2 ± 9.59||169.0 ± 10.1|
|Weight (kg)||79.2 ± 17.9||77.5 ± 16.6||80.0 ±18.4||81.2 ± 20.1|
|BSA (m 2 )||1.90 ± 0.23||1.89 ± 0.22||1.91 ± 0.24||1.91 ± 0.27|
|HR (beats/min)||68.1 ± 11.8||66.9 ± 10.8||68.7 ± 12.1 ∗||69.0 ± 14.8|
|SBP (mm Hg)||132.9 ± 18.8||130.6 ± 18.4||134 ± 18.8 ∗||141.4 ± 20.6 ∗|
|DBP (mm Hg)||78.5 ± 10.5||77.3 ± 10.6||79.0 ± 10.3 ∗||82.5 ± 10.8 ∗|
Transthoracic CFI and ASCi
Table 3 summarizes the relationship between TTE CFI and shunting with ASCi. Assessment of the atrial septum for PFO using CFI demonstrated high specificity but low sensitivity for the detection of shunting with ASCi. Left-to-right shunting visualized by CFI was found to have a positive predictive value of 77% and a negative predictive value of 68%.