Background
Contrast injections during transesophageal echocardiography for patent foramen ovale (PFO) detection may be false negative. The characteristics of false-negative injections were studied retrospectively.
Methods
Contrast transesophageal echocardiography was analyzed for the presence or absence of two characteristics: leftward bulging of the interatrial septum and dense contrast filling of the region of the right atrium adjacent to the interatrial septum.
Results
Two hundred forty-seven injections were administered to 14 patients with PFOs, and 130 (53%) were false negative. The absence of either characteristic during a single injection resulted in low sensitivity for PFO detection of 7%. When the two characteristics were present at the same time, the sensitivity for PFO detection after a single injection was as high as 95%.
Conclusions
The simultaneous presence of both leftward bulging of the interatrial septum and dense contrast filling of the region in the right atrium, adjacent to the interatrial septum, is a prerequisite for PFO detection, and if either characteristic is missing, the injection is inconclusive.
Patent foramen ovale (PFO) is a remnant of the fetal circulation, which although a common and normal finding has been associated with a variety of pathologic conditions. Associated pathologies include cryptogenic stroke, decompression illness, migraine, and oxygen desaturation in obstructive sleep apnea, as well as obstructive pulmonary disease. During the first years of life, the thin septum primum fuses with the left atrial side of the rims of the foramen ovale. A PFO will persist when fusion is incomplete along part of the circumference. The septum primum acts as a flap valve, so the PFO is mostly functionally closed because the pressure in the left atrium normally is a few millimeters of mercury higher than in the right atrium. When the pressure relationship is inverted, with higher pressure in the right atrium, the PFO opens and allows some venous blood to flow from the right to the left atrium, bypassing the filter function of the lungs.
To find a PFO when it actually is present is clinically important because of the clinical significance of a PFO. Most of our knowledge regarding the clinical importance of a PFO originates from case-control studies, in which high and stable sensitivity for PFO detection is crucial. However, contrast injections during transesophageal echocardiography (TEE) are frequently false negative. Varying PFO prevalence in different studies may indicate that this is a common phenomenon. A well-controlled autopsy study by Hagen et al. showed an overall prevalence of 27%. However, a contrast transesophageal echocardiographic study of 1,000 consecutive clinical exams found a prevalence of only 9%.
Another remnant of the fetal circulation is the Eustachian valve, located at the entrance of the inferior vena cava (IVC) into the right atrium. As it directs blood flow toward the foramen ovale, it may contribute to keeping the foramen ovale patent. The prevalence of the Eustachian valve in the population is about 15% to 30%, and its presence is associated with PFO, large shunts, and cryptogenic stroke. However, the Eustachian valve may also be a source of false-negative contrast ultrasound images after injection in an arm vein. During release of Valsalva, contrast-free blood from the IVC surges in toward the PFO and sweeps away the contrast. For this reason, femoral vein injection has been advocated, but this is more invasive than an antecubital injection. Another method is color Doppler with low aliasing velocity.
We have previously shown that contrast injections may well be false negative and that repeated injections up to five times increase the sensitivity for PFO detection. To further explore why contrast injections during TEE are frequently false negative, we retrospectively analyzed a large number of negative contrast injections in patients subsequently shown to have PFOs. We searched for image criteria under which a negative injection might be related to faulty technique and therefore not a reliable indicator of PFO absence. Because a PFO is diagnosed when the shunted blood contains contrast, we analyzed the contrast filling of the region in the right atrium adjacent to the PFO. Because the shape of the septum depends on the interatrial pressure relation, we hypothesized that leftward bulging of the septum would be a prerequisite for PFO detection. The aim of this study was to evaluate the effect of the simultaneous presence or absence of these two characteristics on sensitivity for PFO detection.
Methods
Study Population
Thirty subjects with obstructive sleep apnea were screened for the presence or absence of PFO using contrast TEE. In 14 patients, PFOs were detected, and these 14 constituted the study population (see Table 1 for patient characteristics). All participants gave written informed consent. The study was approved by the human research regional ethics review board in Gothenburg, Sweden.
| Characteristic | Value |
|---|---|
| Age (years) | 61 (53–68) |
| Men | 9 (64%) |
| Body mass index (kg/m 2 ) | 30.9 ± 4.3 |
| Hypertension | 9 (64%) |
| Diabetes | 4 (29%) |
| Sinus rhythm | 13 (93%) |
Transesophageal Echocardiography
Examinations were performed and evaluated according to published recommendations. Left atrial pressure was estimated from the pulmonary vein and mitral valve flow pattern. The systolic maximum tricuspid regurgitation gradient was assessed with and/or without signal amplification with agitated polygeline as echocardiographic contrast. Central venous pressure was estimated from the collapsibility of the IVC on inspiratory sniff.
Contrast TEE
Before the examinations, patients were thoroughly informed, and they also practiced the Valsalva maneuver. All patients blew into a manometer with ≥40 mmHg for ≥8 sec. All patients received premedication with local pharyngeal anesthesia and mild sedation with midazolam 2 mL. All examinations were performed by the same experienced persons. Our extensive research protocol for PFO detection has previously been published and was designed to optimize the detection rate for PFO. The protocol included 15 to 20 contrast injections, regardless of whether a PFO was detected or not. As contrast we used polygeline, agitated with a small amount of air (Haemaccel; Aventis Pharma, Frankfurt am Main, Germany).
The solution was injected as a bolus and flushed with 5 to 10 mL of saline. We used a variety of provocations, such as the Valsalva maneuver, coughing, bed tilt, and premedication with sublingual nitroglycerin. The first two injections of 10 mL each were made in a foot vein, and the others of 2 mL were antecubital. We found the most effective provocation to be a Valsalva maneuver that started a few seconds before the contrast injection and was maintained until the right atrium was filled with contrast (i.e., about 10 sec), but even then, more than two injections were needed to detect all PFOs.
Analysis of PFOs and Characteristics of False-Negative Contrast Injections
As previously reported, a single injection was defined as PFO positive if at least three bubbles were visible in the left atrium, adjacent to the septum, within three heartbeats of opacification of the right atrium. A patient was defined as PFO positive if at least one injection was positive. The total left-to-right excursion of the interatrial septum was measured on digital images, but other transesophageal echocardiographic evaluation was made on S-VHS video. The size of the PFO was estimated according to the maximum number of shunted bubbles after an injection. When ≥20 bubbles were shunted, the PFO was defined as large, and when there were <20 bubbles, the PFO was defined as small.
In this study, we reviewed the video recordings retrospectively and analyzed each contrast injection with regard to the following criteria:
- 1.
The presence or absence of leftward bulging of the interatrial septum during the three beats that followed opacification of the right atrium. Leftward bulging was defined as a leftward shift and leftward convexity of the interatrial septum. It was qualitatively assessed by visual inspection as present or absent. Leftward bulging of the septum during a short moment was sufficient to classify the injection as bulging.
- 2.
The presence or absence of dense contrast filling of the region in the right atrium adjacent to the usual location of a PFO (i.e., the overlap between the septum primum and the septum secundum). When the contrast filling varied during the three beats, the moment when the septum primum bulged over toward the left was chosen for analysis.
- 3.
The presence or absence of a Eustachian valve was noted and the length of the valve was measured.
The presence or absence of septum bulging and right atrial contrast filling was then compared with the previously published results of the PFO analysis of each contrast injection. The PFO analysis was made independently by two persons, and disparities were settled by consensus with a third observer.
Statistical Analysis
The proportions of negative and positive injections exhibiting leftward bulging were compared using the χ 2 test with Yates’s continuity correction. The proportions of small and large PFOs according to left atrial pressure were compared using Fisher’s exact test. Analysis of variance was used for comparing parameters such as the percentage of positive injections and the numbers of shunted bubbles in patients with normal and elevated left atrial pressure, respectively.
Results
The total left-to-right excursion of the interatrial septum measured a mean of 11.8 mm (range, 6–17 mm). A total of 247 injections were given to the 14 PFO-positive patients, and 130 (53%) were negative ( Table 2 ). The absence of leftward bulging was the most frequent characteristic of a false-negative injection and was present in 114 of the 130 negative injections (88%) and only in eight of the 117 positive injections (7%) ( P < .001). The second most frequent characteristic was absent contrast filling, irrespective of leftward bulging. It was found in 30 of 130 negative injections (23%) compared with only one of 117 positive injections. Of the 30 false-negative injections due to absent contrast filling, all but one were seen in patients with concomitant Eustachian valves (valve length ≥ 8 mm). In these patients, 50% of the foot injections were negative.
| Negative injections | Positive injections | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Injections Total in each patient | Bulging | No bulging | Bulging | No bulging Filled | |||||
| Injections | Filled | Nonfilled | Filled | Nonfilled | Injections | Filled | Nonfilled | ||
| 16 | 5 | 0 | 3 | 0 | 2 | 11 | 11 | 0 | 0 |
| 16 | 2 | 0 | 0 | 2 | 0 | 14 | 8 | 0 | 6 |
| 17 | 1 | 0 | 0 | 0 | 1 | 16 | 15 | 0 | 1 |
| 19 | 5 | 0 | 0 | 5 | 0 | 14 | 14 | 0 | 0 |
| 18 | 10 | 0 | 0 | 10 | 0 | 8 | 7 | 0 | 1 |
| 15 | 2 | 0 | 0 | 2 | 0 | 13 | 12 | 1 | 0 |
| 16 | 13 | 0 | 1 | 7 | 5 | 3 | 3 | 0 | 0 |
| 20 | 18 | 0 | 1 | 11 | 6 | 2 | 2 | 0 | 0 |
| 18 | 9 | 0 | 1 | 3 | 5 | 9 | 9 | 0 | 0 |
| 19 | 6 | 1 | 0 | 5 | 0 | 13 | 13 | 0 | 0 |
| 18 | 14 | 0 | 0 | 14 | 0 | 4 | 4 | 0 | 0 |
| 18 | 13 | 1 | 0 | 12 | 0 | 5 | 5 | 0 | 0 |
| 20 | 18 | 3 | 2 | 12 | 1 | 2 | 2 | 0 | 0 |
| 17 | 14 | 1 | 2 | 11 | 0 | 3 | 3 | 0 | 0 |
| Total ( n = 247) | 130 | 6 | 10 | 94 | 20 | 117 | 108 | 1 | 8 |
Transthoracic echocardiography (TTE) showed normal regional wall motion and normal ejection fractions in all patients. As shown in Table 3 , the estimates of systolic pulmonary artery pressure and central venous pressure were within the normal range or only mildly increased. Filling patterns were consistent with elevated left atrial pressure in five patients. One patient had signs of elevated left atrial pressure throughout diastole (pseudonormal filling pattern) and three patients had signs of elevated end-diastolic left atrial pressure but no elevation before atrial contraction. In one patient with atrial fibrillation, the pulmonary venous flow pattern was consistent with elevated left atrial pressure. As shown in Table 3 , elevated left atrial pressure was associated with a smaller number of shunted bubbles and with a nonsignificant smaller proportion of leftward bulging.
| Variable | Filling pattern consistent with normal left atrial pressure | Filling pattern consistent with elevated left atrial pressure | P |
|---|---|---|---|
| Patients | 9 | 5 | |
| LVMI (g/m) | 110 ± 24 | 121 ± 26 | >.20 |
| SPAP (mm Hg) | 24 ± 3 (19–27) | 28 ± 7 (19–36) | >.20 |
| CVP (mm Hg) | 5 ± 0 | 6 ± 2 (5–10) | .19 |
| Bulging (%) | 56 ± 30 | 43 ± 17 | >.20 |
| Positive injections (%) | 56 ± 32 | 34 ± 25 | >.20 |
| Small PFOs ∗ | 0 | 3 | .03 |
| Large PFOs | 9 | 2 | |
| Mean number of shunted bubbles per injection | 58 ± 50 | 10 ± 16 | .06 |
| Mean number of shunted bubbles after foot injection | 86 ± 55 | 3 ± 7 | .006 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
