DOI:10.2214/AJR.06.0489
AJR 2007; 188:844-849
© American Roentgen Ray Society
Dynamic Contrast-Enhanced MRI Before and After Transcatheter Occlusion of Patent Foramen Ovale
Oliver K. Mohrs1,2,
Steffen E. Petersen3,
Damir Erkapic4,
Anja Victor5,
Thomas Schlosser6,
Bernd Nowak7,
Guenter Kauffmann8,
Thomas Voigtlaender7 and
Hans-Ulrich Kauczor2
1 Darmstadt Radiology, Department of Cardiovascular Imaging at Alice-Hospital,
Dieburger Strasse 29-31, 64287 Darmstadt, Germany.
2 German Cancer Research Center, 69120 Heidelberg, Germany.
3 Centre for Clinical Magnetic Resonance Imaging, University of Oxford, Oxford
OX3 9DU, United Kingdom.
4 Kerckhoff-Klinik, Bad Nauheim D-61231, Germany.
5 Institute of Medical Biostatistics, Epidemiology and Informatics, Mainz
D-55101, Germany.
6 Department of Radiology, University of Essen, Essen D-45122, Germany.
7 Cardiovascular Center Bethanien, Frankfurt/Main D-60389, Germany.
8 Department of Radiology, University of Heidelberg, Heidelberg D-69120,
Germany.
Received April 5, 2006;
accepted after revision July 31, 2006.
Address correspondence to O. K. Mohrs
(mohrs{at}radiologie-darmstadt.de).
Abstract
OBJECTIVE. The purpose of this study was threefold: to evaluate the
diagnostic accuracy of dynamic contrast-enhanced MRI compared with
transesophageal echocardiography (TEE) in the detection of patent foramen
ovale (PFO) and of residual shunts after occlusion of PFO, to define cutoff
values for semiquantitative analysis of signal intensity-time curves, and to
compare the diagnostic accuracy of visual detection with that of
semiquantitative analysis.
SUBJECTS AND METHODS. Forty-three patients (18 women, 25 men; mean
age, 51 ± 14 years) who underwent TEE for suspicion of PFO (n
= 26, 19 patients with and seven without PFO) or for routine assessment for
residual shunt after transcatheter PFO occlusion (n = 17, nine
patients with and eight without residual shunt) were consecutively enrolled to
undergo contrast-enhanced MRI (saturation recovery steady-state free
precession sequence). The images were analyzed both visually and
semiquantitatively for arrival of contrast agent in the left atrium before
arrival in the pulmonary veins during a Valsalva maneuver. TEE results were
used as the clinical reference.
RESULTS. With an area under the signal intensity-time curve of 0.85,
height of the first initial peak in signal intensity in the left atrium proved
to be the best discriminator in right-to-left shunt detection. For a cutoff
value of 129% (from baseline signal intensity) for this parameter, sensitivity
and specificity were 90% (17/19) and 100% (7/7) in patients without PFO
devices and 56% (5/9) and 88% (7/8) in patients with PFO devices. The
diagnostic accuracy of both visual assessment and semiquantitative analysis
was consistently superior before PFO device implantation than after device
implantation. The diagnostic accuracy of visual shunt assessment was better
than that of semiquantitative shunt assessment in patients with PFO occluders
(sensitivity, 67% [6/9] correctly diagnosed; specificity, 88% [7/8]) and those
without PFO occluders (sensitivity, 95% [18/19]; specificity, 100% [7/7]).
CONCLUSION. At present, MRI cannot replace TEE for the exclusion of
potential embolic sources, such as thrombus in the left atrial appendage.
However, MRI can be an attractive alternative noninvasive technique if TEE is
technically unfeasible or is declined by patients.
Keywords: cardiac imaging • cardiovascular imaging • congenital anomaly • dynamic MRI • MRI • patent foramen ovale
Introduction
Patent foramen ovale (PFO) is a known cause of cerebral stroke and
transient ischemic attacks due to paradoxic embolism, even in young patients.
Percutaneous transcatheter occlusion is an attractive, less invasive treatment
than surgical approaches [1].
Diagnosis and follow-up assessment are important for preventing further
cerebral events.
Contrast-enhanced transesophageal echocardiography (TEE) is considered the
clinical reference for detection of PFO
[2]. This method, however, is
partially invasive and requires sedation in many patients. Traditional MRI
techniques for assessment of atrial and ventricular septal defects, such as
spin-echo sequences for morphologic assessment and phase-contrast sequences
for flow assessment, are of limited value in PFO evaluation because PFO is
very small and has minimal or intermittent shunt flow. Dynamic
contrast-enhanced MRI has been introduced for the diagnosis of PFO and atrial
septal aneurysm [3]. It remains
speculative, however, whether qualitative visual assessment or
semiquantitative analysis of right-to-left-shunt is superior. In addition, the
role of contrast-enhanced MRI in follow-up assessment after transcatheter
occlusion of PFO has not been studied.
The purpose of this study was threefold: to evaluate the diagnostic
accuracy of contrast-enhanced MRI compared with TEE in the detection of PFO
and of residual shunt after occlusion of PFO, to define cutoff values for
semiquantitative analysis of signal intensity-time curves, and to compare the
diagnostic accuracy of visual detection with that of semiquantitative
analysis.
Subjects and Methods
Study Population
A total of 43 patients were consecutively included in this study from July
2003 to December 2005. Fifteen of these 43 patients had participated in a
previously described pilot study
[3]. Twenty-six of the patients
underwent TEE for initial evaluation of possible PFO. TEE was prompted by
clinical symptoms of conditions such as cerebrovascular events, including
transient ischemic attacks, amaurosis fugax, and strokes, and recurrence of
peripheral emboli and suspicion of PFO during transthoracic echocardiography.
Contrast-enhanced TEE showed evidence of PFO in 19 of the 26 patients. Among
the 17 patients with PFO closure devices, contrast-enhanced TEE showed
residual shunt in nine. Of the 19 patients found to have PFO, seven underwent
PFO closure (PFO confirmed during angiography) and received follow-up with TEE
and contrast-enhanced MRI. Thus, a total of 50 scans were obtained for 43
patients. The following percutaneous transcatheter occlusion devices were
used: 16 CardiaStar (Cardia), five Amplatzer (AGA Medical), two Intrasept
(Cardia), and one Helex (W. L. Gore & Associates). TEE was performed 9
± 7 months (range, 4-37 months) after device implantation to exclude
residual right-to-left shunt. The maximum time period between TEE and
contrast-enhanced MRI was 1 week. The study was approved by the local ethics
committee, and each patient gave written informed consent.
TEE
TEE was performed with a 5-MHz phased multiplane probe (Vingmed System
Five, GE Healthcare). Local pharyngeal anesthesia was achieved with 0.02 mg of
orally administered lidocaine spray (Xylocaine, AstraZeneca). The contrast
agent D-galactose (Echovist, Schering) was administered as a 10-mL bolus into
an antecubital vein during a Valsalva maneuver. Right-to-left shunting was
evaluated by an experienced cardiologist blinded to patient information.
MRI
Contrast-enhanced MRI was performed with a 1.5-T system (Magnetom Sonata
Maestro Class, Siemens Medical Solutions). The combination of a six-channel
body phased-array coil and a two-channel spinal phased-array coil was used for
signal detection. The ECG signal was received from an external system
(Magnitude 3150, InVivo Research).
Contrast-enhanced perfusion studies were performed during a Valsalva
maneuver with 10 mL of gadopentetate dimeglumine (Magnevist, Schering)
followed by a 20-mL saline solution administered into an antecubital vein. The
infusion rate was 6 mL/s. Contrast injection and image acquisition were
started at the same time. The Valsalva maneuver was practiced with patients
before MRI and was done throughout the inspiratory breath-hold. Two slices of
a saturation-recovery steady-state free precession sequence (true fast imaging
with steady-state free precession; TE, 2.7; inversion time, 217 milliseconds;
flip angle, 50°; temporal resolution, 832 milliseconds; matrix size, 144
x 256; voxel size, 1.8 x 1.4 x 6.5 mm) were positioned in
the horizontal long and short axes. These slices were chosen as the optimal
views on cine studies performed with multiple parallel slices in the
horizontal long and short axes. For each slice, 40 consecutive images were
acquired, one per heart cycle. Metal artifacts were present with PFO devices
but were small. All data sets were of sufficient quality to be analyzed.
Contrast-enhanced perfusion studies were analyzed by a radiologist
experienced in cardiovascular MRI and blinded to the TEE diagnosis. The
analysis consisted of visual assessment and interpretation of signal
intensity-time curves for the diagnosis of PFO. Arrival of contrast material
in the left atrium before arrival in the pulmonary veins during a Valsalva
maneuver indicated right-to-left shunt at visual assessment. Signal
intensity-time curves were generated with a mean-curve program (Syngo Mean
Curve, Siemens Medical Solutions).
Two regions of interest (ROIs) were placed in the left atrium and the
better-suited pulmonary vein (i.e., the one with minimal motion and
accommodating a large ROI). ROIs were manually fitted to every image without
changes in size. Atrial ROIs were drawn close to the atrial septum, but
inclusion of pixels from the right atrium was carefully avoided. The signal
intensity-time curves were normalized to baseline signal intensity for each
ROI (second image because the signal in the first image was not fully
saturated after the saturation-preparation pulse). The single-intensity data
points represented percentage of baseline signal intensity. Typical signal
intensity-time curves show a first signal peak in the left atrium caused by
right-to-left shunt flow before the contrast agent arrives in the pulmonary
veins. This initial peak in the left atrium is followed by a second higher
peak representing the bolus of contrast agent arriving through the pulmonary
circulation [3].
Statistical Analysis
As measures of agreement (corrected for agreement by chance only), kappa
values with 95% CIs were calculated for qualitative visual assessment. TEE
results served as the reference method for the diagnosis of PFO and of
residual right-to-left shunt after percutaneous occlusion of PFO. Kappa values
were interpreted according to the method of Landis and Koch
[4]. The sensitivity and
specificity of visual assessment of contrast-enhanced MRI with TEE as the
clinical reference method were determined and expressed only as a percentage
if the denominator was greater than 10. For continuous measurements (such as
relative height of and time to peaks), median, minimum, and maximum values
were obtained. Cutoff values for left atrial peak were chosen on the basis of
receiver operating characteristic curves. Cutoff values were determined with
data acquired during the first scan of all 43 patients (the seven follow-up
examinations were excluded). All analyses were performed with SPSS version 12
(SPSS).
Results
Diagnostic Accuracy of Visual Assessment of Contrast-Enhanced MR Images
PFO was identified in 18/19 (95%) patients and was excluded correctly in
7/7 (100%) patients on initial TEE evaluation. Residual shunt was identified
in 6/9 (87%) patients and excluded correctly in 13/15 (87%) patients,
including the seven patients who underwent follow-up examinations after
transcatheter closure. Tables 1
and 2 summarize the diagnostic
accuracy of visual assessment and semiquantitative assessment in shunt
detection in patients with and those without PFO occluder devices.
View this table:
[in this window]
[in a new window]
|
TABLE 1: Absolute Numbers of Right-to-Left Shunts Diagnosed with MRI Compared
with Transesophageal Echocardiography
|
|
The patient in the case of false-negative detection of a PFO shunt was a
40-year-old woman with medium-grade PFO (25-microbubble shunt flow). In the
three cases of false-negative detection of a residual shunt after PFO
occlusion, the patients were a 70-year-old woman with a CardiaStar occluder
device and 22- and 60-year-old men with Intrasept occluder devices. All three
had medium-grade shunt flow on TEE (10-20 microbubbles). The two
false-positive diagnoses occurred in a 63-year-old man with a CardiaStar
occluder device and a 55-year-old woman with an Amplatzer occluder device.
The kappa value for agreement of TEE and contrast-enhanced MRI in the
detection of right-to-left shunt was good (0.76; 95% CI, 0.58-0.94) for all 50
observations. Assessment in patients without PFO occlusion (n = 26)
yielded a very good kappa value (0.91; 0.73-1.00), whereas assessment after
occlusion (n = 24) showed a moderate kappa value (0.54;
0.20-0.89).
Diagnostic Accuracy of Semiquantitative Analysis of Contrast-Enhanced MR Images
All 28 patients with either PFO or residual shunt had signal intensity-time
curves with an early first peak in the left atrium of a median 157% of
baseline signal intensity at a median time of 7 seconds between contrast
injection and the start of image acquisition compared with patients without
PFO (n =7) or residual shunt (n = 8) with a median maximum
signal intensity of 101% of baseline signal intensity (second image after
contrast injection). This early first peak in the left atrium was followed by
a median peak in the pulmonary vein of 522% at a median time of 17 seconds
(patients without shunt, 484% at 17 seconds). The initial peak and drop back
to baseline in the left atrium (median, 114% in patients with PFO vs 109% in
those without PFO and residual shunt) was followed by a second higher peak of
555% at a median time of 17 seconds (median, 569% at 18 seconds in patients
without right-to-left shunt on TEE).
Figure 1A,
1B,
1C,
1D,
1E,
1F,
1G,
1H illustrates right-to-left
shunting through a PFO before transcatheter occlusion (Figs.
1A,
1B,
1C,
1D) and exclusion of residual
shunt after transcatheter occlusion (Figs.
1E,
1F,
1H). Signal intensity-time
curves before and after PFO occlusion (Fig.
2A,
2B) confirmed the findings of
visual assessment. Figure 3A,
3B,
3C,
3D shows a case of residual
shunt after transcatheter PFO occlusion. As shown in
Figure 4, the relative values
of the first peak in the left atrium were lower in patients with residual
shunt than in those with initially diagnosed PFO. Variations from baseline
signal intensity were slightly greater in patients with exclusion of residual
shunt than in patients without PFO.
View larger version (171K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A —42-year-old man with right-to-left shunt before and absence
of residual shunt after occlusion of patent foramen ovale (PFO). Temporal
sequence of contrast-enhanced dynamic perfusion images during Valsalva
maneuver before transcatheter occlusion of PFO. Images show baseline signal
intensity without enhancement, (RA = right atrium, RV = right ventricle, LA =
left atrium, LV = left ventricle, PV = pulmonary vein, PA = pulmonary artery)
(A), enhancement of right atrium, slight enhancement of pulmonary
artery, and enhancement of entire left atrium due to right-to-left-shunt
(arrow) before enhancement of pulmonary vein (B), decrease in
signal intensity in left atrium representing dip back to baseline after first
initial peak (C), and enhancement of pulmonary vein and second signal
peak in left atrium (D).
|
|
View larger version (181K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B —42-year-old man with right-to-left shunt before and absence
of residual shunt after occlusion of patent foramen ovale (PFO). Temporal
sequence of contrast-enhanced dynamic perfusion images during Valsalva
maneuver before transcatheter occlusion of PFO. Images show baseline signal
intensity without enhancement, (RA = right atrium, RV = right ventricle, LA =
left atrium, LV = left ventricle, PV = pulmonary vein, PA = pulmonary artery)
(A), enhancement of right atrium, slight enhancement of pulmonary
artery, and enhancement of entire left atrium due to right-to-left-shunt
(arrow) before enhancement of pulmonary vein (B), decrease in
signal intensity in left atrium representing dip back to baseline after first
initial peak (C), and enhancement of pulmonary vein and second signal
peak in left atrium (D).
|
|
View larger version (168K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C —42-year-old man with right-to-left shunt before and absence
of residual shunt after occlusion of patent foramen ovale (PFO). Temporal
sequence of contrast-enhanced dynamic perfusion images during Valsalva
maneuver before transcatheter occlusion of PFO. Images show baseline signal
intensity without enhancement, (RA = right atrium, RV = right ventricle, LA =
left atrium, LV = left ventricle, PV = pulmonary vein, PA = pulmonary artery)
(A), enhancement of right atrium, slight enhancement of pulmonary
artery, and enhancement of entire left atrium due to right-to-left-shunt
(arrow) before enhancement of pulmonary vein (B), decrease in
signal intensity in left atrium representing dip back to baseline after first
initial peak (C), and enhancement of pulmonary vein and second signal
peak in left atrium (D).
|
|
View larger version (177K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1D —42-year-old man with right-to-left shunt before and absence
of residual shunt after occlusion of patent foramen ovale (PFO). Temporal
sequence of contrast-enhanced dynamic perfusion images during Valsalva
maneuver before transcatheter occlusion of PFO. Images show baseline signal
intensity without enhancement, (RA = right atrium, RV = right ventricle, LA =
left atrium, LV = left ventricle, PV = pulmonary vein, PA = pulmonary artery)
(A), enhancement of right atrium, slight enhancement of pulmonary
artery, and enhancement of entire left atrium due to right-to-left-shunt
(arrow) before enhancement of pulmonary vein (B), decrease in
signal intensity in left atrium representing dip back to baseline after first
initial peak (C), and enhancement of pulmonary vein and second signal
peak in left atrium (D).
|
|
View larger version (178K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1E —42-year-old man with right-to-left shunt before and absence
of residual shunt after occlusion of patent foramen ovale (PFO). MR images
show results after transcatheter occlusion with Amplatzer device (AGA
Medical). Artificial loss of signal intensity caused by occluder device
(circle, E) is evident. In contrast to examination before
occlusion, enhancement of left atrium before enhancement of pulmonary vein due
to right-to-left shunt can be excluded.
|
|
View larger version (176K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1F —42-year-old man with right-to-left shunt before and absence
of residual shunt after occlusion of patent foramen ovale (PFO). MR images
show results after transcatheter occlusion with Amplatzer device (AGA
Medical). Artificial loss of signal intensity caused by occluder device
(circle, E) is evident. In contrast to examination before
occlusion, enhancement of left atrium before enhancement of pulmonary vein due
to right-to-left shunt can be excluded.
|
|
View larger version (178K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1G —42-year-old man with right-to-left shunt before and absence
of residual shunt after occlusion of patent foramen ovale (PFO). MR images
show results after transcatheter occlusion with Amplatzer device (AGA
Medical). Artificial loss of signal intensity caused by occluder device
(circle, E) is evident. In contrast to examination before
occlusion, enhancement of left atrium before enhancement of pulmonary vein due
to right-to-left shunt can be excluded.
|
|
View larger version (188K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1H —42-year-old man with right-to-left shunt before and absence
of residual shunt after occlusion of patent foramen ovale (PFO). MR images
show results after transcatheter occlusion with Amplatzer device (AGA
Medical). Artificial loss of signal intensity caused by occluder device
(circle, E) is evident. In contrast to examination before
occlusion, enhancement of left atrium before enhancement of pulmonary vein due
to right-to-left shunt can be excluded.
|
|
View larger version (12K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A —42-year-old man with right-to-left shunt. Same patient as in
Figure 1A,
1B,
1C,
1D,
1E,
1F,
1G,
1H. With patent foramen ovale
(PFO) (A) and after transcatheter occlusion of PFO (B). Signal
intensity-time curve (A) shows early initial signal peak
(arrow) (167% of baseline signal) in left atrium (squares)
followed by second higher peak before peak in pulmonary vein
(circles) before PFO occlusion. Signal intensity-time curve after PFO
occlusion (B) shows only slight signal variability (up to 118% of
baseline signal) compared with baseline signal intensity.
|
|
View larger version (12K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B —42-year-old man with right-to-left shunt. Same patient as in
Figure 1A,
1B,
1C,
1D,
1E,
1F,
1G,
1H. With patent foramen ovale
(PFO) (A) and after transcatheter occlusion of PFO (B). Signal
intensity-time curve (A) shows early initial signal peak
(arrow) (167% of baseline signal) in left atrium (squares)
followed by second higher peak before peak in pulmonary vein
(circles) before PFO occlusion. Signal intensity-time curve after PFO
occlusion (B) shows only slight signal variability (up to 118% of
baseline signal) compared with baseline signal intensity.
|
|
View larger version (158K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A —62-year-old man with medium-grade residual shunt on
transesophageal echocardiography after occlusion of patent foramen ovale with
CardiaStar device (Cardia). MR image shows baseline signal without
enhancement.
|
|
View larger version (177K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B —62-year-old man with medium-grade residual shunt on
transesophageal echocardiography after occlusion of patent foramen ovale with
CardiaStar device (Cardia). MR image shows enhancement of right atrium,
pulmonary artery, and entire left atrium due to residual shunt before
enhancement of pulmonary vein.
|
|
View larger version (172K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3C —62-year-old man with medium-grade residual shunt on
transesophageal echocardiography after occlusion of patent foramen ovale with
CardiaStar device (Cardia). MR image shows decrease in signal intensity in
left atrium representing dip back to baseline after first initial peak.
|
|
View larger version (189K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3D —62-year-old man with medium-grade residual shunt on
transesophageal echocardiography after occlusion of patent foramen ovale with
CardiaStar device (Cardia). MR image shows enhancement of pulmonary vein and
second signal-intensity peak in left atrium.
|
|
View larger version (20K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4 —Median signal intensity normalized to baseline signal
intensity (second image after contrast injection and start of image
acquisition) for patients with and without patent foramen ovale (PFO) and with
and without residual shunt after PFO device implantation. Dashed line at 129%
signal intensity of baseline signal represents proposed cutoff value for
diagnosis of shunt for first peak in left atrium (LA). PV = pulmonary
vein.
|
|
The relative height of the first peak was the best discriminator of PFO
shunt detection in patients with and those without PFO occluders, as indicated
by the largest area under the signal intensity-time curve (AUC, 0.85),
followed by the relative level after the initial signal peak (AUC, 0.79), as
shown in Figure 5. For a
cutoff value of 129%, sensitivity was 79% (22/28) and specificity was 93%
(14/15) for discriminating a true initial peak representing right-to-left
shunt from heterogeneous baseline signal intensity in patients without
right-to-left shunt. The diagnostic accuracy for shunt detection was better in
patients without PFO devices than in those with PFO devices. For patients
without PFO devices, the 129% cutoff value gave 90% (17/19) sensitivity and
100% (7/7) specificity. In the detection of residual shunt, however, use of
the cutoff value of 129% resulted in a sensitivity of only 56% (5/9) and a
specificity of 93% (14/15). Four patients with residual shunt on TEE had
contrast-enhanced MRI peak values as low as 100%, 104%, and 107% (two
patients).
View larger version (13K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5 —Receiver operating characteristic curves show relations of
sensitivity to 1 - specificity for different values of relative peak
(solid line) in left atrium and relative level (dashed line)
after first peak. Ideal of 100% sensitivity and 100% specificity would be
reached in upper left corner.
|
|
Contrast-Enhanced MRI Before and After Percutaneous Occlusion of PFO
In seven patients, TEE and contrast-enhanced MRI were performed before and
after transcatheter occlusion of PFO. Visual assessment based on early
contrast enhancement due to intracardiac right-to-left shunt in the left
atrium led to a correct diagnosis in all seven cases before intervention but
to one false-positive finding after occlusion. On the basis of the presence or
absence of an early first signal-intensity peak in the left atrium over the
cutoff value of 129% in all seven patients, PFO was correctly diagnosed and
residual shunt correctly excluded compared with the reference TEE
findings.
Discussion
Our main findings are, first, that contrast-enhanced MRI is a feasible
noninvasive method of assessing right-to-left shunting before and after
transcutaneous occlusion of PFO. Second, the diagnostic accuracy of both
visual assessment and semiquantitative analysis is consistently superior
before PFO device implantation than after device implantation. Third, visual
assessment by an experienced reviewer appears superior to semiquantitative
analysis in PFO shunt detection. Diagnostic accuracy was better for visual
shunt assessment in patients with and those without PFO occluders and for the
entire patient population in the study. In addition, visual assessment
obviates laborious postprocessing. However, less experienced observers may
want to confirm the visual assessment with a semiquantitative assessment. We
identified the height of the first initial peak in signal intensity in the
left atrium as the best discriminator for all subgroups of patients with and
without PFO occluders. For detection of right-to-left shunt through a PFO, we
suggest use of a cutoff value of 129% for the initial peak in signal intensity
in the left atrium. However, a larger study is needed to confirm the cutoff
level.
Results of a recent pilot study
[3] suggest that
contrast-enhanced MRI is reliable in the diagnosis of PFO and atrial aneurysm.
Atrial septal defects can be detected with MRI directly with morphologic
visualization using spin-echo techniques or indirectly with quantitative shunt
flow measurements [5,
6]. Although it is a potential
pathway for thrombi to bypass the pulmonary circulation, a PFO is too small to
be directly imaged with MRI. In addition, right-to-left shunt volume is too
small to be detected with flow quantification.
Contrast-enhanced TEE is the current clinical reference for the diagnosis
of PFO. The findings have correlated well in autopsy studies
[7,
8]. The main disadvantages of
TEE are its partially invasive nature and the inability to acquire images in
more than one plane during administration of the contrast agent.
Semi-quantitative methods may be an important adjunct for inexperienced
reviewers. An advantage of signal intensity-time curve analysis is clear-cut
differentiation of contrast enhancement due to right-to-left shunt in the left
atrium from contrast enhancement due to pulmonary flow. This differentiation
is not possible with TEE and can cause false-positive results in the rare but
important case of pulmonary arteriovenous malformation
[9].
We suspect that the discrepancies between TEE and contrast-enhanced MRI may
reflect the intraindividual variability of the Valsalva maneuver, which is the
most effective way to induce right-to-left atrial shunt
[10]. If a Valsalva maneuver
is not performed, PFO can be missed
[2]. Therefore, deep sedation
may alter the patient's ability to perform the maneuver and thus lead to
misinterpretation. This problem can occur among patients who cannot tolerate
the transesophageal probe or contrast-enhanced MRI without IV sedation. In
addition, a Valsalva maneuver during a long breath-hold for MRI can be
problematic. A false-positive result in a patient with residual shunt on MRI
may be the correct diagnosis if right-to-left shunt is clearly detectable
visually and with computation of signal intensity-time curves. The error may
reflect an insufficient Valsalva maneuver on TEE.
The ability to immediately review images online on the MRI unit screen
helps examiners react on arrival of contrast agent in the left atrium. If a
patient ends the Valsalva maneuver without exhaling, intrathoracic pressure is
reduced, and venous flow into the right atrium is maximized. The result is a
better bolus of contrast agent. Although this procedure is possible on TEE, we
used a power injector with a high flow rate to yield a similar effect.
D-galactose, known as a microbubble contrast agent in sonography, does not
pass the pulmonary capillaries; therefore, additional injections are possible
on TEE. In contrast, an MRI contrast agent passes the pulmonary capillaries
and is present in circulating blood after the first administration. Another
injection is useless because the small amount of contrast agent shunting
through the PFO would not be detected. Thus, patients have to be well informed
about the performance of the Valsalva maneuver.
We used a saturation recovery sequence with steady-state free precession.
To minimize the difference in spatial resolution of TEE and MRI to a ratio of
3:1, we prolonged the TR, which increased the size of occluder-related
artifacts. The presence of occluder-related artifacts may explain the worse
diagnostic results for detection of residual shunt compared with PFO. The
smaller peaks in patients with residual shunt than in patients with initially
diagnosed PFO may be another explanation. Contrast agents with higher
relaxivity probably enhance the peak signal intensity in the left atrium
because of right-to-left shunt flow and probably produce sharper cutoff values
even in patients with a small residual shunt
[11].
MRI at present cannot replace TEE to exclude potential embolic sources such
as thrombus in the left atrial appendage
[12]. Compared with TEE, MRI
is more expensive and requires a longer examination time. MRI, however, may be
an attractive alternative non-invasive technique if TEE is technically
unfeasible or is declined by patients. With technical improvement, the
technique we describe may be part of a single MRI examination.
In conclusion, this study showed that MRI can be used for reliable
diagnosis of PFO with either visual or semiquantitative assessment. The
diagnostic accuracy in detection of residual shunt after transcatheter
occlusion has to be improved.
Acknowledgments
This article contains parts of the doctoral thesis of D. Erkapic. We thank
Christine Rubel and Rainer Schraeder for their support.
References
- Meier B, Lock JE. Contemporary management of patent foramen ovale.
Circulation 2003;107
: 5-9[Free Full Text]
- Schneider B, Zienkiewicz T, Jansen V, Hofmann T, Noltenius H,
Meinertz T. Diagnosis of patent foramen ovale by transesophageal
echocardiography and correlation with autopsy findings. Am J
Cardiol 1996; 77:1202
-1209[CrossRef][Medline]
- Mohrs OK, Petersen SE, Erkapic D, et al. Diagnosis of patent
foramen ovale using contrast-enhanced dynamic MRI: a pilot study.
AJR 2005; 184:234
-240[Abstract/Free Full Text]
- Landis JR, Koch GG. The measurement of observer agreement for
categorical data. Biometrics 1977;33
: 159-174[CrossRef][Medline]
- Beerbaum P, Korperich H, Barth P, Esdorn H, Gieseke J, Meyer H.
Noninvasive quantification of left-to-right shunt in pediatric patients:
phase-contrast cine magnetic resonance imaging compared with invasive
oximetry. Circulation 2001;103
: 2476-2482[Abstract/Free Full Text]
- Hundley WG, Li HF, Lange RA, et al. Assessment of left-to-right
intracardiac shunting by velocity-encoded, phase-difference magnetic resonance
imaging: a comparison with oximetric and indicator dilution techniques.
Circulation 1995;91
: 2955-2960[Abstract/Free Full Text]
- Hausmann D, Mugge A, Becht I, Daniel WG. Diagnosis of patent
foramen ovale by transesophageal echocardiography and association with
cerebral and peripheral embolic events. Am J Cardiol1992; 70:668
-672[CrossRef][Medline]
- Chen WJ, Kuan P, Lien WP, Lin FY. Detection of patent foramen ovale
by contrast transesophageal echocardiography. Chest1992; 101:1515
-1520[Abstract/Free Full Text]
- Mohrs OK, Voigtlander T, Heussel CP, Bardeleben S, Duber C,
Kreitner KF. Morphologic and functional assessment of vascular abnormalities
of the pulmonary vasculature by breath-hold MR techniques [in German].
Rofo 2002; 174:467
-473[Medline]
- Pfleger S, Konstantin Haase K, Stark S, et al. Haemodynamic
quantification of different provocation manoeuvres by simultaneous measurement
of right and left atrial pressure: implications for the echocardiographic
detection of persistent foramen ovale. Eur J
Echocardiogr 2001; 2:88
-93[Abstract/Free Full Text]
- Goyen M, Lauenstein TC, Herborn CU, Debatin JF, Bosk S, Ruehm SG.
0.5 M Gd chelate (Magnevist) versus 1.0 M Gd chelate (Gadovist):
dose-independent effect on image quality of pelvic three-dimensional
MR-angiography. J Magn Reson Imaging2001; 14:602
-607[CrossRef][Medline]
- Mohrs OK, Nowak B, Petersen SE, et al. Thrombus detection in the
left atrial appendage by means of contrast-enhanced MRI: a pilot study.
AJR 2006; 186:198
-205[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
M.-P. Revel, J.-B. Faivre, T. Letourneau, H. Henon, D. Leys, V. Delannoy-Deken, M. Remy-Jardin, and J. Remy
Patent Foramen Ovale: Detection with Nongated Multidetector CT
Radiology,
October 1, 2008;
249(1):
338 - 345.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
