| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Clinical Observations |
1 Department of Radiology, Hôpital Marie Lannelongue, 133 ave. de la
Résistance, Le Plessis-Robinson, France 92350.
2 Department of Thoracic Surgery, Hôpital Marie Lannelongue, Le
Plessis-Robinson, France 92350.
Received June 8, 2005;
accepted after revision September 18, 2005.
Address correspondence to J.-F. Paul
(pauljf{at}ccml.fr).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CONCLUSION. The sensitivity, specificity, and accuracy of MDCT in prediction of the presence of an endarterectomy plane were 99%, 80%, and 96%. Bilateral absence of an endarterectomy plane correlated with postoperative mortality according to Fisher's exact test results (p = 0.004). Submillimeter 16-MDCT therefore may be useful in predicting operability in chronic thromboembolic pulmonary hypertension.
Keywords: cardiopulmonary imaging • chronic thromboembolism • embolism • endarterectomy • MDCT • pulmonary hypertension
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CT is a valuable imaging technique complementary to conventional angiography for diagnosis of the chronic thromboembolic pulmonary hypertension, for determination of technical operability, and for prediction of surgical success [4-7]. Other factors, including patient status, hemodynamic data, and extent of the central clot are also predictive of technical success. Helical CT has been reported to be more sensitive than conventional angiography in the detection of central chronic thromboembolic pulmonary hypertension because the presence of clot sometimes is inferred with conventional angiography rather than directly visualized with CT. Helical CT is as sensitive as MR angiography in the detection of disease down to the segmental level [8].
To our knowledge, the use of CT for detection of an endarterectomy plane has not been described, probably because of lack of spatial resolution of earlier CT scanners. Submillimeter acquisition now is possible, improving spatial resolution and allowing more accurate delineation of an endarterectomy plane. We tested the hypothesis that preoperative 16-MDCT at 0.7-mm collimation can be used for accurate prediction of the presence of an endarterectomy plane in patients undergoing thromboendarterectomy for chronic thromboembolic pulmonary hypertension.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
Imaging Protocol and Analysis
Imaging parameters were as follows: 120 kV; 180 mAs; collimation, 0.7 mm;
slice thickness, 1 mm with 0.7-mm increments or 0.7 mm with 0.5-mm increments;
medium filter (B40). All patients underwent caudocranial scanning from the
lung bases to the lung apices during a single breath-hold. A total volume of
90 mL of nonionic contrast medium (Iopamiron 300 [iopamidol], Schering) was
administered IV at a rate of 3.5 mL/s with an automatic injector device
(Envision CT, Medrad) through an 18-gauge IV catheter. Scanning was
automatically initiated 5 seconds after detection of contrast enhancement
(threshold, 100 H) in the main pulmonary artery. The average time of
acquisition was 15 seconds.
Two independent experienced chest radiologists (more than 10 years of experience each) reviewed thin-slab multiplanar views interactively and recorded the presence or absence of abnormal thickening of the pulmonary artery wall of 2 mm or more down to the lobar arteries [1, 2]. This finding was considered predictive of the presence of an endarterectomy plane at surgery. Fifteen-millimeter thin-slab images in the long axis of both main pulmonary arteries were acquired with both maximum intensity projection (MIP) and native multiplanar display modes. The radiologists selected the planes interactively by moving the plane of orientation to display the entire length of each pulmonary artery.
The slice thickness and window settings of MIP images were adjusted to improve visualization of proximal chronic thromboemboli, if any, with careful attention to the origins of segmental pulmonary arteries (Figs. 1A, 1B and 2). Chronic thromboemboli appeared as dark gray areas along the pulmonary wall. From these images, additional images were acquired by moving the plane of orientation slightly to reveal the origins of each segmental artery.
The right and left main pulmonary arteries were assessed separately. After surgical intervention, all CT scans were analyzed in consensus by the two radiologists who were blinded to the surgical reports. The presence or absence of an endarterectomy plane for each pulmonary artery was recorded in the surgical reports, and CT findings were compared with these records. Preoperative MDCT was performed a mean of 2.2 days (range, 0-6 days) before surgery. Although some patients had severe respiratory insufficiency, breath-holding was possible in all cases because of rapid acquisition. All examinations achieved diagnostic quality for visualization of an endarterectomy plane.
Statistical Analysis
The sensitivity, specificity, and accuracy of CT depiction of pulmonary
mural thickening in prediction of the presence of an endarterectomy plane at
surgery were calculated for each lung separately. Bilateral absence of
endarterectomy planes and postoperative mortality were compared by use of
Fisher's exact test. Statistical significance was considered p
<0.05.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Table 1 shows the comparison of surgical and MDCT results. For the lungs separately, 69 of 70 endarterectomy planes were correctly identified. There were one false-negative finding (failure to detect an endarterectomy plane) and two false-positive findings (abnormal thickening but no endarterectomy plane at surgery). Eight of the 10 cases of absence of an endarterectomy plane at surgery were identified with preoperative MDCT (sensitivity, 99%; specificity, 80%; accuracy, 96%). The two cases of absence of an endarterectomy plane on either side were correctly identified with MDCT. Three of the 40 patients died during the operation: two of these three patients had no left or right endarterectomy plane on MDCT (p = 0.004). These two patients had persistent suprasystemic hypertension at the end of the surgical procedure.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A suitable endarterectomy plane is defined as intimal thickening of the pulmonary artery starting at the pulmonary artery trunks no further than the origin of the lobar arteries [1]. Patients in whom pulmonary artery thickening starts distal to the lobar origin are not candidates for endarterectomy. An endarterectomy plane cannot be attained with a median sternotomy approach. In our series, bilateral absence of an endarterectomy plane correlated with operative mortality because access could not be gained to segmental and subsegmental structures. In both patients who died, pulmonary hypertension was due to predominantly distal obstruction. In both cases, persistent severe pulmonary hypertension immediately after surgical intervention led to death. Reliable preoperative detection of this condition may lead to use of alternative treatments, such as lung or cardiopulmonary transplantation and medical management [1].
Accurate detection of an endarterectomy plane is possible because of improvements in MDCT technology, especially regarding spatial resolution. We have differentiated two main CT patterns of chronic thromboembolic hypertension: an obstructive form, characterized by large amounts of pulmonary endovascular thrombotic material and large and complete occlusions in the central vascular bed, and a parietal form, which is associated with thickening of the walls of the pulmonary arteries and obstruction mainly at the origin of segmental arteries.
We found that the parietal form of chronic thromboembolic hypertension was associated with severe obstruction despite relatively small amounts of endovascular thrombotic material. Detection of the parietal form was possible because of submillimeter spatial resolution of imaging acquisition. We arbitrarily chose a threshold of 2 mm for assessment for an endarterectomy plane to ensure consistency in image analysis. This threshold was defined on the basis of results of previous analysis of radiologic and surgical findings at our institution. Images were reconstructed with a 0.7-mm slice thickness (rather than 1 mm) for the parietal form of hypertension to allow more sensitive detection of minor changes in the wall of the pulmonary artery. A medium filter was found helpful for improving identification of thickening of the arterial wall.
Analysis of chronic thromboembolic involvement of the pulmonary artery requires 3D multiplanar reformatting of multiple images. Thin-slab MIP views, especially in the sagittal and coronal planes, also are useful for visualization of interruption of the vascular bed. Calcifications along the walls of the pulmonary arteries were rare in our study and did not hamper analysis. In a previous study, in which helical CT was used for prediction of the outcome of pulmonary endarterectomy, only 5-mm axial images were used [7]. Use of 5-mm collimation with earlier CT technology does not generate accurate multiplanar views, so an endarterectomy plane could not be accurately detected.
Improvements in CT technology, especially spatial resolution, have made it possible to evaluate abnormalities in the pulmonary arterial wall. However, we had one false-negative and two false-positive results. In the false-negative case, only a poor endarterectomy plane was found at surgery, and only a small amount of endovascular thrombotic material was removed. The two false-positive cases involved the left pulmonary artery. In both cases, wall thickening found with MDCT started at the concavity of the left pulmonary artery arch. We presume that this area is technically difficult to access.
A limitation of this study was that additional information available with MDCT, such as degree of collateral systemic artery development [9] and perfusion data based on areas of hypoattenuation on parenchymal windows [10], was not assessed in this study. This information may have prognostic value, especially in association with the data described. In addition, unlike other investigators [11], we did not compare our findings with hemodynamic changes after surgery.
We did not compare our results with those of conventional pulmonary angiography because we were considering pulmonary wall changes not directly visible with angiography. Nevertheless, associating angiographic and MDCT data may be of value in decision making about operability because subsegmental evaluation is difficult with MDCT data alone. Some very experienced centers rely only on conventional arteriography rather than CT.
Five of the 45 consecutively examined patients did not undergo surgery, partly because MDCT showed no substantial endovascular thrombotic material in the proximal part of the pulmonary arteries, making complete evaluation impossible for this study. A last limitation was that some arteries were found difficult to assess with MDCT. Some parietal forms of chronic thromboembolic hypertension are associated with subtle thickening of the pulmonary artery wall, necessitating higher spatial resolution. The improved spatial resolution of the new generation of CT scanners may be helpful in these difficult cases.
In conclusion, MDCT is a useful technique for prediction of surgical operability in cases of chronic thromboembolic pulmonary hypertension because it accurately depicts an endarterectomy plane. At our center, MDCT is used as an additional tool for selection of patients who would most benefit from thromboendarterectomy.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  E. Castaner, X. Gallardo, E. Ballesteros, M. Andreu, Y. Pallardo, J. M. Mata, and L. Riera CT Diagnosis of Chronic Pulmonary Thromboembolism RadioGraphics, January 1, 2009; 29(1): 31 - 50. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
