| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Cardiac Imaging |
1 From the Department of Diagnostic Radiology, Chonnam National University Medical School, Gwangju, Korea (Y.H.K.); Department of Radiology, University of Texas M.D. Anderson Cancer Center, 1515 Holcombe Blvd, Box 57, Houston, TX 77030 (E.M.M.); Departments of Biostatistics and Bioinformatics (J.E.H.) and Radiology (H.P.M.), Duke University Medical Center, Durham, NC. Received December 30, 2003; revision requested March 2, 2004; revision received June 16; accepted July 20. Address correspondence to E.M.M. (e-mail: emarom@di.mdacc.tmc.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Institutional review board waived patient consent requirement and approved the study. Thin-section contrast material–enhanced spiral chest CT scans in 104 patients, 68 women and 36 men (age range, 19–86 years; mean, 49 years) referred to exclude pulmonary embolism, were retrospectively reviewed. Short-axis diameter and cross-sectional area of the four major pulmonary veins (right inferior and superior, left inferior and superior) were measured at a workstation by using oblique reconstructions. Each vein was measured at six locations, 5 mm apart, starting at atrial ostium. Each measurement was performed three times by an experienced thoracic radiologist, and the mean value was recorded. Roundness was estimated by comparing the ratio of the calculated cross-sectional area to that measured. Mixed effects model was used to compare men and women relative to the distribution of diameters and surface areas and to compare roundness of the right and left veins.
RESULTS: Mean pulmonary vein diameters at the ostia were variable: right superior, 11.4–12.4 mm; left superior, 9.6–10.5 mm; right inferior, 12.3–13.1 mm; and left inferior, 9.0–9.9 mm. Diameter and cross-sectional area of the left superior pulmonary vein were significantly larger in men than in women (P < .005). As expected, the caliber of three of the four veins gradually increased as they approached the left atrium. Caliber of the left inferior pulmonary vein decreased as it entered the left atrium. None of the veins were round; all were ovoid. Left-sided veins and venous ostia were less round than right-sided veins (P < .001).
CONCLUSION: Pulmonary vein diameter, cross-sectional area, and shape vary. Particular care must be taken when the left inferior pulmonary vein is evaluated for stenosis, as it normally narrows as it enters the left atrium.
© RSNA, 2005
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Imaging and Image Review
All CT scans were obtained with a scanner (QXi LightSpeed; GE Medical Systems, Milwaukee, Wis). Images were obtained during a single breath hold at full inspiration, from the top of the left diaphragm to the top of the aortic arch, with the following parameters: collimation, 2.5 mm; table speed, 15 mm per rotation; and rotation, 0.8 second. Images were reconstructed at 1.0-mm intervals. A total of 150 mL of noniodinated contrast material was administered by means of a power injector (Percupump II; E-Z-Em, Westbury, NY) at a rate of 4 mL/sec through an 18- or 20-gauge catheter in an antecubital vein; scanning commenced after a 20–25-second delay. An experienced (6 years) thoracic radiologist (Y.H.K.) retrospectively reviewed the studies at a freestanding workstation (Vitrea 2; Vital Images, Minneapolis, Minn).
Pulmonary vein measurements were obtained as follows. First, oblique transverse, oblique sagittal, and oblique coronal images (with reformatted data) were reconstructed for each pulmonary vein along the long axis of the vein. This was done by drawing a line along the center of the vein as viewed on the transverse image and reformatting the images along this line. This was the long-axis view of the vein. Then, the short-axis view for each vein was produced by creating a plane perpendicular to the long-axis view. This allowed all diameter and cross-sectional area measurements to be obtained in a plane perpendicular to the long axis of the vein (Fig 1). Reformation of the views along the long axis of the pulmonary veins also enabled better visualization of the exact location of the venous ostia by using the reformatted transverse and coronal views. Four veins were analyzed in each subject: the right superior pulmonary vein, the right inferior pulmonary vein, the left superior pulmonary vein, and the left inferior pulmonary vein. All measurements were performed by using window width and window level settings of 600 and 60 HU, respectively. The diameters and cross-sectional areas of the pulmonary veins were measured at six locations at 5-mm intervals, starting from the pulmonary venous ostium (level 1) to a point 25 mm beyond the ostium (level 6). Each measurement was performed three times, and the mean values were recorded.
|
|
|
· r2. Then the roundness was estimated by dividing the estimated cross-sectional area by the measured cross-sectional area. If the veins were perfectly round in cross section, this ratio was 1.0. Ratios less than 1.0 indicated a progressively more oval shape.
Statistical Analysis
The mean value, standard deviation, and 95% confidence intervals for each diameter, cross-sectional area, and roundness measurement were calculated. For each of the four veins, a mixed effects model that accounted for correlation of measurements within a patient was used to compare men and women relative to the vein diameter and surface area. A two-sample t test was used to compare men and women relative to age. The mixed effects model was used to compare the roundness of the right veins with that of the left veins, with comparisons of the right veins with the left superior pulmonary veins and of the right veins with the left inferior pulmonary veins. A paired t test was used to compare the mean diameters and surface areas of the proximal 15 mm to the distal 10 mm of the left inferior pulmonary vein. A software package (SAS, version 8; SAS, Cary, NC) was used to perform the statistical analysis.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
Our measurements stratified according to sex are shown in Table 2. The mixed effects model showed that the diameter of the left superior pulmonary vein was larger for men than it was for women throughout the vein (P < .006). Similarly, the cross-sectional area of the left superior pulmonary vein was significantly greater for men than it was for women (P < .003).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Several aspects of this work deserve further comment. First, because ablation is currently achieved with guided perimetric mapping of the pulmonary veins by using a multipolar loop catheter (Lasso; Biosense Webster, Diamond Bar, Calif), the size and shape of the pulmonary venous ostium is an important consideration in selection of the optimal multipolar loop catheter diameter. For example, an oval ostium may affect the position and stability of a circular multipolar loop catheter because it may tend to orient itself in a canted position and cause a misleading activation sequence (10). Furthermore, there is evidence that arrhythmogenic foci are more likely to occur in abnormally large venous ostia (11). Therefore, pulmonary venography is frequently performed prior to ablation to determine the shape, position, and size of the venous ostia (10). Assessment of these features at venography, however, may be suboptimal because of variable magnification and because it may not be possible to visualize each pulmonary vein in the optimal transverse (perpendicular to the vascular lumen) plane. This concern may be particularly apropos in regard to determination of the shape of the orifice.
We (12), and others (7,13,14), have previously shown that cross-sectional imaging techniques such as CT or magnetic resonance (MR) imaging can be useful for assessment of the number, position, and size of the venous ostia prior to ablation. Our current data show that the pulmonary veins, particularly the left-sided veins, are almost never perfectly round. In fact, the ostia of the left-sided veins may be quite elongated in the craniocaudal dimension. These data are in accord with the findings of a recent study in which MR imaging was used (10) to show that the venous ostia may sometimes be quite narrow despite a normal appearance at venography. These data suggest that measurements obtained at venography may cause either an overestimation or an underestimation of the cross-sectional area of the venous ostia (particularly those that are more oval) and that assessment of pulmonary vein size, shape, and cross-sectional area may be better evaluated by using CT (or MR imaging) than by using planar venography.
Second, because of the normal variability in the size and shape of the pulmonary veins, the diagnosis of postablation stenosis may be difficult without either a reference examination for comparison or a consultation with a table of normal values, such as those we have developed. Our study findings confirm that the veins generally increase in caliber as they approach the left atrium, with the notable exception of the left inferior pulmonary vein. This vein normally narrows as it approaches and enters the atrium. Our assumption is that this feature is possibly caused by compression of the proximal portion of the vein between the aorta and left atrium as the vein enters the atrium. Furthermore, the ostium of this vein is typically more ovoid than most, with the long axis in the craniocaudal direction. This anatomic narrowing of the proximal left inferior pulmonary vein should be borne in mind when CT scans are interpreted for possible venous stenosis. In cases where stenosis is suspected, other supporting evidence at CT, such as soft-tissue infiltration and secondary changes in the affected lobe of the lung, should be sought (9).
As previously noted, preablation imaging often is performed to assess the size, shape, location, and number of the pulmonary veins. In addition to venography, potential imaging modalities include echocardiography, CT, or MR imaging. Transesophagal echocardiography can be used to assess the superior pulmonary veins. It is often difficult or impossible, however, to completely assess the inferior pulmonary veins with echocardiography (15). MR imaging has been shown to be superior to echocardiography for preablation assessment of the pulmonary veins and is probably better than planar venography (16). MR imaging studies, however, can be difficult to perform because of the time required for imaging, artifacts caused by arrhythmia, and patient claustrophobia. On the other hand, almost all patients easily tolerate CT, with scanning times with the latest generation of multi–detector row CT scanners usually less than 10 seconds. By using a workstation, the often complex pulmonary venous anatomy can be easily deciphered with multiplanar reconstructions and displayed for review by referring clinicians. The pulmonary veins also can be displayed and accurately measured in the most appropriate plane. By obtaining this detailed information prior to ablation, the overall procedure time and thus patient radiation exposure may be substantially decreased (5,17,18). It is also possible that more accurate preablation assessment of the size and shape of the ostium may improve the success of the procedure.
Although our study was not designed to evaluate the optimal CT technique for visualization of the pulmonary veins, our experience suggests that optimal visualization requires fairly thin sections, 2.5 mm or less, and adequate intravenous contrast material enhancement. Contrast material is not absolutely necessary, however, and can be eliminated for purposes of pretreatment planning in patients with severe allergies. Contrast material is likely of greater importance when assessment for possible pulmonary vein stenosis after ablation is required. Because the soft tissues adjacent to the stenotic pulmonary vein are often inflamed and infiltrated, the severity of stenosis may not be accurately assessed without contrast material in the lumen of the vein (9).
Our study had some limitations. Because we primarily evaluated patients who were undergoing CT to exclude pulmonary embolism, our results might not be generalized to the whole population. There is no reason to believe that pulmonary venous anatomy should differ significantly between patients who are suspected of having a pulmonary embolism and members of the general population. However, patients with chronic atrial fibrillation and an enlarged left atrium may have larger pulmonary veins than do patients with paroxysmal atrial fibrillation and a normal-sized atrium (19). Thus, our results in regard to size and cross-sectional area may not be generalized to all patients with atrial fibrillation, particularly those with enlarged left atria.
The choice of window width and window level settings for display of the CT angiographic data can affect measured vein diameter and cross-sectional area. We chose settings that were a compromise between standard lung and mediastinal window settings. While one might quibble about this choice, it should be noted that there is substantial disagreement in the literature regarding the optimal window width and window level settings for display of CT angiographic data (20–28). Nevertheless, it is clear that our data may not apply to interpretation of CT data sets displayed at different window width and window level settings.
We had no diagnostic standard for comparison with our CT data. Postmortem examination is arguably the best standard for this type of study, but such data were not available. Venography, as we have noted, is probably not an optimal diagnostic standard because of its limited ability to image each vein in the transverse plane. Furthermore, pulmonary venography was not performed in any of our patients. Moreover, because the anatomy is often complex, cross-sectional imaging techniques such as CT or MR imaging combined with multiplanar or three-dimensional reconstruction capabilities should best depict pulmonary venous anatomy. All of our scans were obtained at end inspiration. Because the caliber of the pulmonary veins may change during respiration, our measurements will likely not apply if the scan is obtained during expiration.
Our study included only patients with four distinct pulmonary veins ("textbook" anatomy). We thus excluded patients with conjoined veins (eg, single left-sided venous ostium), which are uncommon (12). Our results will clearly not apply to patients with conjoined veins; whether the presence of a conjoined venous ostium on one side affects vein size on the contralateral side is unknown. We did not exclude, however, patients with accessory veins (eg, separate right middle lobe vein draining independently into the left atrium). Such accessory veins are usually small and often drain only one pulmonary segment. We do not expect that the presence of an accessory vein would significantly affect the measured size of the dominant pulmonary veins.
Finally, we identified a statistically significant difference between men and women in regard to only one of the pulmonary vein sizes. It is possible that, had we corrected for body surface area, differences might have become apparent in additional veins. Unfortunately, such data were not available.
In conclusion, we reported the mean diameters and cross-sectional areas of the pulmonary veins in a large group of apparently normal patients, as imaged at CT. These data should facilitate diagnosis of pulmonary vein stenosis in patients treated with radiofrequency ablation for atrial fibrillation. Furthermore, we found that the pulmonary veins, particularly the left-sided veins, and ostia were typically oval and not round. This suggests that planar pulmonary venography may not be an optimal technique for determining the size of the multipolar loop catheter. CT or MR imaging with multiplanar reformatted images should better depict the cross-sectional area and profile of the venous ostium. We also noted that the left inferior pulmonary vein normally narrows as it enters the atrium; thus, particular care must be taken when this vein is evaluated for stenosis at CT.
|
FOOTNOTES |
|---|
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, H.P.M., E.M.M., Y.H.K.; study concepts, H.P.M., E.M.M.; study design, E.M.M., J.E.H.; literature research, E.M.M.; clinical studies, E.M.M., Y.H.K.; experimental studies, Y.H.K.; data acquisition, Y.H.K.; data analysis/interpretation, all authors; statistical analysis, J.E.H.; manuscript preparation, Y.H.K., E.M.M., H.P.M.; manuscript definition of intellectual content, E.M.M., H.P.M.; manuscript editing, H.P.M.; manuscript revision/review, H.P.M., J.E.H., E.M.M.; manuscript final version approval, H.P.M., E.M.M., Y.H.K.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. J. Francois, D. Tuite, V. Deshpande, R. Jerecic, P. Weale, and J. C. Carr Pulmonary Vein Imaging with Unenhanced Three-dimensional Balanced Steady-State Free Precession MR Angiography: Initial Clinical Evaluation Radiology, March 1, 2009; 250(3): 932 - 939. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Schroeder, S. Achenbach, F. Bengel, C. Burgstahler, F. Cademartiri, P. de Feyter, R. George, P. Kaufmann, A. F. Kopp, J. Knuuti, et al. Cardiac computed tomography: indications, applications, limitations, and training requirements: Report of a Writing Group deployed by the Working Group Nuclear Cardiology and Cardiac CT of the European Society of Cardiology and the European Council of Nuclear Cardiology Eur. Heart J., February 2, 2008; 29(4): 531 - 556. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Rademaker, F. Saremi, and S. Krishnan Invited Commentary * Authors' Response RadioGraphics, November 1, 2007; 27(6): 1566 - 1567. [Full Text] [PDF]
|
|
|
|
 |
|
 P. Ou, D. S Celermajer, G. Calcagni, F. Brunelle, D. Bonnet, and D. Sidi Three-dimensional CT scanning: a new diagnostic modality in congenital heart disease Heart, August 1, 2007; 93(8): 908 - 913. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. D. Bahnson Clinician's Commentary Radiology, April 1, 2005; 235(1): 49 - 50. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
