|
|
||||||||
Eur J Cardiothorac Surg 2005;27:988-993
© 2005 Elsevier Science NL
a Department of Cardiac Surgery, University of Heidelberg, Im Neuenheimer Feld 110, 69120 Heidelberg, Germany
b Department of Medicine and Biology Informatics, DKFZ (German Cancer Research Centre), Im Neuenheimer Feld 280, 69120 Heidelberg, Germany
c Department of Anaesthesiology, University of Heidelberg, Im Neuenheimer Feld 110, 69120 Heidelberg, Germany
Received 15 November 2004; received in revised form 9 January 2005; accepted 10 January 2005.
* Corresponding author. Tel.: +49 6221 566395; fax:+49 6221 5655895. (E-mail: r.de.simone{at}urz.uni-heidelberg.de).
| Abstract |
|---|
|
|
|---|
Key Words: Right ventricle Ejection fraction Monitoring Transesophageal 3D echocardiography Pulmonary artery catheter Thermodilution Cardiac surgery
| 1. Introduction |
|---|
|
|
|---|
Up to now right ventricular function can be assessed by different methods: biplane angiography [4], radionuclide techniques [5], echocardiography [6,7], magnetic resonance imaging [8] and thermodilution [9,10].
However, only two techniques, like echocardiography and thermodilution can be applied to monitor right ventricular and global heart function during cardiac surgery and at the bedside. Aim of this study was to evaluate these two techniques for intraoperative assessment of right ventricle based on transesophageal three-dimensional echocardiography [11] and to compare these measurements to volumetric thermodilution obtained by using a new generation of fast response thermistor pulmonary artery catheters [912].
| 2. Methods |
|---|
|
|
|---|
2.2. Transesophageal echocardiography
Two-dimensional echocardiographic data were obtained using a 5.0MHz multiplane transesophageal echo-probe (Model 21369 A, connected to SONOS 5500 Imaging System, Philips, Andover, Massachusetts, USA). The probe was advanced to mid-oesophageal level to visualize a modified transversal four-chamber view, which entirely visualizes right ventricle. At each point of measurement, three cardiac cycles were recorded for later evaluation. Multiplane echocardiographic data were processed by the EchoAnalyzer®, a multitask system developed at our institution for three-dimensional reconstruction and measurements [1315].
Right and left ventricular volumes were calculated by using the Coons-patches technique [16] for segmentation of heart cavities, which was implemented into the EchoAnalyzer®. The proposed method enables to quickly extend segmentations performed on intersecting slices of arbitrary orientation to a three-dimensional surface model by means of interpolation with specialized Coons-patches (Fig. 1). The described approach considerably reduces time required for user interaction to obtain the object's shape and volume.
|
2.3. Volumetric thermodilution
Preoperatively, an arterial catheter was placed in the left radial artery, with pressure transducers being positioned at mid-thoracic level. Additionally, we introduced the new pulmonary artery catheter (7.5F CEDV-PAC, model 744H, Baxter Healthcare Corporation, Irvine, CA, USA), which interfaced with a computerized monitoring system (VigilanceTM CCO/SvO2/CEDV Monitor, Baxter Healthcare Corporation, Irvine, CA, USA) and an additional ECG monitor in order to assess RR-intervals. With this new pulmonary artery catheter, modified pulsed warm thermodilution technique is clinically applicable: a computer-controlled thermal filament, mounted on the catheter, is placed in the right ventricle and randomly transfers boli of heat (limited to 7.5W and 44.0°C) to the blood. The resulting changes in blood temperature are recorded with a distally mounted fast response thermistor. Based on simultaneously assessed RR-intervals, diastolic plateaus of the thermodilution curve are identified. This is necessary as the total end-diastolic heat mass in the right ventricle is the result of the heat mass at the end of the previous systole and of cooling effects due to the influx of colder blood during diastolic filling. These relations are summarised in the equation:
Where:
RVEF can be calculated having measured blood temperature before heat application and the decrease in blood temperature during a predefined period (T1 to T2).
Significant tricuspid valve regurgitation and intracardiac shunt, which may affect thermodilution measurements were ruled out by transesophageal echocardiography.
2.4. Study protocol
Measurements were carried out at three time points: before skin incision, 30min after weaning from cardio-pulmonary bypass, and postoperatively 2h after the admission to the intensive care unit. Next to right ventricular ejection fraction, heart rate (HR), central venous pressure (CVP), mean arterial blood pressure (MAP), pulmonary capillary wedge pressure (PCWP), mean pulmonary artery pressure (MPAP) and pulmonary vascular resistance (PVR) were measured.
2.5. Statistics
All results are presented as mean±SD, and P values <0.01 were considered statistically significant. Linear regression analysis was used to describe the correlations between different methods for assessing right ventricular volumes and function. Differences between groups with and without impaired cardiac function were assessed by Student's t test for unpaired data. The method of Bland and Altman [17] was used for the agreement of different diagnostic methods. Differences were considered statistically significant at a value of P<0.01.
| 3. Results |
|---|
|
|
|---|
|
|
|
30% were assigned to the group with impaired function. Eleven patients with normal left ventricular function had an ejection fraction >30%. The group of patients with impaired function showed significantly lower right ventricular ejection fraction (44.1±4.6 vs. 55.1±3.9%; P<0.01) than the group of patients with normal function (Fig. 3).
|
| 4. Discussion |
|---|
|
|
|---|
Due to its geometrical complexity, the assessment of right ventricle, is still today a very difficult task. Whereas left ventricular cavity approximates to an ellipsoid model the main body of right ventricle is crescentic with separate inflow and outflow portions. Since right ventricle cannot be represented by any geometric model, its volume can poorly be calculated by any approximation.
In this study we demonstrated that intraoperative assessment of right ventricular volume by means of thermodilution and three-dimensional echocardiography is feasible and that these techniques are able to provide a useful estimation of right ventricular function, which is promptly available in the operating room.
The lack of correlation between right ventricular volumes obtained by thermodilution and those obtained by three-dimensional echocardiography does not surprise, since it is fundamentally due to the fact that thermodilution can only estimate cardiac volumes by approximation. This is confirmed by the extremely high values of end-diastolic and end-systolic volumes measured by thermodilution technique.
Despite the lack of correlation between right ventricular volumes, the presence of significant agreement between the right ventricular ejection fraction shows that these two methods can be reliably used for serial measuring of right ventricular function during heart surgery, which makes this valuable information available for surgical decision making.
In order to assess the differences of right ventricular ejection fraction in patients with impaired and those with preserved left cardiac function, we classified the patients into two groups as defined by angiographic measurements of left ventricular ejection fraction. Three-dimensional echocardiography revealed that in the group of patients with impaired left function, right ventricular ejection fraction also was lower than in those with preserved left function (Fig. 3). Serial intraoperative assessments of right ventricular function are needed to investigate the prognostic value of this parameter and its impact on surgical decision making in the different surgical procedures.
It is now generally accepted that magnetic resonance imaging is the gold standard technique for non-invasive estimation of ventricular volumes, however, its availability is limited by the induction of strong magnetic fields that prevent its application in the operating room. On the contrary, transesophageal echocardiography is now routinely available in the operating room for intraoperative preoperative assessment of the heart surgery [22]. The interpretation of our results is based on the finding that left ventricular volumes by three-dimensional echocardiography showed good correlations to those measured by magnetic resonance imaging (Table 2) and on the assumption that the assessment of right ventricular volumes by three-dimensional echocardiography showed the same good correlation. In fact, we assumed that the right ventricular volumes measured by three-dimensional echocardiography was the reference method to compare volumetric thermodilution to. As a matter of fact, the lack of a gold standard for direct comparison of right ventricular volumes to magnetic resonance imaging is the main limitation of this study. The postulation that three-dimensional echocardiography is able to measure right ventricle as accurately as left ventricle, even if very suggestive, is based on indirect evidence (Table 2). The statistical analysis shows a significant correlation and agreement, but in reality the fact that thermodilution overestimates right ventricular volumes is rather based on the comparison to findings from the literature [79], and on the anatomical knowledge that enddiastolic volume of right ventricle is closer to the value measured by three-dimensional echocardiography (135.1±23.5ml) than that obtained by thermodilution (195.0±35ml).
A direct comparison of three-dimensional echocardiography to magnetic resonance imaging was beyond the design of this study, which main purpose was to compare intra- and peri-operative techniques for measuring right ventricular function, whereas magnetic resonance imaging cannot be applied in the operating theatre.
Another factor that may affect the assessment of right and left ventricular function is the influence of loading conditions. In the present study each pair of measurements (echocardiography and thermodilution) were obtained simultaneously, under the same loading conditions.
An important innovative approach of this study was the application of the Coons-patches technique which allowed the quick three-dimensional reconstruction of transesophageal echocardiographic data. The Coons-patches technique is a method primarily developed for Computer-Aided Design at the Massachusetts Institute of Technology [16]. This method has been modified and integrated in our system EchoAnalyzer, a multitask developed for analyzing three-dimensional data at our institution [1315]. The fundamental advantages of this technique are the ability of interactive segmentation which may lead to improvements by additional slices; the segmentation on non-parallel slices avoids inconsistencies; the interpolation with Coons-patches requires a processing time shorter than 1s; a significant reduction of number of slices can be achieved by this technique (Fig. 4). Except for very rare cases, automatic methods cannot guarantee to provide the correct segmentation, i.e. the identification of endocardial borders. Display and processing two-dimensional data (slices) are more precise than in three-dimensional datasets. One of the major issue of all three-dimensional methods of obtaining volumes is the inconsistence between the different slices obtained by tracing endocardial contours manually. This limitation has been overcome by the Coons-patches technique, since the use of contours in more than one orientation helped to avoid inconsistencies between neighbouring slices. In addition, this technique enables to rebuild and improve an existing three-dimensional volume by adding new additional slices (Fig. 4). The proposed method enabled to quickly extend segmentations performed on intersecting slices of arbitrary orientation to a three-dimensional surface model by means of interpolation with specialized Coons patches. In summary, the described approach considerably reduced the time required for user interaction to obtain an accurate approximation of the complex right ventricular shape. The concept of segmenting in two-dimensional, non-parallel slices therefore has the potential of a wide user acceptance. One of the major advantage of the Coons-patches technique is the ability of obtaining ventricular volume and ejection fraction by a very fast processing time. The time required by this technique is almost exclusively due to the acquisition of a complete three-dimensional dataset; in the next future this acquisition will be further fastened by the advent of realtime three-dimensional echocardiography, which will reduce the acquisition to the time needed to obtain one single heart beat.
|
| 5. Conclusion |
|---|
|
|
|---|
| Appendix A. Conference discussion |
|---|
|
|
|---|
Dr De Simone: In patients with impaired right ventricular function the dilatation of tricuspid annulus causes tricuspid regurgitation. I think that this does not influence the measurement of right ventricular volume.
I would like to add that we are trying to approach the quantification of tricuspid regurgitation in another way: since we are able to measure the flow in the pulmonary artery, as we can do with another method developed at our institution, we can measure the regurgitation as the difference of two stroke volumes. We are confident that with this method we can also achieve a more exact quantification of tricuspid regurgitation.
Dr A. Wechsler (Philadelphia, PA, USA): First, I wanted to be sure I understood. I saw you compared MRI versus the 3-D echo for the left ventricle. I didn't see it for the right ventricle. So I'm assuming that you're defining the gold standard as the echo. I don't know what you compared it to.
Second quick question is, is the timing predictable enough that with the Millar catheter in the right ventricle you could construct pressure volume loops if you were interested in studying contractility?
Dr De Simone: The second question first: of course, we are interested in studying volume pressures loops. We still need to synchronize the pressure signals to the volume measurements. I think that this technique has a good potential in order to make this.
Then, the first question: the comparison that we did with thermodilution is not possible since thermodilution is not the gold standard for measuring volumes. Of course, we assumed that on the basis of our previous validation on left ventricle volumes with MRI. And I also pointed out that, for example, one end-diastolic volume of 190ml, as measured by thermodilution, is much too high to be comfortable with.
| Acknowledgments |
|---|
| Footnotes |
|---|
Presented at the joint 18th Annual Meeting of the European Association for Cardio-thoracic Surgery and the 12th Annual Meeting of the European Society of Thoracic Surgeons, Leipzig, Germany, September 1215, 2004. | References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
T. J. Pegg, J. B. Selvanayagam, T. D. Karamitsos, R. J. Arnold, J. M. Francis, S. Neubauer, and D. P. Taggart Effects of Off-Pump Versus On-Pump Coronary Artery Bypass Grafting on Early and Late Right Ventricular Function Circulation, April 29, 2008; 117(17): 2202 - 2210. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. De Simone, I. Wolf, S. Mottl-Link, R. Hoda, B. Mikhail, F.-U. Sack, H.-P. Meinzer, and S. Hagl A clinical study of annular geometry and dynamics in patients with ischemic mitral regurgitation: new insights into asymmetrical ring annuloplasty Eur. J. Cardiothorac. Surg., March 1, 2006; 29(3): 355 - 361. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |