| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2004;78:658-665
© 2004 The Society of Thoracic Surgeons
a Departments of Cardiothoracic and Vascular Surgery, Cardiothoracic Centre, All India Institute of Medical Sciences, New Delhi, India
b Department of Cardiology, Cardiothoracic Centre, All India Institute of Medical Sciences, New Delhi, India
c Department of Nuclear Medicine, Cardiothoracic Centre, All India Institute of Medical Sciences, New Delhi, India
Accepted for publication January 9, 2004.
* Address reprint requests to Dr Chowdhury, Department of Cardiothoracic and Vascular Surgery, All India Institute of Medical Sciences, Ansari Nagar, New Delhi-110029, India
e-mail: ujjwalchow{at}rediffmail.com
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and methodsResultsCommentAcknowledgmentsReferences |
|---|
METHODS: One hundred five patients (105), aged 11 months to 35 years old, who had undergone univentricular repair, underwent first-pass and multigated acquisition scan 1 month to 10 years after univentricular repair. Forty-five patients with evidence of Fontan failure underwent radionuclide studies using Technetium-99 m as well as cardiac catheterization (group 1). The remaining sixty randomly selected patients with excellent functional status received radionuclide studies alone (group 2). The receiver operating characteristic curve analysis was done to quantify the diagnostic accuracy of the first-pass study.
RESULTS: There was paradoxical filling of the right lung after femoral injection in all cases of tunnel or conduit obstruction. A first-pass transit time of 16 to 25 seconds (mean ± standard deviation [SD] = 18.82 ± 2.69) was always associated with Fontan failure and high right atrial pressure (range = 20 to 24 mm Hg, mean ± SD = 22.02 ± 1.58). A first-pass transit time of 16 seconds was associated with a sensitivity of 100% and a specificity of 93.33%. The predictive accuracy of a positive or negative result was 91.8% and 100% respectively. The area measured under the receiver operating characteristic curve indicates that 99.41% (SE ± 0.0035) of the time, the value of first-pass time is higher for the Fontan failure group (group 1) compared to the normal group (group 2; p = 0.000).
CONCLUSIONS: Our data indicate that Fontan circuit can be reliably evaluated for both anatomic and functional flaws by radionuclide studies; radionuclide first-pass time may be used to predict the chances of Fontan failure postoperatively as well as its presence; and in the presence of atrial fibrillation with fast ventricular rate, analysis using first-pass radionuclide may be impossible and gated equilibrium radionuclide angiocardiography may be the preferred method. Inspection of the systemic ventricular time-activity curve is of crucial importance in this regard.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentAcknowledgmentsReferences |
|---|
|
Patients and methods |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentAcknowledgmentsReferences |
|---|
12 mm Hg, systemic ventricular ejection fraction 
0.35). Subsequently, these patients were grouped according to Texas Heart Institute Fontan risk score [4]. Those with 0 to 3 risk points were considered to be at low risk for a Fontan operation (n = 60). Patients with 4 to 5 points were at moderate risk (n = 45). Patients with 6 or more risk points, or 3 points in any single category of Texas Heart Institute Fontan risk score received a bidirectional Glenn as an interim palliation [2, 3].
Patient characteristics
Between January 1992 and June 2002, 105 of 450 patients undergoing univentricular repair at All India Institute of Medical Sciences in New Delhi, India, underwent a first-pass and multigated acquisition scan (MUGA) 1 month to 10 years after a Fontan procedure. The patients were allocated to one of two groups:
Group 1 patients (n = 45) were between 11 months and 29 years old (mean ± standard deviation [SD] = 83.52 ± 76.76 months). These survivors had evidence of Fontan failure, such as persistent low cardiac output, cyanosis, hepatomegaly, and "significant pleural effusion." Radionuclide studies were requested to evaluate noninvasively the postoperative Fontan hemodynamics. Cardiac catheterization and angiocardiography were subsequently performed in all these cases. The results of radionuclide studies were compared with those of hemodynamic and angiographic evaluation.
Group 2 patients (n = 60) included survivors of the Fontan procedure more than 1 year after repair without symptoms of congestive heart failure, significant atrioventricular valve regurgitation (> moderate degree), significant arrhythmias (Lown grade >2), and significant resting ventricular outflow obstruction (10 mm Hg).
These patients were between 11 months and 35 years old (mean ± SD = 74.80 ± 78.25 months). There was no selection bias within the group because all patients were randomly selected. The data were analyzed independently without the knowledge of clinical characteristics of the patients. These patients were in excellent functional status postoperatively (New York Heart Association [NYHA] class I or II) and received radionuclide studies only.
Descriptive characteristics and underlying cardiac diagnoses are summarized in Table 1. Tricuspid atresia accounted for 38% of diagnoses (40 of 105 patients). Other diagnoses occurred in varying frequency. For the purpose of statistical analysis, patients with diagnosis other than tricuspid atresia were grouped together. Twenty-one percent (21%) of patients had associated atrial isomerism, and 23.8% had anomalous systemic venous connection; 37.1% of patients (39 of 105) had had previous palliative procedures. All patients underwent preoperative cross-sectional echocardiogram, cardiac catheterization, and angiocardiogram.
|
There were significant differences between groups in regard to the date of operation (p < 0.0001), cardiac morphology (tricuspid atresia vs nontricuspid atresia p < 0.01), right atrial isomerism (p = 0.001), mean pulmonary artery pressure more than 15 mm Hg (p < 0.001), systemic ventricular end-diastolic pressure 12 mm Hg (p < 0.001), systemic ventricular ejection fraction (p = 0.000), pulmonary vascular resistance (p < 0.001), systemic ventricular hypertrophy (p < 0.001), prior pulmonary artery banding (p < 0.01), concomitant pulmonary arterioplasty (p = 0.05), and fenestrated versus nonfenestrated total cavopulmonary connection (p < 0.0001; Table 1).
A combination of both first-pass and multigated acquisition scans were used to reveal the details regarding ventricular function, residual right-to-left shunt, and systemic or pulmonary venous pathway obstruction. Fifty patients received radionuclide study twice: first within 2 weeks of surgery, and then after 6 months.
Patients with total cavopulmonary connection (TCPC) required two injections. The dynamic studies of Glenn anastomosis was obtained by right antecubital vein injection of technetium 99m pertechnetate and a lower-limb vein injection of technetium 99m pertechenetate was necessary to evaluate the inferior vena cava-pulmonary artery connection (intra or extra cardiac).
All patients with bilateral superior venae cavae (n = 20) underwent bilateral bidirectional Glenn as part of the Fontan circuit. The radiotracer was injected rapidly in the right antecubital vein as a very tight bolus, ie, less than 2 seconds in duration. It is more important that the radionuclide be injected smoothly and that it arrives in the central circulation on a single front.
Surgical procedures
Two basic modifications of the univentricular repair were utilized and varied according to the preoperative anatomy. Modifications included creation of an intraatrial lateral tunnel made of polytetrafluoroethylene patch in 93 patients and an extracardiac polytetrafluoroethylene conduit in 12 patients. Fenestration was done using a 4-mm or 5-mm coronary punch in the Gore-Tex baffle (W.L. Gore and Associates, Flagstaff, AZ) [2, 3]. The details of the operative and additional prcedures are outlined in Table 1. Cardiopulmonary bypass and myocardial preservation techniques were the same over the entire study period.
Radionuclide techniques
First-pass study
First-pass radionuclide angiography was performed with the patient in supine position using a single-crystal gamma scintillation camera equipped with a parallel hole low energy all purpose collimator. Technetium 99m pertechnetate (99 mTcO4–) was used as radiotracer at a dose of 0.2 mCi (7.4 MBq)/kg body weight in 0.2-mL to 0.5-mL volume. This was injected rapidly as a bolus through a large bore needle in a right antecubital vein and in a femoral vein on two different occasions.
Quality control of the injection technique was achieved quantitatively by analysis of a time-activity curve over the superior caval vein or by visual inspection of serial images of the cardiac flow activity. A broken bolus was detected as the arrival of two waves of radioactivity in the central circulation.
Images in the anterior projection were obtained and stored at 20 frames/second (0.05 second/frame) using a dedicated online digital computer. The data were reformatted at a frame rate of 0.5 seconds/frame and the resulting images were viewed in real time and slow motion.
All postoperative patients with supraventricular arrhythmias (n = 7) were treated with digoxin to control the irregular heart rhythm. Processing of the first-pass study for calculation of the transit time was essentially the same for all patients. For the calculation of ejection fraction in the presence of numerous irregular beats, analysis with a first-pass study was impossible and equilibirium radionuclide angiocardiography was the preferred method.
Multiple-gated acquisition scan/equilibrium radionuclide angiocardiography
All patients underwent an electrocardiogram-gated radionuclide cardiac angiography (multiple-gated acquisition [MUGA] scan). The MUGA scan was done after in vivo red cell labeling. For in vivo labeling, stannous pyrophosphate (2 to 3mg) was first injected intravenously. After 15 to 20 minutes of first injection, technetium 99m pertechnetate (99 mTcO4–) was injected intravenously at a dose of 0.2 mCi (7.4 MBq)/kg body weight, with a minimum dose of 2 mCi (74 MBq) and a maximum dose of 20 mCi (740 MBq).
Cardiac imaging was acquired on a dual-head gamma camera (Ecam; Siemens Medical, Iselin, NJ) fitted with a low energy all purpose collimator and interfaced with a computer and electrocardiogram (ECG) gating device. A 20% gamma camera energy window was set symmetrically over the 140 KeV photon peak of technetium 99m. Equilibrium radionuclide angiocardiography (ERNA) studies were acquired on computer using a 64x64 pixel matrix. The acquisition was synchronized to the patients R-wave on the electrocardiogram and the R-R interval was divided into 24 equal frames. Images in each patient were acquired in the best septal view, which was 40 degrees left anterior oblique in most patients. Each study was acquired for approximately 5 minutes, collecting data from 300 to 400 cardiac cycles.
In order to selectively reduce or eliminate irregular heart beats, the method of online postbeat filtration was used. The initial dominant R-R intervals (15 to 30 beats) were averaged and taken as reference. A window of 10% to 15% allowable intervals was selected. If the R-R interval was within these predetermined acceptable limits, the computer proceeded to acquire the next cardiac cycle. In case the R-R interval was outside these limits, no further cycles were acquired until the R-R interval returned to the acceptable range.
Initially, each study was examined in a cine mode to note any patient movements. Studies were processed further if there was no patient's motion or other artifacts. Stroke volume, paradox, and phase images were created to help define chamber boundaries. Ejection fraction was calculated by automated standard techniques from computer by putting the regions of interest on the systemic ventricular end-diastole and end-systole. Cine-images and regional ejection fraction were also analyzed for regional wall motion abnormalities. A systemic ventricular ejection fraction of more than 50% with no abnormality in regional wall motion was considered a normal MUGA study.
Lung perfusion scan
Lung perfusion scan was performed on 39 patients with prior palliative surgical procedures (systemic-pulmonary artery shunt [n = 34], pulmonary artery band [n = 5]) and on 9 patients with Fontan pathway obstruction. With the patient supine, 2 to 4 mCi of technetium 99m macroaggregated albumin was injected slowly (more than 5 to 10 seconds) with the patient taking moderately deep breaths. Acquisition was started almost immediately and two views (anterior and posterior) were acquired for assessment of lung perfusion. In 9 patients with tunnel obstruction, in whom pulmonary embolism had to be ruled out, 8 views (anterior, posterior, both laterals, and both posterior and anterior obliques) were acquired.
Assessment of operative outcome
Functional status
The patients clinical course (survival, need for cardiac medications, occurrence of late complications) after hospital discharge were monitored. The primary outcome of interest, NYHA classification of functional status was assigned to the Fontan survivors. This outcome was then collapsed to form two groups, "good" functional status (class I or II) and "poor" functional status (class III or IV).
Echocardiograms
The following echocardiographic variables were compared before and after the operation: atrioventricular valve regurgitation, presence of absence of flow through the fenestration, and systemic ventricular function. In patients in which the presence of thrombus was questionable on transthoracic echo, a transesophageal echocardiogram was performed for confirmation.
Analysis of rhythm changes
Ambulatory electrocardiograms were obtained by 24-hour Holter studies when 12-lead electrocardiogram showed dysrhythmias, or when patients had symptoms suggestive of dysrhythmias [2, 3]. A Mortara H-12 Holter unit (Mortara Instruments Inc., Milwaukee, WI) was used for monitoring and the recordings were analyzed on a standard Mortara Holter analyzer by a senior consultant in electrophysiology [2, 3].
Statistical methods and analysis
All data were analyzed with BMDP software (Biomedical Data Processing Statistical Software; University of California Press, Berkeley, CA). Interval-related data were expressed as the mean ± standard deviation and categorical variables were expressed as percentages. Noncategorical variables were analyzed using paired t-test. A p value of less than 0.05 was considered significant.
The receiver operating characteristic (ROC) curve analysis was done to evaluate the sensitivity, the specificity and predictive accuracy of an abnormal first-pass study in predicting the chances of Fontan failure and to determine the optimal cut-off point of the transit time to predict the same [5, 6].
The ROC curve was constructed using intercooled Stata 7.0 software (Stata Corp, College Station, TX). To quantify the diagnostic accuracy of first-pass study, the area under the ROC curve was statistically analyzed [6].
|
Results |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentAcknowledgmentsReferences |
|---|
|
|
Twelve patients demonstrated pulmonary perfusion abnormalities on lung-perfusion scan. Eight of them had prior Blalock-Taussig's shunt, 2 patients had prior Waterston-Cooley's shunt, and the remaining 2 had prior pulmonary artery band. Residual narrowing was seen at the site of the repair with decreased perfusion. However, other hemodynamic and functional variables were within normal limits.
Systemic ventricular function
Postoperatively, the systemic ventricular ejection fraction of patients with good clinical results (n = 60) ranged from 45% to 60% (mean ± SD = 52.60 ± 4.60). All patients with persistent hepatomegaly and fluid retention following Fontan operation (n = 45) had low ejection fraction ranging between 20% and 25% (mean ± SD = 23.82 ± 1.54). By paired sample comparisons, the preoperative and postoperative differences of ejection fraction of both groups of patients were statistically significant (p = 0.000; Table 1).
Serial evaluation of both groups of patients at 2 weeks and 6 months did not demonstrate any deterioration of systemic ventricular ejection fraction (mean ejection fraction ± SD = 53.18 ± 4.82 and 25.12 ± 2.52, respectively). It remained either stable or improved from the postoperative baseline.
Effect of supraventricular tachyarrhythmias on first-pass study and MUGA scan
The overall incidence of postoperative supraventricular arrhythmias was 6.6% (7 of 105). It was more common among group 1 patients when compared with group 2 (13.3% vs 1.6%, p = 0.04; Table 1). All patients with atrial fibrillation were treated with digoxin to control the ventricular rate. The transit time was estimated by the first-pass study and systemic ventricular ejection fraction was estimated by MUGA scan. Processing of the first-pass study for calculation of the transit time was essentially the same for all patients. The radiotracer remained visible in the intracardiac tunnel for 16 to 24 seconds (mean ± SD = 20.7 ± 2.49 seconds) and the ventricular ejection fraction ranged between 20% and 27% (mean ± SD = 24.28 ± 2.49) in all patients with atrial fibrillation (n = 7).
Analysis of the ROC curve
A first-pass transit time of 16 to 25 seconds (mean ± SD = 18.82 ± 2.69) was always associated with Fontan failure and high right atrial pressure (mean ± SD = 22.02 ± 1.58, range 20 to 24 mm Hg). Using first-pass transit time of 16 seconds as the optimal cut-off point of Fontan failure, the sensitivity was 100% and the specificity was 93.33%. The predictive accuracy of a positive or negative result was 91.8% and 100%, respectively (Table 2; Fig 1). The results obtained from the area analysis indicate that 99.41% (standard error [SE] ± 0.0035) of the time, the value of first-pass transit time is higher for the Fontan failure group (group 1) compared with the normal group (group 2), which is highly significant (p = 0.000; Fig 1).
|
Comment |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentAcknowledgmentsReferences |
|---|
Evaluation of Fontan failure
The incidence of Fontan failure after a modified Fontan operation has declined dramatically over the years, despite the application of the Fontan operation in increasingly high-risk candidates [1–4]. Previous reports from this center have documented a 14% incidence of Fontan failure and 27% incidence of significant pleural effusions [2, 3].
The results obtained from the ROC curve analysis indicate that first-pass transit time predictably separates patients with Fontan failure and high right atrial pressure from the uncomplicated Fontan group with normal right atrial pressure (Table 2; Fig 1). The diagnosis of Fontan failure could be determined with acceptable sensitivity and specificity. A first-pass transit time of 16 to 25 seconds (mean ± SD = 18.82 ± 2.69) was always associated with Fontan failure and high right atrial pressure (range = 20 to 24 mm Hg, mean ± SD = 22.02 ± 1.58) A first-pass transit time of 16 seconds was associated with a sensitivity of 100% and a specificity of 93.33%. The predictive values of positive or negative results were 91.8% and 100%, respectively. This finding indicates that first-pass transit time may be used to predict the chances of Fontan failure postoperatively as well as its presence (Table 2; Fig 1).
The results obtained from the area analysis under ROC curve indicate that 99.41% (SE ± 0.0035) of the time, the value of first-pass transit time is higher for the Fontan failure group (group I) compared to the normal group (group 2), which is highly significant (p = 0.000; Fig 1).
Evaluation of right-to-left shunt
The first-pass examination detected persistent interatrial right-to-left shunting in all patients with a functioning fenestration. These findings were confirmed using Doppler echocardiography. The value of the radionuclide techniques for detecting these findings has been addressed by other investigators [7, 8].
Evaluation of the Fontan pathway obstruction
The reported incidence of thrombus formation after the Fontan procedure has ranged from 3% to 20% and does not appear to be related to any of the modifications of Fontan circulation [1–3, 8–11]. There does not appear to be a consensus in published reports regarding candidate selection and the optimal type or duration of anticoagulation [1–3, 8–11].
Thrombosis of the Fontan pathway is often silent and may remain undetected until pathway obstruction results in significant elevation of venous pressures [1–3, 8–11]. Radionuclide studies would be one way of serial evaluation of the Fontan patients on follow-up for early detection and treatment of the insidious complication before it becomes advanced.
In our study population, despite adequate oral anticoagulation, there were seven instances of tunnel thrombosis and two instances of neointimal peel formation. Radionuclide angiocardiography detected tunnel or conduit obstruction in all patients, in particular, by paradoxical intense visualization of both lungs after femoral injection. These obstructions allowed inferior vena caval blood to flow upwards into the superior vena cava-pulmonary artery connection through the paravertebral and epidural veins and azygous system. This may have reduced blood flow in the conduit and favored its thrombosis. Associated low cardiac output probably contributed to the event in 5 patients. Significant left atrioventricular valve regurgitation with pulmonary hypertension was the cause in 1 patient, and extracardiac prosthetic conduit thrombosis was the cause in the last patient. Four similar reports of paradoxical filling of the right lung after a lower-limb injection of microspheres have been described by other authors in patients with conduit obstruction [8–10]. Above findings were confirmed by Doppler echocardiography, magnetic resonance imaging, and angiocardiography before revision surgery.
Evaluation of pulmonary blood flow
The lung perfusion scan revealed asymmetric pulmonary perfusion in 10 of 34 patients with prior systemic-to-pulmonary artery shunts, and 2 of 5 patients with prior pulmonary artery band. Treves and colleagues [7] and Alderson and colleagues [12] reported their experience with perfusion lung scans in the evaluation of patients following repair of tetralogy of Fallot. Lung perfusion abnormalities on the side of the systemic-to-pulmonary artery shunts were common and felt to be secondary to one of the three possible causes: pulmonary vascular disease in the shunted lung, stricture or kinking at the anastomotic site, or intraoperative damage to the pulmonary artery [12].
Thrombotic and thromboembolic events are of major concern in the postoperative assessment of patients after a Fontan procedure and pulmonary embolism is a very important cause of abnormal lung perfusion after such an operation [1–3, 8–11]. In this series, although there were seven instances of tunnel thrombosis with infrahepatic extension, none had extension of the thrombus cephalad to the Glenn connection. The lung perfusion scan in these patients revealed no evidence of pulmonary embolism.
Evaluation of systemic ventricular function
In this study population, all patients with Fontan failure (group 1) had low ejection fraction preoperatively (mean ± SD = 37.24 ± 3.85, range 30% to 47%). After operation, there was significant depression of ejection fraction of both groups of patients (p = 0.000; Table 1).
Radionuclide-derived systemic ventricular ejection fraction has been demonstrated to be highly accurate, extremely reproducible, and substantially better than that of echocardiographic estimation [8, 13, 14]. Varying ventricular morphologies (right, left, indeterminate, or biventricular) in the Fontan cohort do not lend themselves well to calculation of ejection fraction by conventional angiocardiograms [8, 13, 14]. Radionuclide-derived ejection fraction is probably more accurate because it is devoid of geometrical assumptions inherent in conventional volume calculations using contrast angiocardiography. Additionally, the ability to obtain serial measurement of ejection fraction after a single isotope injection makes this modality particularly appealing [8, 13, 14].
Supraventricular tachyarrhythmias: effect on image display and calculation of systemic ventricular ejection fraction
Literature documents that supraventricular arrhythmias in Fontan patients impacts unfavorably on long-term outcome and that ventricular function is adversely affected [1–4, 11, 15]. Modifications of the Fontan procedure have been reported to diminish the frequency of such arrhythmias but have not abolished them [1–4, 11, 15].
In the presence of atrial fibrillation with fast ventricular rate, analysis using first-pass radionuclide studies is usually impossible and ERNA may be the preferred method [14, 16]. The latter technique offers several options to selectively reduce or eliminate irregular heart beats falling outside the user-specified limits [14, 16]. Inspection of the systemic ventricular time-activity curve is of crucial importance in the setting of atrial fibrillation with fast ventricular rate. Only when a predominant cycle length exists (ie, presence of a well defined trough in the time-activity curve) can a meaningful systemic ventricular ejection fraction be calculated from such data. In such patients a narrow R-R interval window would prevent averaging of cardiac cycles with varying length.
Supraventricular arrhythmias did not affect imaging transit times and ejection fraction estimation in this study population. Contrary to what is generally believed, ERNA-derived systemic ventricular ejection fraction could be calculated in most patients with atrial fibrillation. Control of the heart rate by using digoxin, use of arrhythmia filtration to eliminate irregular beats outside the user-specified limits, and selecting the systemic ventricular time-activity curve with predominant cycle-length for estimation of ejection fraction are the possible contributing factors for the findings of this study.
Limitations
Neither first-pass radionuclide angiography nor multigated radionuclide angiography can be applied to all clinical situations. There are inherent limitations of the above-mentioned testing techniques. Several technical issues are relevant to the performance of first-pass studies.
First, it is necessary to position the patient carefully during radionuclide injection, and the quality of bolus injection is important. The patient must be able to remain still beneath the detector during the period of data acquisition. The injection technique must be impeccable; it is necessary to have a compact radionuclide bolus without streaming. Injections can be made from either the jugular or the antecubital venous systems. Injections at more peripheral sites are not suitable.
Separate injections of a radionuclide in the upper and lower limbs are needed for each view of intervention to be imaged. In addition, data processing is more demanding, time consuming and operator dependent. If the patient cries or performs the Vasalva maneuver at the time of injection of radionuclide in the lower extremity veins, the inferior vena caval pressure is transiently elevated and may result in tracer passage to the right lung through the mediastinal collateral veins. This may erroneously suggest pathologic elevation of the inferior vena caval pressure due to various postoperative complications [8–10]. Finally, the interpretive accuracy of this imaging method is assured only when the postoperative venous anatomy in each patient is clarified.
Implications and recommendations
The results of the study suggest that radionuclide-determined transit time using the first-pass technique is an useful noninvasive test to predict the chances of Fontan failure postoperatively as well its presence. The technique is reproducible and optimizes several technical variables associated with echocardiographic and contrast angiocardiographic examination.
In the presence of atrial fibrillation with controlled ventricular rate, the transit time could be accurately estimated by first-pass method. For calculation of ejection fraction in the presence of numerous irregular beats, multigated acquisition scan is the preferred method.
Because radionuclide angiography can be performed serially with a high degree of reproducibility, it may be used for late postoperative assessment, obviating the need for frequent cardiac catheterization.
In addition, we propose that routine and serial utilization of this modality may be the investigation of choice to detect asymptomatic thrombotic occlusion of the systemic venous pathway.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentAcknowledgmentsReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
