Standard Orthotopic Heart Transplantation Versus Total Orthotopic Heart Transplantation
A Transesophageal Echocardiography Study of the Incidence of Left Atrial Thrombosis
Background After standard orthotopic heart transplantation (Sd HT), the enlarged resultant atria may promote atrial thrombosis. The purpose of this study was to compare the incidence of spontaneous echo contrast and left atrial thrombosis after Sd HT and total orthotopic (Tot HT) heart transplantation.
Methods and Results Transesophageal echocardiography (TEE) was performed in 75 patients with Sd HT and in 20 patients with Tot HT. Despite the use of antiplatelet therapy, an acute arterial embolism occurred in 11 (15%) of the 75 patients with Sd HT but in none of the 20 Tot HT patients. All patients were in sinus rhythm. Left ventricular ejection fraction was similar in Sd HT and Tot HT patients. Left atrial diameter was smaller in Tot HT patients than in Sd HT patients (41±4 versus 58±6 mm, P<.001). In Sd HT patients, spontaneous echo contrast was present in 43 patients (57%) and was associated with left atrial thrombus in 20 patients (on the left atrial appendage in 12 patients, on the posterior wall in 6, and on the suture in 2). No thrombus was detected by transthoracic echocardiography; all thrombi were detected by TEE. On the other hand, no left atrial thrombus was observed in Tot HT patients, and only 1 patient had spontaneous echo contrast. Of the 11 Sd HT patients who experienced an arterial embolism, 5 had both spontaneous echo contrast and left atrial thrombus and 5 had only spontaneous echo contrast.
Conclusions This study demonstrates a high rate of left atrial thrombus after Sd HT and emphasizes the role of TEE in the follow-up of these patients. The therapeutic implications are the need for a preventive anticoagulant therapy in the high-risk population receiving Sd HT diagnosed with TEE and the consideration of Tot HT as a better surgical approach as far as thrombotic complications are concerned.
Heart transplantation is the ultimate therapeutic option in the management of end-stage heart failure. “Standard” orthotopic heart transplantation is performed worldwide according to the initial technique described by Shumway et al1 and is relatively simple, requiring only two atrial anastomoses between the posterior and lateral portions of the recipient and donor atria.
However, recent transesophageal echocardiographic studies2 3 4 have pointed out anatomic and physiological imperfections of this technique because it leads to enlarged resultant atria and therefore promotes stasis and atrial thrombosis.
Dreyfus et al5 and Banner et al6 have proposed an alternative technique of orthotopic heart transplantation, the “total” atrioventricular transplantation, in which complete excision of the recipient’s atria is undertaken to allow atrial as well as ventricular transplantation.
The aim of this study was to compare, by the use of transthoracic and transesophageal echocardiography, the standard orthotopic heart transplantation with the total atrioventricular transplantation.
This study enrolled 95 patients (75 men, 20 women; mean age, 51±9 years) among 185 heart transplant recipients from two institutions in France: Charles Nicolle Hospital in Rouen and La Timone Hospital in Marseille. (Heart transplantations were performed in Rouen by F. Bouchard, A. Tabley, J.P. Bessou, and R. Soyer and in Marseille by B. Kreitman, A. Riberi, and D. Metras.) Of the 95 heart transplant recipients, 75 were transplanted by the standard technique (group A) and 20 by the total atrioventricular transplantation technique (group B). Standard transplantation was performed according to the initial technique described by Shumway et al.1 Total atrioventricular transplantation was performed after total excision of the right atrium, leaving the origins of the superior and inferior venae cavae for bicaval anastomosis, and after excision of the left atrium, leaving only two cuffs of tissue including both ostia of the pulmonary veins for separate pulmonary venous anastomosis.
All patients received immunosuppressive therapy including cyclosporine, azathioprine, and prednisone in addition to antiplatelet therapy (250 mg aspirin in 51 patients and 300 mg dipyridamole in 44 patients). No patient had evidence of rejection at endomyocardial biopsy performed on the same day as echocardiography. Written, informed consent was obtained from all the patients for the study.
Transthoracic and transesophageal echocardiography were performed with an Acuson XP 128 echocardiograph (Acuson Inc) equipped with a 2- to 2.5-MHz phased-array transducer for transthoracic echocardiography and a high-frequency biplane phased-array transducer for transesophageal echocardiography. Transthoracic echocardiography enabled measurements of cardiac chamber size (left atrial, left ventricle end-systolic and end-diastolic diameters). Left atrial diameter was obtained by taking the mean of three different measurements: the first by the M-mode echocardiogram recorded in the parasternal short-axis view at end systole and the second and third measurements obtained from the horizontal and vertical left atrial sizes in the two-dimensional apical four-chamber view. Left ventricular ejection fraction was calculated using the single-plane area-length method. Transmitral pulsed Doppler flow echocardiography enabled measurements of transmitral flow velocities: early (E) and late (A) peak velocities and their subsequent ratio, E/A, as well as the isovolumetric relaxation time.
Transesophageal echocardiography was carried out after lidocaine pharyngeal anesthesia and intravenous midazolam had been administered, with no complications or failures. Both atria, interatrial septum, and left atrial appendage were carefully scanned. Color flow Doppler echocardiography was used to detect and quantify valvular regurgitation and interatrial shunt. Contrast solution was infused during a Valsalva maneuver to detect patent foramen ovale.
Spontaneous echo contrast (SEC) was defined as the presence of swirling smoke–like echoes within the atrial cavity, with appropriate gain settings to distinguish SEC from echoes due to excessive gain. SEC was subgrouped into marked and mild, marked SEC being an intense echo visible in the entire left atrium at a normal gain level and mild SEC a discrete echo occupying some parts of the left atrium at a high gain level.7
All echocardiograms were reviewed by two independent observers to determine the presence or absence of SEC and left atrial thrombi. Differences in interpretation were resolved by consensus. Thrombi were considered to be present only when the images depicted as thrombi were clear, obvious, and unquestionable for both the observers.
In cases of total heart transplantation, particular attention was paid to the vascular anastomoses: superior and inferior venae cavae and pulmonary vein anastomoses were carefully scanned. When the region of anastomosis was identified, systematic evaluation with color and pulsed Doppler echocardiography was performed to detect any stenosis.
Endomyocardial biopsy and right heart catheterization were performed on the same day as transesophageal echocardiography in 68 of 75 patients in group A and 12 of 20 patients in group B. Right heart pressures including right atrial, right ventricular, pulmonary arterial, and pulmonary capillary wedge pressures were measured by a Swan-Ganz thermodilution catheter that was introduced via right jugular puncture. Cardiac output was determined by thermodilution method, and the mean of five successive measurements was taken.
All values were expressed as mean±SD. ANOVA was used to compare subgroups of patients with mild SEC, marked SEC, and without SEC. An unpaired Student’s t test was used to compare the two groups of patients with and without thromboembolic complications. A value of P<.05 was considered statistically significant. Multiple logistic regression analysis (maximal likelihood ratio method) was used to determine independent predictors of SEC and left atrial thrombus in the entire population; the presence of SEC or atrial thrombus served as dependent variables, whereas left atrial size (≥55 mm), recipient age (≥55 years), time elapsed since transplantation (≥6 months), left ventricular ejection fraction (≤55%), and number of previous acute rejection episodes (≥2) were used as independent variables.
Among the 95 patients enrolled in this study, 75 had been transplanted by the standard technique (group A) and 20 by the total atrioventricular transplantation technique (group B). Our experience with heart transplantation began in 1987 with the standard technique and continued until 1993 when we systematically performed total transplantation in our two institutions. Our two groups are comparable for age, pretransplantation diagnosis of heart disease, and donor age as shown in Table 1⇓. The time from heart transplantation was significantly longer in group A than in group B (27±19 versus 5±4 months, P<.001) because total transplantation has been performed more recently in our two institutions. All patients were in regular sinus rhythm at the time of examination.
Despite the use of antiplatelet therapy in all patients since their surgeries, an acute arterial embolic event occurred in 11 (15%) of the 75 enrolled patients of group A, who received standard orthotopic transplants (0.42 embolic events per patient and per month or 3.5% embolic events per patient and per year), but in none of the 20 patients of group B, who received total atrioventricular transplants. The overall incidence of arterial embolization in the total population transplanted (185 patients) over the entire time period was 5.9%. At the time these embolic events occurred in group A patients, the mean time elapsed since transplantation had been 22±23 months: 5 embolisms within the first 6 months, 3 embolisms within the first 2 years, and 3 embolisms during the fourth year after transplantation. These embolic events consisted of 6 strokes, 3 lower limb artery embolisms, and 2 mesenteric artery embolisms. Because of the embolic events, these 11 patients underwent transesophageal study and were subsequently started on oral anticoagulant therapy.
Baseline transthoracic echocardiographic characteristics of patients of groups A and B are listed in Table 1⇑. Left ventricular size and ejection fraction were similar in both groups. However, the left atrial diameter was markedly greater in patients of group A (58±6 versus 41±4 mm; P<.001). On Doppler echocardiography, a normal and regular pattern of left ventricular filling was seen on the Doppler mitral flow in group B patients with normal E and A waves, whereas Doppler mitral flow showed great variations from cycle to cycle in group A patients, reflecting the asynchronous atrial electrophysiological activity. SEC and left atrial thrombus were not detected by transthoracic examination and were found only by transesophageal echocardiography.
Presence of SEC
The presence of SEC was documented by transesophageal echo in the left atrium of 43 of 75 patients (57%) who received standard transplantation (group A). However, SEC was observed in only 1 of the 20 patients who had total transplantation (group B) (P<.001). Patients in group A were divided into three subgroups: no SEC (32 patients), presence of mild SEC (31 patients), or presence of marked SEC (12 patients) (Table 2⇓). There was no significant difference among the three subgroups regarding recipient or donor age, left ventricular ejection fraction, hemodynamic data, and particularly the time from transplantation. Left atrial diameter was slightly but significantly greater in patients with mild (59±5 mm) and marked (61±5 mm) SEC than in patients with no SEC (56±5 mm, P<.02 mild SEC subgroup versus no SEC subgroup and P<.001 marked SEC subgroup versus no SEC subgroup).
To reduce the discrepancy between group A and group B concerning time elapsed from transplantation, we compared the transesophageal studies performed within the first postoperative year in 26 patients in group A (mean time since standard transplantation, 9±3 months) and in 19 patients in group B (mean time since total transplantation, 5±4 months). The incidence of SEC was 61% (16 of 26 patients) in group A, whereas it was 5% (1 of 19 patients) in group B.
Left Atrial Thrombus
Left atrial thrombi were observed by transesophageal echo in both the transverse and longitudinal views in 20 of 75 patients (27%) of group A. No thrombi were found in the 20 patients of group B. The thrombi were localized in the donor left atrial appendage in 12 patients, on the posterior wall of the left atrium in 6 patients, and on the suture in 2 patients. They occurred only in patients who showed SEC (5 thrombi among the 12 patients with marked SEC and 15 thrombi among the 31 patients with mild SEC). No such thrombi were observed in the 32 patients who had no SEC. Baseline characteristics of the patients in group A with and without left atrial thrombus are listed in Table 3⇓. The two subgroups of patients were not statistically different except for the cardiac index, which was significantly lower in patients with left atrial thrombus (2.75±0.5 versus 3.1±0.6 L · min−1 · m−2, P<.05), and for time since heart transplantation, which was significantly higher in patients with left atrial thrombus (34±26 versus 24±16 months, P<.05). When only the patients studied during the first postoperative year were considered, a left atrial thrombus was found in 23% of patients in group A (6 of 26 patients) and in none of the patients in group B.
Multiple logistic regression analysis was used to compare left atrial diameter, time elapsed since transplantation, recipient age, left ventricular ejection fraction, and number of previous acute rejection episodes concerning their value for identification of patients with SEC or left atrial thrombus among the entire population, and selected left atrial diameter (≥55 mm) as the only independent predictor (P<.001) (adequacy of fit via χ2 P=.653).
Association of SEC and Left Atrial Thrombus to Thromboembolic Episodes
Eleven patients among the 75 patients in group A (15%) had a previous history of arterial embolism, but none of the 20 patients in group B had such history. Because of the embolic events, these 11 patients received oral anticoagulant therapy. SEC was present in 10 of these 11 patients (mild in 6, marked in 4), and left atrial thrombus was detected in 5 of these patients (4 in the left atrial appendage and 1 on the posterior wall). There was no significant difference between patients in group A with or without thromboembolism regarding the recipient or donor age, time from transplantation, left atrial diameter, left ventricular ejection fraction, or hemodynamic data (Table 4⇓).
The atrial septum was perfectly visualized in all the patients by transesophageal echo. Two cases of atrial septum aneurysm were found: one in group A that was associated with mild SEC in a patient who had had a stroke 12 months after surgery and the other in a patient in group B. A patent foramen ovale was found on the donor component of the interatrial septum after contrast infusion in 7 patients in group A. The left atrial suture line was found to be particularly prominent at the left atrial free wall in most patients (mean length, 11±3 mm). This protruding suture line just above the left atrial appendage created a niche on which 7 of the 12 thrombi located in the left atrial appendage were found. Protruding atherosclerotic plaques were observed in the aortic arch of 2 patients: 1 patient in group A who also had had mild SEC and presented with a stroke 3 years after surgery and 1 in group B. Mitral regurgitation was a frequent finding (55 of 75 patients in group A and 15 of 20 in group B) and was graded as mild in most of them, except that in 1 patient with standard transplantation it was graded as moderate. Mild tricuspid regurgitation was detected in all of the patients. Scanning of the vascular anastomoses in group B was easy for left pulmonary veins and superior vena cava that we examined in all the patients. We could locate the inferior vena cava anastomoses in only 5 patients and the right pulmonary vein anastomoses in only 11 patients. In no case did we detect either stenosis or thrombus on the site of the vascular anastomoses in group B patients.
This transthoracic and transesophageal echocardiography study was designed to compare the incidence of thromboembolic complications within the two techniques of orthotopic heart transplantation. The standard technique, which is practiced worldwide, was described by Shumway and coworkers.1 This technique, which is efficient and attractive because of its simplicity, involves mainly a ventricular substitution, because the posterior and lateral portions of the recipient atria are left in situ and anastomosed to the donor atria. Angermann et al2 recently demonstrated by transesophageal echocardiography the anatomic and physiological abnormalities that result from this technique. The anatomic abnormalities include enlargement of the resultant atrial cavities with abnormal “hourglass” shape and prominent sutures between the receiver and donor components. The physiological abnormalities include asynchronous contractions of the recipient and donor atria as a consequence of the persisting independent electrical activity in both the components, leading to an aneurysmal-like behavior of the recipient’s interatrial septum. Another technique more recently proposed is a total heart transplantation that includes total excision of both the right and left atria of the recipient and orthotopic transplantation of an intact donor heart with its complete atria. This method, described by Dreyfus et al5 and Banner et al,6 is more technically demanding, necessitating separate anastomoses of the superior vena cava, inferior vena cava, and pulmonary veins. However, this technique offers anatomic and physiological advantages because the atrial cavities are of physiological size, they have no protruding suture lines, and there are no asynchronous atrial contractions.
The atrial anatomic changes that result from standard orthotopic heart transplantation in our patients lead to pathological consequences such as a high incidence of SEC and of atrial thrombi that correspond to a high embolic event rate in this group of patients. These abnormalities could be found only by transesophageal study that we routinely perform in our two institutions within the first year after heart transplantation. SEC was found in 57% (43 of 75) of our patients transplanted by the standard technique and in only 5% (1 of 20) of our patients transplanted by the total atrioventricular technique. A left atrial thrombus was found in 20 of the 75 patients with the standard technique, associated with SEC in all patients, and in none of the 20 patients with the total transplantation technique. It is very unlikely that this marked difference in the incidence of SEC and left atrial thrombosis between the two techniques can be explained by the longer time elapsed from transplantation to the transesophageal study because SEC and left atrial thrombus were also found frequently (61% and 23%, respectively) in the 26 patients with standard transplantation enrolled within the first postoperative year.
Until now, there has been little information on the incidence of both SEC and atrial thombosis after heart transplantation despite the large number of heart transplants worldwide, with a long follow-up. The frequency with which SEC and thombi are detected by transesophageal echocardiography has been reported in previous studies but from small series of heart transplant patients operated on by the standard method. Among a population of 20 heart transplant recipients, Angermann et al2 described SEC in 5 patients (25%) and a left atrial thrombus in 3 patients (15%) with 1 of these 3 patients having a peripheral arterial embolism. Polanco et al3 found a higher incidence of SEC (47%) in a study of 30 heart transplant recipients, all of whom were receiving antiplatelet therapy but no anticoagulant therapy. Thrombus formation occurred in only 2 patients (6%) in this series and was detected by both transthoracic and transesophageal studies: one was localized in the right atrium and the other was in the left atrium, which embolized, caused obstruction of the left ventricular outflow tract, and led to the patient’s death.
Our findings from this larger series confirm these previous reports and emphasize the important role of early transesophageal echocardiography in the detection of atrial thrombus, which involves a high risk of acute arterial embolism in heart transplant recipients. The reasons for the formation of SEC and left atrial thrombi after standard transplantation are still a matter of debate, since heart transplant atria constitute a unique example of anatomic substrate of SEC and atrial thrombosis without the usually involved factors that favor atrial thrombosis, such as mitral valve stenosis, prosthetic mitral valve, left ventricular dysfunction, or atrial fibrillation.8 9 10 11 Other than the anatomic substrates for the atrial thrombi formation (such as enlarged left atrium and prominent suture line), there could be several other factors that may be involved (ie, intracardiac hemodynamic alterations caused by the asynchronous contractions of the recipient and donor atrial components, the presence of arrhythmias either in the course of an acute rejection episode or preexisting in the recipient atrial component, and possibly alterations in hemostasis such as an increased platelet aggregation related to inhibition of the endothelial prostacyclin synthesis mediated by cyclosporine).12
This comparative study between two surgical techniques clearly emphasizes the important role of the surgical technique since thromboembolic complications occurred only after standard transplantation and not after total transplantation. However, no data are yet available that follow patients who have received total transplantation for more than 6 months, especially for the incidence of complications such as stenosis or thrombus on the site of vascular anastomoses.
In our series, an arterial embolism was documented in 11 of the 75 patients receiving standard transplantation (15%) enrolled for transesophageal study. This high rate of embolic events is questionable and may be explained by the fact that these 11 patients had a transesophageal echocardiogram for the purpose of looking for a cardiac source of embolism and were therefore included in the study. Moreover, it must be noted that the overall incidence of arterial embolism was 5.9% (11 of the 185 patients) in the total group of heart transplant recipients from both our two institutions) and 9% when we considered all the patients who had had a standard heart transplantation (11 of 122 patients). However, ischemic or anoxic changes, including cerebral infarcts, were found during more than 60% of 31 postmortem examinations in heart and heart-lung transplant recipients.13 Andrews et al14 reviewed the neurological complications that occurred among a series of 90 patients treated with standard orthotopic cardiac transplantation and found that within the first 60 days of postoperative care, 2 patients (2%) had suffered ischemic cerebral infarctions that could be related to an emboli from cardiac origin.
In our series, we decided to use oral anticoagulant therapy in the patients who had a definite left atrial thrombus. However, we chose not to institute oral anticoagulant therapy in patients with isolated spontaneous contrast in their left atria despite a few studies that have suggested that the presence of atrial smoke identifies patients at an increased thromboembolic risk.7 9 We think that more data are necessary from heart transplant recipients to determine whether spontaneous contrast is associated with a higher rate of atrial thrombus. Our data also demonstrate that transesophageal echocardiography is the procedure of choice for the evaluation of both SEC and left atrial thrombi. At the present time, we perform routine transesophageal echocardiography in all heart transplant recipients 6 months after surgery; however, since the first thromboembolic episode occurred within the first month after transplantation, we may consider performing transesophageal echocardiography before discharging the patient, at 1 month after transplant.
Limitations of the Study
Because of the recent introduction of this surgical technique, the number of patients with total transplantation was smaller than that of patients with standard transplantation, and the follow-up was clearly shorter. Larger series with longer follow-up times would therefore be necessary to enable one to draw conclusions on the potential complications of this new technique, to make sure that no late formation of thrombus could occur (particularly on the site of anastomoses), and to establish the true benefits of total transplantation as regards the rate of thromboembolic complications.
The initial technique of orthotopic heart transplantation introduced by Shumway and coworkers is the most commonly used. However, the prominent sutures at the site of atrial anastomoses, the enlargement of the resultant atria, and also the asynchronous contraction of the recipient and donor atria contribute to explain the high incidence of atrial thrombus in these patients, thereby necessitating consideration of oral anticoagulant therapy in the group of patients with left atrial thrombus. On the other hand, total orthotopic heart transplantation seems from this preliminary report to avoid these thromboembolic complications, and therefore, although technically more demanding, it should become the operative procedure of choice if our results are confirmed in larger series.
Reprint requests to Geneviève Derumeaux, MD, Service de Cardiologie, Hôpital Charles Nicolle, Centre Hospitalier et Universitaire de Rouen, 1 rue de Germont, 76000 Rouen, France.
- Copyright © 1995 by American Heart Association