Section I: Chronic constrictive pericarditis
Definition
Chronic constrictive pericarditis is an inflammatory process that involves the pericardium, leading to pericardial thickening and compression (constriction) of the ventricles. The resultant impairment in diastolic filling reduces cardiac function but with relatively preserved global systolic function.
Historical note
It is said that Galen, in AD 160, described cicatricial thickening of the pericardium in an animal and surmised that the same condition might occur in humans. The first formal account of the condition in humans was that of Lower, who described both acute and chronic constrictive pericarditis in 1669. Other early descriptions were those by Bonetus in 1679 and Vieussens in 1715. , Lancisi apparently understood the pathology of the condition; and, in 1728 he described at autopsy a patient with a small heart encased by a thick adherent pericardium in association with marked swelling of the abdomen and jugular veins.
As a result of the observations of Morgagni in 1760 and Laennec in 1819 that pericardial adhesions were rarely associated with symptoms, little attention was paid to the possible clinical significance of chronic pericarditis for nearly a century because of ignorance of the difference between adhesive pericarditis and constrictive pericarditis. , In 1842, Corrigan described the pericardial knock. , The early literature contains only four reports, largely ignored, stressing that chronic constrictive pericarditis could be clinically important.
Interest was refocused on the condition by Pick’s report in 1896 of three patients with chronic constrictive pericarditis whose clinical course had been thought in life to be due to cirrhosis of the liver. About that time, surgeons were becoming more expert and aggressive, and Weill in 1895 and Delorme in 1889 suggested that pericardiectomy be used to treat this condition. , Brauer, in 1902, suggested removing the bony precordium as a method of relief. The first operation directed against chronic constrictive pericarditis was carried out by Hallopeau. White signaled the modern era of diagnosis and treatment in 1935. , Surgical experience was expanded by Harrington and Barnes at the Mayo Clinic and by Heuer and Stewart at New York Hospital. ,
Animal experiments began to clarify some of the perplexing problems that persisted despite the advent of surgical treatment. In 1929, Beck reproduced the syndrome by injecting Dakin solution into the pericardial cavity of dogs. He demonstrated that simple obliteration of the pericardial cavity by adhesions did not produce the syndrome; only a thick, dense scar around the heart did so, thus solving the riddle of 100 years earlier. He then demonstrated in animals that the syndrome could be relieved by pericardiectomy. The pathogenesis was further elucidated by the cardiac catheterization studies of Sawyer and colleagues and by the ingenious experiments of Isaacs and colleagues. ,
The hemodynamic characteristics of constriction were reported by Bloomfield in 1946 when he demonstrated elevated right atrial and right ventricular diastolic pressure with an early diastolic dip in a patient who had constrictive pericarditis. , Hansen subsequently reported the right ventricle dip and plateau pressure-pulse pattern in 1951. , The condition of effusive-constrictive pericarditis was popularized by Hancock in 1971, while Kendall reported constrictive pericarditis as a complication of cardiac surgery in 1972. , ,
Isner discussed the value of computed tomography scanning in constrictive pericarditis in 1982, and Soulen wrote of magnetic resonance imaging studies in the disease in 1984. , , D’Cruz first described the unique two-dimensional echocardiogram finding of a decreased angle between the left atrial posterior wall and the left ventricle in the parasternal long-axis view. , Additional investigation, technologic advancement, and refining of techniques have culminated in better and less invasive diagnostic and surgical methods. ,
Morphology
The pericardium is composed of two parts, fibrous and serous, arranged in three layers. The fibrous pericardium layer is a tough connective tissue sac that surrounds but which is unattached to the heart. The serous pericardium consists of the remaining two layers: the inner visceral layer adheres to the heart. It forms the epicardial covering, while the outer parietal layer adheres to the internal surface of the fibrous pericardium.
Normally, the potential space between the inner and outer layers of the visceral pericardium contains a thin layer of fluid. A demonstrable amount of fluid normally accumulates only over the atrioventricular junctions. This has provoked controversy about the pressures normally present between the fibrous pericardium and the inner layer of the visceral pericardium (epicardium).
As chronic constrictive pericarditis develops, the fibrous parietal pericardium and one or both layers of the visceral pericardium are involved to some extent, but details of the pathologic process vary. If the two layers of the visceral pericardium remain separate, the pericardial space contains variable amounts of fluid, often with extensive and sometimes hemorrhagic fibrinous deposits on both surfaces. This entire fibrous and fluid mass can constrict the heart. When the process is far advanced, the two layers of the visceral pericardium thicken and fuse and, along with the fibrous pericardium, encase the heart in a thick, solid, fibrous, and often calcified envelope that is adherent to the myocardium.
In addition, cardiac muscle fiber atrophy occurs in many cases. , Atrophy may appear relatively early in the course of the disease. Myocardial fibrosis also complicates the late stages of chronic constrictive pericarditis.
Clinical features and diagnostic criteria
Pathophysiology of cardiac compression
Cardiac compression occurs in chronic constrictive pericarditis. But cardiac compression is also a feature of acute cardiac tamponade and effusive constrictive pericardial disease—a condition characterized by fluid in the pericardial space and markedly reduced compliance of the pericardium.
Normal.
Pericardial pressure is subatmospheric under normal circumstances, similar to intrapleural pressure. Both intrapericardial and intrapleural pressures become more negative during inspiration. There are small fluctuations of intrapericardial pressure related to the cardiac cycle, and the transpericardial pressure (pericardial minus pleural pressure) is highest at end-diastole, the period of largest ventricular volume. , Pericardial pressure rises as ventricular volume is increased beyond normal limits by the rapid infusion of fluid. Under such circumstances, the effects of pericardial restraining are predominantly on the right ventricle.
Pressure-volume relationships and stress-strain characteristics of the normal pericardium are such that there is little increase in intrapericardial pressure when a small amount of fluid is placed intrapericardially. With rapid addition of more fluid, the pressure slope rises progressively. , At this stage, adding small amounts of fluid causes a large increase in intrapericardial pressure, and conversely, removing small amounts causes a large decrease in intrapericardial pressure (the rationale of pericardiocentesis in acute pericardial tamponade). Furthermore, because of pericardial hysteresis, intrapericardial pressure at a given volume during fluid removal is lower than during the addition of fluid.
Intrapericardial events—normal and abnormal—affect both cardiac filling and cardiac output. Such events are reflected in phasic and overall atrial and venous pressures. The effect on systemic venous pressure is particularly important because it is easily observed in the jugular venous pressure. Normally, the inferior and superior vena caval pressures exceed atmospheric pressure by only a few millimeters of mercury (mmHg). The jugular venous (and caval) pulse consists sequentially of three positive upward waves and two downward movements. First is the a wave, generated by atrial systole. The c wave follows, caused by the displacement of the tricuspid valvar apparatus toward the right atrium during isovolumic ventricular systole. A negative x descent is next, generated in part by descent of the closed tricuspid valve apparatus at the beginning of ventricular ejection and in part by decreased intrapericardial pressure resulting from reduced ventricular volume as the ventricle ejects. A positive v wave is generated by passive filling of the right atrium from the cavae and coronary sinus. Finally, as the tricuspid valve opens, a negative y descent occurs as blood flows rapidly from the right atrium to the right ventricle.
Acute cardiac tamponade.
Rapid increase in intrapericardial fluid (e.g., blood) produces acute cardiac tamponade. Intrapericardial pressure may rise as high as 20 to 30 mmHg. Such a pressure would be incompatible with life if it were not for reflex venoconstriction, catecholamine release, and the immediate retention of sodium and water by the kidneys as part of the total body response to reduced cardiac output. As a result, venous pressure rises to the level of intrapericardial pressure, and cardiac output is augmented, although usually at a reduced level from normal. This process has led to one definition of cardiac tamponade as a condition in which right atrial and systemic venous pressure are determined by elevated intrapericardial pressure.
When intrapericardial pressure first rises, it tends to exceed left as well as right atrial pressure, and in patients who survive, both left and right atrial pressures rise in response to the neurohumoral compensatory mechanisms mentioned earlier. At this stage, right and left atrial pressures, right and left ventricular diastolic pressures, pulmonary artery diastolic pressure, and pulmonary artery wedge pressure are identical to intrapericardial pressure. Untreated, the patient dies when cardiac output decreases despite compensatory mechanisms.
In this classic setting of acute cardiac tamponade, the heart is small and quiet, venous pressure is elevated, and systemic arterial blood pressure is depressed—a group of signs known as the Beck triad . Elevation of venous pressure may be mild, or it may reach 20 mmHg or more. Jugular venous pulse waves are altered because the tamponade effect is least during ventricular ejection when ventricular volume is smallest. No cardiac filling occurs during diastole; thus, there is no y descent. All filling occurs during systole, so the x descent is preserved and exaggerated.
Unless hypotension is extreme, the condition is also characterized by pulsus paradoxus, an inspiratory decrease in arterial systolic blood pressure exceeding 10 mmHg during quiet respiration. The mechanism underlying pulsus paradoxus in acute cardiac tamponade is complex. During inspiration, caval flow into the right atrium increases, just as in the normal situation; indeed, the percentage of increase is greater than normal. The resultant increase in right-sided heart volume raises intrapericardial pressure still further, and pericardial transmural (intrapericardial minus intrapleural) pressure rises. Left ventricular volume is decreased as the ventricular septum is displaced leftward by the increased right ventricular volume. Left ventricular inflow is diminished because the somewhat decreased right ventricular output is easily accommodated by the expanding lung blood volume during inspiration, with less transmitral flow. These phenomena result in decreased left ventricular stroke volume and diminished arterial blood pressure. Another contributor to pulsus paradoxus is the delay in passage of the increased caval flow of early inspiration to the left ventricle so that by the time it has occurred, respiration has generally shifted to the expiratory phase. Also, the inspiratory decrease in intrathoracic pressure tends to decrease aortic and arterial pressure, and inspiration directly decreases left ventricular contraction.
Chronic constrictive pericarditis.
The basic pathophysiology of chronic constrictive pericarditis has been debated for over half a century. By 1949, Holman and Willett concluded that constriction of the caval orifices and atria was important in its pathogenesis. For this reason, they adopted the median sternotomy approach for its surgical correction. In 1951, Burwell concluded from cardiac catheterization study that both right and left ventricular function were impaired, and constriction of caval orifices or atria played no role. In 1952, Isaacs and colleagues showed in dogs that a change in the pressure-volume curves of the two ventricles resulted from experimentally produced constrictive pericarditis, which was the fundamental pathophysiologic change associated with the disease ( Fig. 18.1 ). These investigators also demonstrated, during the development of constriction, an increase in right and left ventricular diastolic pressure and a decrease in stroke volume. In their experimental animals, a small increase in volume resulted in a considerable increase in end-diastolic pressure. These studies indicated that lack of ventricular diastolic distensibility, and thus inability to generate an adequate preload (see Chapter 4 ), was a characteristic of hearts with chronic constrictive pericarditis. These considerations influenced Scannell and colleagues to adopt a left anterolateral thoracotomy as their surgical approach of choice by 1952.
Pressure-volume curves of left and right ventricles (LV, RV) in dogs, made immediately after death by introducing 10-mL boluses of saline solution via aorta or pulmonary artery into the closed ventricle. Note that in the normal dog heart (Before), compliance of right ventricle is greater than left. In hearts of dogs with generalized pericardial constriction (After), compliance is low and similar in both ventricles.
(Data from Isaacs JP, Carter BN 2nd, Haller JA Jr. Experimental pericarditis: the pathologic physiology of constrictive pericarditis. Bull Johns Hopkins Hosp . 1952; 90:259.)
Several features of clinical cases of chronic constrictive pericarditis derive from these basic abnormalities of diastolic function. Ventricular filling is impaired, and ventricular stroke volume is reduced because of decreased compliance of the fused cardiac and pericardial mass. Phasic aspects of ventricular filling are also altered. For a brief period in early diastole, ventricular filling is rapid. However, the limit of ventricular distensibility is reached rapidly, and the right ventricular pressure pulse displays an early diastolic dip and then a high diastolic plateau (square root sign) . There is nearly complete diastolic ventricular filling during the first 50 milliseconds of diastole.
Systemic venous pressures are correspondingly abnormal; mean venous pressure is elevated. The x descent is steep and deep, corresponding to the beginning of ejection. The y descent is also steep and deep, corresponding to the early diastolic dip of right ventricular pressure. This differs from events during acute cardiac tamponade, where the y descent is absent. The normal inspiratory increase in vena caval flow and decrease in pressure are diminished and often absent.
Pulsus paradoxus is said to be infrequent in chronic constrictive pericarditis, in contrast to the situation with acute cardiac tamponade. However, the frequency of its recognition is influenced by cardiac rhythm; it is usually present when there is sinus rhythm but impossible to detect when there is atrial fibrillation (a frequent accompaniment of chronic constrictive pericarditis).
Ventricular end-diastolic volumes are small in this disease, as are end-systolic volumes and stroke index. The rate of increase of left ventricular systolic pressure and ejection fraction is not altered. Thus, systolic left ventricular function under these circumstances is normal, but this does not necessarily indicate normal contractility.
Effusive constrictive pericardial disease.
Although seen in a number of settings, effusive pericardial disease is common in nephrogenic pericarditis. In effusive constrictive pericarditis, increased volume of pericardial fluid produces the characteristic clinical picture of acute cardiac tamponade, with the absence of a y descent and a preserved and prominent x descent in the jugular venous pulse. However, because of coexisting pericardial thickening, aspiration of pericardial fluid does not return the situation to normal. Instead, the thickened pericardium begins to restrain the heart, but only after the rapid filling phase of the ventricles is over. Thus, the y descent is again present and is prominent, occurring when the right atrium is in free communication with the right ventricle through the open tricuspid valve and simultaneously with the early diastolic dip of ventricular pressure. In effusive constrictive pericardial disease, after the fluid is removed, there is no respiratory variation in the right atrial and venous pressures, just as in constrictive pericarditis.
Etiology
In most patients the etiology of chronic constrictive pericarditis is not known. McCaughan and colleagues were able to identify a specific etiologic factor in only 27% of their patients and Blake and colleagues in only 34%. , In the current era (>1990), nearly half of cases have a nonidiopathic etiology (postoperative 32.3%, radiation 11.4%). In about 10% of cases, documented acute pericarditis precedes the development of chronic constrictive pericarditis. Prior to its effective treatment, tuberculosis was the etiology of chronic constrictive pericarditis in up to 17% of cases. , Currently, a prominent cause is mediastinal radiation for malignant disease. Rheumatoid disease and sarcoidosis occasionally are causes. Trauma is another uncommon cause, with hemopericardium usually present as the precursor of pericardial thickening and constriction. ,
Prior cardiac surgery is an increasingly common cause of constrictive pericarditis, and it may be more common after coronary artery bypass grafting than after other operations. The interval between the original cardiac operation and the development of evidence of pericardial constriction is highly variable, ranging from 1 month to nearly 10 years. Mean interval is about 2 years.
Clinical presentation
Classically, symptoms of chronic constrictive pericarditis are delayed for several years after the clinical or subclinical episode of acute pericarditis. The interval may, however, be as short as 3 to 4 weeks in those rare instances in which pericarditis develops after cardiac surgery or 4 to 12 months after trauma or acute nonspecific pericarditis.
Initial symptoms may be only fatigue with or without modest effort breathlessness, and neck vein distention may be noticed. Insidiously, however, hepatomegaly and ascites develop, initially with or without peripheral edema. Even within the context of such evidence of appreciable fluid retention, breathlessness may occur only on exertion and not at rest; although in severe cases, there may be orthopnea. Paroxysmal nocturnal dyspnea occurs infrequently.
Clinical findings
When constriction is not severe, clinical findings may be limited to modest but persistent elevation of jugular venous pressure and slight liver enlargement with or without intermittent ankle edema. As constriction increases, there is a progressive increase in venous pressure and hepatomegaly, with the eventual development of persistent peripheral edema, ascites, and pleural effusion. Venous pressure fails to decline during inspiration (Kussmaul sign), but this is not specific, in that the same findings may accompany right ventricular failure, restrictive myocardial disease, or tricuspid valve stenosis. By this stage, pulsus paradoxus is to be expected if sinus rhythm persists, and pulse pressure often is reduced. The apex beat is usually not palpable, but there is often systolic retraction in the left parasternal region. This retraction may be followed by a visible and palpable forward thrust extending toward the expected site of the cardiac apex. This impulse, which results from forceful ventricular filling with the onset of diastole, may be mistaken for the apex beat and used to argue against the existence of pericardial constriction. Rapid ventricular filling in early diastole is also associated with an unusually early, often loud, third heart sound that is sometimes referred to as a pericardial knock, but usually, there are no murmurs.
As in other forms of heart failure, salt and water retention are present. Anand and colleagues found important increases in total body water, extracellular volume, plasma volume, and exchangeable sodium in their study of patients with proven constrictive pericarditis. However, renal plasma flow was only moderately decreased, and glomerular filtration rate was normal. Norepinephrine, renin activity, aldosterone, and cortisol were also increased, as was plasma atrial natriuretic hormone, although not to levels usually seen in other heart failure syndromes. The ratio of left atrial to aortic diameter measured by echocardiography was only minimally increased, indicating that in constrictive pericarditis, the atria are prevented from expanding. The restricted distensibility of the atria may limit the secretion of atrial natriuretic hormone, thus reducing natriuretic and diuretic effects, resulting in retention of sodium and water greater than that occurring in patients with edema from myocardial disease.
Laboratory investigation
Protein-losing enteropathy occurs in some patients with chronic constrictive pericarditis who develop ascites and hepatomegaly. They may have severe hypoproteinemia, with depression of albumin and gamma globulin, and an increased rate of leakage of plasma protein into the gastrointestinal tract. This syndrome also develops after other conditions that chronically elevate inferior or superior vena caval pressure, such as the Fontan operation and atrial switch operations with inferior vena caval obstruction (see Chapter 52 ). The preoperative model for end-stage liver disease (MELD) and MELD-XI score provides important morbidity and mortality risk stratification in constrictive pericarditis patients.
Chest radiography
The chest radiograph may be unremarkable, although about one-third of patients show moderate to marked enlargement of the cardiac silhouette, and some have pleural effusions. Pericardial calcification is evident in about 25% and radiologic evidence of compression in about 60%.
Electrocardiography
The electrocardiogram (ECG) is usually abnormal, with nonspecific ST-segment and T-wave changes in 90% of cases. In about 40% of patients with surgically verified chronic constrictive pericarditis, the QRS complexes have low voltage, and an atrial arrhythmia is present in 30%.
Imaging studies
Transthoracic echocardiography.
Transthoracic echocardiography is the primary diagnostic modality in the assessment of constrictive pericarditis. A standard evaluation should include two-dimensional Doppler assessments and, where appropriate, speckle tracking. When thorough and complete, transthoracic echocardiography is diagnostic in over 70% of patients. The Mayo Clinic Criteria include three variables independently associated with constrictive pericarditis: respiration-related ventricular septal shift, preserved or increased medial mitral anular e’ velocity, and prominent hepatic vein expiratory diastolic flow reversals. The criteria are sensitive, specific, and provide effective differentiation among the confusing mimickers of restrictive cardiomyopathy and tricuspid valve regurgitation.
Computed tomography.
Computed tomography (CT) provides important anatomic information about pericardial thickening and/or effusion, associated coronary artery disease, and location of important structures (e.g., areas of pericardial calcification, great vessels, and patent bypass grafts). , Anatomic findings on CT study may have diagnostic importance when physiologic phenomena of restriction to ventricular diastolic filling are demonstrated. , Oren and colleagues, using CT, demonstrated that the abnormally rapid early diastolic filling of the left ventricle characteristic of constrictive physiology, coupled with a measured pericardial thickness greater than or equal to 10 mm, can distinguish constrictive from normal or restrictive physiology. It is important to note, however, that pericardial thickness may be normal in up to 18% of patients with surgically proven constrictive pericarditis.
Magnetic resonance imaging.
Magnetic resonance imaging (MRI) also can provide measurements of pericardial thickness and depict characteristic right atrial dilation and right ventricular compression. Not only is MRI able to define pericardial thickness, but it also effectively characterizes features that represent constrictive physiology. Masui and colleagues found the sensitivity, specificity, and accuracy of MRI in diagnosis of constrictive pericarditis to be 88%, 100%, and 93%, respectively. Pericardial inflammation can also be identified with MRI. The issue of inflammation may aid in the medical management and in the timing of pericardiectomy.
Cardiac catheterization
Characteristically, end-diastolic pressures are elevated and equal in the right atrium, pulmonary artery, and left atrium; this is the hallmark of chronic constrictive pericardial disease. , In the report by McCaughan and colleagues, such findings were obtained in all patients coming to catheterization. Intraventricular pressure pulse contours characteristically demonstrate an early rapid fall in diastolic pressure in the right ventricle, followed by a rapid rise to an elevated diastolic plateau (square root sign). , Left ventricular pressure pulse usually has a similar contour.
Mean right atrial pressure fails to decrease normally during inspiration, or it may rise slightly. There is a transient increase in pulmonary blood volume and a slight reduction in right ventricular afterload, resulting in a fall in pulmonary arterial and right ventricular systolic pressure and a decline in pulmonary venous pressure and left ventricular diastolic pressure as well.
Vaitkus and Kussmaul identified the predictive accuracy of three different hemodynamic criteria for differentiating constrictive from restrictive disease ( Table 18.1 ): (1) equalization of right and left ventricular end-diastolic pressure favors constriction; (2) constriction is associated with more modest elevation of right ventricular systolic pressure (≤50 mmHg); in restriction, it exceeds that amount; (3) in constriction, right and left ventricular end-diastolic pressure usually are greater than one-third of right ventricular systolic pressure; in restriction, the ratio is characteristically less than one-third.
TABLE 18.1
Predictive Accuracy of Individual Hemodynamic Criteria for Constrictive Pericarditis and Restrictive Cardiomyopathy
Data from Vaitkus PT, Kussmaul WG. Constrictive pericarditis versus restrictive cardiomyopathy: a reappraisal and update of diagnostic criteria. Am Heart J . 1991;122:1431.
| Criterion | Constrictive Pericarditis (%) | Restrictive Cardiomyopathy (%) | Overall Predictive Value (%) |
|---|---|---|---|
| LVEDP—RVEDP ≤5 mmHg | 92 | 70 | 85 |
| RV systolic pressure ≤50 mmHg | 90 | 24 | 70 |
| RVEDP/RV systolic pressure ≥0.33 | 95 | 32 | 76 |
LV, Left ventricular; RV, right ventricular; EDP, end-diastolic pressure.
The Mayo Clinic group has further identified two dynamic criteria that greatly aid in the diagnosis of constrictive pericarditis. , , The first is dissociation of intrathoracic and intracardiac pressures during inspiration with a significant decrease in the gradient between the pulmonary artery wedge pressure and left ventricular diastolic pressure. The second is ventricular interdependence with discordance between right and left ventricular tracings during inspiration. Enhanced ventricular interaction can also be assessed with the comparison of ejection times in the pulmonary artery and ascending aorta. This technique simplifies the assessment for constriction and differentiating constriction from restrictive cardiomyopathy and severe tricuspid valve regurgitation.
When hemodynamic studies are equivocal, rapid infusion (in 6 to 8 minutes) of 1000 mL of normal saline solution produces diagnostic features of occult chronic constrictive pericardial disease. , These features include not only striking elevations of filling pressures but also development of typical pressure pulse morphologic characteristics of constriction, loss or reversal of the respiratory variation of right atrial pressure, and precise diastolic equilibration of cardiac pressures.
Endomyocardial biopsy
When diagnosis is unclear, myocardial biopsy may be useful. Normal myocardium, nonspecific changes (e.g., from irradiation), or myocarditis on biopsy must be considered nondiagnostic, because they may be present with either restrictive or constrictive disease. The finding of amyloid disease is diagnostic of a restrictive etiology.
Natural history
Knowledge of the natural history of surgically untreated patients with chronic constrictive pericarditis is incomplete. The interval between an etiologic event and onset of clinical evidence of constriction varies between a few months and many years. Factors that determine rate of progression of the disease and its symptoms are unknown. Atrial fibrillation commonly occurs at some stage and can result in sudden deterioration in circulatory status.
Somerville has estimated that once signs and symptoms of chronic constrictive pericarditis develop, a semi-invalid life can be led over an interval of 5 to 15 more years. When the clinical syndrome includes ascites, progression is more rapid, particularly in children.
Technique of operation
Because patients with chronic constrictive pericarditis coming to operation are often seriously ill, an arterial catheter is inserted into the radial artery for pressure recording, in addition to usual preparations in the operating room. A central venous pressure line and often a pulmonary arterial catheter are also inserted. Approach may be through a median sternotomy or left anterolateral thoracotomy.
Median sternotomy approach
The median sternotomy approach may be used with or without cardiopulmonary bypass (CPB). Although unnecessary for most patients, CPB facilitates dissection of the heart from the pericardium especially when there is calcification that burrows into the myocardium. Extracorporeal circulation is also useful for support when dissection and manipulation of the ventricles lead to hemodynamic disturbance. Perfusion is maintained at normothermia, and arrest of the heart is rarely necessary. When CPB is used, it may be convenient to use the femoral vessels for both venous and arterial cannulation (see Special Situations and Controversies of Chapter 2 ).
The pericardium is opened vertically anteriorly in the midline ( Fig. 18.2 ). Opening both pleural spaces facilitates exposure and identification of the phrenic nerves. The pericardial flaps are then dissected laterally with low voltage electrocautery, superiorly and inferiorly. To the right, dissection passes across the atrioventricular groove and proceeds across the anterior and lateral walls of the right atrium if the cleavage plane is easily developed. If it is not, as previously stated, this portion of thickened pericardium can be left in situ as constriction is a disease of the ventricles, and freeing the atria is not necessary. In the former instance, the pericardial flap is excised about 1 to 1.5 cm anterior to the right phrenic nerve. To the left, dissection proceeds across the front of the ventricles and then over the lateral left ventricular wall to about 1 to 1.5 cm in front of the left phrenic nerve.
Pericardiectomy for chronic constrictive pericarditis through a left anterolateral incision. (A) Positioning of patient and proposed incision. Padding beneath left arm is important. (B) Phrenic nerve has been mobilized away from pericardium. An initial longitudinal incision has been made in thickened parietal pericardium. (C) Flaps have been dissected back, completely liberating both left and right ventricles, leaving where necessary small calcific plaques in situ. First portion of pulmonary trunk has been completely unroofed. (D) Appearance at end of procedure. Thickened pericardium has been removed from diaphragm.
The anterior pericardium is excised from the great vessels to the diaphragm ( Fig. 18.3 ). Dissection continues over the diaphragm inferiorly and, when it can be safely performed, posterior to the left phrenic nerve back to the left atrioventricular groove ( Fig. 18.4 ). It is sometimes possible to remove the thickened, often calcified, outer pericardial layer, because there is generally a cleavage plane between pericardium and the overlying thickened pleura containing the phrenic nerve. Most often the phrenic nerve can be left attached to a strip of pericardium. The inferior pericardium can usually be separated easily from the underlying fibrous, central diaphragm ( Fig. 18.5 ). The coronary arteries should be visible on the surface of the heart. If the epicardium is thin and relatively normal, it need not be disturbed. If it is thickened, it must be removed either in its entirety or in a sufficient number of areas to allow more normal diastolic filling of the ventricles. , Failure to do this compromises results of operation.
The pericardium and the right and left phrenic nerves are shown through a median sternotomy exposure. The pericardium is incised near the midline.
(From Hemmati A. Contemporary techniques of pericardiectomy for pericardial disease. Cardiol Clin . 2017;35(4):559-566.)
After sharp dissection toward the left and right phrenic nerves, the anterior pericardium is removed, exposing the epicardial coronary arteries. The pericardium beneath each phrenic nerve is preserved to avoid injury.
(From Hemmati A. Contemporary techniques of pericardiectomy for pericardial disease. Cardiol Clin. 2017;35(4):559-566.)
The heart is freed from the diaphragmatic aspect of the pericardium and the pericardium posterior to the left phrenic nerve. The pericardium is then dissected from the diaphragm as shown.
(From Hemmati A. Contemporary techniques of pericardiectomy for pericardial disease. Cardiol Clin . 2017;35(4):559-566.)
Left anterolateral thoracotomy approach
The patient is positioned supine, with a roll beneath the left scapula. The left hand is secured beneath the left buttock, with the elbow padded and positioned on the left side of the table ( Fig. 18.2 A). A curving left anterolateral skin incision is made beneath the breast anteriorly and more laterally over the fifth interspace. Incision is carried through the pectoralis major anteriorly, and the fifth interspace is opened. The interspace incision is extended well anteriorly. The internal thoracic vessels can be ligated and divided. The incision can be taken across the sternum if exposure is inadequate. The rib spreader is inserted, and the interspace incision extended laterally with electrocautery as the spreader is gradually opened.
The left phrenic nerve is identified and freed from the pericardium if possible. Because the nerve is often densely adhered to the pericardium, most commonly the nerve can be mobilized with a narrow strip of pericardium to avoid injury. The pericardium is incised through an area of minimal calcification posterolateral, if possible, over what is presumed to be left ventricle ( Fig. 18.2 B). On occasion, this initial incision through the abnormal pericardium takes the dissection immediately onto the myocardium; in other cases, it enters a fluid-filled space (see “ Morphology ” earlier in this section).
When a space is entered, the initial longitudinal incision is carried anteriorly and posteriorly from its superior and inferior extremes. The anterior pericardial flap is dissected as far as the right atrioventricular groove, beneath the elevated thymus and prepericardial fat, and resected ( Fig. 18.2 C). The posterior flap is dissected far posteriorly to the level of the left atrioventricular groove and then excised. Dissection must be carried superiorly onto the pulmonary trunk, because failure to relieve pericardial bands across it can result in postoperative gradients and severe right ventricular hypertension. The piece of pericardium left inferiorly is dissected off the diaphragm. Fibrous plaques adherent to the epicardium are then dissected off through the entire area of resection.
If no pericardial space is found, the entire longitudinal incision and its anterior and posterior extensions are made only through the fibrous pericardium. Then the incision is deepened in an area that seems to be over myocardium rather than over the interventricular or atrioventricular groove. Slowly and carefully, the posterior flap is dissected off the left ventricular myocardium. At first, this dissection is done only in areas in which it proceeds reasonably well, leaving the epicardium on the myocardium wherever it is thin and normal. When dissection in this plane is not possible, such as in an area of calcification or dense scarring, islands of calcification and scarring may be left attached to the myocardium but separated from other areas. Dissection moves to the anterior pericardial flap whenever progress ceases posteriorly and vice versa. In the setting of constrictive pericarditis, it is not necessary to dissect across the interventricular groove containing the coronary vessels. But it is important to be certain that all constrictions on the atrioventricular groove are removed, because they can result in gradients between atrium and ventricle.
No special effort is made to free the atria, venae cavae, or cavoatrial junctions because constriction does not occur in these areas. When dissection is complete, the pericardial flaps, as well as the diaphragmatic portion of the pericardium, are excised ( Fig. 18.2 D).
Two pleural drainage catheters are inserted, the tip of one being placed posteriorly and inferiorly and that of the other anteriorly and superiorly. The interspace incision is closed with heavy pericostal and perichondrial absorbable sutures, and the muscle layers are closed with continuous absorbable sutures. The skin is closed with a continuous subcuticular suture.
Choice of surgical approach
The main advantage of median sternotomy is perceived improved ability to free the heart from the pericardium and remove pericardium from the right ventricle. , The main advantages of the left anterolateral approach are cosmetic, exposure afforded for liberation of the left ventricle, and in special situations when sternotomy might be hazardous (e.g., patent left internal mammary artery (LIMA) graft at high risk of injury).
All things being equal, median sternotomy is a better approach when CPB is used, although CPB can be used with the anterolateral approach by cannulating the femoral vessels (see Special Situations and Controversies of Chapter 2 ). The Mayo group favors median sternotomy ( Fig. 18.3 , Fig. 18.4 , Fig. 18.5 , Fig. 18.6 ) and liberal use of use of CPB to assist in performing complete pericardiectomy. , ,
View of the surgical field after removal of the diaphragmatic pericardium and a portion of pericardium posterior to the left (Lt.) phrenic nerve. Note the small strip of residual pericardium beneath the phrenic nerves and in the oblique sinus posterior to the left atrium. IVC , inferior vena cava; Rt. , right.
(From Hemmati A. Contemporary techniques of pericardiectomy for pericardial disease. Cardiol Clin . 2017;35(4):559-566.)
An additional benefit of the median sternotomy approach is the ability to address associated cardiac comorbidity. Pericardiectomy and release of constriction may lead to right ventricular and tricuspid anular dilation with worsening of functional valve regurgitation. , Intraoperative transesophageal echocardiography is an important adjunct to address this issue. If there is moderate or worse tricuspid valve regurgitation after separation from CPB, then tricuspid valve repair is indicated, as untreated significant tricuspid valve regurgitation is associated with reduced long-term survival. ,
Complementary techniques
A high-speed burr or ultrasonic dissector may help to define the epicardial layer and dissect the adherent calcified pericardium away from the myocardium. , Complete resection of all thickened and constrictive epicardium is as important for achieving a good result with complete removal of the parietal pericardium. However, in some patients (particularly those with postoperative or postirradiation constriction), it is impossible to develop a consistent plane of dissection. In cases in which the epicardial peel is exceptionally adherent, a cross-hatching “waffle” procedure described by Heimbecker and colleagues or multiple incisions of the peel (turtle cage operation) allow myocardial expansion and restoration of adequate hemodynamics. , ,
Special features of postoperative care
Postoperative care is given as described in Chapter 4 . Low cardiac output early postoperatively occurs commonly in patients who have advanced disability, fluid retention, and ascites preoperatively. Low cardiac output may require maintaining left atrial pressure at a relatively high level (15 mmHg or more) and using catecholamine infusions for prolong periods of time. Intraaortic balloon pulsation is effective in these patients when cardiac output is low and unresponsive to other measures.
Results
Survival
Early (Hospital) death.
Hospital mortality after pericardiectomy for chronic constrictive pericarditis does not approach zero even in the current era. In an earlier era, early (hospital) mortality was 10% to 15%. , More recently, it has been about 5% or less. The Mayo Clinic group reported a 2.5% operative mortality in 355 patients operated from 1993 to 2013. Others have reported similar results. ,
Time-related survival.
Patients operated on for chronic constrictive pericarditis have a time-related survival that is less than, or in favorable cases similar to, that of an age-gender-ethnicity–matched population ( Fig. 18.7 ). Survival, including hospital deaths, at 1, 5, 10, and 20 years is about 90%, 75%, 65%, and 55%, respectively. ,
