Discussion of the myriad causes of pericarditis and pericardial effusion is beyond the scope of this chapter, but suffice it to say that virtually any pathologic process can affect the pericardium¹ and cause a detectable pericardial effusion whenever the rate of accumulation of pericardial contents (transudate, exudate, pus, blood, or gas) exceeds the reabsorption ability of the pericardium. The most common causes of pericardial effusion include idiopathic pericarditis (presumably viral), trauma (including iatrogenic catheter injury to the cardiac chambers or vessels), malignancy, postmyocardial infarction, uremia, connective tissue disease, autoimmune disorders, myxedema, and radiation.²,³ Infectious or purulent pericarditis is caused most commonly by Staphylococcus aureus, followed by other Gram-positive organisms, fungi or mycobacteria.⁴,⁵ Myocarditis and pericarditis frequently coexist, and elevations of cardiac troponin levels can be seen in both.⁶ Acute pericardial injury of any type can also trigger an autoimmune process that can lead to continued or recurrent effusion.

Patients with acute pericarditis often experience sharp, aching, or pressurelike precordial pain that is worsened by cough, inspiration, and recumbency. Pain may radiate to the shoulders⁷ and may be confused with angina or acute myocardial infarction, particularly since the ECG may show diffuse concave ST-segment elevation. One distinguishing feature, however, is the concomitant depression of the PR segment and the absence of reciprocal ST depression. A pericardial friction rub and fever may be present, but rigors or spiking fevers should raise the concern of purulent pericarditis. The mainstay of diagnosis is echocardiography, which clearly shows an effusion. Fluid first accumulates posteriorly and then extends anteriorly. Small effusions have <10-mm clear space between the heart and parietal pericardium, moderate effusions have a 10- to 20-mm gap, and large effusions have a >20-mm gap. The size of the effusion correlates roughly with prognosis.⁸ Effusions may be loculated (partitioned) by nonuniform fibrous adhesions that form between the parietal and visceral pericardium, a pattern typically observed after cardiac surgery. Echocardiography also reveals the extent to which the pericardial pressure is compromising cardiac function, showing early diastolic collapse of the right ventricle as pericardial pressure transiently exceeds intracavitary pressure. In idiopathic pericarditis, the effusion usually resolves spontaneously or after symptomatic treatment with nonsteroidal anti-inflammatory agents and sometimes with colchicine⁹ or corticosteroids.

Chronic idiopathic effusions may also persist without symptoms or signs of tamponade despite effusion, and volumes >500 mL may be followed conservatively with serial echocardiograms if asymptomatic.¹⁰ On the other hand, significant pericardial effusion with even early signs of hemodynamic compromise is usually an indication for prompt drainage (pericardiocentesis, or surgical subxiphoid window placement). The pericardial pressure is normally subathmospheric (−5 to +5 mmHg) and it tracks the intrathoracic pressure during the inspiratory/expiratory cycle. As discussed in details in Chapter 23, during inspiration the pericardial pressure falls more than the systemic venous pressure. The net result is an increase in right atrial filling gradient leading to augmentation of atrial filling and right ventricular stroke volume. If an excess of pericardial fluid accumulates in the pericardial space beyond its limited capacity to stretch, the pericardial pressure rises and begins to progressively compress the underlying cardiac chambers. The diastolic y descent becomes absent, owing to impairment of rapid atrial emptying secondary to compression by the pericardial effusion. By the time the classic bedside findings of jugular venous distension and pulsus paradoxus (fall in systolic arterial pressure on normal inspiration) develop, only a small further accumulation of fluid separates the patient from frank hemodynamic collapse. As shown in Figure 44.1, the abnormal physiology of cardiac tamponade is relieved when (a) the pericardial pressure falls to a level ≤0 mmHg and separates from the right atrial pressure and (b) the right atrial pressure falls to the normal range and exhibits return of a normal diastolic y descent, indicating restoration of normal rapid atrial emptying and early ventricular diastolic filling.

The main pericardial procedure performed in the catheterization laboratory is needle puncture and catheter drainage of pericardial fluid—pericardiocentesis (Chapter 38). Diagnostic pericardiocentesis may be performed to evaluate the etiology of pericarditis, particularly for suspected purulent or tuberculous pericarditis, persistence or recurrence of a large effusion, or a high suspicion of malignant effusion without a tissue diagnosis from the primary site. The diagnostic yield for pericardial fluid analysis is quite low, with as few as 6% of diagnostic procedures and 29% of therapeutic procedures yielding useful information.¹¹, ¹², ¹³ Diagnostic yield tends to be higher in large pericardial effusions.¹⁴

Pericardial fluid always should be sent for cell count; Gram, acid-fast bacteria (AFB), and special stains; cultures (aerobic, anaerobic, AFB, fungal); and cytology. Although differentiation between an exudate and a transudate can be accomplished with determination of fluid protein and LDH levels, such differentiation is of dubious value because the timing of sampling and effects of treatment may alter values and because sensitivity and specificity are poor with significant overlap between the exudative and transudative groups.¹³ Since most effusions are serosanguineous or hemorrhagic, fluid appearance is not useful in differentiating various etiologic groups with the exception of purulent fluid, which is specific but not sensitive for infection. As long as adequate samples are obtained, fluid cytology has 92% to 95% sensitivity and 100% specificity for malignant pericardial effusion, although nonmalignant effusions are also common in patients with underlying malignancy. There are no specific pericardial fluid findings for postpericardiotomy syndrome, radiation or uremic pericarditis, hypothyroidism, or trauma. Special circumstances may require additional analysis for the following: viral cultures for viral infection; fluid cholesterol level in myxedema; fat studies for chylopericardium; latex fixation for rheumatoid antigen; gamma globulin complexes and fluid complement levels for rheumatoid arthritis; fluid antinuclear antibody levels for systemic lupus erythematosus; and tuberculosis stains and/or cultures, fluid adenosine deaminase determination,¹⁵ or polymerase chain reaction¹⁶ for tuberculosis (see also Chapter 38). Purulent pericarditis may be associated with low pH, high protein levels, glucose levels <35 mg/dL, and elevated leukocyte counts in the range of 6,000 to 240,000/mm³.⁴,¹⁷

The diagnostic yield of pericardiocentesis may be increased by retrieval of pericardial tissue by a surgical pericardial biopsy performed via thoracotomy, subxiphoid incision, or thoracoscopy. Percutaneous pericardial biopsy (assisted by pericardioscopy) has led to a specific diagnosis in 49% to 53% of patients in several limited series.¹⁸,¹⁹ The pericardioscopy technique is not widely available and requires surgical or 16F percutaneous pericardial access. In hemorrhagic effusion, obtaining adequate parietal pericardial visualization may require 2 to 3 days of active drainage, replacement of pericardial effusion with saline, and instillation of 100 to 300 mL of air.¹⁹ Several small uncontrolled series suggest that this approach may increase the likelihood of obtaining a definitive diagnosis in patients with large recurrent pericardial effusions and in patients in whom there is a strong clinical suspicion of malignant pericarditis or tuberculous pericarditis. In a prospective series of 141 patients with unexplained pericardial effusions who underwent 142 surgical pericardioscopy including cytological fluid analysis, visualization of the pericardium, and guided biopsy, a specific etiologic cause (neoplastic, infected purulent, or sterile radiation-induced effusion) was identified in 49%, whereas 51% were considered idiopathic. Of note, an unrecognized cause not detected by pericardioscopy-biopsy was subsequently discovered in 4%.¹⁹ No deaths were attributable to pericardioscopy but in-hospital mortality was 5.6% related to underlying disease.

In another series, 35 patients with pericardial effusion owing to inflammatory disease underwent pericardioscopy, and pericardial biopsies were performed. Diagnoses of viral pericarditis, lymphocytic perimyocarditis, bacterial pericarditis, and antibody-mediated autoreactive pericarditis were obtained; however, it was unclear if this resulted in a change of management strategy.²⁰ The same authors reported a series of 14 patients with idiopathic pericarditis
and 15 patients with malignant pericarditis who underwent percutaneous pericardioscopy with both pericardial and epicardial biopsy from a registry of 136 patients undergoing pericardiocentesis. In this study, subxiphoid pericardiocentesis and sampling of fluid for cytology, immunologic studies, and culture were performed first, followed by evacuation of pericardial fluid, placement of warmed clear sterile saline in the pericardial sac, and introduction of both rigid and flexible pericardioscopes. Both epicardial biopsies and pericardial biopsies were obtained with a resterilizable bioptome, after site selection by both pericardioscopy and biplane fluoroscopy, before the saline was evacuated. In this series of patients with proven malignant pericarditis, fluid cytology was diagnostic in 71% and epicardial biopsy in 80%.²¹

In a more recent study by Seferovic, 49 patients with a large pericardial effusion underwent parietal pericardial biopsy using fluoroscopy or pericardioscopy guidance (for more extensive sampling). Diagnostic efficiency was improved by extensive sampling (4-20 biopsies) as compared with a single biopsy¹⁹ (40% versus 8.3%, P < 0.05). In our experience and in several series of pericardiocentesis in patients with malignant effusion, cytologic examination is positive in about 80% to 85% of cases and false-negative cytologic analysis is rare in carcinomatous pericarditis (as opposed to malignant involvement by lymphoma or mesothelioma). In the era of molecular diagnostic tools, a clear role for diagnostic pericardioscopy and directed pericardial and/or epicardial biopsy is not yet defined relative to cytologic examination of recovered fluid.

Therapeutic pericardiocentesis is indicated for any sign or symptom of tamponade, limiting dyspnea,¹⁴ symptoms of compression of surrounding structures such as lung or esophagus, and acute hemopericardium with circulatory compromise following a catheter-based or surgical intervention. The diagnosis of early tamponade is usually made echocardiographically. Hemodynamically, the diagnosis of tamponade is made if there is identical elevation of left- and right-side diastolic pressures with loss of the y descent in a patient with pericardial effusion. However, it may be noted that pericardiocentesis is contraindicated for hemopericardium or tamponade in the presence of a diagnosed ascending aortic dissection, as it can accelerate bleeding and shock,²² making immediate surgery preferable in such situations. Another particular example in which emergency pericardiocentesis is required is when catheter injury to a cardiac chamber or vessel produces acute hemopericardium. This may happen as the result of temporary pacemaker placement in the right ventricle, passage of a stiff right heart catheter, endomyocardial biopsy, trans-septal puncture, retrograde crossing of a stenotic aortic valve with a straight guidewire, coronary atherectomy or high-pressure stent dilation, or passage of a stiff hydrophilic guidewire into a small coronary branch in a patient receiving a glycoprotein IIb/IIIa platelet antagonist.

The patient may complain of chest pain, followed by progressive hypotension and tachycardia. Bradycardia may also appear owing to stimulation of the vagal nerve by blood in the pericardium, but the hypotension persists even after the bradycardia is resolved by atropine administration. Careful fluoroscopy of the right and left heart borders may show that the normal pulsations of the heart have been replaced by an immobile tense pericardial shadow, and right heart catheterization may show the classic hemodynamic findings described below as well as inability to advance the right heart catheter fully to the right atrial border. Urgent pericardiocentesis may be lifesaving in this situation. When the cause of bleeding into the pericardium is injury to a coronary artery, either placement of a fabric-covered stent or coil embolization of a small bleeding distal branch may be required. Ongoing bleeding after initial drainage and reversal of anticoagulation, however, usually warrants emergency surgery.

While percutaneous pericardiocentesis is the procedure commonly used for acute pericardial tamponade, surgical or balloon pericardiotomy is performed for recurrent effusions. The techniques of pericardiocentesis, balloon pericardiotomy, and the pericardial approach to epicardial ablation for ventricular arrhythmias and for left atrial appendage exclusion are described in Chapter 38.