Pericardial Effusion

Pericardial effusion develops in 5% to 15% of patients with cancer and is sometimes the initial manifestation of malignancy. Most pericardial effusions in cancer patients result from obstruction of the lymphatic drainage of the heart secondary to metastases. The typical presentation is that of a patient with known cancer who is found to have a large pericardial effusion without signs of inflammation. Bloody pericardial fluid is not a reliable sign of malignant effusion.

The most common malignant causes of pericardial effusions are lung and breast cancers, leukemias (specifically acute myelogenous, lymphoblastic, and chronic myelogenous leukemia [blast crisis]), and lymphomas. In one report from Kühn et al, at Children’s Hospital Boston, 39% of children with moderate to large pericardial effusions had malignant effusions.

Not all pericardial effusions associated with cancer are malignant, and cases with negative cytology may represent as many as half of cancer-associated pericardial effusions. Many effusions that initially have negative cytology will become positive over time. Such effusions are more common in patients with mediastinal lymphoma, Hodgkin lymphoma, or breast cancer. Other nonmalignant causes include drug-induced (eg, sirolimus [Rapamune] or docetaxel) or postirradiation pericarditis, tuberculosis, collagen diseases, uremia, and congestive heart failure. As reported by Fukada and colleagues, pericardial effusions can be seen in up to 35% of patients after chemoradiotherapy for esophageal cancer, but only 8% of patients will develop symptoms.

Tamponade occurs when fluid accumulates faster than the pericardium can stretch. Compression of all four heart chambers ensues, with tachycardia and diminishing cardiac output. Fluid loading can counteract intrapericardial pressure temporarily. Reciprocal filling of right- and left-sided chambers with inspiration and expiration, secondary to paradoxical movement of the ventricular septum, is a final mechanism to maintain blood flow before death.

Diagnosis

A high index of suspicion is required to make the diagnosis of pericardial effusion.

Signs and symptoms

Dyspnea is the most common symptom, but it is very nonspecific. Patients may also complain of chest pain or discomfort, easy fatigability, cough, and orthopnea, or they may be completely asymptomatic. Signs include distant heart sounds, and pericardial friction rub. With cardiac tamponade, progressive heart failure occurs, with increased shortness of breath, cold sweats, confusion, hypotension, jugular venous distention, and pulsus paradoxus. Pulsus paradoxus is often misunderstood as a “paradoxical” decrease in systolic blood pressure with inspiration. In fact, decrease in systolic blood pressure is a normal physiologic phenomenon with inspiration but is typically less than 10 mm Hg. If pulsus paradoxus is greater than 13 mm Hg, pericardial effusion should be suspected.

Chest x-ray

Chest radiographic evidence of pericardial effusion includes cardiomegaly with a “water bottle” heart; an irregular, nodular contour of the cardiac shadow; and mediastinal widening.

Electrocardiogram

ECG shows nonspecific ST- and T-wave changes, tachycardia, low QRS voltage, electrical alternans, and atrial dysrhythmia.

Pericardiocentesis and echocardiography

An echocardiogram not only can confirm a suspected pericardial effusion but also can document the size of the effusion and its effect on ventricular function. Vignon reported on the accuracy of echocardiography performed by noncardiologist residents with limited training in an ICU and concluded that brief and limited training of noncardiologist ICU residents with no prior training in ultrasound methods appears “feasible and efficient” to address simple clinical questions about using echocardiography and was specifically useful in the diagnosis of pleural and pericardial effusions. A pericardial tap with cytologic examination (positive in 50% to 85% of cases with associated malignancy) will confirm the diagnosis of malignant effusion or differentiate it from other causes of pericardial effusion. Serious complications, including cardiac perforation and death, can occur during pericardiocentesis, even when performed with echocardiographic guidance by experienced clinicians.

Tumor markers/staining and cytogenetics

Tumor markers or special staining and cytogenetic techniques may improve the diagnostic yield, but ultimately an open pericardial biopsy may be necessary. Szturmowicz et al, from Warsaw, Poland, studied pericardial fluid carcinoembryonic antigen (CEA) and cytokeratin fragment (CYFRA) 21-1 levels in 84 patients with pericardial effusion. There were significant differences in patients with malignant vs benign effusions with both tests. With cutoff points of > 100 ng/mL for CYFRA 21-1 and > 5 ng/mL for CEA, 14 of 15 patients who had malignant pericardial effusion with negative cytologic results had a positive result on one or both tests.

CT and MRI

CT and MRI as diagnostic adjuncts may provide additional information about the presence and location of loculations or mass lesions within the pericardium and adjacent structures. Restrepo et al have published a comprehensive, well-illustrated description of CT features of pericardial tamponade.

Cardiac catheterization

This may occasionally be of value to rule out superior vena caval obstruction, diagnose microvascular tumor spread in the lungs with secondary pulmonary hypertension, and document constrictive pericarditis before surgical intervention. Right atrial and pulmonary capillary wedge pressures may also be measured. In most cases of effusion, catheterization does not yield information beyond the echocardiogram.

Pericardioscopy

This allows visualization and biopsy at the time of subxiphoid or thoracoscopic pericardiotomy and can improve the diagnostic yield.

Prognosis

In general, cancer patients who develop a significant pericardial effusion have a high mortality, with a mean time to death of 2.2 to 4.7 months. However, about 25% of selected patients treated surgically for cardiac tamponade enjoy a 1-year survival.

Investigators in Barcelona, Spain studied the effects of volume expansion in patients with large pericardial effusions and pericardial tamponade. They administered 500 mL of normal saline over 10 minutes and measured hemodynamic and echocardiographic parameters. A total of 57% had tamponade on physical examination, and 20% were hypotensive. Volume expansion resulted in increases in mean arterial, intrapericardial, right atrial, and left ventricular end-diastolic pressures. The cardiac index increased by > 10% in 47% of patients and remained unchanged in 22%, but actually decreased in 31%. No patient had clinical complications. Predictors of improved hemodynamics were a pressure below 100 mm Hg and a low cardiac index. Sagrista-Sauleda et al reported in 2008 that in approximately half of patients with cardiac tamponade, particularly those with low blood pressure, cardiac output will increase after volume overload. Therefore, the administration of fluids should be guided by the patient’s clinical status rather than used routinely in patients with suspected tamponade.

Treatment

General concepts

As is the case with malignant pleural effusion, it is difficult to evaluate treatments for pericardial effusion because of the many variables. Because malignant pericardial effusion is less common than malignant pleural effusion, it is more difficult to collect data in a prospective manner. Certain generalizations can, however, be derived from available data:

• All cancer patients with pericardial effusion require a systematic evaluation and should not be dismissed summarily as having an untreatable and/or terminal problem.

• Ultimately, both the management and natural course of the effusion depend on (1) the underlying condition of the patient, (2) the extent of clinical symptoms associated with the cardiac compression, and (3) the type and extent of the underlying malignant disease.

General treatment approaches

Asymptomatic, small effusions may be managed with careful follow-up and treatment directed against the underlying malignancy. On the other hand, cardiac tamponade is a true oncologic emergency. Immediate pericardiocentesis, under echocardiographic guidance, may be performed to relieve the patient’s symptoms. A high failure rate is anticipated because the effusion rapidly recurs unless steps are taken to prevent it. Therefore, a more definitive treatment plan should be made following the initial diagnostic/therapeutic tap.

In patients with symptomatic, moderate-to-large effusions who do not present as an emergency, therapy should be aimed at relieving symptoms and preventing recurrence of tamponade or constrictive pericardial disease. Patients with tumors responsive to chemotherapy or radiation therapy may attain longer remissions with appropriate therapy.

There are two theoretical mechanisms for control of pericardial effusion: (1) creation of a persistent defect in the pericardium, allowing fluid to drain out and be reabsorbed by surrounding tissues; or (2) sclerosis of the mesothelium, resulting in the formation of fibrous adhesions that obliterate the pericardial cavity.

Postmortem studies have demonstrated that both of these mechanisms are operative. The fact that effusions can recur implies that there is either insufficient damage to the mesothelial layer or that rapid recurrence of effusion prevents coaptation of visceral and parietal pericardium and prevents the formation of adhesions. This, in turn, would suggest that early closure of the pericardial defect can result in recurrence.

Treatment methods. Various methods can be used to treat malignant pericardial effusion.

• Observation—Observation alone may be reasonable in the presence of small asymptomatic effusions.

• Pericardiocentesis—Pericardiocentesis is useful in relieving tamponade and obtaining a diagnosis. Echocardiographic guidance considerably enhances the safety of this procedure. About 90% of pericardial effusions will recur within 3 months after pericardiocentesis alone.

• Pericardiocentesis and percutaneous tube drainage—Pericardiocentesis and percutaneous tube drainage can now be performed with low risk and are recommended by some clinical groups. Marcy et al, of Nice, France, reviewed multiple, well-illustrated percutaneous methods for management of malignant pericardial effusions. Problems that may occur include occlusion or displacement of the small-bore tubes, dysrhythmia, recurrent effusion, and infections. Mayo Clinic cardiologists recommend initial percutaneous pericardiocentesis with extended catheter drainage as their technique of choice.

Kunitoh et al, from the National Cancer Center Hospital in Tokyo, performed a randomized controlled trial in 80 patients who had undergone pericardial drainage for malignant pericardial effusion. These patients were then randomized to either observation alone (A) after drainage or intrapericardial bleomycin instillation (15 mg followed by 10 mg every 48 hours [B]). Drainage tubes were removed when daily drainage was 20 mL or less. The results, published in 2009, showed that survival with control of malignant pleural effusion at 2 months was 29% in arm A and 46% in arm B (P = .08); the median survival was 79 days vs 119 days.

• Intrapericardial sclerotherapy and chemotherapy—Intrapericardial sclerotherapy and chemotherapy following percutaneous or open drainage have been reported to be effective treatments by some groups. Problems include pain during sclerosing agent treatments and recurrence of effusions. Good results have been reported with instillation of a number of agents, including bleomycin (10 mg), cisplatin (30 mg), mitomycin (2 mg), thiotepa (1.5 mg), and mitoxantrone (10 to 20 mg). Agents are selected based on their antitumor or sclerosing effect.

Martinoni et al, from Milan, Italy, reported on the use of intrapericardial administration of thiotepa (15 mg on days 1, 3, and 5) following placement of a pericardial drainage catheter in 33 patients with malignant pericardial effusion. There were three recurrent effusions (9.1%). The median survival time was 115 days. They concluded that this protocol is safe, well tolerated, and improves the quality and duration of life.

• Pericardiocentesis and prolonged catheter drainage—Simple pericardiocentesis alone has unacceptable recurrence rates of up to 90%. However, using a guidewire, a pigtail catheter may be inserted into the pericardial space and left in place until the drainage becomes minimal. Percutaneous catheter drainage has a low complication rate, but higher rates of recurrence than pericardial window. In their experience with 246 patients, McDonald and colleagues reported a 16.5 % rate of recurrent symptomatic effusion after percutaneous catheter drainage vs 5% after pericardial window. However, for patients with malignant effusions and limited life expectancy, results may be comparable, as recently reported by Patel and others.

• Balloon pericardial window—After percutaneous placement of a guidewire following pericardiocentesis, a balloon-dilating catheter can be placed across the pericardium under fluoroscopic guidance and a window created by balloon inflation.

At the National Taiwan University, cardiologists performed percutaneous double-balloon pericardiotomy in 50 patients with cancer and pericardial effusion and followed their course using serial echocardiograms. Success without recurrence was achieved in 88%. Fifty percent of patients died within 4 months, and 25% survived to 11 months.

Sidebar: Ruiz-García et al, from Madrid, Spain, treated 16 patients with malignant pericardial effusions, using percutaneous balloon pericardiotomy as the initial and definitive treatment. All patients had been hemodynamically compromised on echocardiography. There were no acute complications and all cases were initially successful. There were three later failures, requiring two pericardial window surgeries and one repeat percutaneous balloon pericardiotomy. The authors consider percutaneous balloon pericardiotomy to be a simple, safe technique that can be effective in preventing recurrence in many patients with severe malignant pericardial effusion (Ruiz-García J et al: Rev Esp Cardiol 66:357–363, 2013).

• Subtotal pericardial resection—Subtotal pericardial resection is seldom performed today. Although it is the definitive treatment, in that there is almost no chance of recurrence or constriction, higher morbidity and longer recovery time render this operation undesirable in patients who have a short anticipated survival time. Its use is restricted to patients with good prognosis and constrictive pericarditis rather than pericardial effusion.

• Limited pericardial resection—Limited pericardial resection (pericardial window) via anterior thoracotomy, thoracoscopic, or subxiphoid approach has a low morbidity. There is a low risk of recurrence. Cardiac herniation is possible if the size of the opening in the pericardium is not carefully controlled. A pericardial drain is typically placed at the time of the procedure. If necessary, a sclerosing agent may also be administered. Subxiphoid pericardial window may be performed with the under local anesthesia or combined with endoscopic instrumentation.

At City of Hope, Cullinane et al reported on 62 patients with malignant disease who had a surgical pericardial window created for management of pericardial effusion. Windows were created either thoracoscopically (32) or by subxiphoid (12) or limited thoracotomy (18) approaches. Primary tumors included NSCLC, breast, hematologic, and other solid-organ malignancies. Three recurrent effusions (4.8%) required reoperations. Eight patients (13%) died during the same admission as their surgical procedure. The median survival was much shorter for patients with NSCLC (2.6 months) than for patients with breast cancer (11 months) or hematologic malignancy (10 months). The surgical pericardial window is a safe and durable operative procedure that may provide extended survival in certain subgroups of cancer patients.

• Development of a subxiphoid pericardioperitoneal window—Development of a subxiphoid pericardioperitoneal window through the fused portion of the diaphragm and pericardium allows continued drainage of pericardial fluid into the peritoneum. This may be done laparoscopically in stable patients, but we advise caution. As described in a case series by Romano and Glass, carbon dioxide pneumoperitoneum may adversely affect the cardiopulmonary hemodynamics due to increased intra-abdominal pressure, caval compression, and decreased venous return. There will also be increased afterload. In addition, there is the potential for prolonged hypercarbia, acidosis, and hypoxemia.

Toth et al, from Miskolc, Hungary, reported in 2012 on a new technical method using a Chamberlain, parasternal mediastinoscopic approach to create a pericardial window in 22 patients with malignant pericardial effusions. There were no operative deaths, and one patient (4.5%) experienced recurrence of a pericardial effusion.

• Technical factors—Prior pleurodesis for malignant pleural effusion makes an ipsilateral transpleural operation difficult or impossible. In lung cancer patients, major airway obstruction may preclude single-lung anesthesia and, thus, thoracoscopic pericardiectomy. Prior median sternotomy may prohibit the use of a subxiphoid approach.

• Complications—A 30-day mortality rate of 10% or higher has been reported for all of these modalities but is related more to the gravity of the underlying tumor and its sequelae. A small percentage of patients will develop severe problems with pulmonary edema or cardiogenic shock following pericardial decompression. The mechanisms of these problems are poorly understood.

Wagner et al, from Memorial Sloan-Kettering Cancer Center, retrospectively studied 179 consecutive pericardial windows for malignant effusions over a 5-year period. These included lung (44%), breast (20%), hematologic (10%), and gastrointestinal cancers (7%). The overall survival of the whole group was poor, with a median of 5 months survival. They defined paradoxical hemodynamic instability (PHI) as hypotension and shock in the immediate postoperative period. PHI occurred in 19 (11%) of patients. Patients most likely to have PHI showed evidence of tamponade on echocardiogram (89% vs 56% without tamponade; P = .005), had a positive cytology or pathology (68% vs 41%; P = .03), and had larger volumes drained. Most important, 58% of patients with paradoxical hemodynamic instability did not survive the initial hospitalization.

Late neoplastic pericardial constriction can occur following initially successful partial pericardiectomy. Patients with combined malignant pericardial and pleural effusions will often have relief of recurrent pleural effusion following control of pericardial effusion, perhaps because reducing systemic venous pressure results in reduced production of pleural fluid. Simultaneous pleurodesis in the left side of the chest following a pericardial window procedure might increase the incidence of recurrent pericardial effusion and should be avoided.

• Radiotherapy—External-beam irradiation is utilized infrequently in this clinical setting but may be an important option in specialized circumstances, especially in patients with radiosensitive tumors who have not received prior radiation therapy. Responses ranging from 66% to 93% have been reported with this form of treatment, depending on the type of associated tumor.

• Chemotherapy—Systemic chemotherapy is effective in treating pericardial effusions in patients with lymphomas, hematologic malignancies, or breast cancer. Long-term survival can be attained in these patients. If the pericardial effusion is small and/or asymptomatic, invasive treatment may be omitted in some of these cases. Data are limited regarding the effectiveness of systemic chemotherapy or chemotherapy delivered locally in prevention of recurrent pericardial and pleural effusion. Several studies have reported on the effectiveness of intrapericardial instillation of chemotherapy (most commonly cisplatin) for the treatment of malignant effusion. It is unclear whether the results are due to cytotoxic effect on malignant cells or the sclerosing effect of the drugs.

Biologic therapy with various agents is in the early stages of investigation.