Pleural Effusion Secondary to Diseases of the Heart
In this chapter, the pleural effusions that occur after coronary artery bypass graft (CABG) surgery, those that occur after cardiac injury (Dressler’s syndrome), those that occur concomitantly with pericardial disease, and those that occur after heart transplantation are discussed.
POST-CORONARY ARTERY BYPASS SURGERY
More than 500,000 CABG procedures are now performed annually in the United States (1). Because pleural effusions complicate many of these procedures, pleural effusions that occur after CABG are one of the most common types of effusion.
Incidence
In the period immediately following CABG, there is a very high incidence of pleural effusion. In one study of 152 patients who had undergone CABG surgery, the incidence of pleural effusion on routine chest radiographs 7 days postoperatively was 42% (2). In a subsequent study, patients underwent chest ultrasound on the 7th, 14th, and 30th postoperative day (3). In this latter study, the incidence of pleural effusion was 89.4% on the 7th postoperative day, 76.6% on the 14th postoperative day, and 57.4% on the 30th postoperative day (3). In a more recent prospective study, the prevalence of pleural effusions in 349 patients was 63% 30 days postoperatively (4). The prevalence of pleural effusion appears similar whether the patient has valve surgery or not in addition to the CABG (4).
In patients who have undergone CABG, there is a substantial incidence of large pleural effusions although most effusions are small (5). In a recent study of 349 patients, the incidence of pleural effusion that occupied more than 25% of the hemithorax 30 days after CABG was 9.7% (4). In this study, the incidence of larger effusions was higher in the patients who received internal mammary grafts (10.9%) than it was in patients who received only saphenous vein grafts (4.5%) (4). Hurlbut et al. (6) reported that 4% of 100 patients who had undergone CABG developed moderate-to-large effusions. If 10% of all the patients undergoing CABG develop a moderate-to-large pleural effusion, then the exudative pleural effusion following CABG is one of the most common types of exudative pleural effusion.
Pathogenesis and Pathologic Features
In general the larger pleural effusions, those occupying more than 25% of the hemithorax, that occur after CABG surgery can be divided into those that reach their maximal size within the first 30 days of surgery and those that reach their maximal size more than 30 days after surgery (7). The etiology of the large early pleural effusion after CABG surgery is probably related to trauma to the pleura during surgery (3). Patients undergoing internal mammary artery (IMA) grafting are more likely to have a pleural effusion than those undergoing only saphenous vein grafting (SVG) (3). Patients undergoing bilateral IMA bypasses are more likely to have a pleural effusion than those undergoing unilateral IMA bypasses (8). Patients with a pericardial effusion postoperatively are more likely to have a pleural effusion, but it is likely that both are a result of trauma rather than one being responsible for the other (3). The effusions
occurring early in the postoperative period are frequently bloody. In one series, the mean pleural fluid red blood cell count in 45 patients with large pleural effusions within the first 30 days of surgery exceeded 2,000,000/mm3 (7), which is equivalent to a hematocrit of 20%.
occurring early in the postoperative period are frequently bloody. In one series, the mean pleural fluid red blood cell count in 45 patients with large pleural effusions within the first 30 days of surgery exceeded 2,000,000/mm3 (7), which is equivalent to a hematocrit of 20%.
The etiology of the effusions that occur more than 30 days after CABG is not known. The fluid is an exudate with predominantly lymphocytes (7). Because the fluid is an exudate, the effusion is probably not due to congestive heart failure. The presence of the lymphocytes suggests an immunologic basis. Pleural biopsies obtained within the first few months of surgery demonstrate an intense lymphocytic pleuritis (9). Immunohistochemical staining demonstrates that the lymphocytes in the pleural tissue are both T lymphocytes and B lymphocytes with a predominance of B lymphocytes (9). The effusions have been attributed to the post-cardiac injury syndrome (PCIS) (10). This explanation, however, is unsatisfactory because patients with PCIS usually have fever, chest pain, pericarditis, and pneumonitis in addition to the pleural effusion. Patients with the late pleural effusions after CABG usually do not have fever, chest pain, pericarditis, or pneumonitis (4). Possibly, the late pleural effusion after CABG is a variant of or a limited variety of the PCIS (7).
The administration of topical hypothermia through iced slush during CABG surgery appears to be associated with a higher prevalence of pleural effusion. In one study of patients receiving only saphenous venous grafts, the prevalence of pleural effusion was 50% in 50 patients receiving topical hypothermia, but only 18% in 50 patients not receiving topical hypothermia (11). In a second study of 505 nonrandomized, consecutive patients undergoing CABG surgery, 60% of the 191 patients who received topical hypothermia had a pleural effusion, whereas only 25% of the 314 patients who did not receive topical hypothermia had a pleural effusion (12). The explanation for the association between iced slush and the presence of pleural effusion is not known, but it has been speculated that cold injury to the phrenic nerve may cause atelectasis (5).
It has been hypothesized that the development of the pleural effusion post-CABG is due, at least in part, to the patients being on cardiopulmonary bypass. This is not definitely the case, however, because in two small series (13,14) the prevalence of pleural effusion was actually higher in patients who had off-pump coronary artery bypass surgery than in those who had on-pump surgery. However, in a series from Nashville, the prevalence of effusion at 30 days postoperatively that occupied more than 25% of the hemithorax was only 3% in the off-pump group (15) whereas in a previous study from the same hospital it had been 10% in the on-pump group (4).
Clinical Manifestations
Dyspnea is the only symptom that most patients with pleural effusions experience after CABG (4). Pleuritic chest pain, chest wall tenderness, fever, pneumonitis, and pericarditis are all unusual. In one series of 29 patients with large pleural effusions, 75.9% complained of dyspnea, 10.3% complained of chest pain, and only 1 (3.4%) complained of fever (4).
The pleural effusions that occur after CABG surgery tend to be unilateral on the left side. In the study using ultrasound in which 42 of 47 patients had pleural effusions on the seventh postoperative day, 17 (40%) of the effusions were unilateral on the left, 24 (57%) were bilateral, and 1 (2%) was unilateral on the right (3). By the 30th postoperative day, there were 27 patients with effusions, and 18 (67%) of these were unilateral left sided, 8 (30%) were bilateral, and 1 (4%) was unilateral right sided (3).
In studies of patients who have larger pleural effusions, the effusions are usually left sided, or if they are bilateral, they are larger on the left. In the study by Sadikot et al. (7) of 71 patients with post-CABG pleural effusions who underwent thoracentesis, 42 of the effusions (59%) were unilateral left sided, 18 (25%) were bilateral and usually larger on the left, and 11 (15%) were unilateral on the right.
As mentioned in the preceding text, the larger pleural effusions that occur after CABG can be divided into those that occur within the first 30 days of the surgery and those that occur more than 30 days after surgery (7,16). The late effusions do not appear to evolve from the early effusions. The characteristics of the pleural fluid in the two situations are quite different. The pleural fluid with the early effusions is bloody, with a mean red blood cell count of approximately 2,000,000/mm3 (7,16). The pleural fluid is frequently eosinophilic, with a mean eosinophil percentage of greater than 40% (7). The pleural fluid eosinophilia is probably due to the blood in the pleural space. Patients with eosinophilic pleural effusions post-CABG also tend to have peripheral eosinophilia (17). There is a significant correlation between the percentage of eosinophils in the pleural fluid and the serum, although the percentage of eosinophils in the serum is lower (17). In these patients, the eosinophilia is correlated with the
levels of interleukin-5 and eotaxin-3 in the pleural fluid, which are higher than the corresponding levels in the serum (17). The mean pleural fluid lactate dehydrogenase (LDH) with the bloody effusions is approximately twice the upper limit of normal for serum (7). It is likely that much of the pleural fluid LDH is LDH-1, which is the LDH from the red blood cells. The pleural fluid protein is in the exudative range, and the pleural fluid glucose level is not reduced (7).
levels of interleukin-5 and eotaxin-3 in the pleural fluid, which are higher than the corresponding levels in the serum (17). The mean pleural fluid lactate dehydrogenase (LDH) with the bloody effusions is approximately twice the upper limit of normal for serum (7). It is likely that much of the pleural fluid LDH is LDH-1, which is the LDH from the red blood cells. The pleural fluid protein is in the exudative range, and the pleural fluid glucose level is not reduced (7).
In contrast to the bloody exudates that were discussed in the preceding text, the pleural fluid that occurs more than 30 days after CABG is a clear yellow lymphocyte-predominant exudate. The mean lymphocyte percentage for 26 late effusions in one series was 61%, whereas the mean eosinophil percentage was only 2% (7). The pleural fluid LDH tends to be lower with the late effusions than with the early effusions and averages about the upper limit of normal for serum (7). As with the early effusions, the pleural fluid protein is in the exudative range and the pleural fluid glucose level is not reduced (7). A small percent of patients will develop dyspnea from a pleural effusion more than 90 days post-CABG surgery (18). Most effusions occurring this long after surgery are transudates due to heart failure (18).
Diagnosis
The diagnosis of pleural effusion secondary to CABG is one of exclusion. In the days immediately after CABG, the main diagnoses to exclude are congestive heart failure, pulmonary embolus, parapneumonic effusion, and chylothorax. Congestive heart failure is excluded if the patient has an exudative pleural effusion. Chylothorax is excluded if the fluid is clear yellow or if the triglyceride levels are low. Pulmonary embolus is more difficult to exclude, and a computed tomography (CT) angiography is necessary in some cases (see Chapter 17). However, the pleural effusion with pulmonary embolus usually occupies less than 25% of the hemithorax and disappears spontaneously within a couple of weeks. Patients with parapneumonic effusions are usually febrile, and the pleural fluid differential white blood cell (WBC) count reveals predominantly neutrophils and a very low percentage of eosinophils.
The differential diagnosis is somewhat different for the late pleural effusion occurring after CABG, and the main diagnoses to consider are congestive heart failure, chylothorax, tuberculosis, malignancy, constrictive pericarditis, and pulmonary embolus. As with the early effusion, the diagnosis of congestive heart failure is eliminated if the patient has an exudative pleural effusion and the diagnosis of chylothorax is excluded if the patient’s pleural fluid is clear or has a low triglyceride level. With a lymphocyte-predominant pleural effusion, one must exclude tuberculosis. Because the adenosine deaminase (ADA) level is less than 40 IU/L in patients with pleural effusions after CABG (19) and is above this level in patients with tuberculous pleuritis, demonstration of an ADA below 40 IU/L virtually excludes the diagnosis of tuberculous pleuritis. Patients with constrictive pericarditis will usually have other signs and symptoms such as bilateral pedal edema and ascites.
Treatment
Most patients with smaller pleural effusions post-CABG require no treatment as the effusion gradually disappears (5). When a patient is identified with a large pleural effusion after CABG (occupying more than 25% of the hemithorax), a thoracentesis should be performed to exclude the other diagnoses in the differential outlined earlier. Because most of these patients are dyspneic (4) and the therapy of choice for these effusions is a therapeutic thoracentesis, it is recommended that the initial thoracentesis be a therapeutic thoracentesis (5).
If the other diagnostic possibilities are excluded and the fluid recurs, a second and then a third therapeutic thoracentesis are indicated. Many patients are also given nonsteroidal anti-inflammatory agents (NSAIDs) or oral prednisone, but there are no controlled studies documenting the efficacy of this approach. There is one study that evaluated the effectiveness of diclofenac, an NSAID, in preventing pleural effusion in the immediate postoperative period (20). Niva et al. (20) randomized patients to receive 50 mg diclofenac (22 patients) or placebo (19 patients) orally every 8 hours in the postoperative period. They reported that the control group had a higher incidence of pleural effusion (42.1%) at discharge than did the diclofenac-treated group (22.7%) (20). In a second study, Imazio et al. (20a) studied whether colchicine would reduce the incidence of effusions post cardiac surgery in a double-blind randomized study of 360 patients. They reported that the incidence of pleural effusion was significantly less (12.2%) in the group that received colchicine starting on the third postoperative day and continuing for a month than it was in the group that received placebo (22.8%) (21).
Some patients have been managed successfully with chemical pleurodesis. Before one becomes too aggressive in managing this condition, it is important to realize that most patients will do well with no more than a couple of thoracenteses. In our prospective study, we followed 30 patients with pleural effusions occupying more than 25% of the hemithorax for 12 months. During this period, 8 (27%) received no invasive treatment for the pleural effusion, 16 (53%) received a single thoracentesis, 2 (7%) received two thoracenteses, and 4 (13%) received three or more thoracenteses. Only one patient was still receiving periodic thoracenteses 12 months after CABG (4). Twenty-two of the 25 patients who underwent thoracentesis reported that their dyspnea was alleviated with the thoracentesis (4).
On occasion, the effusion persists despite several therapeutic thoracenteses. We have subjected eight such patients to thoracoscopy in recent years. At thoracoscopy, several patients had thin sheets of fibrous tissue that coated the lung and prevented it from expanding (9). It is likely that this sheet of fibrous tissue “trapped” the lung and prevented it from reexpanding. After the fibrous tissue coating the visceral pleura was removed, the lung expanded and the effusion did not recur (9). However, because most had a mechanical or a chemical pleurodesis at the same time, one cannot be certain that the decortication was responsible for the effusion not recurring. Recurrent pleural effusions post-CABG surgery have also been managed successfully with video-assisted thoracic surgery with talc insufflation (21). No mention was made in this latter paper about membranes encasing the visceral pleural (21).
In view of the series mentioned in the preceding text, thoracoscopy is recommended for an effusion after CABG that continues to recur for several months despite several therapeutic thoracenteses. At thoracoscopy, any fibrous tissue coating the visceral pleura should be removed and the parietal pleura should be abraded to create a pleurodesis.
Pleural Effusion After Ventricular Assist Device Placement
Patients who receive ventricular assist devices are being used more frequently as bridges to cardiac transplantation or to facilitate patient transfers. It appears that most placements of this device are associated with the development of a pleural effusion (22). Guha et al. (22) reviewed the placement of 22 of these devices and reported that every patient had a pleural effusion after placement (22). Six of the 22 patients (18%) had effusions before surgery. Nine patients required thoracentesis to relieve dyspnea and the median time from placement to thoracentesis was 23 days. All the pleural fluids examined were blood-tinged exudates (22).
POST-CARDIAC INJURY (DRESSLER’S) SYNDROME (PCIS)
PCIS is characterized by the onset of fever, chest pain, pleuropericarditis, and parenchymal infiltrates in the weeks following injury to the pericardium or myocardium (23,24). There is no universal agreement on the definition of the PCIS (22). Mott et al. (25) defined noncomplicated PCIS as the presence of a temperature greater than 100.5°F, patient irritability, pericardial friction rub, and a small pericardial effusion with or without pleural effusion following cardiac trauma (25). They defined complicated PCIS as a noncomplicated PCIS plus the need for hospital readmission with or without the need for pericardiocentesis or thoracentesis (25). I prefer the proposed definition of Imazio et al. (26) who define the PCIS as the presence of at least two of the following: fever without alternative explanation, pleuritic chest pain, pericardial friction rub, new or worsening pleural effusion, and new or worsening pericardial effusion. This syndrome has been described following myocardial infarction, cardiac surgery, blunt chest trauma, percutaneous left ventricular puncture, pacemaker implantation, angioplasty, and repair of pectus excavatum (27).