Transient Systolic Dysfunction after Relief of Cardiac Tamponade

Case Reports

Case 1

A 46-year-old white woman was hospitalized because of increasing shortness of breath. Adenocarcinoma of the right breast had been detected 12 years previously. After a radical mastectomy, she received eight cycles of cyclophosphamide, methotrexate, and 5-fluorouracil (CMF) with adjunctive radiation therapy. She had done well until 5 months before the current hospitalization, at which time she was diagnosed with deep venous thrombosis and pulmonary emboli. She received anticoagulant therapy and inferior vena caval filter placement. The echocardiographic study done at that time showed a small pericardial effusion, a moderate increase in right ventricular dimensions, and normal left ventricular function. She was asymptomatic after therapy until 2 weeks before admission, at which time she developed accelerating shortness of breath.

Physical examination at admission revealed a pulse of 140 beats/min; a blood pressure of 90/50 mm Hg, with a pulsus paradoxus of 30 mm Hg; and a respiratory rate of 20 to 24 breaths/min. Auscultation revealed scattered rhonchi on chest examination and distant heart sounds without murmurs or gallops on cardiac examination. An echocardiographic study revealed a large circumferential pericardial effusion with right atrial and ventricular collapse during diastole, suggesting a compressive effect of the pericardial fluid. Removal of 650 mL of serosanguineous fluid by pericardiocentesis was accompanied by a decrease in intrapericardial pressure from 17 to 0 mm Hg, a decrease in heart rate from 146 to 115 beats/min, and an increase in systolic blood pressure from 90 to 110 mm Hg. Cytologic examination of the pericardial fluid was consistent with the presence of adenocarcinoma.

The following morning, bibasilar rales, a displaced apical cardiac impulse on cardiac palpation, and a soft third heart sound were noted. A repeated echocardiogram showed a small residual pericardial effusion, no evidence of chamber collapse, and remarkably diffuse hypokinesis of the left ventricle. Gradual improvement in left ventricular function was noted during the next week, with resolution of pulmonary congestion and third heart sound. She showed improved functional capacity, progressing from bed immobilization to ambulation on the ward with relatively mild dyspnea. Serial echocardiograms during this time documented substantially improved left ventricular systolic function, with the ejection fraction returning to baseline by day 7 of hospitalization (Figure 1).

Case 2

A 50-year-old woman was seen for possible cardiac tamponade after the development of hypotension and pulsus paradoxus during hospitalization. Six years previously she had undergone right modified radical mastectomy for estrogen receptor-positive adenocarcinoma. Three years later, she underwent left modified radical mastectomy, and adjunctive radiation and chemotherapy with CMF and tamoxifen were administered when metastatic disease was noted. She had done well until 2 weeks before admission; at that time, she noted the onset of nausea, vomiting, anorexia, and dyspnea with any exertion. A chest roentgenogram and a computed tomographic scan of the chest showed a large pericardial effusion and bilateral pleural effusions. One liter of serosanguineous fluid was removed by left-side thoracentesis without significant relief of dyspnea. She complained of dyspnea at rest, and the systolic blood pressure decreased from 110 to 90 mm Hg, with a pulsus paradoxus of 20 mm Hg. A transthoracic echocardiogram showed a large circumferential pericardial effusion with right atrial and right ventricular collapse throughout most of diastole, with filling occurring only during the presystolic moment. The left ventricle was markedly reduced in size but had apparently normal wall motion. Pericardiocentesis was done and 650 mL of bloody fluid was removed, resulting in a reduction in right atrial pressure from 22 mm Hg to 7 mm Hg and a reduction in intrapericardial pressure from 22 mm Hg to 1 to 2 mm Hg. Cytologic examination of the pericardial fluid showed malignant cells.

Despite stabilization of the systemic blood pressure after pericardiocentesis, the patient continued to have dyspnea with any significant exertion. A third heart sound was evident. An echocardiogram obtained 36 hours after pericardiocentesis showed no significant pericardial effusion but did show a severe reduction in global left ventricular systolic function; the estimated ejection fraction was 25% and antero-apical akinesis and apical dyskinesis were present. Serial echocardiograms taken during a period of 10 days documented substantial improvement in left ventricular systolic function and marked improvement in wall motion (see Figure 1). She was discharged and recovered her baseline ambulatory function within 2 weeks.

Discussion

Experimental and clinical studies have shown that cardiac tamponade does not appear to impair left ventricular systolic function and that left ventricular systolic function does appear to improve after relief of tamponade, despite significant increases in systolic and diastolic volumes [2-4]. We described two women with recurrent breast cancer and malignant pericardial effusions in whom global left ventricular dysfunction developed after relief of tamponade and who spontaneously recovered during a 2-week period. Neither patient had previous evidence of left ventricular dysfunction. None of the antineoplastic therapies they received have been associated with the clinical syndromes described [5]. Direct myocardial involvement of tumor was never observed on multiple imaging examinations, and recovery of function without intervening antineoplastic therapy argues against such a cause.

Increased intrapericardial pressures and altered ventricular hemodynamics accompanying tamponade affect both diastolic and systolic volumes and thus stroke volume [6]. Depression of left ventricular systolic function because of decreased coronary blood flow during tamponade was documented in early canine models of tamponade [3, 7, 8], as were microscopic ischemic lesions and subendocardial hemorrhage [7]. In their model, tamponade resulted in a disproportionately greater decrease in coronary blood flow when compared with a canine model of hemorrhagic shock producing the same degree of circulatory impairment [7]. Interestingly, in our patients, left ventricular systolic function appeared to be normal and no wall motion abnormalities were detected during tamponade. Only after pericardiocentesis was depressed was ventricular function noted, suggesting either that occult dysfunction was masked by the reduction in chamber sizes and associated tachycardia was seen during tamponade or that dysfunction was triggered by the hemodynamic alterations that followed removal of pericardial fluid and alleviation of the tamponade. The interplay between the sympathetic and the parasympathetic branches of the autonomic nervous system are also likely to influence cardiac function before and after pericardiocentesis. Martins and colleagues [8] showed that although exogenous catecholamine administration improved coronary blood flow in patients with tamponade, filling pressures remained unchanged and cardiac index was only modestly increased. They proposed that the sympathetic nervous system was already activated and that elevated endogenous catecholamine levels muted any potential benefit. The removal of the stimulus for sympathetic outflow (that is, the alleviation of tamponade by pericardiocentesis) might well lead to the unmasking of left ventricular dysfunction that was compensated for by high endogenous catecholamine levels. A parasympathetic-vagal cardioprotective effect in the early stages of tamponade has been suggested by some investigators [9]. Malignant effusions account for an increasing percentage of patients who develop cardiac tamponade. Coincident congestive heart failure could be significant in terms of prognosis and candidacy for further chemotherapy, given the potential cardiotoxicity of several antineoplastic agents. It may be particularly important to be aware of the possibility of transient left ventricular systolic dysfunction after relief of tamponade. Because both our patients had breast cancer, we cannot be sure that breast cancer is unrelated to this phenomenon. Further study is needed to determine whether this occurs in patients with nonmalignant pericardial effusions.