pericarditis cured, all treated with amphotericin B, with or without flucytosine, combined with pericardiectom~ Two other patients treated with amphotericin B but without surgery died, and they were considered microbiologic failures. The remaining patients were untreated. Pericardial drainage is considered essential for successful therapy of bacterial pericarditis. One or more pericardiocenteses may suffice, but open surgical drainage, either by a pericardial window or pericardiectomy, is often needed. 2•6 The literature regarding Candida pericarditis is too sparse for definitive recommendations with respect to need and technique of pericardial drainage. The single patient whose case is reported herein demonstrates that a single pericardiocentesis, which did not completely evacuate the pericardial space and was associated with technical difficulties possibly due to loculations, in combination with anti-Candida chemotherapy can result in cure, and that thoracotomy is not necessary in all cases. The patient's immunocompetence and overall good health combined with adequate anti-Candida chemotherapy likely contributed to the salutary outcome. Apparently purulent pericarditis caused by bacteria such as Staphylococcus aureus usually requires effective drainage whereas that caused by Mycobacterium tuberculosis and possibly Candida frequently does not require drainage. 6 Reasons for differing requirements for surgical drainage are not well-de6ned. Speculatively, pericarditis caused by mycobacteria and fungi is often subacute; the relatively slow clinical course may allow sufficient time for antimicrobials and the immune system to effect cure without courting tamponade or persistent sepsis. The pharmacokinetics of amphotericin B are poorly understood, especially with respect to pericardial penetration. 7 The amphotericin B level in pericardial fluid was 50 percent of serum in a single patient with Candida pericarditis. 3 Serum levels in the range of 1.0 to 2.4 mWL have been measured after commonly used doses of amphotericin B.7 Since the MIC of our isolate to amphotericin B was less than 0.5 mWL, it is likely that inhibitory concentrations were present in pericardial fluid for at least part of the dosing interval. The use of flucytosine likely increased the potency of the anti-Candida chemotherap~ Although prohlems with antifungal susceptibility testing are recognized, it is of interest and possibly of therapeutic import that the Candida isolate in this report was highly resistant to several azoles, including ketoconazole and fluconazole, 8 and suggests need for caution in use of azoles for treatment of deep-seated Candida infections. REFERENCES 1 Kraus WE, Valenstein ~ Corey GR. Purulent pericarditis caused by Candida: report of three cases and identification of high-risk populations as an aid to early diagnosis. Rev Infect Dis 1988; 10: 34-41 2 Rubin RH, Moellering RC. Clinical, microbiologic, and therapeutic aspects of purulent pericarditis. Am J Med 1975; 59:68-
78
3 Eng R, Sen ~ Browne K, Louria DB. Candida pericarditis. Am J Med 1981; 70:867-69 4 Klacsmann PG, Bulkley BH, Hutchins GM. The changed spectrum of purulent pericarditis. Am J Med 1977; 63:666-72 5 Gronemeyer PS, Weissfeld AS, Sonnenwirth AC. Purulent pericarditis complicating systemic infection with Candida tropicalis.
Am J Clin Patholl982; 77:471-75 6 Savoia MC, Oxman MN. Myopericarditis, pericarditis, and mediastinitis. In: Mandell GL, Douglas HG, Bennett JE, eds. Principles and practice ofinfectious diseases. New York: Churchill Livingstone, 1990:721-32 7 Gallis HA, Drew RH, Pickard ww. Amphotericin B: 30 years of clinical experience. Rev Infect Dis 1990; 12:308-29 8 Korting HC, Ollert M, Georgii A, Froschl M. In vitro susceptibilities and biotypes of Candida albicans isolates from the oral cavities of patients infected with human immunodeficiency virus. J Clin Microbioll989; 26:2621-31
Nonbacterial Thrombotic Endocarditis· Assessment by Transesophageal Echocardlography Daniel G. Blanchard, M.D.;t Robert S. Ros8, M.D.; and Howard C. Dittrich, M.D., F.C.C.l
Nonbacterial thrombotic endocarditis (NBTE) is difJicult to detect antemortem and is often not suspected until embolic events occur. Transthoracic echocardiography is useful in diagnosing NBTE, but it may be limited by suboptimal imaging and resolution. Herein we describe the 8rst reported case of NBTE diagnosed by transesophageal echocardiograph~ As early detection and treatment may avert signi6cant embolic complications, transesophageal echocardiography should be strongly considered if other techniques are nondiagnostic and clinical suspicion of NBTE remains high. (Chat 1992; 102:964-56)
=
DIe disseminated intravascular coagulation· NBTE bacterial thrombotic endocarditis '
=non-
N to detect antemortem, and
onbacterial thrombotic endocarditis (NBTE) is difficult
it is usually discovered at Because new cardiac murmurs are an infrequent 6nding, 1 NBTE may not be suspected until embolic events occur. Unfortunatel~ arterial embolization is common in NBTE and may cause significant morbidity. As early diagnosis and treatment may prevent or reduce embolic complications, 2 accurate antemortem detection of thrombotic valvular vegetations is necessar~ Herein, we describe an unusual example of NBTE, the mst reported case (to our knowledge) diagnosed by transesophageal echocardiography. autops~
CASE REpORT A 64-year-old woman initially presented with right ophthalmoplegia and right cranial nerve palsies. Computed tomographic (CT) scans and cerebral angiography revealed bilateral internal carotid artery aneurysms at the junctions ofthe cavernous and precavernous segments. She was referred for neurosurgical clipping and repair. Her physical examination revealed right cranial nerve III, ~ ~ and *From the Division of Cardiology, Department of Internal Medicine, University of California, San Diego School of Medicine San ~~ ' tRecipient of an American Heart Association Research Fellowship Grant, California Chapter, Burlingame, Calif: Reprint requests: Dr. Dittrich, UCSD Medical Center. 225 Dickinson
Street, San Diego 92103
'
Nonbacterial Tht'Ombotic Endocarditis (Blanchard. Ross, DIttrich)
FlCURE 1. Transesopbageal eehocardiographic image of the mitral valve, demonstrating large vegetation (arrowhead) on the anterior and posterior leaflets (LA =left atrium; LV =left ventricle). A pulmonary arterial catheter is visible in the right ventricle (RV). VI palsies and was otherwise nonnal. Laboratory studies were notable ror nonnal coagulation test results, white blood cell count, bleeding time, and electrocardiogram. As the planned procedure involved cardiopulmonary bypass, a transthoracic eehocardiogram was perfonned and demonstrated nonnal cardiac size, function, and valves. Craniotomy with aneurysm clipping was perfonned, and involved cardiopulmonary bypass and hypothennia ror 7 h and total circulatory arrest for 2 h. Postoperatively, progressive thrombocytopenia and prolongation ofthe prothrombin time occurred, despite platelet transfusion and vitamin Ie administration. Fibrinogen levels dropped and 6brin degradation products were present in the patient's serum, consistent with disseminated intravascular coagulation (Die). Subsequently, adult respiratory distress syndrome developed, and urine output dropped markedly. Empiric broadspectrum antibiotic therapy was started, despite multiple negative blood cultures, and intravenous pressors were used to maintain adequate blood pressure. One week after surgery, acute right hand ischemia developed. A transthoracic echocardiogram revealed new mitral regurgitation and a possible mitral vegetation. To better assess cardiac and valvular function, biplane transesophageal eehocardiography (Hewlett-Packard Sonos 1000, Andover, Mass) was perfonned. Large mitral vegetations (Fig 1) and severe mitral regurgitation were present. In addition, lea8et thickenin~ and multiple discrete vegetations were noted on the aortic, tricuspid. and pulmonic valves in several difJerent eehocardiographic views. Mild tricuspid insufficiency was present, and ventricular function was otherwise nonnal. A repeated head CT scan showed new areas of low attenuation consistent with ischemic injury as well as a midline shift toward the right. Intravenous heparin therapy was begun, but the patient remained neurologically unresponsive and died ten days after surgery. At autopsY, 6brinous, thrombotic vegetations were noted on all cardiac valves and were largest on the mitral valve (Fi~ 2). The vegetations were loosely adherent, and the underlyin~ valvular structures showed no evidence ofdestruction or erosion. No bacteria were seen on microscopy and all cultures were negative. Thrombotic material and emboli were present in the renal and cerebral vasculature. DISCUSSION
Originally described more than a century ago, 3 NBTE has long been associated with malignant neoplasms (especially adenocarcinoma) and hypercoagulable states, and occurs in 10 to 15 percent of patients with DIC.· Definitive antemor-
FICURE 2a (upper). Anterior mitral valve leaflet at autopsy: large thrombotic vegetations are present (arrow). FIGURE 2b (lower). Microscopic section thmu~ mitral valve with adherent v~etation (M = mitral valve; V =ve~etation).
tem diagnosis has been difficult, as the thrombotic vegetations are generally small and nondestructive.' The first reported identification of NBTE by M-mode echocardiography was in 1976, and since then, several cases of diagnosis by two-dimensional echocardiography have been reported. s Earlier in this century, NBTE was thouw.t to be an incidental autopsy finding of limited importance. Over time, however, it has become clear that significant complications are associated with thrombotic valvular vegetations: systemic embolization, particularly to the cerebral vasculature, is common,· and may be avoided in some cases by prompt and aggressive anticoagulation.' Transthoracic echocardiography is an excellent noninvasive diagnostic technique, but it may be limited by suboptimal image quality and resolution, especially when vegetations are under 2 to 3 mm in size or the patient cannot be properly positioned. Transesophageal echocardiography overcomes these limitations and provides superior images and detail of cardiac chambers and valves; in recent studies involvin~ infective endocarditis, this technique was more effective than transthoracic imaging in detecting bacterial vegetations and abscesses. 8 In the present case, fulminant DIC occurred in the perioperative period, probably secondary to release of brain tissue factors into the systemic circulation and prolonged cardiopulmonary bypass. Transthoracic echocardiography was adequate before surgery, but because of suboptimal positioning, complete visualization of the cardiac valves could not be performed postoperatively. Excellent images of the myocardium and all valves were obtained by transeCHEST 1102 I 3 I SEPTEMBER. 1992
955
sophageal echocardiography, and, through measurement of pulmonary venous Doppler flow characteristics, severity of mitral regurgitation was assessed. Postmortem examination of the heart confirmed all echocardiographic findings. This case of NBTE is unusual in several respects. First, all four valves were affected by the process. While the mitral valve was most involved by thrombus, discrete vegetations were seen on all four valves by biplane transesophageal echocardiography and this was confirmed at autopsy. Pathologically, these vegetations were fibrinous and nondestructive. In a recent review of NBTE, the incidence of quadrivalvular involvement in a compilation of 747 cases was 0.1 percent. I Second, the large vegetations developed rapidly. The preoperative echocardiogram showed no valvular abnormalities, but with. the development of DIe postoperativel~ large amounts of thrombus quickly formed on the cardiac valves, as demonstrated by transesophageal echocardiography on the seventh day after surger~ There is little information available on the time course of vegetation growth in NBTE, and no systematic prospective studies have been performed. This report demonstrates that thrombus formation may be rapid and extensive. Third, this case ofNBTE caused significant valvular insufficienc~ In general, because inflammatory reactions are not present in NBTE vegetations, leaflet destruction or perforation is rare, and valvular dysfunction seldom occurs. 4 In this patient, however, the large thrombotic vegetations caused marked mitral valvular distortion, abnormal leaflet coaptation, and severe regurgitation. In summar~ this case illustrates several points. NBTE vegetations may develop rapidly, and the presence of significant valvular dysfunction does not rule out the diagnosis. When clinically indicated, transesophageal echocardiography may be quite helpful in demonstrating valvular vegetations, especially if the lesions are small and transthoracic imaging is technically difficult. Although the procedure is semi-invasive, it is quite safe, 7 and should be considered if other techniques are nondiagnostic and clinical suspicion of NBTE remains high. REFERENCES 1 Lopez JA, Ross RS, Fishbein MC, Siegel RJ. Nonbacterial thrombotic endocarditis: a review Am Heart J 1987; 113:773-84 2 Rogers LR, Cho E-S, Kempin S, Posner JB. Cerebral infarction from non-bacterial thrombotic endocarditis: clinical and pathologic study including the effects of anticoagulation. Am J Med 1987; 83:746-56 3 Ziegler R. Uber den Bau die Entstehung Endocaritischer Effiorescenzen. Werh Dtsch Kong Intern Med 1888; 7:399 4 MacDonald RA, Robbins SL. The significance of nonbacterial - thrombotic endocarditis: an autopsy and clinical study of78 cases. Ann Intern Med 1957; 46:255-73 5 Lopez JA, Fishbein MC, Siegel RJ. Echocardiographic features of nonbacterial thrombotic endocarditis. Am J Cardioll987; 59: 478-80 6 Mugge A, Daniel WG, Frank G, Lichtlen PR. Echocardiographic diagnosis of infective endocarditis: reassessment of prognostic implications of vegetation size determined by the transthoracic and the transesophageal approach. J Am Coli Cardioll989; 14: 631-38 7 Daniel WG, Erbel R, Kasper ~ et ale Safety of transesophageal echocardiography: a multicenter survey of 10,419 examinations. Circulation 1991; 83:817-21
958
HIV-Associated Perlcardial
Effusions·
Mark] Eisenberg, M.D.; Amy Nelson B. Schiller, M.D.
s. Gordon; and
Following a case of cardiac tamponade in a patient with the acquired immunode6ciency syndrome (AIDS), we examined the frequency and clinical spectrum of pericardial etTusions associated with human immunode6ciency virus infection (HIV) at our institution. or 187 hospitalized patients documented to have pericardial effusions over a one-year period, 14 (7 percent) were known to be HIVpositive at the time of their ecbocardiograms. One patient presented with a large effusion and cardiac tamponade, three had moderate effusions, and ten had small eff'usions. The probable effusion etiology was established in four cases and included endocarditis (2), lymphoma (1), and myocardial infarction (1). In hospital mortality was 29 percent (4 of 14). From our study, as well as a growing number of reports in the literature, we conclude that HIV-associated pericardia) eff'usioDS are frequently seen and that their clinical spectrum is broad. (Cheat 1992; 102:956-58) the 1 human immunodeficiency virus I nfectionhaswithbeen associated with a variety of cardiac (HIV)
type
complications, including myocarditis, endocarditis, and dilated cardiomyopath~ In addition to these complications, a number of reports have described pericardial effusions associated with HIV infection. 1•14 Following a recent case of cardiac tamponade in an AIDS patient, we examined the frequency and clinical spectrum of HIV-associated pericardial effusions at our institution. CASE REPORT A 36-year-old man with AIDS presented with four days of fevers, chills~
night sweats, and progressive dyspnea. He had been treated
for pulmonary tuberculosis during the previous six months but had discontinued his medications one month before admission. On physical examination, the patient was in moderate distress with a heart rate of 112 bpm~ a blood pressure of 94/76 mm Hg, a respiratory rate of281m in, and a temperature of38.1°C. Examination was also remarkable for oral thrush, bilateral basilar rales, elevated jugular venous pressure, and distant heart sounds. Initial laboratory test results were unremarkable. A chest roentgenogram, however, revealed an enlarged cardiac silhouette, bilateral pleural effusions, and apical scarring. Subsequently, an echocardiogram showed a large pericardia1 effusion associated with right ventricular collapse, right atrial collapse, inferior vena cava plethora with blunted response to respiration, and respiratory variation of mitral inflow velocities. Because of the patient's hemodynamic instability, a pericardiocentesis was performed. More than 1,000 ml of serosanguineous fluid was drained, and blood pressure rose to 115185 mm Hg. The next day, a pericardial window was created but subsequent laboratory examination of pericardial tissue and fluid showed only ·~"?m the Cardiovascular Research Institute, the Cardiology DiVISIon of ~e Department of Medicine, and The John Henry Mills
Echocanbograpliy Laboratory of the University of California San Francisco. ' Supported by an Institutional National Research Service award HL 07192, Training Program in Heart and Vascular Disease~ NHLBI, Bethesda. Reprint requests: Dr. Eisenberg, Box 1214 Moffitt-Umg Hospital, 505lbmassus Avenue, San Francisco 94143 HIV-Associated PericardiaI Effusions (EIsenberg, Gordon, SChiller)
78
3 Eng R, Sen ~ Browne K, Louria DB. Candida pericarditis. Am J Med 1981; 70:867-69 4 Klacsmann PG, Bulkley BH, Hutchins GM. The changed spectrum of purulent pericarditis. Am J Med 1977; 63:666-72 5 Gronemeyer PS, Weissfeld AS, Sonnenwirth AC. Purulent pericarditis complicating systemic infection with Candida tropicalis.
Am J Clin Patholl982; 77:471-75 6 Savoia MC, Oxman MN. Myopericarditis, pericarditis, and mediastinitis. In: Mandell GL, Douglas HG, Bennett JE, eds. Principles and practice ofinfectious diseases. New York: Churchill Livingstone, 1990:721-32 7 Gallis HA, Drew RH, Pickard ww. Amphotericin B: 30 years of clinical experience. Rev Infect Dis 1990; 12:308-29 8 Korting HC, Ollert M, Georgii A, Froschl M. In vitro susceptibilities and biotypes of Candida albicans isolates from the oral cavities of patients infected with human immunodeficiency virus. J Clin Microbioll989; 26:2621-31
Nonbacterial Thrombotic Endocarditis· Assessment by Transesophageal Echocardlography Daniel G. Blanchard, M.D.;t Robert S. Ros8, M.D.; and Howard C. Dittrich, M.D., F.C.C.l
Nonbacterial thrombotic endocarditis (NBTE) is difJicult to detect antemortem and is often not suspected until embolic events occur. Transthoracic echocardiography is useful in diagnosing NBTE, but it may be limited by suboptimal imaging and resolution. Herein we describe the 8rst reported case of NBTE diagnosed by transesophageal echocardiograph~ As early detection and treatment may avert signi6cant embolic complications, transesophageal echocardiography should be strongly considered if other techniques are nondiagnostic and clinical suspicion of NBTE remains high. (Chat 1992; 102:964-56)
=
DIe disseminated intravascular coagulation· NBTE bacterial thrombotic endocarditis '
=non-
N to detect antemortem, and
onbacterial thrombotic endocarditis (NBTE) is difficult
it is usually discovered at Because new cardiac murmurs are an infrequent 6nding, 1 NBTE may not be suspected until embolic events occur. Unfortunatel~ arterial embolization is common in NBTE and may cause significant morbidity. As early diagnosis and treatment may prevent or reduce embolic complications, 2 accurate antemortem detection of thrombotic valvular vegetations is necessar~ Herein, we describe an unusual example of NBTE, the mst reported case (to our knowledge) diagnosed by transesophageal echocardiography. autops~
CASE REpORT A 64-year-old woman initially presented with right ophthalmoplegia and right cranial nerve palsies. Computed tomographic (CT) scans and cerebral angiography revealed bilateral internal carotid artery aneurysms at the junctions ofthe cavernous and precavernous segments. She was referred for neurosurgical clipping and repair. Her physical examination revealed right cranial nerve III, ~ ~ and *From the Division of Cardiology, Department of Internal Medicine, University of California, San Diego School of Medicine San ~~ ' tRecipient of an American Heart Association Research Fellowship Grant, California Chapter, Burlingame, Calif: Reprint requests: Dr. Dittrich, UCSD Medical Center. 225 Dickinson
Street, San Diego 92103
'
Nonbacterial Tht'Ombotic Endocarditis (Blanchard. Ross, DIttrich)
FlCURE 1. Transesopbageal eehocardiographic image of the mitral valve, demonstrating large vegetation (arrowhead) on the anterior and posterior leaflets (LA =left atrium; LV =left ventricle). A pulmonary arterial catheter is visible in the right ventricle (RV). VI palsies and was otherwise nonnal. Laboratory studies were notable ror nonnal coagulation test results, white blood cell count, bleeding time, and electrocardiogram. As the planned procedure involved cardiopulmonary bypass, a transthoracic eehocardiogram was perfonned and demonstrated nonnal cardiac size, function, and valves. Craniotomy with aneurysm clipping was perfonned, and involved cardiopulmonary bypass and hypothennia ror 7 h and total circulatory arrest for 2 h. Postoperatively, progressive thrombocytopenia and prolongation ofthe prothrombin time occurred, despite platelet transfusion and vitamin Ie administration. Fibrinogen levels dropped and 6brin degradation products were present in the patient's serum, consistent with disseminated intravascular coagulation (Die). Subsequently, adult respiratory distress syndrome developed, and urine output dropped markedly. Empiric broadspectrum antibiotic therapy was started, despite multiple negative blood cultures, and intravenous pressors were used to maintain adequate blood pressure. One week after surgery, acute right hand ischemia developed. A transthoracic echocardiogram revealed new mitral regurgitation and a possible mitral vegetation. To better assess cardiac and valvular function, biplane transesophageal eehocardiography (Hewlett-Packard Sonos 1000, Andover, Mass) was perfonned. Large mitral vegetations (Fig 1) and severe mitral regurgitation were present. In addition, lea8et thickenin~ and multiple discrete vegetations were noted on the aortic, tricuspid. and pulmonic valves in several difJerent eehocardiographic views. Mild tricuspid insufficiency was present, and ventricular function was otherwise nonnal. A repeated head CT scan showed new areas of low attenuation consistent with ischemic injury as well as a midline shift toward the right. Intravenous heparin therapy was begun, but the patient remained neurologically unresponsive and died ten days after surgery. At autopsY, 6brinous, thrombotic vegetations were noted on all cardiac valves and were largest on the mitral valve (Fi~ 2). The vegetations were loosely adherent, and the underlyin~ valvular structures showed no evidence ofdestruction or erosion. No bacteria were seen on microscopy and all cultures were negative. Thrombotic material and emboli were present in the renal and cerebral vasculature. DISCUSSION
Originally described more than a century ago, 3 NBTE has long been associated with malignant neoplasms (especially adenocarcinoma) and hypercoagulable states, and occurs in 10 to 15 percent of patients with DIC.· Definitive antemor-
FICURE 2a (upper). Anterior mitral valve leaflet at autopsy: large thrombotic vegetations are present (arrow). FIGURE 2b (lower). Microscopic section thmu~ mitral valve with adherent v~etation (M = mitral valve; V =ve~etation).
tem diagnosis has been difficult, as the thrombotic vegetations are generally small and nondestructive.' The first reported identification of NBTE by M-mode echocardiography was in 1976, and since then, several cases of diagnosis by two-dimensional echocardiography have been reported. s Earlier in this century, NBTE was thouw.t to be an incidental autopsy finding of limited importance. Over time, however, it has become clear that significant complications are associated with thrombotic valvular vegetations: systemic embolization, particularly to the cerebral vasculature, is common,· and may be avoided in some cases by prompt and aggressive anticoagulation.' Transthoracic echocardiography is an excellent noninvasive diagnostic technique, but it may be limited by suboptimal image quality and resolution, especially when vegetations are under 2 to 3 mm in size or the patient cannot be properly positioned. Transesophageal echocardiography overcomes these limitations and provides superior images and detail of cardiac chambers and valves; in recent studies involvin~ infective endocarditis, this technique was more effective than transthoracic imaging in detecting bacterial vegetations and abscesses. 8 In the present case, fulminant DIC occurred in the perioperative period, probably secondary to release of brain tissue factors into the systemic circulation and prolonged cardiopulmonary bypass. Transthoracic echocardiography was adequate before surgery, but because of suboptimal positioning, complete visualization of the cardiac valves could not be performed postoperatively. Excellent images of the myocardium and all valves were obtained by transeCHEST 1102 I 3 I SEPTEMBER. 1992
955
sophageal echocardiography, and, through measurement of pulmonary venous Doppler flow characteristics, severity of mitral regurgitation was assessed. Postmortem examination of the heart confirmed all echocardiographic findings. This case of NBTE is unusual in several respects. First, all four valves were affected by the process. While the mitral valve was most involved by thrombus, discrete vegetations were seen on all four valves by biplane transesophageal echocardiography and this was confirmed at autopsy. Pathologically, these vegetations were fibrinous and nondestructive. In a recent review of NBTE, the incidence of quadrivalvular involvement in a compilation of 747 cases was 0.1 percent. I Second, the large vegetations developed rapidly. The preoperative echocardiogram showed no valvular abnormalities, but with. the development of DIe postoperativel~ large amounts of thrombus quickly formed on the cardiac valves, as demonstrated by transesophageal echocardiography on the seventh day after surger~ There is little information available on the time course of vegetation growth in NBTE, and no systematic prospective studies have been performed. This report demonstrates that thrombus formation may be rapid and extensive. Third, this case ofNBTE caused significant valvular insufficienc~ In general, because inflammatory reactions are not present in NBTE vegetations, leaflet destruction or perforation is rare, and valvular dysfunction seldom occurs. 4 In this patient, however, the large thrombotic vegetations caused marked mitral valvular distortion, abnormal leaflet coaptation, and severe regurgitation. In summar~ this case illustrates several points. NBTE vegetations may develop rapidly, and the presence of significant valvular dysfunction does not rule out the diagnosis. When clinically indicated, transesophageal echocardiography may be quite helpful in demonstrating valvular vegetations, especially if the lesions are small and transthoracic imaging is technically difficult. Although the procedure is semi-invasive, it is quite safe, 7 and should be considered if other techniques are nondiagnostic and clinical suspicion of NBTE remains high. REFERENCES 1 Lopez JA, Ross RS, Fishbein MC, Siegel RJ. Nonbacterial thrombotic endocarditis: a review Am Heart J 1987; 113:773-84 2 Rogers LR, Cho E-S, Kempin S, Posner JB. Cerebral infarction from non-bacterial thrombotic endocarditis: clinical and pathologic study including the effects of anticoagulation. Am J Med 1987; 83:746-56 3 Ziegler R. Uber den Bau die Entstehung Endocaritischer Effiorescenzen. Werh Dtsch Kong Intern Med 1888; 7:399 4 MacDonald RA, Robbins SL. The significance of nonbacterial - thrombotic endocarditis: an autopsy and clinical study of78 cases. Ann Intern Med 1957; 46:255-73 5 Lopez JA, Fishbein MC, Siegel RJ. Echocardiographic features of nonbacterial thrombotic endocarditis. Am J Cardioll987; 59: 478-80 6 Mugge A, Daniel WG, Frank G, Lichtlen PR. Echocardiographic diagnosis of infective endocarditis: reassessment of prognostic implications of vegetation size determined by the transthoracic and the transesophageal approach. J Am Coli Cardioll989; 14: 631-38 7 Daniel WG, Erbel R, Kasper ~ et ale Safety of transesophageal echocardiography: a multicenter survey of 10,419 examinations. Circulation 1991; 83:817-21
958
HIV-Associated Perlcardial
Effusions·
Mark] Eisenberg, M.D.; Amy Nelson B. Schiller, M.D.
s. Gordon; and
Following a case of cardiac tamponade in a patient with the acquired immunode6ciency syndrome (AIDS), we examined the frequency and clinical spectrum of pericardial etTusions associated with human immunode6ciency virus infection (HIV) at our institution. or 187 hospitalized patients documented to have pericardial effusions over a one-year period, 14 (7 percent) were known to be HIVpositive at the time of their ecbocardiograms. One patient presented with a large effusion and cardiac tamponade, three had moderate effusions, and ten had small eff'usions. The probable effusion etiology was established in four cases and included endocarditis (2), lymphoma (1), and myocardial infarction (1). In hospital mortality was 29 percent (4 of 14). From our study, as well as a growing number of reports in the literature, we conclude that HIV-associated pericardia) eff'usioDS are frequently seen and that their clinical spectrum is broad. (Cheat 1992; 102:956-58) the 1 human immunodeficiency virus I nfectionhaswithbeen associated with a variety of cardiac (HIV)
type
complications, including myocarditis, endocarditis, and dilated cardiomyopath~ In addition to these complications, a number of reports have described pericardial effusions associated with HIV infection. 1•14 Following a recent case of cardiac tamponade in an AIDS patient, we examined the frequency and clinical spectrum of HIV-associated pericardial effusions at our institution. CASE REPORT A 36-year-old man with AIDS presented with four days of fevers, chills~
night sweats, and progressive dyspnea. He had been treated
for pulmonary tuberculosis during the previous six months but had discontinued his medications one month before admission. On physical examination, the patient was in moderate distress with a heart rate of 112 bpm~ a blood pressure of 94/76 mm Hg, a respiratory rate of281m in, and a temperature of38.1°C. Examination was also remarkable for oral thrush, bilateral basilar rales, elevated jugular venous pressure, and distant heart sounds. Initial laboratory test results were unremarkable. A chest roentgenogram, however, revealed an enlarged cardiac silhouette, bilateral pleural effusions, and apical scarring. Subsequently, an echocardiogram showed a large pericardia1 effusion associated with right ventricular collapse, right atrial collapse, inferior vena cava plethora with blunted response to respiration, and respiratory variation of mitral inflow velocities. Because of the patient's hemodynamic instability, a pericardiocentesis was performed. More than 1,000 ml of serosanguineous fluid was drained, and blood pressure rose to 115185 mm Hg. The next day, a pericardial window was created but subsequent laboratory examination of pericardial tissue and fluid showed only ·~"?m the Cardiovascular Research Institute, the Cardiology DiVISIon of ~e Department of Medicine, and The John Henry Mills
Echocanbograpliy Laboratory of the University of California San Francisco. ' Supported by an Institutional National Research Service award HL 07192, Training Program in Heart and Vascular Disease~ NHLBI, Bethesda. Reprint requests: Dr. Eisenberg, Box 1214 Moffitt-Umg Hospital, 505lbmassus Avenue, San Francisco 94143 HIV-Associated PericardiaI Effusions (EIsenberg, Gordon, SChiller)
