Eur J Cardiothorac Surg 1998;13:165-169
© 1998 Elsevier Science NL
Pericardial effusion and AIDS: benefits of surgical drainage
Pierre Gounya,
Christophe Lancelina,
Pierre-Marie Girardb,
Claire Hocquet-Cheynela,
Willy Rozenbaumb,
Oscar Nussaumea
a Vascular and Thoracic Surgery Department, Assistance Publique Tenon Hospital, 7, rue de la Chine, 75020 Paris, France
b Infectious Disease Department, Assistance Publique Rothschild Hospital, 33, Boulevard de Picpus, 75012 Paris, France
Received 25 August 1997;
received in revised form 1 December 1997;
accepted 9 December 1997.
Corresponding author. Tel.: +33 1 40307816; fax: +33 1 40307987.
 |
Abstract
|
|---|
Objectives: During the last few years, AIDS has been the main cause of large pericardial effusions in urban settings. We have therefore had to perform surgical pericardial drainage for diagnostic and/or therapeutic purposes in AIDS patients. This study was designed to establish the diagnostic and therapeutic yield of pericardial drainage for these patients. Methods: We retrospectively reviewed the data of the 13 AIDS patients with a pericardial effusion, referred to our surgical department between December 1989 and December 1996 for surgical drainage and pericardial biopsy. Results: Cytological studies and searches for bacteria, mycobacteria and parasites were all negative. The histology of the 13 pericardial biopsies disclosed three pericardial locations of a Kaposi’s sarcoma (all three patients had a pre-existent extra-cardiac location of this sarcoma) and one pericardial location of an already known immature mediastinal teratoma. In the nine other cases, the lesions were aspecific. Four patients died of multivisceral failure within 30 days of surgery. For the survivors, surgical drainage afforded relief and there were no clinical signs of recurrent effusion. Conclusions: The cause of pericardial effusion in AIDS is still often unknown, even after pericardial biopsy. Here, aspecific pericarditis was the most common diagnosis. Although the prognosis of such effusion in these patients is known to be poor, surgical drainage provided relief for those who survived the post-operative period.
Key Words: AIDS • Pericardial effusion • Pericarditis • Pericardial drainage • Kaposi’s sarcoma
|
Introduction
|
|---|
Acquired immunodeficiency syndrome (AIDS) is nowadays highly prevalent among patients with large pericardial effusion
[1]
[2]
[3]. There are various etiologies of this effusions including Kaposi’s sarcoma
[4]
[5], lymphoma
[6] and pathogenic microbial infection
[7]
[8]. Treatment by pericardiocentesis is not always effective and the results of puncture fluid analysis are controversial
[2]
[6]
[9]. Therefore, surgical drainage may be envisaged to treat pericardial effusion effectively and to allow biopsy of the pericardium. Does surgical drainage actually achieve these purposes? We report here our experience in pericardial drainage in AIDS patients.
|
Patients and methods
|
|---|
We reviewed the records of all the AIDS patients referred to our surgical department between December 1989 and December 1996, for surgical drainage of a pericardial effusion. They comprised 13 patients (12 men and 1 woman Table 1) whose mean age was 36 years (extremes: 27 and 44 years). All patients underwent 2-dimensional and M-mode echocardiography (ATL Ultramark 9 computed sonography). The pericardial drainage was performed under general anaesthesia with endotracheal intubation in all cases. Using the abdominal subxiphoid route, the pericardium was incised. The peritoneum was not opened. The fluid was drained and sampled for microbiological and cytological analyses. A piece of pericardium of 
3x3 cm was taken and sent in a saline solution for microbiological and histological analyses. A 28 F chest tube was placed in the pericardial space and drawn out through a separate incision. The pericardial incision was left open and the abdominal incision was closed with running absorbable sutures. Negative pressure suction was maintained post-operatively for as long as it was productive.
We determined the status of HIV infection, the associated pathologies, the characteristics of the pericardial effusion, the operative findings, the results of the analyses of operative samples and the length of survival. None of the patients was lost to follow-up.
|
Results
|
|---|
The 13 patients were all in the AIDS stage. The mean period that elapsed between the discovery of their seropositivity and pericardial drainage was 61 months (extremes: 36 and 84 months). The mean interval between the first opportunistic infection and the occurrence of pericardial effusion was 18 months (extremes: 0 and 48 months). Six patients had Kaposi’s sarcoma: three of them had isolated cutaneous tumours, two had cutaneous and palatine tumours, and one had cutaneous and digestive tumours. One patient had non Hodgkin’s lymphoma, two had a history of infection by Pneumocystis Carinii and five had an evolutive infection, caused in three cases by Cytomegalovirus and in two cases, by Mycobacterium avium intracellulare. Four patients had digestive manifestations of microsporidiosis. The clinical manifestations of pericardial effusion were tamponade in two cases, incapacitating isolated dyspnea in eight, dyspnea with hepatalgia in two and ascites with dyspnea in one case (Table 2). A total of 11 patients had fever. A pleural effusion was associated with the pericardial effusion in eight cases. No correlation was found between the symptoms and the volume of pericardial effusion. The echocardiograms quantified this effusion on the basis of its estimated volume and its repercussions on the right cardiac cavities. Four patients had diastolic collapse of the right atrium and right ventricle, one patient had isolated diastolic collapse of the right atrium, one had isolated diastolic collapse of the right atrium and eight patients had an isolated pericardial effusion with no cardiac repercussion. A large effusion of >800 cc was diagnosed in five cases and a moderate effusion was found in the eight others.
For two patients who had pericardiocentesis during first intention echography, the small amount of fluid recovered and the early reconstitution of the effusion required surgical drainage. The 11 other patients had first intention surgical drainage. The mean volume of fluid drained was 600 ml (100–1600 ml). The fluid was serosanguinous in seven cases, and straw-coloured in the six others. In all cases, it was an exudate. Cytological studies and searches for bacteria, viruses, mycobacteria and parasites were all negative. The 13 pericardial biopsies disclosed three pericardial locations of Kaposi’s sarcoma, one pericardial location of an already known immature mediastinal teratoma and nine aspecific lesions. None of the three pericardial locations of Kaposi’s sarcoma was isolated, and the patients concerned all had cutaneous, palatine or digestive manifestations of this sarcoma. No granulomatous lesions suggesting the presence of mycobacteria were found in the biopsies.
In most cases, the pericardial drain was removed 3 days after the operation (extremes: days 2 and 4). Pericardial drainage led to an immediate clinical improvement in 12 patients. One patient with persistent dyspnea had a large left pleural effusion without pericardial effusion 3 days after drain removal. Pleural drainage resulted in an immediate improvement of the patient's condition.
The mean hospital stay in the surgical department was 4 days (extremes: 3 and 7). A total of 11 patients returned to the medical department after drain removal. There were no complications involving cardiac rhythm disturbances or post-drainage hypovolemic shock, but two patients were transferred to the resuscitation unit for incipient respiratory distress. Postoperative mortality was 30%, as four patients died of multivisceral failure within <30 days of surgery. Three of the nine others died 2 months after drainage, respectively of Kaposi lung sarcoma, disseminated mycobacteriosis and candidal septicemia, thus raising mortality to 54% at 2 months. Of the six other patients, three died of cachexia at 4, 5 and 8 months, respectively and one of immature teratoma at 4 months. The two remaining patients survived for 19 months. They died of a cardiac and a cutaneous lymphoma, respectively (Table 3). Only five echocardiograms were performed respectively at 3, 4 and 12 months after drainage. There was no recurrence of the pericardial effusion.
|
Discussion
|
|---|
The incidence of cardiac manifestations in the course of AIDS is at present increasing, although these manifestations are often not diagnosed clinically
[2]
[10]
[11]
[12]. They include epicardial, myocardial and pericardial locations of Kaposi’s sarcoma or lymphoma, as well as cardiac locations of opportunistic infections. Infections by Mycobacterium tuberculosis, Mycobacterium avium intracellulare, Cryptococcus Neoformans, Herpes simplex or Cytomegalovirus have been reported in pericardial fluid
[7]
[13]
[14]
[15]
[16]. However, culture of pericardial fluid is often unrevealing. Pericarditis is the most common cardiac finding in patients with AIDS (17). Thus, in a series of 115 autopsied patients with HIV, 59 subjects (50%) had cardiac lesions and serious pericardial effusions were found in 37 of these lesions (63%). Akhras
[18] studied 124 homosexual men using Doppler echocardiography: 101 had AIDS (group A), and 23 had had HIV infection without opportunistic infection (group B); pericardial effusion was found in 44 group A patients (44%) and two group B patients (9%). In the course of AIDS, the presence of a stable pericardial effusion is fairly common, and its severity does not usually increase
[10]
[19]
[20]
[21]. Moreover, AIDS is at present the most frequent underlying disease
[2] in large pericardial effusion. In the series reported by Reynolds
[2], 28% of patients undergoing pericardiocentesis had AIDS. Kwan
[10] found that 35% of a series of patients with cardiac tamponade had AIDS. Therefore, AIDS should always be suspected in the presence of a large pericardial effusion, although pericarditis is seldom the first manifestation of AIDS
[10].
The etiology of pericardial effusion in the course of AIDS is difficult to establish, and cytological analysis of the fluid is not the examination of choice
[6]
[17]. Even pericardial biopsy does not always reveal a specific etiology. In the series reported by Zakowki
[6], as well as in ours, non specific pericardial effusions were much more common than neoplasia or infection. Such effusion may occur in the course of inflammatory pericarditis, or may be isolated, without true pericarditis. As pericardial effusions with a normal pericardium are well known, Grody even distinguished between pericarditis and pericardial effusion
[8]. An isolated effusion may accompany myocardiopathy. HIV has also been suggested to play a direct role in the occurrence of an isolated effusion
[8]
[20]. However, infectious forms of pericarditis are difficult to characterize. Among patients living in Africa, the tuberculous etiology predominates. Most of the patients concerned were found to have evolutive systemic tuberculosis
[3]. On the other hand, in the United States and Europe, where tuberculosis is less prevalent, there is no predominant infectious etiology for pericarditis
[9]
[19]. The micro-organisms are seldom identified and diagnosis is then based on a series of postulates. In cases of mycobacterial infection, demonstration of the presence of granulomatous lesions in the pericardial biopsy is an essential factor of diagnosis.
In our series, Kaposi’s sarcoma was identified in three patients and all of them had a pre-existent extra-cardiac location of this sarcoma. However, pericardial locations of isolated Kaposi’s sarcoma have also been reported
[5]. Although a lowering of the incidence of visceral lesions due to Kaposi’s sarcoma has been observed
[1]
[22], it was an important cause of pericardial effusion in our series.
As regards the clinical outcome after drainage, four patients died during the post-operative period. None of these deaths was directly related to the procedure, but they were caused by the underlying illnesses. Of the nine survivors, three died within 2 months. This proportion of 54% mortality at 2 months is consistent with the mortality reported by others
[23]. It indicates that the patients concerned were in the advanced stage of AIDS at which pericardial effusions usually occur. Nevertheless, two patients survived for 19 months. In these patients, as well as in others whose survival was shorter, pericardial drainage led to an immediate clinical improvement and there was no recurrence of the pericardial effusion. Thus, the procedure seemed effective, as it is in nonimmunocompromised patients
[24]. We did not use any form of pericardial window—either pleural or pericardial—because we believe that the pericardial adhesions induced by suction drainage prevent recurrent effusions. This was confirmed in the present small series, although the duration of drainage was somewhat shorter than is usually advised
[23]
[24].
Large or poorly tolerated effusions are often treated by an initial pericardial puncture, without previous comparison of the respective advantages of puncture and surgical drainage. Surgical drainage has only had a restricted place in the different series published since 1990 (Table 4). However, examination of the puncture fluid is not always sufficiently conclusive to permit diagnosis, clinical improvement after puncture is slow, several punctures are often needed, and it is not possible to perform biopsies. In our series, the diagnostic yield of pericardial biopsy was 30%. The surgical procedure allowed complete evacuation of the effusion, thus leading to an immediate clinical improvement. For the patients fit enough to survive the postoperative period, it provided a significant relief. As the drainage never lasted more than a few days, it did not lead to any infectious complication. There were no accidental wounds among the surgical team. Drainage requires the same precautions as any other invasive technique. Pericardiocentesis, which requires the use of a hollow-bore needle is probably more dangerous than drainage in this respect.
|
References
|
|---|
- Lewis W. AIDS: Cardiac findings from 115 autopsies. Prog Cardiovasc Dis 1989;32:207-215.[Medline]
- Reynolds M.M., Hecht S.R., Berger M., et al. Large pericardial effusions in the acquired immunodeficiency syndrome. Chest 1992;102:1746-1747.[Abstract/Free Full Text]
- Cegielski J.P., Ramaiya K., Lallinger G.J., et al. Pericardial disease and human immunodeficiency virus in Dar es Salaam, Tanzania. Lancet 1990;335:209-212.[Medline]
- Steigman C.K., Anderson D.W., Macher A.M., et al. Fatal cardiac tamponade in acquired immunodeficiency syndrome with epicardial Kaposi’s sarcoma. Am Heart J 1988;116:1105-1107.[Medline]
- Stotka J.L., Good C.B., Downer W.R., Kapoor W.N. Pericardial effusion and tamponade due to Kaposi’s Sarcoma in acquired immunodeficiency syndrome. Chest 1989;95:1359-1361.[Abstract/Free Full Text]
- Zakowski M.F., Ianuale-Shanerman A. Cytology of pericardial effusions in AIDS patients. Diagn Cytopathol 1993;9:266-269.[Medline]
- D’Cruz I.A., Sengupta E.E., Abrahams C., et al. Cardiac involvement, including tuberculous pericardial effusion, complicating acquired immune deficiency syndrome. Am Heart J 1986;112:1100-1102.[Medline]
- Grody W.W., Cheng L., Lewis W. Infection of the heart by the human immunodeficiency virus. Am J Cardiol 1990;66:203-206.[Medline]
- Hsia J., Ross A.M. Pericardial effusion and pericardiocentesis in human immunodeficiency virus infection. Am J Cardiol 1994;74:94-95.[Medline]
- Kwan T., Karve M.M., Emerole O. Cardiac tamponade in patients infected with HIV. A report from an Inner-City Hospital. Chest 1993;104:1059-1062.[Abstract/Free Full Text]
- Acierno L.J. Cardiac complications in acquired immunodeficiency syndrome (AIDS): A review. J Am Coll Cardiol 1989;13:1144-1154.[Abstract]
- Hecht S.R., Berger M., Van Tosh A., Croxson S. Unsuspected cardiac abnormalities in the acquired immune deficiency syndrome. An echocardiographic study. Chest 1989;96:805-808.[Abstract/Free Full Text]
- Nathan P.E., Arsura E.L., Zappi M. Pericarditis with tamponade due to Cytomegalovirus in the acquired immunodeficiency syndrome. Chest 1991;99:765-766.[Abstract/Free Full Text]
- Dalli E., Quesada A., Juan G., et al. Tuberculous pericarditis as the first manifestation of acquired immunodeficiency syndrome. Am Heart J 1987;114:905-906.[Medline]
- Woods G.L., Goldsmith J.C. Fatal pericarditis due to Mycobacterium avium intracellulare in acquired immunodeficiency syndrome. Chest 1989;95:1355-1357.[Abstract/Free Full Text]
- Turco M., Seneff M., McGrath B.J., Hsia J. Cardiac tamponade in the acquired immunodeficiency syndrome. Am Heart J 1990;120:1467-1468.[Medline]
- Kaul S., Fishbein M.C., Siegel R.J. Cardiac manifestations of acquired immune deficiency syndrome: A 1991 update. Am Heart J 1991;122:535-543.[Medline]
- Akhras F., Dubrey S., Gazzard B., Noble M.I.M. Emerging patterns of heart disease in HIV infected homosexual subjects with and without opportunistic infections; a prospective colour flow Doppler echocardiographic study. Eur Heart J 1994;15:68-75.[Abstract/Free Full Text]
- Dacso C.C. Pericarditis in AIDS. Cardiol Clin 1990;8:697-699.[Medline]
- Levy W.S., Simon G.L., Rios J.C., Ross A.M. Prevalence of cardiac abnormalities in human immunodeficiency virus infection. Am J Cardiol 1989;63:86-89.[Medline]
- Eisenberg M.J., Gordon A.S., Schiller N.B. HIV-associated pericardial effusions. Chest 1992;102:956-958.[Abstract/Free Full Text]
- Silver M.A., Macher A.M., Reichert C.M., et al. Cardiac involvement by Kaposi’s sarcoma in acquired immune deficiency syndrome. Am J Cardiol 1984;53:983-985.[Medline]
- Flum D.R., McGinn J.T., Jr., Tyras D.H. The role of the ‘pericardial window’ in AIDS. Chest 1995;107:1522-1525.[Abstract/Free Full Text]
- Moores D.W., Allen K.W., Faber L.P., et al. Subxiphoid pericardial drainage for pericardial tamponade. J Thorac Cardiovasc Surg 1995;109:546-552.[Abstract/Free Full Text]
Top
Abstract
Introduction
