Download Purulent pericarditis: Review of a 20

JACC Vol. 22, No . 6
November 05, 1993- 1 66 1 -5
1661
PERICARLUIS
Purulent Perl'=~_l ~ IN: Review of a
General Flosph''
Y
Experience
*111
ia
JAUME SAGRISTA-SAULEDA, MD, JOSE A . BARRABIftS, MD,
GAIETA PERMANYER-MIRALDA, MD, JORDI SOLER-SOLER, MD, FACC
Barcelona, Spain
Objectives . The purpose of this study was to review the features
of purulent pericarditis in patients from a general hospital during
a recent 20-year period.
fiaekgmund. Although studies published from 1974 to 1977
suggested a changing spectrum for purulent pericarditis, this view
has not been proved .
Methods. We retrospectively evaluated the records of 33 patients from one general hospital who had a diagnosis of purulent
pericarditis during the period 1972 to 1993 . All autopsy protocols
from the sate period were also reviewed . In 19 patients (group 1),
the condition was diagnosed during life; in 14 (group 11), it was
identified at autopsy .
Results . In group I, the possible sources of pericardial infection
were identified in 17 patients ; pneumonia (6 patients) was the most
common source . Empyema was present in 10 patients; 15 had
cardiac tomponade . The most common microorganisms were
streptococci, pneumococci and staphyloc&-cl . Six patients developed constrictive perkitis and rep' peri,ardiectonly. TOM
patients died, 1 patient was lost to follow up and 15 patients hw a
favorable outcome at a mean follow-up interval of 35 months .
In group 11, the clinical diagnoses included pneumonia (five
patients) among other infections, with enipyema in six patients .
Purulent pericarditis was probably the direct cause of death in two
patients.
Conclusions . In our experience, the spectrum of purulent
pericarditis has not changed in recent years . Many patients do not
have the classical findings of pericarditis, and diagnosis is made
only at autopsy or after tamponade has developed. Empyema
remains a common predisposing condition . Purulent perie2rditis
is still a severe disease, but its prognosis is excellent in patients
who can be discharged from the hospital .
Although purulent pericarditis has been known for many
years (1), its features have been reported to have changed
over the past decades (2-5) . Despite medical progress, this
disease still carries a high mortality rate (2-4,6,7) . The
diagnosis is missed in a high proportion of patients and is
often identified only at autopsy .
Because no major review of purulent pericarditis has
been published in the past 15 years, the present report was
based on a review of experience with the disease in patients
admitted to the Hospital General Universitari Vail d'Hebron
over the past 2 decades (1972 to 1991) .
patients fulfilling the following criteria were included in the
study : 1) purulent or cloudy pericardial exudate with a predominance of polymorphonuclear neutrophils ; or 2) growth of
bacteriri ;-, the culture of pericardial fluid . In addition, one
patient with extensive polymorphonuclear infiltrate in pericardial tissue and acute constrictive pericarditis in the context of
an intrathoracic infection was included .
Pericardial exudate was investigated in the microbiology
laboratory. Smears for microscopic examination (Gram
stain) were prepared as were cultures for aerobic and anaerobic organisms .
Patients were assigned to two groups . Group I included
those whose diagnosis of purulent pericarditis was made
during life . Group 11 comprised patients whose disease was
first identified at autopsy .
Methods
Patients. The clinical records were reviewed of all patients who had a clinical diagnosis of purulent pericarditis in
the Hospital General Universitari Vail d'Hebron between
January 1972 and December 1991 . The protocols of all autopsies performed during that period were also reviewed. All
From the Servei de Cardiologia, Hospital General Universitari Vail
d'Hebron, Barcelona, Spain .
Manuscript received December 31, 1992 ; revised manuscript received
May 26, 1993, accepted June 2, 1993 .
Address for correspondence : Dr. Jaume Sagrista-Sauleda, Servei de
Cardiologia, Hospital General Universitari Vail d'Hebron, Pg . Vail d'Hebron,
s/n, 08035 Barcelona, Spain .
©1993 by the American College of Cardiology
(J Am Coll Cardiol 1993,22:1661-5)
Results
During the 20-year period, 33 patients with purulent
pericarditis were identified in a hospital population of
593,601 patients . The diagnosis was made during life in 19 of
the 33 patients and at autopsy in the remaining 14 . No patient
had features of acquired immunodeficiency syndrome (AIDS) .
Group I . Group I comprised 19 patients (12 male, 7
female patients) aged 7 to 72 years . Only nine were admitted
to the cardiology service . Five patients had an underlying
07311097/913100
1662
JACC Vol . 92, No . 6
November 15 .1993 : 1 66 1 -5
PURULENT PEICARDITIS L
Table 1 . Possible Source of Infection and Clinical Background in
the 19 Group I Patients
Patients (no.)
Pneumonia*
Periodontal infection, floor of mouth abscess,
mediastinitis
Peritonsillar abscess, cervical abscess, mediastinitis
Sepsis (originating from skin, oral cavity,
parenteral nutrition, colon cancer)
Peritonitis of bile origin
Subphrenic abscess after traumatic rupture of the
colon
Possible urinary tract infection
Unknown
6 (4 with empyema)
3 (2 with empyema)
I (with empyema)
4 (3 with empyema)
I
I
I
2
*
left lung was affected in two patients, the right lung in three and both
lungs in one . Group I - patients whose diagnosis of purulent pericarditis was
made during life .
chronic debilitating disease (alcohol abuse, rheumatoid arthritis and ulcerative colitis) . The possible source of infection and clinical background in the patients in group I are
reported in Table 1 . Empyema was found in the vast
majority . Clinical features of pericarditis developed between
2 and 30 days (mean 10) after onset of the infectious disease .
The symptoms and signs in the 19 patients are shown in
Table 2 . All patients had leukocytosis . The electrocardiogram (ECG) showed diffuse ST segment elevation in 10
patients ; 3 had diffusely negative T waves ; 7 had a low QRS
complex amplitude . 4 had paroxysmal atrial fibrillation and 2
had normal ECG findings . No patient had electrical alternans . In all patients the cardiac silhouette on the chest
radiograph was mildly or moderately enlarged . In addition, 6
patients had pneumonia and 12 had pleural effusion, An
echocardiogram, recorded in 13 patients, showed moderate
to severe pericardial effusion in 12 . The patient without
effusion had constrictive pericarditis that had developed
suddenly during a pyopneumothorax of 19 days' duration .
The six patients without an echocardiogram had clinical
evidence of cardiac tamponade.
The diagnosis was confirmed through the removal of
Tie 2 . Clinical Manifestations of Pericarditis in the 19 Group
I Patients
Patients
No .
Fever
19
Peticardial chest pain
7
Dy17
PeticarM friction rub
7
Ptrkardial effusion
18*
t
15
o
12(:
100
37
89
37
94
79
63
*One patient had already developed constriction when he was seen in our
center. tClinicai diagnosis . tTen patients had empyema. Group I = patients
whose diagnosis of purulent pericarditis was made during life .
purulent pericardial effusion . The indications for this procedure were cardiac tamponade (15 patients) ; no tamponade
but the presence of mediastinitis and empyema and echocardiographic evidence of pericardial effusion (2 patients) ; no
tamponade but a recent urinary tract infection and pericardial effusion with marked toxicity and leukocytosis (I patient), and pyopneumothorax with subsequent acute constrictive pericarditis requiring pericardiectomy (I patient) .
Pericardiocentesis was carried out in 14 patients . In all
instances, purulent fluid (150 to 1,150 ml) was obtained . In 16
patients the pericardium was surgically drained (in 4 as a first
therapeutic procedure and in 12 after pericardiocentesis had
demonstrated purulent pericardial effusion [6 had again
developed tamponadel) .
Microbiology . Table 3 shows the results of the cultures .
Culture of pericardial fluid yielded negative findings in 10
patients, but 6 had received antibiotic agents before the
culture was obtained .
Clinical outcome and complications . Six patients developed severe constrictive pericarditis during their initial hospital admission and required pericardiectomy . Five of the six
recovered uneventfully, and one died of heart failure 7 days
after operation . Three other patients developed features of
cardiac constriction that disappeared spontaneously . In two
of the three, constriction was mild (xY pattern in jugular
venous pulse and early septal diastolic notch) . The third
patient had severe constriction with overt venous congestion, but progressive clinical improvement occurred, with
normalization of the echocardiogram and external pulse
recordings after 4 months .
Three patients died : one because of renal failure and two
of causes primarily related to the pericardial disease .
Sixteen patients (84%) were discharged from the hospital .
One was lost to follow-up ; the remaining 15 patients were
followed up for I to 156 months (mean 35) . None of the 15
developed constrictive pericarditis .
Group II . Among 3,419 necropsy studies, purulent pericarditis was identified in 14 patients (9 men, 5 women, aged
43 to 80 years [mean 61]) . Nine patients had an underlying
chronic disease (neoplasia, chronic renal failure, hepatic
cirrhosis, alcohol abuse and liver transplantation) . Thirteen
patients were admitted because of an acute infectious disease ; the remaining patient had a temporary pacemaker . The
clinical diagnoses and microbiologic findings are enumerated
in Table 4 .
No patients in group II were admitted to the cardiology
service . In only two patients was a clinical diagnosis of
pericardial disease made, but tamponade was not present
and pericardiocentesis was not performed . In two patients
neck vein distension was noted before death, but pericardiocentesis was not carried out . In another patient, ST
segment elevation had been present on the ECG . An
echocardiogram had been recorded in only one patient in
group II .
Death occurred I to 25 days after the onset of the disease .
At autopsy, purulent pericardial exudate, 50 to 200 ml, was
JACC Vol. 22, No . 6
SAGRIS
PURE"
November 15,1993 :1661-5
'LEDA ET AL .
ARDITts
Table 3 . Results of Microbiologic Studies in the 19 Patients With Purulent Pericarditis in Group I
Pt
No.
I
ro
Origin of
Fericarditis
3
Unknown
Pneumonia
Pneumonia
4
5
6
1
Urinary tract
Pneumonia
Unknown
Periodontal
8
9
10
11
12
Pneumonia
Skin sepsis
Peritonitis
Pneumonia
Periodontal
13
14
15
16
17
18
19
Pericardial
Fluid
Negative
Negative
Peptococctes
sp.,
G- anaerobes
Negative
Negative
Negative
Negative
Pleural
Fluid
Cervical
Exudate
Blood
Culture
Negative
Negative
Proteus mirabilis
Protests sp .
Streptococcus viridans.
Fusobacterium sp„
Peprococcus sp .
Streptococcus pneramoniae
Staphylococcus aureu.,
PN sepsis
Periodontal
Petitonsillar abscess
Negative
Negative
Sr eptococcus miller!
S . pneumonic
Streptococcus anitis,
llacteroides sp .,
G -- anncrobes
S . aureus
Negative
Negative
Oral sepsis
Pneumonia
Colon cancer
Subphrenic abscess
®
S. pneumoaiae
Clostridium septictan
P . aeruginosa
S . attilleri
Negative
S . ntitis, llacteroldes sp .,
U- anaerobes
lfaenaopltilus influenzcae
S . aureas
Bacteroides sp .,
ti- anaerobes
Enterococcaas faecahs
S . vindwts, Streptococcus sp .
Pseudoanoaws aeruginosa,
Enterobacter cloacae
H . influenzae
S . milleri
G- = gram-negative ; Group I = patients whose diagnosis of purulent pericarditis was made during life ; PN = parenteral n, .trition ; Pt = patient .
present in all cases . Pericarditis was considered the immediate cause of death in two patients ; the others probably died
of underlying diseases because the amount of pericardial
fluid found was considered insufficient to impair hemodynamic function .
Discuss
o
The present series is the largest series of patit fits with
purulent pericarditis to Rae ieput Icd front a single tertidr~
hospital during the past 15 years . Our experience indicates
Table 4 . Clinical Diagnoses and Microbiologic Findings in the 14 Group 11 Patients
Pt
No .
Clinical
Diagnosis
I
2
3
4
5
II
Pneumonia, empyema
Pneumonia, empyema
Purulent cltolecystitis
Pneumonia, empyema
AMI, right ventricular perforation
by temporary pacemaker, sepsis
Meningitis, sepsis
Mastoiditis, pleuritis
Empyema
Sepsis by catheter
Floor of mouth abscess,
mediastinitis, empyema
Bile tract infection, sepsis
12
13
14
Urinary tract infection, sepsis
Liver transplantation, sepsis
Pneumonia
6
7
8
9
10
Pericardial
Fluid
Other
Cultures*
Escherichia coil
Staphylococcus aureus
Negative
Negative
Pseudomonas aeruginosa
S . aureus, Streptococcus
mills, P . aeruginosa
Klebsiella sp ., Enteracoccus
faccalis, Proteus sp .
E . coli, Klebsiella sp .
P. aeruginosa
E . coli, anaerobes
S. aureus
Streptococcus sp .
P . aeruginosa
*Extrapericardial cultures : pleural fluid (Patients 4 and 10) ; mastoidal exudate (Patient 7) ; blood (Patient 9) ; urine
(Patient 13) . AMI = acute myocardial infarction ; Group Ill = patients whose diagnosis of purulent pericarditis was
made at autopsy ; Pt = patient .
1664
JACC Vol . 22. No. 6
November 15, 1933 ;1661-5
SAGRISTA-SAULEDA ET AL .
PURULENT PERICARDITIS
that the features of the disease have not substantially
changed during that period . Purulent pericarditis is still a
critical disease, potentially lethal and frequently marked by
severe complications, either from the pericarditis itself or
from the underlying disease . The diagnosis was missed in 14
of the 33 pai .icnts, a problem that has been documented in
other series (3-5,8) . There are two basic reasons for this low
index of suspicion . 1) Purulent pericarditis is usually not
associated with the characteristic features of acute pericarditis (4,9,10) . Thus, only 37% of our group I patients
reported characteristic chest pain, and a friction rub was
heard in only 37% . In group II, characteristic chest pain was
not recorded in any patient, and a rub was auscultated in
only one, 2) Other clinical features, such as fever, dyspnea
or tachycardia, may be attributed to the underlying infectious disease . A third contributing factor may be that patients with purulent pericarditis are rarely admitted to a
cardiology service (11) but are admitted to other services
where the index of suspicion is lower .
Purulent pericarditis almost invariably develops in the
context of a severe infectious disease . In the present series
of 33 patients, the most common source of infection was
pneumonia, followed by generalized sepsis from several
sources and by mediastinitis due to downward propagation
of a mouth abscess. In only three patients (all in group I) was
a source of infection not identified on admission . Remarkably, empyema was still a commonly associated finding in
our series (15 of 33 patients) . In contrast to other series (4),
we did not find cases secondary to thoracic or cardiac
surgery . Purulent pericarditis was not diagnosed in any of
the 458 patients with AIDS admitted to our hospital during
the review period .
In addition to the clinical context of the infection, clinical
data suggested either purulent pericarditis or the need to
drain the pericardium in almost all group I patients . Most
important, 15 (79%) of the 19 patients developed clinical
cardiac tampo
e.
Purulent pericarditis was not clinically suspected in any
of the 14 patients from group II, and only 2 had a diagnosis
of pericardial disease . Remarkably, however, it is probable
that only 2 of the 14 patients died of pericardial disease itself .
It is therefore obvious that in most group I patients the
diagnosis was made at a time when pericarditis had already
resulted in severe hemodynamic compromise . However, it
may be that in group II patients the diagnosis went unnoticed
because there were few findings suggesting severe pericarditis in most . In any case, the present series confirms that the
d' osis of purulent pericarditis is rarely made before
hemodynamic compromise develops (4,9).
A wide variety of bacterial organisms have been reported
as causative agents (12,13) . Although some reports (5) suggest an inc
incidence of gram-negative organisms, in
the present series the most commonly isolated organisms
were gram-positive cocci .
and co
. It should be emphasized that
9 of the 19 patients in group I developed constrictive
pericarditis . Constriction developed suddenly, early in relation to the onset of pericardial disease . Seven patients had
severe constriction, and six required emergency pericardiectomy . Evolution to acute cardiac constriction has been
observed in some series (4) but not in others (5,9) .
Three patients in our series developed cardiac constriction that later disappeared spontaneously . This type of
evolution is similar to the transient cardiac constriction
reported previously by our group (14) in patients with acute
idiopathic pericarditis, but it has not been reported in
purulent pericarditis .
Death was the consequence of pericarditis in only 2 of 3
patients in group I and in 2 of 14 in group II . The other
patients died of underlying disease . This fact illustrates that
purulent pericarditis usually develops in the context of a
severe infectious disease, which in many cases is a determinant of the poor prognosis .
The patients who were discharged from the hospital had a
good late outcome . Thus, the complications of acute pericarditis appeared early in the clinical course, but no sequelae
or late constriction was obscrved in any patient dining a
mean follow-up period of 35 months .
Conclusions
Purulent pericarditis is still a severe disease that is
recognized in most patients only at necropsy or after severe
hemodynamic compromise has developed . Empyema remains a common underlying condition and pncumococcus a
not exceptional offender . In virtually all cases, purulent
pericarditis develops in the context of intrathoracic or subphrenic infection or generalized sepsis . In these situations,
any suspicion of pericardial effusion or hemodynamic compromise makes echocardiography mandatory . If the
echocardiogram suggests moderate or severe effusion, appropriate measures for pericardial drainage should be undertaken. The possibility of constrictive pericarditis, usually
acute, should be kept in mind for the surveillance of the
patients. The severity of the disease depends both on the
pericarditis itself and on the underlying infectious disease .
However, the prognosis is .xcellent for patients who can be
discharged from the hospital .
References
1 . Galen C . De Anatomicus Administrationibus (Singer C, translator) . New
York: Oxford University Press, 1956 :192 .
2. Boyle JD, Pearce ML, Guze LB . Purulent pericarditis : review of literature and report of eleven cases . Medicine 1961 ;40:119-44.
3 . Gould K, Barnett JA, Sanford JP . Purulent pericarditis in the antibiotic
era . Arch Intern Med 1974;134:923-7 .
4. Rubin RH, Moellering RC Jr . Clinical, microbiologic and therapeutic
aspects of purulent pericarditis . Am J Med 1975 ;59:68-78 .
5 . Klacsmann PG, Bulkley BH, Hutchins GM . The changed spectrum of
purulent pericarditis. An 86 year autopsy experience in 200 patients . Am
J Med 1977;63:666-73 .
6. Israel L, Cavaillds J, Kovrilsky E . Bernard E. Pdricardite suppurde a
streptocoques gudrie aprds drainage chinirgical . Coeur Med Interne
1%3 ;4 :477-82.
JACC Vol. 22, No . 6
November 15,1993 :1661-1
7. k1orand PEI, Tadei A, Baba Arneur A . Les p6denrdhes purulentes . Sea'
Hop Paris 1966 ,.45 :821-37,
8 . Bark SL, Rice PA, Reynolds CA, Finland M . Pneumococcai pl;oicarditis : a persisting problem in contemporary diagnosis . Am J Med 19M70 :
247-51 .
9 . Majid AA, Omai A . Diagnosis and management of purulent pericarditis .
Exp-n-icnce with pericardiectomy, J Thorac Cardiovasc Surg 1991 :102 :
413-7,
10 . KaulTman CA, Walanakunakorn C, Phair JP, Purulent pneumococcal
pericarditis . A continuing problem in the antibiotic era. Am J Med
1973 ;54:743-50 .
SAG myA-SAULEDA ET AL .
1665
PURULENT PERICARDIT IS
IL Permanyer-Miralda G, Sagrist .1-Sauleda J, Solvc-SukT 3 . Primary acute
pericardial disease : a prospective series of 231 consecutive patients . Am
J Cardiol 190 ;5b :623-3012 . Lorell BIt, BraunwaW E . Pericardial disease . in : Braunwa?d E, editor
heart Disease . A Textbook of Cardiovascular Medicine . Philadelphia :
Saunders, 1992 ;1465-516.
13 . Hall IP . Purulent pericarditis . Postgrad Med J 1989 -,65 :444-8 .
14. SagFisW-Sauleda J, Frrmanyer-Nlirdida G, Candell-Riera J, Ao&A J .
Soler-Soler J . Transient cardiac constriction : an unrecognized pattern of
evolution in effusive acute idiophatic pericarditis . Am 3 Cardiol 1987 -,59 :
9646 .