Вы находитесь на странице: 1из 12

Cardiovascular Pathology 10 (2001) 157 – 168

Surgical pathology of the parietal pericardium:

A study of 344 cases (1993–1999)$
Kirstine Y. Oha, Michio Shimizub, William D. Edwardsc,
Henry D. Tazelaarc,*, Gordon K. Danielsond
Mayo Medical School, Mayo Clinic, Rochester, MN, USA
Department of Pathology, Saitama Medical School, Saitama, Japan
Division of Anatomic Pathology, Hilton 11, Mayo Clinic, 200 First Street Southwest, Rochester, MN 55905, USA
Division of Cardiovascular Surgery, Mayo Clinic, Rochester, MN, USA
Received 5 February 2001; received in revised form 15 May 2001; accepted 17 May 2001


Among 344 cases with surgically resected parietal pericardium, ages ranged from 1 to 87 years (mean, 55), and 64% were male. Causes of
pericardial disease included neoplastic (33%), idiopathic (30%), iatrogenic (23%), and others (14%). Pericardial constriction (Group 1)
represented the largest group (143 cases, 76% male). Maximal pericardial thickness was 1 – 17 mm (mean, 4). Fibrotic thickening occurred in
96%. Chronic lymphoplasmacytic inflammation affected 73% (mild or moderate in 97%). Calcification was uncommon (gross in 28%,
microscopic in 8%), and granulomas were rare (4%, none tubercular). Constriction was idiopathic in 49% and iatrogenic (postpericardiotomy
or postirradiation) in 41%. Neoplasms and cysts (Group 2) represented the second largest group (96 cases). Among 43 cases with secondary
pericardial involvement, carcinomas accounted for 53% and lymphomas 21%. Forty cases (Group 3) had pericardial effusions (75% chronic),
which were idiopathic in 28% and postpericardiotomy in 23%. Thirty-three cases (Group 4) had acute or recurrent pericarditis clinically,
which was idiopathic in 70%. Lastly, 32 cases (Group 5) had pericardial resection for conditions unrelated to primary pericardial disease. In
conclusion, pericardial constriction tended to be nontubercular (100%), nongranulomatous (96%), idiopathic or iatrogenic (90%), and
noncalcific (64%), and it could occur with normal pericardial thickness (4%). Because considerable overlap in the gross and microscopic
features existed among cases with noncalcific pericardial constriction (Group 1), pericardial effusions (Group 3), and pericarditis (Group 4),
clinical information was necessary to provide an accurate clinicopathologic interpretation. D 2001 Elsevier Science Inc. All rights reserved.

Keywords: Constrictive pericarditis; Pericardial neoplasm; Pericarditis; Pericardial effusion; Pericardium; Surgical pathology

1. Introduction 2. Methods

Although valvular and ischemic diseases result in the 2.1. Study group
greatest number of surgically excised specimens from the
heart, pericardial disorders now represent an appreciable All parietal pericardial specimens surgically excised at
proportion of the cases at our institution. The reasons for Mayo Clinic Rochester from January 1, 1993 through
pericardial resection are numerous, but have been described December 31, 1999 were reviewed, as approved by the
in only one surgical pathology series [1]. Given these Institutional Review Board. Of the 358 cases initially
considerations, the current study was undertaken to evaluate identified, 14 were excluded due to insufficient pericardium
the pathology of pericardial disease in a relatively large histologically, thus leaving a study group of 344 cases.
number of surgical cases from a single institution.
2.2. Clinical features
Presented in part as a poster on 14 October 2000 at the annual From medical charts, the age and gender of the patient
meeting of the American Society of Clinical Pathologists in San Diego, CA.
* Corresponding author. Tel.: +1-507-284-1192; fax: +1-507-284-
and the duration, clinical diagnosis, and etiology of the
1599. pericardial disease were recorded. Duration was defined as
E-mail address: tazelaar.henry@mayo.edu (H.D. Tazelaar). acute ( < 1 week), subacute (1 week – 1 month), chronic

1054-8807/01/$ – see front matter D 2001 Elsevier Science Inc. All rights reserved.
PII: S 1 0 5 4 - 8 8 0 7 ( 0 1 ) 0 0 0 7 6 - X
158 K.Y. Oh et al. / Cardiovascular Pathology 10 (2001) 157–168

Table 1 ies, echocardiography, computerized tomography (CT), elec-

Demographic features in 344 surgical cases of pericardial disease
tron beam (ultrafast) CT, or magnetic resonance imaging.
Age (year) Gender Cases Pericardial effusions were classified by duration. Pericarditis
Clinical category Mean Range M F M/F No. % (including acute and recurrent types) was distinguished from
Constriction 57 12 – 81 109 34 3.2 143 42 pericardial effusions by the presence of the syndrome of
Neoplasm or cyst 56 4 – 80 55 41 1.3 96 28 chest pain, friction rub, fever, shortness of breath, cough,
Effusion 57 3 – 87 25 15 1.7 40 12 elevated erythrocytic sedimentation rate, and electrocardio-
Pericarditis 47 11 – 81 18 15 1.1 33 9
graphic changes.
Other 51 1 – 80 14 18 0.8 32 9
Total 55 1 – 87 221 123 1.8 344a 100 Surgical procedures were classified as radical, subtotal,
or partial pericardiectomy or as pericardial window, resec-
M = males; F = females.
Among the 341 patients, an 11-year-old boy had two operations for tion (for neoplasms), biopsy, or unknown, as recently
pericarditis and one operation for pericardial constriction, and a 64-year- reviewed by Ling et al. [2]. Radical pericardiectomy implied
old man had two operations for a pericardial neoplasm, thereby producing removal of the pericardium anteriorly (between the right and
344 cases. left phrenic nerves, from the level of the great arteries to the
diaphragmatic surface), left laterally (posterior to the left
( > 1 month), or unknown. Neoplasms and recurrent peri- phrenic nerve), inferiorly (along the diaphragmatic surface),
carditis were considered chronic processes. and posteriorly (to the atrioventricular junction). In contrast,
Each case was assigned to one of five categories, based on a subtotal pericardiectomy indicated resection only of the
the most prevalent type of pericardial disease: constriction pericardium between the two phrenic nerves. A partial
(Group 1), neoplasm or cyst (Group 2), effusion (Group 3), pericardiectomy represented any other major removal of
pericarditis (Group 4), or other (Group 5). A preoperative the pericardium, including a ‘‘completion’’ pericardiectomy,
diagnosis of constriction was established on the basis of in which a portion had been previously resected. Pericardial
clinical features of right heart failure and one or more of the windows were small resections performed to drain the
following: cardiac catheterization with hemodynamic stud- pericardial space. All nonpericardial cardiac operations were

Table 2
Clinical features in 344 surgical cases of pericardial disease
Clinical diagnostic category Total
Clinical feature Constriction Neoplasm or cyst Effusion Pericarditis Other No. %
Acute 0 0 1 2 0 3 1
Subacute 1 0 4 3 0 8 2
Chronic 137 96 30 27 0 290 84
Unknown 5 0 5 1 32a 43 13
Total 143 96 40 33 32 344 100

Neoplasm or cyst 0 96 9b 0 7b 112 33
Idiopathic 70 0 11 23 0 104 30
Postpericardiotomy 43 0 9 3 0 55 16
Postirradiation 16 0 3 3 3c 25 7
Autoimmune 9d 0 1e 2f 0 12 3
Infection 1g 0 1h 2i 2j 6 2
Sarcoidosis 2 0 2 0 0 4 1
Other 2k 0 4l 0 20m 26 8
Total 143 96 40 33 32 344 100
Patients had no primary pericardial disease.
Neoplasm was only in adjacent structures, not in the pericardium.
Without pericardial constriction.
Six with rheumatoid arthritis, and three with Dressler’s syndrome.
Dressler’s syndrome.
One with scleroderma, and one with Dressler’s syndrome.
Coxsackie B viral pericarditis.
Mixed bacterial and fungal pericarditis.
One bacterial and one fungal infection.
Two aspergillus infections.
One chest trauma and one pericarditis following atrioventricular nodal ablation with right ventricular perforation.
One chest trauma, one uremia, one ascending aortic dissection, and one right ventricular perforation during pacemaker placement.
Twelve congenital heart disease (including nine Ebstein’s anomaly), two valve surgery, one ascending aortic dissection, one ascending aortic aneurysm
in Marfan’s syndrome, one restrictive cardiomyopathy, one chest trauma, one incidental pericardial inflammatory nodule at operation, and one unknown.
K.Y. Oh et al. / Cardiovascular Pathology 10 (2001) 157–168 159

recorded, as was the presence of coexistent ischemic, Pericardial disease was chronic in 84% of the cases. It
valvular, and congenital heart disease; heart failure; and was neoplastic in 33%, idiopathic in 30%, iatrogenic in
pleural effusions. 23%, and of other cause in 14% (Table 2). The most
common surgical procedure was a radical pericardiectomy,
2.3. Pathologic features which was performed in 30% of the cases (Table 3).

Surgical pathology reports and corresponding micro- 3.1.2. Pathologic features

scopic slides were reviewed in each case. From the report On average, four pieces of parietal pericardium were
or the slide, the maximal pericardial thickness was recorded evaluated microscopically from each case (Table 4). Peri-
to the nearest millimeter. The following, if present, were cardial thickening was due to fibrosis in 67%. It occurred in
also recorded: calcification (gross or microscopic), inflam- all five groups, but was greatest in patients with constriction
mation (acute or chronic, and graded semiquantitatively as or pericarditis. Mild to moderate chronic inflammation was
mild, moderate, or marked), granulomas (caseating or non- present in 56% of the cases. Granulomas were observed in
caseating, with the results of special stains for organisms), only 3%, and none were due to tuberculosis.
and fibrosis, granulation tissue, fibrin deposition, hemosi-
derin, and mesothelial hyperplasia. The presence of neutro- 3.1.3. Coexistent heart disease
phils, primarily within small vessels, was considered Among the 341 patients, 83 (24%) had coronary artery
surgery-related, whereas prominent extravascular neutro- disease, 51 (15%) had valvular disease, and 20 (6%) had
philic infiltrates were considered disease-related. If a neo- congenital heart disease. Of the 78 patients with heart
plasm was present, no other microscopic information was failure, none were New York Heart Association (NYHA)
recorded for that case. Class I, 4 were Class II, 18 were Class III, and 24 were
Class IV; the NYHA class was not specified in 32 patients,
of whom 7 had right heart failure. At the time of pericardial
3. Results resection, 48 patients also had resection of a neoplasm
outside the pericardium, 43 underwent valve repair, 33
3.1. General findings had coronary artery bypass grafting, and 20 underwent
repair of congenital cardiac anomalies.
3.1.1. Clinical features
Among the 341 patients, one had two pericardial oper- 3.2. Pericardial constriction (Group 1)
ations and another had three, resulting in 344 cases. Ages at
operation ranged from 1 to 87 years (mean, 55), and 3.2.1. Clinical features
children were present in each of the five clinical categories Patients with symptoms related to pericardial constriction
(Table 1). Males accounted for 221 (64%) of the cases and represented the largest group (143 cases), of which 76%
were particularly prone to develop constriction. were male. In addition, another 16 patients with features of

Table 3
Types of surgical procedures for parietal pericardial disease in 341 patientsa
Clinical diagnostic category Total
Surgical procedure Constriction Neoplasm or cyst Effusion Pericarditis Other No. %
Radical 75 0 14 21 0 110 30
Partial 36b 3 7c 4 14 64 18
Resection 0 61 0 0 2d 63 17
Biopsy 1 34 7 5 10 57 16
Subtotal 31 0 3 1 1 36 10
Window 0 12 13 3 4 32 9
Unknown 0 0 1 0 1 2 <1
Total 143 110e 45f 34g 32 364a 100
One patient had three operations and one patient had two operations (accounting for 344 cases), and 20 patients had two procedures during the same
operation, resulting in 364 procedures. Not included are 64 focal epicardial resections (from 50 patients with constriction, 3 with neoplasms, 6 with effusions, 4
with pericarditis, and 1 with other pericardial disease).
Includes 11 completion pericardiectomies (for six prior resections for constriction, and five prior windows).
Includes two completion pericardiectomies (one prior window, and one prior 40% excision).
Resection of two pericardial masses, found at operation to be non-neoplastic.
Fourteen patients had two pericardial procedures during the same operation.
Five patients had two pericardial procedures during the same operation.
One patient had two pericardial procedures during the same operation.
160 K.Y. Oh et al. / Cardiovascular Pathology 10 (2001) 157–168

Table 4
Pathologic features in 344 surgical cases of pericardial disease
Clinical diagnostic category Total
Pathologic feature Constriction (n = 143) Neoplasm or cyst (n = 96) Effusion (n = 40) Pericarditis (n = 33) Other (n = 32) No. %
No. of specimens (microscopic)
Mean 5 3 5 4 4 4
Range 1 – 16 1–7 1 – 14 1 – 10 1 – 14 1 – 16

Maximal thickness (mm)

Mean 4 3 4 3 3 4
Range 1 – 17 1–5 1–9 1 – 13 1–6 1 – 17

Acute inflammation
Absent 131 94 30 26 26 307 89
Mild 9 1 5 3 4 22 6
Moderate 3 0 4 2 0 9 3
Marked 0 1 1 2 2 6 2

Chronic inflammation
Absent 39 80 5 9 11 144 42
Mild 82 13 21 18 13 147 43
Moderate 19 2 13 6 5 45 13
Marked 3 1 1 0 3 8 2

Absent 137 95 38 33 31 334 97
Noncaseating 5 1 2 0 0 8 2
Caseating 1 0 0 0 1 2 1

Absent 91 95 37 30 26 279 81
Gross and microscopic 40 0 1 2 6 49 14
Microscopic only 12 1 2 1 0 16 5

Other features
Fibrotic thickening 137 11 33 26 23 230 67
Fibrin deposition 48 5 26 9 5 93 27
Hemosiderin pigment 40 3 15 5 7 70 20
Granulation tissue 20 4 18 7 6 55 16
Mesothelial hyperplasia 5 2 10 2 1 20 6

constriction were listed in different clinical categories (3 without detectable calcium. Importantly, 4% showed no
with neoplastic involvement and 13 with prominent effu- calcification, fibrosis, or increased thickness. Inflammation
sions), and are discussed with Groups 2 and 3, respectively. was chronic and lymphoplasmacytic in 73% and was mild
Pericardial constriction was chronic in 96% (Table 2). to moderate in 97% of these. Acute inflammation coex-
There was a history of previous pericarditis in 25 (17%) and isted in 8%. Six patients had granulomas (none with
pericardial effusions in 20 (14%). The surgical pericardiec- organisms detected by special stains; two due to sarcoi-
tomy was radical in 52% and subtotal or partial in 47% dosis). Fibrin was detected in 34%, and hemosiderin (due
(Table 3). Eleven patients had had previous pericardial to previous hemorrhage) was seen in 28%.
operations (six resections for constriction, and five pericar-
dial windows). 3.2.3. Causes of constriction
There was no identifiable cause for pericardial con-
3.2.2. Pathologic features striction in 49% of the patients (Table 2). The two most
The maximal pericardial thickness ranged from 1 to 17 prevalent known causes were both iatrogenic and included
mm (mean, 4) (Table 4); thickening was present in 96% previous pericardiotomy (30%) and irradiation (11%).
and was due to fibrosis (Fig. 1). Calcification was present Radiotherapy had been used to treat 11 lymphomas, 3
in only 36% (grossly visible or palpable in 28% and breast cancers, 1 lung cancer, and 1 unknown malignancy,
detectable only by microscopy in 8%). Pericardial thick- and the time interval between radiation and pericardiec-
ness ranged from 1 to 12 mm (mean, 4) in specimens with tomy ranged from 2 to 42 years (mean, 23). Eleven
calcification, 2– 8 mm (mean 4) in those with only micro- patients had had incomplete prior pericardial resections
scopic calcification, and 1– 17 mm (mean 5) for cases without relief of symptoms and then underwent a com-
K.Y. Oh et al. / Cardiovascular Pathology 10 (2001) 157–168 161

Fig. 1. Photomicrographs from patients with constrictive pericarditis. (A – C) Low-power views of the pericardium (taken at the same magnification),
showing fibrotic thickening (A), plate-like calcification (B), and normal thickness (C, between arrows). (D,E) Chronic nongranulomatous
lymphoplasmacytic inflammation, of mild degree at the border between the pericardium and adjacent adipose tissue (D), and of moderate degree within
fibrotic pericardium (E).

pletion pericardiectomy. There were no cases of posttu- theliomas, one metastatic adenocarcinoma of the lung).
bercular constriction. Seven patients had both pericardial effusions and pericardial
neoplasms. Fifteen patients also had pleural effusions.
3.3. Neoplasms and cysts (Group 2)
3.3.2. Pathologic features
3.3.1. Clinical features The maximal pericardial thickness ranged from 1 to 5
Neoplasms and cysts represented the second largest mm (mean, 3), excluding two cases with oblique tissue
group and included 96 cases and 110 specimens (Table 3). orientation on the slide. Thickening was tumor-related in
There was no appreciable gender predilection (Table 1). The all cases, with coexistent fibrosis in 11% (Table 4).
two most frequently performed procedures were resection Chronic lymphoplasmacytic inflammation was present in
(55%) and biopsy (31%). An additional seven patients had only 17% and was mild in 81% of these. A noncaseat-
pericardial procedures to rule out neoplastic disease and are ing granuloma was found in pericardial tissue from a
included in Group 5. patient with an unresectable squamous cell carcinoma of
Overall, metastatic disease accounted for 64%, primary the lung.
benign tumors and cysts 23%, and mesotheliomas 13%
(Table 5). Carcinomas and lymphomas were the most 3.4. Pericardial effusions (Group 3)
prevalent forms of secondary involvement. Among the
18 cases with tumor identified only in adjacent tissues 3.4.1. Clinical features
and not in the pericardium directly, there were no lym- Of the 40 patients with pericardial effusions, 63% were
phomas or mesotheliomas. male (Table 1). Effusions were chronic in 75% (Table 2).
Three patients had symptoms of constriction due to Specific gravity was determined in six cases, and all were
neoplastic encasement of the heart (two pericardial meso- transudative. The two procedures performed most often
162 K.Y. Oh et al. / Cardiovascular Pathology 10 (2001) 157–168

Table 5 3.5. Pericarditis (Group 4)

Neoplasms or cysts in 96 patients with pericardial resection
In the In adjacent 3.5.1. Clinical features
Neoplasms or cysts pericardium tissues No. % The smallest group included 32 patients with pericarditis,
Secondary (n = 61) 30% of whom also had a pericardial effusion. There was no
Carcinoma 23a 15b 38 40 gender predilection (Table 1). One patient had two pericar-
Lymphoma 9c 0 9 9
dial operations, and another had two pericardial procedures
Other 11d 3e 14 15
during the same operation, yielding 33 operations and 34
Primary (n = 22) specimens. Of the 33 operations, 82% were for chronic
Cyst 20 0 20 21 recurrent disease (Table 2). A radical pericardiectomy was
Lymphangioma 1 0 1 1 performed in 62% (Table 3).
Lipoma 1 0 1 1
In an 11-year-old boy, a pericardial window was made
Mesothelioma (n = 13) to relieve a septic streptococcal effusion, and a subsequent
Pleural 11 0 11 11 partial pericardiectomy was performed when sterile peri-
Pericardial 2 0 2 2 carditis occurred 1 month after the first operation. Within
another month, he developed pericardial constriction and
Total 78 18 96 100
underwent a completion pericardiectomy (included in
Twenty-one adenocarcinomas (15 lung, 4 breast, 1 esophagus, 1 Group 1).
colon) and two squamous cell carcinomas (one lung, one esophagus).
Seven adenocarcinomas (five lung, two esophagus), four squamous
cell carcinomas (three lung, one esophagus), three non-small cell 3.5.2. Pathologic features
carcinomas of lung, and one mediastinal embryonal carcinoma. Maximal pericardial thickness ranged from 1 to 13 mm
Six B-cell lymphomas, two T-cell lymphomas, and one nodular (mean, 3), and fibrosis was present in 79%. Chronic
sclerosing Hodgkin’s disease.
d lymphoplasmacytic inflammation was observed in 73%
Five malignant thymomas (four Type I, one Type II) and six others
(osteosarcoma, liposarcoma, synovial sarcoma, peripheral nerve sheath and was mild to moderate in all (Table 4) (Fig. 3). Acute
tumor, endometrial mixed mullerian tumor, and mediastinal mixed germ neutrophilic inflammation, often with edema, occurred in
cell tumor with yolk sac and mature teratoma components). 21%, and acute and chronic inflammation coexisted in four
Three others (Type I thymoma, synovial sarcoma, and mediastinal cases. Both examples of severe acute inflammation were in
germ cell tumor).
the 11-year-old boy.

3.5.3. Causes of pericarditis

were radical pericardiectomy (31%) and pericardial window In most cases (68%), the cause of pericarditis was
(29%) (Table 3). unknown (Table 2). Previous pericardiotomy and mediasti-
Eighteen patients had isolated effusions. Of the remain- nal irradiation each accounted for 9%.
ing 22 patients, 9 had both a pericardial effusion and
hemodynamic evidence of constriction due to ventricular
compression by loculated fluid or hematoma. Four patients 3.6. Other pericardial disorders (Group 5)
had effusive – constrictive disease (that is, a pericardial
effusion and constriction by the epicardium rather than 3.6.1. Clinical features
by the parietal pericardium) [3]. The remaining nine cases Thirty-two patients had pericardial resection for causes
(not included in the neoplastic group above) each had both unrelated to primary pericardial disease (Table 1). Twelve
a pericardial effusion and a neoplasm that did not involve patients had surgery to repair congenital cardiac anoma-
the pericardium. lies, during which excessive parietal pericardium was
also removed. For seven patients with clinically suspect-
3.4.2. Pathologic features ed primary pericardial neoplasms, surgery ruled out
The maximal pericardial thickness ranged from 1 to 9 their presence. The other 13 patients are described in
mm (mean, 4), and fibrosis was identified in 83% (Table 4). Table 2. A partial pericardiectomy was performed in
Chronic lymphoplasmacytic inflammation was present in 44% (Table 3).
88% (mild to moderate in 97%) (Fig. 2). Acute inflamma-
tion was observed in 25%, and fibrin deposition in 65%. 3.6.2. Pathologic features
Old postoperative foreign-body granulomas were found in The maximal pericardial thickness ranged from 1 to 6 mm
two cases. (mean, 3), excluding two cases with oblique tissue orienta-
tion on the slide. For patients with congenital heart disease,
3.4.3. Causes of effusions the thickness was only 1 –3 mm (mean, 2). Chronic lym-
Pericardial effusions were idiopathic in 28%. The two phoplasmacytic inflammation was present in 66% and
most common known causes were previous pericardiotomy included six cases with coexistent acute neutrophilic inflam-
and neighboring neoplasms (23% each). mation (Table 4). A pericardial caseating granuloma without
K.Y. Oh et al. / Cardiovascular Pathology 10 (2001) 157–168 163

Fig. 2. Photomicrographs from a patient with pericardial effusion. (A) Low-power view of moderately thickened and fibrotic pericardium, with fibrin along the
inner surface and inflammatory infiltrates near the outer surface. (B – D) Inflammation along the outer border of the pericardium (B), showing both a chronic
lymphoplasmacytic infiltrate with a germinal center (C) and an acute neutrophilic infiltrate (D).

identifiable organisms was observed in a patient with commonly performed procedure was a radical pericardiec-
invasive pulmonary aspergillosis. tomy for constriction or effusion. For the microscopic
evaluation of non-neoplastic pericardial disorders, four or
more representative pieces were usually obtained, decal-
4. Discussion cified if necessary, embedded and cut on edge, and stained
with hematoxylin – eosin.
Much has been written about the clinical features,
causes, treatment, and operative results of pericardial 4.1. Pericardial constriction
diseases [3 –8]. However, little has been published con-
cerning the pathology of surgically resected tissues [9– 15], 4.1.1. Clinical features
and only one consecutive series of all pericardial resections Pericardial constriction comprised the largest group
(35 patients) has been reported [1]. The current investiga- (42%) in the present investigation. Similarly, in other
tion uniquely reviewed the pathology of the parietal studies from North America and Europe, constric-
pericardium in a large number (344 specimens) of recent tion represented 25 – 44% of all pericardial operations
cases from a single institution. [1, 16– 18]. Constriction accounted for 79% of the peri-
As in other studies, pericardial operations in the present cardial excisions in Turkey, where tuberculosis is still
study were performed for constriction, neoplasms, effusions, prevalent [4].
pericarditis, or other disorders, and the surgical procedures Although pericardial constriction usually affects adults,
included radical, subtotal, or partial pericardiectomy and the age range is broad and includes children. There is a
pericardial window, resection (of neoplasms), or biopsy. notable male preponderance, as reflected by a male-to-
Patients ranged in age from 1 to 87 years (mean, 55), and female ratio of 3:1 in the current study and 2:1 –3:1 in
each of the clinical categories included children. The most other reports [2,9,19 – 23].
164 K.Y. Oh et al. / Cardiovascular Pathology 10 (2001) 157–168

Fig. 3. Photomicrographs from patients with chronic or recurrent pericarditis. (A) Pericardium thickened by edema and neovascularization, with minimal
inflammation, from a patient treated with corticosteroids. (B) Granulation tissue with clumps of residual compact fibrin. (C,D) Chronic lymphoplasmacytic
infiltrates within fibrotic pericardium (C) and within edematous pericardium (D).

4.1.2. Pathologic features also reported by others [7]. Thus, symptomatic pericardial
In the current study, the pericardial thickness ranged from constriction can occur in the setting of a noncalcified and
1 to 17 mm (mean, 4). Thickening was due primarily to nonthickened pericardium, a fact that should be borne in
fibrosis, and calcification was grossly visible in only 28% of mind when evaluating patients preoperatively. Conversely,
the cases. Inflammation, when present, was mild and of since patients with pericardial neoplasms, pericardial effu-
chronic lymphoplasmacytic type in 79%. These general sions, or pericarditis also commonly have appreciable
morphologic findings are similar to those described in other thickening, not all patients with a thickened pericardium
studies of pericardial constriction [1,9 –11,13,24]. Pericar- will have features of constriction.
dial neovascularization and mesothelial pseudocysts have
also been described; both were identified in the current 4.1.3. Causes of constriction
study, but were not quantitated. In most cases, the cause of chronic constriction cannot be
Agarwal and Chopra [9] described two microscopic determined. Idiopathic disease accounted for 49% of the
patterns. One was characterized by dense fibrosis, lympho- cases in the present study and 33 – 83% of the cases in other
plasmacytic infiltrates, neovascularization, mesothelial series [2,15,19,22]. Though speculative, viral pericarditis is
pseudocysts, and calcium or bone formation, and the other considered the most likely cause of idiopathic constriction.
consisted primarily of dense fibrosis with prominent gran- Iatrogenic disease is currently the most common known
ulomas and neovascularization. The granulomatous pattern cause of pericardial constriction and is usually the result of
was often associated with tuberculosis. Only the lympho- prior cardiac surgery or mediastinal irradiation. It develops in
plasmacytic pattern was encountered in the current study. 0.3% of all patients undergoing cardiac surgery [25]. Con-
Interestingly, 4% of the cases in the present study showed striction was the result of previous pericardiotomy in 30% of
no fibrosis, calcification, or thickening — an observation the current patients. It accounted for 18% of the cases
K.Y. Oh et al. / Cardiovascular Pathology 10 (2001) 157–168 165

reviewed by Ling et al. [2] and for 11% of those reported by culosis, though once a common cause in Asia, Europe, and
Cameron et al. [22] including 29% of their patients who North America, has steadily decreased in frequency (Table 6)
presented after 1980. The reported time interval between [2,9,11,16,17,19 –23,29 – 32]. There were no cases of tuber-
cardiac surgery and a diagnosis of postoperative pericardial culosis in the current study, from 1993 to 1999. On a
constriction has ranged from 2 weeks to 21 years (mean, worldwide basis, however, tuberculosis is still a relatively
2 years) [19,25]. Multiple pathogenetic theories have been frequent cause of pericarditis and constriction [33].
proposed, including the formation of adhesions due to organ-
ization of fibrinous exudates or pericardial blood, the use of 4.2. Pericardial neoplasms and cysts
povidone or other antiseptic solutions in the pericardial space,
or simply pericardial trauma during the operation [26]. Of the 4.2.1. Clinical and pathologic features
52 patients in the current study with constriction and a history Among various large studies, the pericardium has shown
of prior cardiac surgery, 14 also had symptoms of the diverse neoplastic involvement. Nonetheless, carcinomas of
postpericardiotomy syndrome, another condition that may the lung and breast and lymphomas were the three most
predispose patients to develop constriction [25,27,28]. commonly encountered tumors in our investigation and in
Acute pericarditis with an effusion is often a precursor to others (Table 7) [34 – 41]. Although melanoma frequently
constriction [7]. In contrast, recurrent pericarditis occurs less metastasizes to the heart, most patients also have wide-
frequently and rarely proceeds to constriction, although spread extracardiac metastases that preclude cardiac surgery.
cases have been observed. Among patients with constric- Thus, there were no examples of pericardial melanoma in
tion, a previous history of acute pericarditis was docu- the current surgical study.
mented in 6% of the cases reported by Cameron et al. Effusions occurred in 7 of our 96 patients, but the 3 who
[22], 16% of those by Ling et al. [2], and 17% of those in underwent pericardiocentesis preoperatively had no malig-
the current investigation. It should be emphasized, however, nancy identified cytologically. This is somewhat unusual, as
that the majority of patients in each study had not had the reported specificity of cytologic analysis has ranged
pericarditis. Other causes of pericardial constriction include from 73% to 93% [42,43].
infections (tuberculosis, bacteria, fungi, viruses, and para- Pericardial mesotheliomas represent fewer than 1% of all
sites), autoimmune diseases (systemic lupus erythematosus, malignant mesotheliomas and may produce arrhythmias,
rheumatoid arthritis, and Dressler’s syndrome), neoplasms compression syndromes, or pericardial constriction [34].
(especially lymphomas, carcinomas of the lung and breast, Both cases in the current study produced symptoms of cardiac
and mesotheliomas), drugs (such as the ergot alkaloids), and tamponade. One was diagnosed as effusive – constrictive
miscellaneous disorders (uremia, sarcoidosis, amyloidosis, disease by echocardiography, and the other as possible
asbestosis, trauma, and chylopericardium) [2,3,19,22]. constrictive pericarditis.
Interestingly, the most prevalent causes of constriction It is important to emphasize that patients with underlying
have changed appreciably during the past century. Tuber- malignancies may have pericardial disease due to non-

Table 6
Frequency of pericardial constriction due to tuberculosis (TB) by region and year in 15 surgical studies
Reference (study, by continent) Country Time span Pt. no. % TB
[21] India 1954 – 1985 118 61
[9] India 1960 – 1976 86 38
[4] Turkey 1983 – 1993 105 38
[30] Japan 1952 – 1976 57 14

[20] Denmark 1953 – 1983 34 79
[31] France 1979 – 1989 84 12
[29] Ireland 1958 – 1983 32 9
[19] Germany 1970 – 1990 71 4

North America
[16] USA (TN) 1930 – 1971 45 62
[11] USA (OH) 1930 – 1950 61 28
[32] USA (CA) 1955 – 1982 31 19
[17] USA (GA) 1974 – 1980 26 8
[23] USA (MN) 1936 – 1982 231 6
[22] USA (CA) 1970 – 1985 95 2
[2] USA (MN) 1985 – 1995 133 1
Current study USA (MN) 1993 – 1999 143 0
Pt. no. = number of patients; CA = California; GA = Georgia; MN = Minnesota; OH = Ohio; TN = Tennessee; USA = United States of America.
166 K.Y. Oh et al. / Cardiovascular Pathology 10 (2001) 157–168

Table 7
Three most common sites of primary malignancies among 798 cases of metastatic involvement of the pericardium from seven autopsy or surgical series
Reference Carcinoma of lung Carcinoma of breast Lymphoma Other primary site Total
[35] 8 7 2 5 22
[41] 52 35 38 64 189
[39] 10 6 3 9 28
[36] 66 34 24 93 217
[38] 32 8 2 41 83
[34] 80 61 13 62 216
Current studya 16 4 9 14 43

Number 264 155 91 288 798
Percentage 33 19 11 36 100
Includes only the 43 cases with pericardial involvement (see Table 5).

neoplastic causes. Pericardial effusions or constriction may [40,45 –47]. The most common known cause was neoplastic
be a consequence of mediastinal irradiation, infections, disease, affecting 23% of the 40 patients in the current
drugs, or idiopathic pericarditis [37]. They may cause investigation and 33 – 50% in other series [40,45 – 47].
symptoms alone or in various combinations, with or without Among our nine patients with effusions and a neoplasm in
pericardial metastases. adjacent structures, neither biopsy nor pericardiectomy had
revealed the malignancy in pericardial tissues, and only two
4.3. Pericardial effusion patients had malignant effusions by pericardiocentesis. Sim-
ilar discrepancies between biopsy and cytology have been
4.3.1. Clinical features reported in 30– 58% of such cases [45,48,49].
This complex category includes isolated effusions, con- Radiation-induced effusions should be distinguished
striction or neoplasms with an effusion, and effusive – from malignant effusions. Radiation injury is thought to
constrictive disease. Constriction with an effusion, affecting progress from acute pericarditis to effusion and then to
23% of the patients in the current study, results from constriction, often with many intervening asymptomatic
loculated fluid that produces constrictive hemodynamics years [14,50,51]. Of the 15 patients undergoing operation
[7]. In contrast, effusive– constrictive pericarditis, involving for pericardial effusions reported by Piehler et al. [47],
10% of the current patients with an effusion, is characterized features of constriction were present in only 40% overall
by constriction by the epicardium and by a tense pericardial but in all 3 patients with a history of mediastinal irradiation.
effusion. Its diagnosis relies on both hemodynamic and Uremia caused an effusion in 2% of the current cases and
clinical findings [3]. in 8– 20% in three series of pericardial windows for effu-
sions [45,46,49]. Other causes of effusions for which surgery
4.3.2. Pathologic features may be performed include infections, autoimmune connec-
Other than for patients with malignancy, irradiation, or tive tissue diseases, trauma, sarcoidosis, iatrogenic perfora-
uremia, there are few descriptions of the pericardial histol- tion during pacemaker placement, and other rare disorders.
ogy associated with effusions. In an older study from the
Mayo Clinic, Wychulis et al. [44] described the microscopy 4.4. Pericarditis
from 26 patients who had undergone surgery for effusions
or recurrent pericarditis; the pericardium was normal or only 4.4.1. Clinical features
mildly fibrotic in 18 and showed pericarditis with chronic This category includes both acute pericarditis and relaps-
lymphoplasmacytic inflammation in 8, among whom 2 had ing or recurrent pericarditis. An episode of acute pericarditis
tuberculosis. Olsen et al. [40] found acute and chronic will become recurrent in 15– 32% of cases [7,12]. Patients
inflammation in 76% of patients with benign effusions with pericarditis are frequently treated with nonsteroidal
and in 77% of those with idiopathic effusions. In the present anti-inflammatory agents, with some relief of symptoms, but
study, specimens showed acute or chronic inflammation in often require long-term treatment with corticosteroids. In the
90%, fibrosis in 83%, and fibrin deposition in 65%. It is not current study, 18 of the 23 patients with idiopathic recurrent
surprising that fibrosis is associated with both effusions and pericarditis were steroid-dependent. Indications for pericar-
constriction, since effusion may represent a stage in the diectomy include unresponsiveness to therapy and intract-
development of constriction [44]. able or disabling pain [12].
Pericarditis tends to affect adults, but may occur in
4.3.3. Causes of effusions children and adolescents, and has a male predilection.
Idiopathic disease accounted for 30% of the cases in the Among 215 patients in two large series, ages ranged from
present study and 13– 39% in other large surgical series 3 to 81 years (mean, 42), and 66% were male [6,44,52].
K.Y. Oh et al. / Cardiovascular Pathology 10 (2001) 157–168 167

Similarly, in the current study, ages ranged from 11 to 81 [7] Spodick DH. The pericardium: a comprehensive textbook. In: Fun-
damental and clinical cardiology, vol. 27. New York: Marcel Dek-
years (mean, 47), and 55% were male.
ker 1997.
[8] Tuna IC, Danielson GK. Surgical management of pericardial diseases.
4.4.2. Pathologic features Cardiol Clin 1990;8:683 – 96.
Although histopathologic descriptions of hemorrhagic [9] Agarwal S, Chopra P. Constrictive pericarditis: a histopathological
study of 91 cases. Indian Heart J 1977;29:278 – 82.
and purulent pericarditis are plentiful, those of acute or
[10] Brockington GM, Zebede J, Pandian NG. Constrictive pericarditis.
recurrent noninfectious pericarditis are sparse. Kluge and Cardiol Clin 1990;8:645 – 61.
Hall [13] described smoldering pericarditis with irregular [11] Chambliss JR, Jaruszewski EJ, Brofman BL, Martin JF, Feil H.
mesothelial destruction, and Mambo [1] reported nonspe- Chronic cardiac compression (chronic constrictive pericarditis): a crit-
cific chronic fibrous pericarditis with layers of fibrosis and ical study of sixty-one operated cases with follow-up. Circulation
1951;4:816 – 35.
inflammation and a denuded mesothelium. In 11 patients
[12] Fowler NO. Recurrent pericarditis. Cardiol Clin 1990;8:621 – 6.
with recurrent pericarditis reported by Fowler [12], 5 [13] Kluge T, Hall VK. Surgery in acute and chronic pericarditis: patho-
showed fibrosis with adhesions, 3 had only mild fibrosis, physiology and management. Scand J Thorac Cardiovasc Surg
2 exhibited fibrinous exudates, and 1 had inflammation 1976;10:21 – 30.
without fibrosis. Radiation-induced acute fibrinous pericar- [14] Veinot JP, Edwards WD. Pathology of radiation-induced heart disease:
a surgical and autopsy study of 27 cases. Hum Pathol 1996;27:766 – 73.
ditis may also later become fibrotic [51].
[15] Walsh TJ, Baughman KL, Gardner TJ, Bulkley BH. Constrictive
In the current study, although all patients had acute or epicarditis as a cause of delayed or absent response to pericardiec-
recurrent pericarditis clinically, inflammation was absent tomy: a clinicopathological study. J Thorac Cardiovasc Surg 1982;
microscopically in 18% (and in 25% when the two cases 83:126 – 32.
of infectious pericarditis were excluded) and was of only [16] Dean RH, Killen DA, Daniel RA, Collins HA. Experience with peri-
mild degree in 72% of the noninfected cases. This was cardiectomy. Ann Thorac Surg 1973;15:378 – 85.
[17] Miller JI, Mansour KA, Hatcher CR. Pericardiectomy: current indica-
attributed to the effects of preoperative treatment with tions, concepts, and results in a university center. Ann Thorac Surg
steroids. The relative paucity of fibrin, hemosiderin, and 1982;34:40 – 5.
granulation tissue is also understandable. Thus, it should be [18] Hehrlein FW, Moosdorf R, Pitton M, Dapper F. The role of pericar-
emphasized that previous therapy may appreciably affect the diectomy in pericardial disorders. Eur Heart J 1991;12:7 – 9.
microscopic appearance. [19] Tirilomis T, Unverdorben S, von der Emde J. Pericardiectomy for
chronic constrictive pericarditis: risks and outcomes. Eur J Cardio-
Thorac Surg 1994;8:487 – 92.
4.4.3. Causes of pericarditis [20] Aagaard MT, Haraldsted VY. Chronic constrictive pericarditis treated
The disease was idiopathic in 68% of the current cases with total pericardiectomy. Thorac Cardiovasc Surg 1984;32:311 – 4.
[21] Bashi VV, John S, Ravikumar E, Jairaj PS, Shyamsunder K, Krish-
and in 38 – 86% of patients in other series [52 –54]. Prior
naswami S. Early and late results of pericardiectomy in 118 cases of
pericardiotomy and mediastinal irradiation each accounted constrictive pericarditis. Thorax 1988;43:637 – 41.
for 9% of the current cases. Epidemics of Coxsackie B and [22] Cameron J, Oesterle SN, Baldwin JC, Hancock EW. The etiologic
other viruses have been responsible for episodic increases in spectrum of constrictive pericarditis. Am Heart J 1987;113:354 – 9.
the incidence of pericarditis. Tuberculosis may also cause [23] McCaughan BC, Schaff HV, Piehler JM, Danielson GK, Orzulak TA,
acute pericarditis [33], although the current study contained Puga FJ, Pluth JR, Connolly DC, McGoon DC. Early and late results
of pericardiectomy for constrictive pericarditis. J Thorac Cardiovasc
no such cases. Additional causes of pericarditis that may Surg 1985;89:340 – 50.
lead to surgical intervention include other infections, auto- [24] Roberts WC, Spray TL. Pericardial heart disease: a study of its
immune connective tissue diseases, uremia, neoplasms, and causes, consequences, and morphologic features. Cardiovasc Clin
other rare disorders [6,52,55]. 1976;7:11 – 66.
[25] Cimino JJ, Kogan AD. Constrictive pericarditis after cardiac surgery:
report of three cases and review of the literature. Am Heart J 1989;
118:1292 – 301.
References [26] Ng ASH, Dorosti K, Sheldon WC. Constrictive pericarditis following
cardiac surgery — Cleveland Clinic experience: report of 12 cases and
[1] Mambo NC. Diseases of the pericardium: morphologic study of sur- review. Cleveland Clin Q 1984;50:39 – 45.
gical specimens from 35 patients. Hum Pathol 1981;12:978 – 87. [27] Killian DM, Furiasse JG, Scanlon PJ, Loeb HS, Sullivan HJ. Constric-
[2] Ling LH, Oh JK, Schaff HV, Danielson GK, Mahoney DW, Seward tive pericarditis after cardiac surgery. Am Heart J 1989; 118:563 – 8.
JB, Tajik AJ. Constrictive pericarditis in the modern era: evolving [28] Mehta A, Mehta M, Jain AC. Constrictive pericarditis. Clin Cardiol
clinical spectrum and impact on outcome after pericardiectomy. Cir- 1999;22:334 – 44.
culation 1999;100:1380 – 6. [29] Blake S, Bonar S, O’Neill H, Hanly P, Drury I, Flanagan M, Garrett J.
[3] Fowler NO. Constrictive pericarditis: its history and current status. Aetiology of chronic constrictive pericarditis. Br Heart J 1983;50:
Clin Cardiol 1995;18:341 – 50. 273 – 6.
[4] Arsan S, Merca S, Sarigül A, Atasoy S, Demircin M, Dogan R, [30] Mizuno A. Surgical treatment for chronic constrictive pericarditis. Jpn
Pasaoglu I, Bozer AY. Long-term experience with pericardiectomy: Circ J 1978;42:185 – 90.
analysis of 105 consecutive patients. Thorac Cardiovasc Surg 1994; [31] Nataf P, Cacoub P, Dorent R, Jault F, Bors V, Pavie A, Cabrol A,
42:340 – 4. Gandjbakhch I. Results of subtotal pericardiectomy for constrictive
[5] Fowler NO. Pericardial disease. Heart Dis Stroke 1992;1:85 – 94. pericarditis. Eur J Cardio-Thorac Surg 1993;7:252 – 6.
[6] Ilan Y, Oren R, Ben-Chetrit E. Acute pericarditis: etiology, treatment [32] Robertson JM, Mulder DG. Pericardiectomy: a changing scene. Am J
and prognosis. A study of 115 patients. Jpn Heart J 1991;32: 315 – 21. Surg 1984;148:86 – 92.
168 K.Y. Oh et al. / Cardiovascular Pathology 10 (2001) 157–168

[33] Fowler NO. Tuberculous pericarditis. J Am Med Assoc 1991; [45] MillsMcGoon SA, Julian S, Holliday RH, Vinten-Johansen J, Case
266:99 – 103. LD, Hudspeth AS, Tucker WY, Cordell AR. Subxiphoid pericardial
[34] Burke A, Virmani R. Tumors of the heart and great vessels. Atlas of window for pericardial effusive disease. J Cardiovasc Surg 1989;30:
tumor pathology, 3rd series, fascicle 16. Washington (DC): Armed 768 – 73.
Forces Institute of Pathology, 1996. pp. 171 – 209. [46] Mueller XM, Tevaearai HT, Hurni M, Ruchat P, Fischer AP, Stumpe F,
[35] DeLoach JF, Haynes JW. Secondary tumors of heart and pericardium: von Segesser LK. Long-term results of surgical subxiphoid pericardial
review of the subject and report of one hundred thirty-seven cases. drainage. Thorac Cardiovasc Surg 1997;45:65 – 9.
Arch Intern Med 1953;91:224 – 39. [47] Piehler JM, Pluth JR, Schaff HV, Danielson GK, Orszulak TA, Puga
[36] Fine G. Neoplasms of the pericardium and heart. In: Gould S, editor. FJ. Surgical management of effusive pericardial disease: influence of
Pathology of the heart and blood vessels. 3rd ed. Springfield (IL): extent of pericardial resection on clinical course. J Thorac Cardiovasc
Charles C Thomas, 1968. pp. 851 – 83. Surg 1985;90:506 – 16.
[37] Hancock EW. Neoplastic pericardial disease. Cardiol Clin 1990; [48] Hsu FI, Keefe D, Desiderio D, Downey RJ. Echocardiographic and
8:673 – 82. surgical correlation of pericardial effusions in patients with malignant
[38] Mukai K, Shinkai T, Tominaga K, Shimosato Y. The incidence of disease. J Thorac Cardiovasc Surg 1998;115:1215 – 6.
secondary tumors of the heart and pericardium: a 10-year study. Jpn [49] Van Trigt P, Douglas J, Smith PK, Campbell PT, Wall TC, Kenney
J Clin Oncol 1988;18:195 – 201. RM, O’Connor C, Sheikh KH, Corey G. A prospective trial of sub-
[39] Nakayama R, Yoneyama T, Takatani O, Kimura K. A study of meta- xiphoid pericardiotomy in the diagnosis and treatment of large peri-
static tumors to the heart, pericardium and great vessels: I. Incidences cardial effusion: a follow-up report. Ann Surg 1993;218:777 – 82.
of metastases to the heart, pericardium and great vessels. Jpn Heart J [50] Morton DL, Kagan AR, Roberts WC, O’Brien KP, Holmes EC,
1966;7:227 – 34. Adkins PC. Pericardiectomy for radiation-induced pericarditis with
[40] Olsen PS, Sørensen C, Anderson HØ. Surgical treatment of large effusion. Ann Thorac Surg 1969;8:195 – 206.
pericardial effusions: etiology and long-term survival. Eur J Cardio- [51] Schneider JS, Edwards JE. Irradiation-induced pericarditis. Chest
Thorac Surg 1991;5:430 – 2. 1979;75:560 – 4.
[41] Thurber DL, Edwards JE, Achor RWP. Secondary malignant tumors [52] Zayas R, Anguita M, Torres F, Gimenez D, Bergillos F, Ruiz M,
of the pericardium. Circulation 1962;26:228 – 41. Cuidad M, Gallardo A, Valles F. Incidence of specific etiology and
[42] Malamou-Mitsi VD, Zioga AP, Agnantis NJ. Diagnostic accuracy of role of methods for specific etiologic diagnosis of primary acute peri-
pericardial fluid cytology: an analysis of 53 specimens from 44 con- carditis. Am J Cardiol 1995;75:378 – 82.
secutive patients. Diagn Cytopathol 1996;15:197 – 204. [53] Bolanowski PJP, Sarkaria JS, Quinlan DP, Neville WE. Surgical man-
[43] Wiener HG, Kristensen IB, Haubek A, Kristensen B, Baandrup U. agement of acute pericarditis. Am Surg 1977;43:573 – 8.
The diagnostic value of pericardial cytology: an analysis of 95 cases. [54] Permanyer-Miralda G, Sagristá-Sauleda J, Soler-Soler J. Primary
Acta Cytol 1991;35:149 – 53. acute pericardial disease: a prospective series of 231 consecutive pa-
[44] Wychulis AR, Connolly DC, McGoon DC. Surgical treatment of tients. Am J Cardiol 1985;56:623 – 30.
pericarditis. J Thorac Cardiovasc Surg 1971;62:608 – 17. [55] Shabetai R. Acute pericarditis. Cardiol Clin 1990;8:639 – 44.