CARDIOVASCULAR JOURNAL OF AFRICA • Vol 23, No 3, April 2012
AFRICA
143
The value of serum tumour markers in the prediction
of aetiology and follow up of patients with pericardial
effusion
U BILDIRICI, U CELIKYURT, E ACAR, O BULUT, T SAHIN, G KOZDAG, D URAL
Abstract
Background:
The aim of this study was to evaluate the value
of tumour markers in the differential diagnosis of pericardial
effusions and to assess their changing levels during follow up.
Methods:
Sixty-nine patients who were admitted to hospital
with a diagnosis of pericardial effusion were included in the
study. Serum tumour markers were measured on admission
and after a mean of 18
±
7 months’ follow up. An aetiological
diagnosis was made on clinical evaluation, imaging tech-
niques and biochemical, microbiological and pathological
analysis. The patients were divided into five groups according
to the aetiology of their pericardial effusions.
Results:
Carbohydrate antigen (CA) 12-5 and CA 15-3, and
carcinoembryonic antigen (CEA) levels were significantly
higher in patients with malignancies than in those with viral/
idiopathic pericarditis. With multivariate analysis, CA 15-3
levels were found to be the most significant determinant (
p
=
0.027). In the ROC curve analysis, CA 15-3 values above 25
U/ml predicted a malignancy with 71% sensitivity and 78%
specificity.
Conclusion:
Tumour markers, particularly CA 15-3, may be
useful in the differential diagnosis and prediction of malig-
nancies in patients with pericardial effusion. In patients with
viral/idiopathic aetiology, these serum tumour markers were
slightly elevated in the acute phase, but after a mean of one
year of follow up, their levels returned to normal, contrary to
those with malignancies.
Keywords:
pericardial effusion, CA 125, CA 15-3, CEA
Submitted 1/4/10, accepted 3/6/11
Cardiovasc J Afr
2012;
23
: 143–146
DOI: 10.5830/CVJA-2011-029
Pericardial effusions (PE) are common and produced by a wide
variety of diseases,
1,2
including viral and bacterial infections,
tuberculosis, malignancy, heart failure, chronic renal failure and
rheumatic diseases.
1,2
However, despite all the diagnostic tests
available, the most common cause is idiopathic.
1-3
Malignancy
is not the main cause of PE, but PE may be the first indication
of cancer, and therefore early detection would enable rapid
diagnosis, which is important in improving the survival rate of
cancer patients.
4
Sampling of pericardial fluid (pericardiocentesis) and
pericardial biopsy play an important role in identifying the
underlying aetiology of PE.
5
It is not always possible to perform
pericardial biopsy, however, due to its potential complications.
Although pericardiocentesis is easier to perform than pericardial
biopsy, it has a low probability of detection of malignant cells
(30–50%).
6
High complication rates in mild and moderate
effusions also limit its diagnostic application.
There are a limited number of markers available for the
evaluation and differential diagnosis of PE,
6,7
including
carbohydrate antigen (CA) 125, CA 15.3 and carcinoembryonic
antigen (CEA).
6,8
CEA has been identified as a useful marker
for differentiating PE related to malignant pathology of the
gastrointestinal system.
5,9,10
Increased levels of CA 15-3 in the blood are primarily
observed in breast cancer.
11,12
CA 15-3 has also be used to
evaluate both pleural and pericardial effusions.
13-16
CA 125 levels
may be detected in pericardial fluid secretions related to both
malignant and benign aetiologies.
16,17
Therefore, the prognostic
value of CA 125 for the detection of malignancies is limited.
18
The aim of this study was to evaluate the relationship between
tumour markers and the underlying aetiology in patients with PE.
We also examined the diagnostic value of these tumour markers
in detecting malignancies in patients with PE, and determined
their changing levels during follow up.
Methods
A total of 76 patients with PEwho were admitted or referred to our
hospital between January 2004 and March 2007 were included
in the study. The mean follow-up period was 18
±
7 months
(range 8–27 months). The aetiological evaluation included
complete blood count, measurement of troponin I, erythrocyte
sedimentation rate, viral disease determination (Epstein Barr,
cytomegalo virus, coxsackie A virus, parva virus, hepatitis
A, B, C), thyroid-stimulating hormone, and rheumatological
markers [rheumatoid factor (RF), anti-nuclear antibody,
anti-smooth muscle antibody, anti-double-stranded DNA].
All patients underwent computerised tomography of the
thorax. Patients with heart failure (ejection fraction
<
45%)
and severe pericardial effusion were excluded (as it could have
affected the marker levels). Patients who refused follow-up visits
were also excluded. In total, seven patients were excluded from
the study.
Echocardiography was performed on all patients and CA
markers were checked at the time of hospitalisation and at
the end of the follow-up period. Pericardiocentesis was not
Department of Cardiology, Kocaeli University Medical
Faculty, Kocaeli, Turkey
U BILDIRICI, MD,
U CELIKYURT, MD
E ACAR, MD
T SAHIN, MD
G KOZDAG, MD
D URAL, MD
Seka State Hospital, Kocaeli, Turkey
O BULUT, MD
