Suzanne L. Aquino, MD #{149} W. Richard Webb, MD #{149} Bryan J.

Gushiken, MD
Pleural Exudates and Transudates:
Diagnosis with Contrast-enhanced CT
803
Thoracic Radiology
PURPOSE: To determine the accu-
racy of computed tomography (CT) in
enabling differentiation of pleural
exudates from transudates.
MATERIALS AND METHODS:
Eighty consecutive patients (86 effu-
sions) underwent contrast-enhanced
CT. Thoracentesis was performed to
measure pleural and serum total pro-
tein and lactate dehydrogenase (LDH)
values. Effusions were classified as
exudates with accepted criteria. CT
scans were evaluated for the pres-
ence and appearance of parietal pleu-
ral and extrapleural fat thickening.
RESULTS: Fifty-nine effusions were
exudates and 27 were transudates.
Thirty-six of the 59 exudates (61%)
were associated with parietal pleural
thickening. All cases of empyema
and 56% of the parapneumonic exu-
dative effusions had pleural thicken-
ing. The specificity of this finding in
diagnosing the presence of an exu-
date is 96%.
CONCLUSION: Parietal pleural
thickening at contrast-enhanced CT
almost always indicates the presence
of a pleural exudate. A pleural exu-
date in the absence of pleural thick-
ening occurs most frequently in pa-
tients with malignancy or uncompli-
cated parapneumonic effusion.
Index terms: Pleura, CT #{149} Pleura, diseases,
66.335, 66.693, 66.773 #{149} Pleura, fluid, 66.693,
66.76, 66.773 #{149} Pleura, infection, 66.76 #{149} Pleura,
neoplasms, 66.335
Radiology 1994; 192:803-808
I From the Department of Radiology, Univer-
sity of California, 505 Parnassus Ave. San Fran-
cisco, CA 94143-0628. Received December 29,
1993; revision requested February 28, 1994; revi-
sion received April 5; accepted April 25. Address
reprint requests to S.L.A.
RSNA, 1994
C OMPUTED tomography (CT) is fre-
quently used to assess patients
with pleural abnormalities associated
with neoplasm, asbestos exposure,
pneumonia, and empyema. It has
been recently reported that the find-
ing of parietal pleural thickening at
contrast-enhanced CT can help distin-
guish pleural effusions representing
exudates or empyema from those rep-
resenting transudates (1). Transuda-
tive pleural effusions and a large per-
centage of malignant exudates do not
show parietal pleural thickening or
enhancement. Also, several studies
have described extrapleural fat thick-
ening and edema in association with
pleural thickening or effusion, al-
though it was unclear whether chest-
tube placement had preceded these
findings in many of the cases.
We reviewed the appearance of
the pleura on contrast-enhanced CT
scans obtained in 80 consecutive pa-
tients with pleural effusions to deter-
mine the accuracy of CT in enabling
differentiation of pleural exudates
from transudates on the basis of pari-
etal pleural thickening. The appear-
ance of extrapleural fat was also
assessed.
MATERIALS AND METHODS
Eighty consecutive patients with 86
pleural effusions were evaluated with con-
trast-enhanced CT and diagnostic thora-
centesis. Patients were selected for the
study if they met the following criteria:
(a) presence of a pleural effusion at con-
trast-enhanced CT, (b) pleural lactic dehy-
drogenase (LDH), total protein, glucose,
and serum LDH and total protein values
obtained at thoracentesis, and (c) results
from pleural effusion culture and cytologic
examination. There were 25 female pa-
tients and 55 male patients aged 10-88
years (mean, 58 years). The 86 effusions
represented a variety of diagnoses but
most were pleural effusions associated
with malignancy or pneumonia (Table 1).
No patients had previously undergone
placement of a pleural drainage tube or
sclerotherapy.
Thoracentesis was performed an aver-
age of 4.3 days (range, 0-20 days) before or
after the CT examination. For the pur-
poses of this study, a pleural effusion was
classified as an exudate or a transudate in
accordance with definitions proposed by
Light (2). Modifications of Lights criteria
have been shown to increase the specific-
ity in detecting exudates (3); however, we
chose to adhere to the original criteria,
which is followed at our institution and at
the hospital laboratory. To be classified as
an exudate, an effusion met at least one of
the three following criteria: (a) a pleural
fluid total protein/serum total protein ra-
tio of more than 0.5, (b) a pleural fluid
LDH/serum LDH ratio of more than 0.6,
or (c) pleural fluid LDH greater than two-
thirds of the upper limits of normal for
serum LDH. At our institution, the upper
limit of normal for serum LDH is 200 IU.
Effusions were classified as transudates
if they that did not meet any of the above
criteria. Patients were excluded from the
study if there were insufficient laboratory
data with which to evaluate the nature of
the effusion. On the basis of these criteria,
59 of 86 effusions were classified as exu-
dates. Twenty-seven effusions were classi-
fled as transudates (Table 1).
Pleural effusions were also analyzed for
the presence of infectious organisms by
means of staining (including gram, fungal,
and mycobacterial stains), culture (aerobic,
anaerobic, fungal, and mycobacterial), and
cytologic examination. Parapneumonic
effusions are usually defined as those asso-
ciated with pneumonia, lung abscess, or
bronchiectasis, and an empyema is consid-
ered to be a complicated parapneumonic
effusion in which pus is present and the
culture is positive (2,4). Complicated para-
pneumonic effusions necessitating chest-
tube drainage are also considered to be
those with very high LDH ( > 1,000 LU),
low pH ( < 7.0), or low glucose ( < 40 mg/
dL) values (2). In our study, effusions were
considered to be empyemas only if culture
results were positive. Otherwise, all effu-
sions associated with pulmonary infec-
tions were classified as parapneumonic
effusions. Malignant effusions were de-
fined by the presence of positive cytologic
results. Effusions with negative cytologic
Abbreviation: LDH = lactic dehydrogenase.
Diagnosis
No. of
Effusions
No. of
Exudates
No. of
Transudates
Malignancy
Positive cytologic findings 14 14 0
Negative cytologic findings 17 9 8
Parapneumonic effusion 25 18 7
Empyema 10 10 0
Liver transplantation 4 0 4
Congestive heart failure 3 0 3
Dressler syndrome 2 2 0
Contralateral empyema 2 0 2
Sepsis 1 0 1
FocalLIP 1 1 0
Asbestos exposure 1 1 0
Trauma 1 1 0
Pulmonary embolus 1 1 0
Hypoalbuminemia 1 0 1
Undiagnosed
Total
3 2 1
86 59 27
Note-LIP = lymphocytic interstitial pneumonia.
Table 2
Thickness of Parietal Pleura in 59
Exudative Effusions
Note-LIP = lymphocytic interstitial pneu-
monia.
804 #{149} Radiology September 1994
results were classified as effusions associ-
ated with malignancy, in distinction to
the truly malignant effusions.
Contrast-enhanced CT was performed
in all patients with either a GE 9800 scan-
ner (GE Medical Systems, Milwaukee,
Wis) or an Imatron Cine CT scanner (Ima-
tron, San Francisco, Calif). Contiguous 10-
mm-thick sections were obtained with the
GE scanner, and contiguous 6-8-mm-thick
sections were obtained with the Imatron
scanner. All patients received 60-200 mL
of iohexol (Omnipaque 300 or 350; Win-
throp-Breon Laboratories, New York, NY)
at a rate of 1.7 mL/sec.
CT scans were reviewed blindly by two
observers who reached a consensus. Stud-
ies were evaluated for the presence of pa-
rietal and visceral pleural thickening. Pan-
etal pleural thickening was diagnosed
only if a pleural line was visible internal to
the nibs, in areas in which pleural effusion
was also seen (5). When visible, the thick-
ness of parietal pleura was measured and
its extent and appearance classified as fo-
cal or diffuse and irregular on smooth.
Pleural thickening was considered diffuse
if it was visible in all locations in which
fluid was visible. The presence of visceral
pleural enhancement and thickening adja-
cent to fluid collections was also assessed.
Effusions were described as either cres-
centric and uniloculan or loculated. Cres-
centric, unilocular effusions were defined
as collections of pleural fluid that accu-
mulated in the dependent portion of the
chest seen without septations. Conversely,
loculation was used to describe any effu-
sion that had septations, was compant-
mentalized in the pleural space, or was
accumulated in the fissures, away from
most of the effusion.
The extrapleural fat adjacent to the ribs
was evaluated for visibility, thickness,
asymmetry, and attenuation. Because ex-
trapleural fat can be seen in healthy sub-
jects, for the purposes of this study, the
presence of extrapleural fat was arbitrarily
considered normal if it was less than 2 mm
thick. The attenuation of the extrapleural
fat was estimated to be the same as that of
the chest wall fat, the same as that of the
musculature of the chest wall, or intenme-
diate, between that of fat and musculature.
RESULTS
Fifty-nine exudates were evaluated
with CT (Table 1). Thirty-six showed
focal or diffuse parietal pleural thick-
ening (Table 2). Thus, the sensitivity
of parietal pleural thickening as seen
at CT in the detection of the presence
of a pleural exudate was 61% (confi-
dence interval, 47%-73%) (Table 3).
Pleural thickness measured 2-4 mm
in 30 of the 36 cases with thickening,
and the thickness of the pleura did
not appear to correlate with the diag-
nosis. Of the 36 effusions associated
with parietal pleural thickening, 12
showed diffuse, smooth, parietal
pleural thickening in association with
Table 1
Diagnoses in Patients with 86 Effusions
the pleural effusion (Fig 1) and two
had diffuse, irregular thickening
(Table 4, Fig 2). The remaining 22 ef-
fusions were associated with irregular
or smooth focal pleural thickening.
All effusions were crescentric and uni-
locular. Diffuse, irregular thickening
was seen in the presence of malig-
nancy and asbestos exposure; how-
ever, only two effusions were associ-
ated with this finding (Fig 2). Focal,
irregular, pleural thickening was seen
with malignancy. Twenty-three pleu-
ral exudates did not show parietal
pleural thickening (Table 2). These
included 12 effusions associated with
malignancies, eight parapneumonic
effusions, two of unknown cause, and
one occurring after chest trauma.
Parietal pleural thickening was vis-
ible in all 10 empyemas (Fig 3) but
was present in only 10 (56%) of 18
parapneumonic exudates (Fig 4). All
five complicated parapneumonic effu-
sions (LDH > 1,000 LU or glucose <40
mg/dL, negative findings at culture
for infection) had pleural thickening.
Neither the characteristics of the pleu-
ral thickening nor the thickness of the
parietal pleura were helpful for dis-
tinguishing empyema from an un-
infected parapneumonic exudate
(Tables 2, 4). However, none of the
transudative parapneumonic effu-
sions were associated with pleural
thickening.
Of the 27 transudates, only one
showed parietal pleural thickening.
Thus, the specificity of this finding in
diagnosing the presence of an exu-
date is 96% (confidence interval, 79%-
99%; positive predictive value, 97%;
negative predictive value, 54%) (Table
3). The patient with pleural thicken-
Diagnosis
Thickness
Normal
2
mm
2-4
mm
>4
mm
Malignancy
Positive cyto-
logic findings 10 2 1 1
Negative cyto-
logic findings 2 3 2 2
Pneumonia 8 3 6 1
Empyema 0 3 6 1
FocaILIP 0 1 0 0
Dressler syndrome 0 0 1 1
Asbestos exposure 0 0 1 0
Pulmonary
embolus 0 1 0 0
Trauma 1 0 0 0
Undiagnosed
Total
2 0 0 0
23 13 17 6
ing developed bilateral effusions in
association with congestive heart fail-
ure (Fig 5). CT demonstrated unilat-
eral, diffuse, parietal pleural and ex-
trapleural fat thickening, visceral
pleural thickening, reduction in vol-
ume of the ipsilateral hemithorax, and
collapse of the underlying lung. It
was subsequently determined that
this patient likely had a preexisting
pleural peel resulting from prior em-
pyema. Of the remaining 26 transuda-
tive effusions, 25 were crescentric and
unilocular. One transudative effusion
caused by congestive heart failure
showed unilateral loculation of fluid
in the fissures without pleural thick-
ening. Of the 27 transudates, most
Table 3
Summary of CT Findings in 59 Exudates and 27 Transudates
Parietal Extrapleural Visceral
Group Pleural Thickening Fat Thickening Pleural Thickening
Exudates 36 21 13
Transudates 1 1 1
2.
Figures 1, 2. (1) Contrast-enhanced CT scan of the thorax of a 70-year-old man with Dressier
syndrome who recently underwent open heart surgery. The parietal pleura adjacent to the
left effusion is diffusely thickened. The underlying fat is more than 4 mm thick and is of inter-
mediate attenuation. Results of biopsy and cytologic examination indicated a benign effusion,
consistent with inflammation. (2) Contrast-enhanced CT scan shows a malignant pleural effu-
sion in a 72-year-old woman with a history of breast carcinoma. Irregularly thickened parietal
pleura is adjacent to the effusion, which was proved to be malignant at cytologic examination.
The subpleural fat is thickened and of intermediate attenuation.
Table 4
Characteristics of Parietal Pleural Thickening in 59 Exudates
ns
Diagnosis Normal
Focal
Smooth
Thickening
Irregular
Diffuse
Smooth
Thickening
Irregular
Malignancy
Positive cytologic finding 10 1 1 1 1
Negative cytologic finding 2 4 3 0 0
Pneumonia 8 5 0 5 0
Empyema 0 6 0 4 0
FocalLIP 0 1 0 0 0
Dressler syndrome 0 0 0 2 0
Asbestos exposure 0 0 0 0 1
Pulmonary embolus 0 1 0 0 0
Trauma 1 0 0 0 0
Undiagnosed
Total
2 0 0 0 0
23 18 4 12 2
Volume 192 #{149} Number 3 Radiology #{149} 805
Note-LIP = lymphocytic interstitial pneumonia.
additional cases had an extrapleural
fat layer that was less than 2 mm
thick, which was considered normal.
Only one of the 23 exudates without
parietal pleural thickening was inter-
preted as showing extrapleural fat
thickening; this occurred in a patient
. . with a parapneumonic effusion. Fat
thickening was associated with a tran-
sudate only in the patient with pari-
etal pleural thickening. The specificity
of this finding was 96% (confidence
interval, 79%-99%). Increased attenu-
ation of extrapleural fat (intermediate
or the same as that of soft tissue) was
seen in 10 patients with fat thicken-
ing. Most patients had fat of interme-
diate attenuation, and eight had em-
pyemas or parapneumonic effusion
(Figs 3, 4).
Visceral pleural thickening was vis-
ible in 13 patients with exudates (five
with pneumonia, five with empyema,
and three with malignancy) (Fig 4)
and in the patient with a transudative
effusion and pleural thickening (Fig
5). Thickened visceral pleura mea-
sured less than 2 mm in all patients.
The sensitivity of this finding was
22% and the specificity 96%.
Of the 31 effusions in patients with
malignancies, 23 were exudates (Table
6). Eleven showed parietal pleural
thickening. Seven of the 11 had nega-
tive findings at cytologic examination,
and four had positive findings (Fig 2).
There was no difference in the degree
or appearance of pleural thickening
between those patients with positive
and negative cytologic findings
(Tables 2, 4). Of the 12 exudates asso-
ciated with malignancy that did not
show parietal pleural thickening, 10
had positive cytologic findings (Fig 6).
Of the effusions associated with lung
cancer, none of those associated with
pleural thickening had positive cyto-
logic findings; four of seven effusions
associated with lymphoma, breast
cancer, or other tumors had pleural
thickening and positive cytologic re-
sults. The malignant mesothelioma
diagnosed with open biopsy showed
diffuse nodular thickening. Extrapleu-
ral fat thickening was visible in three
of the 23 exudative effusions.
DISCUSSION
occurred in patients with a history of
malignancy but were associated with
benign pleural fluid at cytologic ex-
amination or were parapneumonic
effusions in the absence of empyema
(Table 1).
Of the 36 exudates associated with
parietal pleural thickening, 20 showed
extrapleural fat thickening; therefore,
the sensitivity was 36% (confidence
interval, 24%-49%) (Table 3). Eight of
these exudates were empyemas and
eight were parapneumonic effusions
(Table 5). The extra pleural fat was
2-4 mm thick in nine cases and at
least 4 mm thick in 12 (Table 5); eight
Exudative effusions can have a vari-
ety of causes. They often reflect the
presence of pleural inflammation, in-
fection, or neoplasm, and in such
cases are thought to be due to an in-
creased permeability of abnormal
pleural capillaries and the release of
high-protein fluid into the pleural
space (2,6). In patients with malig-
Figures 3, 4. (3) Contrast-enhanced CT scan of the thorax of a persistently febrile 48-year-old
woman I month after orthotopic liver transplantation. A pulmonary abscess in the right lower
lobe is in direct communication with the adjacent pleural effusion. The panietal pleura is en-
hanced, and the subpleural fat is thickened and has almost the same attenuation as that of soft
tissue. Enterococci were found at culture of the effusion. (4) CT scan of the thorax of a 29-year-
old man with acquired immunodeficiency syndrome and a partially treated pneumococcal
pneumonia. The posterior parietal pleura is smoothly thickened, and the underlying fat is of
intermediate attenuation. The visceral pleura is also enhanced (arrow). Although the findings
were highly suggestive of empyema, results of multiple cultures of the effusion were negative.
I Table5
Thickness and Attenuation of Extrapleural Fat in 59 Exudates I
2mm 2-4mm >4mm #{149}
Interme- Soft Interme- Soft Interme- Soft
Diagnosis
Malignancy
Normal Fat diate Tissue Fat diate Tissue Fat diate Tissue
Positive cytologic
finding 11 0 2 0 0 0 0 1 0 0
Negative cytologic
finding 6 1 0 0 0 0 0 1 1 0
Pneumonia 6 2 1 1 0 1 2 3 2 0
Empyema 2 0 0 0 4 1 0 1 1 1
FocalLIP 0 1 0 0 0 0 0 0 0 0
Dressler syndrome 0 0 0 0 1 0 0 0 1 0
Asbestos exposure 1 0 0 0 0 0 0 0 0 0 I
Pulmonary embolus 1 0 0 0 0 0 0 0 0 0 :
Trauma 2 0 0 0 0 0 0 0 0 0
Undiagnosed
Total
I 0 0 0 0 0 0 0 0 0
30 4 3 1 5 2 2 6 5 1
Note.-Fat = same attenuation as that of chest wall fat, intermediate = attenuation between that of
fat and soft tissue, soft tissue = same attenuation as that of soft tissue. LIP = lymphocytic interstitial
pneumonia.
806 #{149} Radiology September 1994
nancy, exudates can also reflect the
presence of lymphatic obstruction or
lung disease rather than a pleural ab-
normality (6). Furthermore, it is theo-
rized that some exudates may result
from fluid released directly into the
pleural space by a damaged lung, as
with pneumonia, pulmonary embo-
lism, and lung transplantation (6).
Transudative effusions are not asso-
ciated with pleural disease and are
considered to be the result of systemic
abnormalities that cause an imbalance
in the hydrostatic and osmotic forces
leading to the formation of pleural
fluid. This results in an outpouring of
low-protein fluid from the pleural
capillaries and, occasionally, the pa-
renchymal interstitium into the pleu-
ral space (6). Common causes of a
transudative effusion include conges-
tive heart failure, cirrhosis, and ne-
phrotic syndrome.
Differentiating an exudate from a
transudate can be important in clini-
cal management, particularly in pa-
tients with infection and malignancy.
For example, the presence of a transu-
dative effusion in association with
pneumonia does not warrant further
evaluation or treatment; an exudative
parapneumonic effusion may not re-
quire chest-tube drainage if simple.
However, if the exudate becomes
complicated, chest-tube therapy is
indicated because it is likely that em-
pyema will occur. In a patient with an
underlying malignancy, an exudative
effusion in the absence of infection is
strongly suggestive of pleural metas-
tases or recurrence despite negative
cytologic findings. Often these effu-
sions necessitate drainage and sclero-
therapy when symptomatic or recur-
rent. In a patient with neoplasms, the
presence of a transudate excludes ma-
lignant involvement of the pleura,
although they can be seen in the early
stages of mediastinal involvement
with lymphatic obstruction (7).
Exudates and transudates differ in
many ways, but according to gener-
ally accepted criteria proposed by
Light (2), exudative effusions are con-
sidered to be those with a pleural
fluid total protein/serum total protein
ratio of more than 0.5, a pleural fluid
LDH/serum LDH ratio of more than
0.6, or an absolute pleural fluid LDH
of more than two-thirds of the normal
value of serum LDH. A pleural fluid-
specific gravity exceeding 1.016 or a
pleural fluid protein exceeding 3 g/dL
are other criteria used to diagnose
exudate, but these have a somewhat
lower specificity (8). Although the
classification of a pleural effusion as
an exudate or transudate is usually
based on results of thoracentesis, it is
also important to know the accuracy
of imaging studies in assessing pleural
fluid, as they are commonly obtained
in this setting and are often used to
guide interventional procedures.
Sonography can be useful for de-
tecting pleural fluid characteristics
compatible with an exudative effu-
sion. Yang et al (9) found that all effu-
sions with septation, complex non-
septation, or homogeneous echogeni-
city at sonography were exudative.
Anechoic effusions, however, could
be either transudative or exudative.
The sensitivity of sonography in their
study was 66%, with a specificity of
100% and a positive predictive value
of 100% . In our study, parietal pleural
thickening was shown at contrast-
enhanced CT in 61 % of exudates, and
this finding had a specificity of 96%
and a positive predictive value of 97%.
Waite et al (1) have reported that
CT shows pleural thickening and en-
hancement in almost all patients with
empyema or parapneumonic effusion.
In their study, 24 of 25 empyemas
demonstrated pleural thickening and
enhancement at CT; however, some
patients underwent thoracostomy
tube insertion before undergoing CT.
Table 6
Characteristics of 31 Pleural Effusions Associated with Malignancy
Diagnosis
No. of
Exudates
No. of
Transudates
No. with
Thickened
Panietal Pleura
No. with No
Thickening of
Parietal Pleura
Lymphoma
Positive cytologic findings 4 0 2 2
Negative cytologic findings 3 3 2 4
Breast cancer
Positive cytologic findings 5 0 1 4
Negative cytologic findings I 3 0 4
Lung cancer
Positive cytologic findings 3 0 0 3
Negative cytologic findings 4 0 4 0
Other
Positive cytologic findings 2 0 1 1
Negative cytologic findings
Total
1 2 1 2
23 8 11 20
Note-Other = esophageal cancer, hepatocellular cancer, Kaposi sarcoma, malignant mesothelioma,
ovarian cancer.
Volume 192 #{149} Number 3 Radiology #{149} 807
5. 6.
Figures 5, 6. (5) Contrast-enhanced CT scan of a fibrothorax with superimposed congestive heart failure in a 53-year-old man. The thorax
shows right-sided volume loss with an ipsilateral shift of the mediastinum. The right lung is collapsed, and the parietal and visceral (straight
arrows) pleura are diffusely thickened. The subpleural fat (curved arrow) is thicker than that in the left side. (6) CT scan shows a malignant
pleural effusion in a 77-year-old man with adenocarcinoma of the left lung. The adjacent pleura is normal. The associated pleural effusion was
exudative, and findings at cytologic examination were positive. The thin line with the same attenuation as soft tissue lying between the poste-
nor ribs is normal intercostal musculature (arrow).
In our study, CT was sensitive in de-
tecting pleural thickening in empy-
ema; all 10 empyemas showed pleural
thickening. Conversely, pleural thick-
ening was seen in 56% of the unin-
fected exudative (culture-negative)
parapneumonic effusions. Included in
this group with pleural thickening,
however, are all of the complicated
parapneumonic effusions (ii = 5),
which were treated with chest-tube
drainage. Twenty-eight percent of
parapneumonic effusions were tran-
sudates and did not show pleural
thickening. The thickness of the pan-
- - sions. Of the 23 exudative effusions
associated with neoplasms, 11 (48%)
showed pleural thickening (Table 2).
Surprisingly, only four (28%) of 14
malignant effusions showed pleural
thickening. In the study by Waite et al
(1), 73% of malignant effusions did
not show pleural thickening, and of
the five with thickening, three had
I prior sclerotherapy or infection. Ir-
I regular pleural thickening was seen
I only in association with malignancy
or asbestos exposure (Table 4); the
presence of nodular pleural thicken-
ing has previously been reported as
having a sensitivity of 51% and a
specificity of 94% for the diagnosis of
a malignant pleural process (10).
Thickening and increased attenua-
tion of extrapleural fat is another
finding that is suggestive of pleural
inflammation or infection and has
been reported in patients with empy-
etal pleura or the shape of the effu- ema, malignancy, and asbestos expo-
sion was not helpful in distinguishing sure (1,5,11). Eight of 10 empyemas in
empyemas from uninfected parapneu- our study demonstrated extrapleural
monic effusions because all were cres- fat thickening, with three showing
centric and unilocular (Table 2). abnormally increased fat attenuation
Pleural thickening was less fre- (Fig 3). The two cases that did not
quent when associated with malig- have extrapleural fat thickening had
nancy, and if these cases are ex- fungal empyemas. A prior study (11)
cluded, the sensitivity of parietal has demonstrated increased attenua-
pleural thickening in diagnosing an tion of extrapleural fat in a larger per-
exudate increases to 69%. Of 31 effu- centage of empyemas than was seen
sions in patients with neoplasm, eight in our study. However, pleural drain-
(26%) were transudates and were not age tubes had been placed in 11 of the
associated with pleural thickening; 13 patients. Waite et al (1) found an
none of these were malignant effu- increase in attenuation of extrapleural
808 #{149} Radiology September 1994
fat in 11 of 18 (61%) empyemas with-
out prior tube insertion (1). The some-
what lower frequency of increased fat
attenuation we report may reflect dif-
ferences in the duration of empyema
in our cases. Extrapleural fat thicken-
ing was also common in those pa-
tients with pleural thickening associ-
ated with parapneumonic effusions
(Table 5, Fig 4). Only one of the 23
exudates without associated parietal
pleural thickening was interpreted as
showing extrapleural fat thickening.
Increased attenuation of extrapleural
fat, presumably representing edema
or inflammation, was present in 10
patients with fat thickening; eight
had empyema or parapneumonic ef-
fusion. Four of eight effusions associ-
ated with a fat thickness of less than
2 mm showed increased attenuation;
this finding may indicate inflamma-
tion in the absence of thickening or
could reflect a volume-averaging phe-
nomenon.
CT can play an important role in
the diagnosis of exudative and tran-
sudative effusions. Pleural thickening
associated with a pleural effusion in a
patient with pneumonia indicates the
presence of an exudate, and thoracen-
tesis is warranted. If pleural thicken-
ing is absent, the presence of an em-
pyema or a complicated parapneu-
monic effusion necessitating chest-
tube drainage is highly unlikely. In
patients with malignancy, the pres-
ence of pleural thickening indicates
the presence of an exudate; the ab-
sence of parietal pleural thickening
does not exclude the presence of an
exudate or malignant effusion, and
thoracentesis should be performed
for diagnosis. The degree of pleural
thickening is not helpful for predict-
ing the diagnosis, although the pres-
ence of extrapleural fat thickening is
suggestive of an empyema or a para-
pneumonic effusion. Irregular pleural
thickening is suggestive of a neo-
plasm. #{149}
References
1. Waite RJ, Carbonneau RJ, Balikian JP,
Umali CB, Pezzella AT, Nash G. Parietal
pleural changes in empyema: appearances
at CT. Radiology 1990; 175:145-150.
2. Light RW. Pleural diseases. Philadelphia,
Pa: Lea & Febiger, 1990; 39-73.
3. Romero 5, Candela A, Martin C, Hernan-
dez L, Tnigo C, GilJ. Evaluation of differ-
ent criteria for the separation of pleural
transudates from exudates. Chest 1993;
2:399-404.
4. Light RW. Parapneumonic effusions and
empyema. Clin Chest Med 1985; 6:55-62.
5. Im JG, Webb WR, Rosen A, Gamsu G.
Costal pleura: appearances at high-resolu-
tion CT. Radiology 1989; 171:125-131.
6. Broaddus VC, Light RW. What is the ori-
gin of pleural transudates and exudates
(editorial)? Chest 1992; 102:658-659.
7. Sahn SA. Malignant pleural effusions.
Clin Chest Med 1985; 6:113-125.
8. Jay SJ. Diagnostic procedures for pleural
disease. Clin Chest Med 1985; 6:33-48.
9. Yang PC, Luh KT, Chang DB, Wu HD, Yu
CJ, Kuo SH. Value of sonography in de-
termining the nature of pleural effusion:
analysis of 320 cases. AJR 1992; 159:29-33.
10. Leung AN, Muller NL, Miller RR. CT in
differential diagnosis of diffuse pleural dis-
ease. AJR 1990; 154:487-492.
11. Takasugi JE, Godwin JD, Teefey SA. The
extrapleural fat in empyema: CT appear-
ance. BrJ Radiol 1991; 64:580-583.