Professional Documents
Culture Documents
Rosado-de-Christenson Localized Fibrous Tumors of The Pleura 2003
Rosado-de-Christenson Localized Fibrous Tumors of The Pleura 2003
RadioGraphics
Abbreviations: AFIP ⫽ Armed Forces Institute of Pathology, H-E ⫽ hematoxylin-eosin, LFTP ⫽ localized fibrous tumors of the pleura, PA ⫽ pos-
teroanterior
Index terms: Lung neoplasms, 66.317, 66.3254 ● Pleura, neoplasms, 66.317, 66.3254
The opinions and assertions contained herein are the private views of the authors and are not to be construed as official nor as reflecting the views of
the Departments of the Air Force, Navy, Army, or Defense.
760 May-June 2003 RG f Volume 23 ● Number 3
RadioGraphics
Introduction lesions were evaluated with chest radiography, 78
Localized fibrous tumors are rare mesenchymal with chest CT, 18 with MR imaging, 10 with an-
neoplasms that most commonly affect the pleura giography, and nine with US. Fifty-five patients
but have also been described in a number of other were imaged with chest radiography and CT; 17
locations including the mediastinum and the with chest radiography, CT, and MR imaging;
lung. Extrathoracic localized fibrous tumors have and one with chest radiography and MR imaging.
been reported in the abdomen, the head and Six patients were evaluated only with CT, and
neck, and the central nervous system. Many three with only chest radiography.
names have been used to designate this neoplasm. For 80 patients, detailed clinical histories were
The inconsistent nomenclature that appears in available and were reviewed for age, gender, and
the published literature emphasizes controversies clinical presentation. Surgical and pathology re-
regarding the precursor cell for localized fibrous ports were reviewed to determine tumor size, lo-
tumors and their variable microscopic appearance cation, pleural surface of origin, presence or ab-
and unpredictable biologic behavior. sence of a pedicle, as well as gross and micro-
Patients with localized fibrous tumors of the scopic findings. Results of needle biopsies, when
pleura (LFTP) are typically adults who may performed, were also noted. The diagnosis of lo-
present with symptoms related to local or sys- calized fibrous tumor was confirmed in every case
temic effects produced by the neoplasm or who through the microscopic evaluation of glass slides
may be entirely asymptomatic. Benign and malig- prepared from the resected tissues and reviewed
nant subtypes of LFTP are recognized. At gross by an experienced thoracic pathologist in the De-
examination, these are lobular soft-tissue masses, partment of Pulmonary and Mediastinal Pathol-
which are often described as pedunculated lesions ogy at the AFIP. The lesions were classified as
arising from the visceral pleura. Radiologically, benign or malignant based on established histo-
they are intrathoracic masses of variable sizes, logic criteria, particularly the presence of more
which may not exhibit the classic imaging features than four mitotic figures per 10 high-power fields.
described in extraparenchymal lesions. On cross- All chest radiographs and CT studies were re-
sectional images, they are well-defined lobular viewed by two thoracic radiologists (M.L.R.,
heterogeneous masses. Excision is curative in the G.F.A.), and MR imaging studies were reviewed
majority of patients, although a small but signifi- by three thoracic radiologists (M.L.R., G.F.A.,
cant number of lesions recur, undergo malignant H.P.M.). Findings were recorded by consensus.
transformation, or metastasize. Chest radiographs were evaluated to determine
Although the most common primary pleural lesion size and location within the thorax as well
neoplasm is malignant mesothelioma, radiologists as border characteristics and presence or absence
should also be able to identify the much rarer of associated findings including pleural effusion
LFTP, as these two neoplasms have radically dif- and mass effect.
ferent prognoses. To help familiarize radiologists CT studies were performed at multiple institu-
with the spectrum of radiologic features of LFTP, tions with a variety of scanners and variable scan-
we review a large series of LFTP, with emphasis ning techniques. Sixty-six studies were performed
on the radiographic appearance of these lesions after the administration of intravenous contrast
and their findings at computed tomography material, and 25 were performed without contrast
(CT), magnetic resonance (MR) imaging, an- material (13 of these studies were performed be-
giography, and ultrasonography (US). We also fore and after intravenous contrast material ad-
describe the clinical presentation of patients with ministration). Lesions were evaluated to deter-
LFTP and the pathologic characteristics of these mine location, mobility within the thorax, border
tumors and discuss their therapy and prognosis. characteristics, attenuation, presence or absence
of calcification, and patterns of contrast enhance-
Materials and Methods ment.
A retrospective review of 101 cases of localized MR imaging studies were performed on a vari-
fibrous tumors referred to the Pulmonary and ety of MR imaging equipment, with various com-
Mediastinal Section of the Department of Radio- binations of axial, coronal, and sagittal planes.
logic Pathology at the Armed Forces Institute of Fifteen lesions were imaged with T1-weighted
Pathology (AFIP) between 1987 and 2001 was and T2-weighted or cine gradient recalled echo
performed. Ten multifocal or recurrent LFTP sequences, and three lesions were imaged with
and nine extrapleural localized fibrous tumors T1-weighted sequences only. Six lesions were
occurring in the mediastinum (n ⫽ 6) and in the imaged before and after administration of intrave-
lung (n ⫽ 3) were excluded. The remaining 82 nous gadolinium. All lesions were evaluated to
LFTP form the basis of this review. Seventy-six determine morphologic features, signal intensity
characteristics, and patterns of enhancement and
to exclude involvement of adjacent structures.
RG f Volume 23 ● Number 3 Rosado-de-Christenson et al 761
RadioGraphics
Table 1
Clinical Data for 80 Patients with LFTP
Figure 3. Benign
LFTP. Photograph
of a cut section of a
pedunculated LFTP
demonstrates an
ovoid lobular mass
with a thin pedicle
(arrow) by which it
was attached to the Figure 4. Benign LFTP. Photograph of
visceral pleura. Note a benign LFTP demonstrates a spherical
the firm, yellow-tan lobular mass. Note the site of prior broad
appearance of the attachment (arrows) of the lesion to the
tumor. parietal pleura.
Pathologic Findings
Needle biopsies were performed in 27 lesions (22
benign, five malignant). The diagnosis of LFTP
was established in seven benign and in one malig- Figure 5. Benign LFTP. Photograph of a cut section
of a benign LFTP demonstrates an ovoid lobular mass
nant LFTP based on the microscopic examina-
with extensive internal hemorrhage and necrosis.
tion of the biopsy specimens. The remaining bi-
opsies were inconclusive. There were 64 benign
and 18 malignant LFTP. Pathology reports were (22%) had necrosis, and cystic change was found
available in 81 cases. Forty-nine LFTP (60%) in three (17%).
were located in the right hemithorax, and 33
(40%) in the left. Benign lesions ranged in size Imaging Findings
from 2 to 30 cm (average size, 13.2 cm), and ma-
lignant lesions ranged in size from 3 to 23 cm (av- Chest Radiography.—Seventy-six LFTP (59
erage size, 14.4 cm). Macroscopic descriptions of benign, 17 malignant) were imaged with radiog-
63 benign LFTP noted gross evidence of hemor- raphy. Sixty (79%) lesions extended into or occu-
rhage in 27 (43%), cystic change in 14 (22%), pied the inferior hemithorax (Fig 6). Twenty-one
and necrosis in three (5%) (Fig 5). Seven malig- of these lesions (18 benign, three malignant)
nant LFTP (39%) exhibited hemorrhage, four abutted the ipsilateral hemidiaphragm, con-
formed to its shape, and simulated diaphragmatic
elevation or eventration (Fig 7). The majority of
RG f Volume 23 ● Number 3 Rosado-de-Christenson et al 763
RadioGraphics
Figure 6. Benign LFTP in an asymptomatic 29-year-old woman. Posteroanterior (PA) (a) and lateral (b)
chest radiographs demonstrate an ovoid, slightly lobular mass in the left inferior hemithorax that abuts the left
hemidiaphragm.
Figure 7. Benign LFTP in a 62-year-old woman with right-sided chest pain. PA (a) and lateral (b) chest ra-
diographs demonstrate a rounded well-defined mass of the right inferior hemithorax that conforms to the shape
of the diaphragm and mimics diaphragmatic elevation. The LFTP was discovered at abdominal CT.
764 May-June 2003 RG f Volume 23 ● Number 3
RadioGraphics
Figure 9. Benign LFTP in an asymptomatic 54-year-old man. (a, b) PA (a) and lateral (b) chest radiographs ‹
demonstrate an ovoid mass in the left hemithorax. The inferior border of the lesion is well defined, but its superior
border is ill defined (a). The lesion is aligned along the course of the major fissure on the lateral radiograph (b).
(c, d) PA (c) and lateral (d) chest radiographs obtained 6 years later demonstrate interval growth of the lesion and a
change in its position within the thorax that confirms its pleural location. (e) Contrast-enhanced chest CT scan (lung
window) shows the superior border of the lesion, its extension into the fissure, and a pedicle (arrow) that connected
the mass to the fissural visceral pleura. (f) Contrast-enhanced chest CT scan (mediastinal window) demonstrates the
heterogeneously enhancing lobular mass, which forms obtuse and acute angles with the adjacent pleura. Note associ-
ated ipsilateral pleural thickening or fluid.
RG f Volume 23 ● Number 3 Rosado-de-Christenson et al 765
RadioGraphics
766 May-June 2003 RG f Volume 23 ● Number 3
RadioGraphics
Figure 11. Benign LFTP in an asymptomatic 39-year-old man. Unenhanced chest CT scans (mediastinal
window) demonstrate a soft-tissue mass of the left inferior hemithorax with well-defined lobular borders. The
mass forms acute angles with the adjacent pleural surface and contains a geographic area of low attenuation
(a) and multifocal coarse calcifications (b).
768 May-June 2003 RG f Volume 23 ● Number 3
RadioGraphics Figure 12. Malignant LFTP in a 30-year-old man with chest pain. (a) PA chest radiograph demon-
strates a well-marginated rounded mass in the right paravertebral inferior hemithorax. (b) Unenhanced
chest CT scan (mediastinal window) demonstrates a well-defined heterogeneous mass with focal areas of
low attenuation and subtle punctate calcification. Note that the lesion forms acute angles with the adja-
cent pleura. (c) Selective angiogram demonstrates exuberant tumor vascularity. (d) Photograph of a cut
section of the resected gross specimen demonstrates a heterogeneous spherical mass with extensive ne-
crosis, hemorrhage,and cystic change.
RadioGraphics
RadioGraphics
Figure 16. Benign LFTP in an asymptomatic 62-year-old man. (a) Contrast-enhanced chest
CT scan (mediastinal window) demonstrates a lobular heterogeneous soft-tissue mass with geo-
graphic and focal areas of low attenuation. Note that the lesion forms acute angles with the adja-
cent pleura. (b) Photograph of a cut section of the gross specimen demonstrates a well-circum-
scribed lobular mass with a focal area of necrosis that corresponds to the low-attenuation area seen
at CT.
RadioGraphics
Figure 18. Benign LFTP in an asymptomatic 68-year-old man. (a) Coronal T1-weighted MR
image demonstrates a large lobular heterogeneous mass of intermediate signal intensity with linear
areas of high signal intensity. Note mass effect on the ipsilateral hemidiaphragm and mediastinum.
(b) Coronal T1-weighted MR image obtained after intravenous administration of gadolinium
demonstrates heterogeneous patchy multifocal enhancement within the lesion.
Figure 19. Benign LFTP in a 48-year-old man with chest pain. (a) Sagittal T1-weighted MR
image demonstrates a lobular ovoid mass of intermediate signal intensity located in the right infe-
rior hemithorax. Note mass effect on the liver and focal chest wall invasion. (b) Sagittal T2-
weighted MR image at the same level demonstrates heterogeneous high signal intensity with flow
void areas within the lesion that represent vessels and intrinsic low-signal-intensity septa (arrow).
MR Imaging.—Eighteen LFTP (13 benign, five 18). Chest wall involvement was exhibited by one
malignant) were evaluated with MR imaging. benign (8%) and one malignant LFTP (20%)
Coronal or sagittal images were obtained in all (Fig 19), and diaphragm involvement was evident
cases. Eleven (85%) benign and four (80%) ma-
lignant LFTP manifested as lobular masses (Fig
772 May-June 2003 RG f Volume 23 ● Number 3
RadioGraphics
Figure 20. Malignant LFTP in a 53-year-old woman with abdominal pain. (a) Axial T1-weighted MR
image shows a mass of intermediate to low signal intensity in the right diaphragmatic region, which pro-
duces mass effect on the diaphragm and indents the liver. (b) Sagittal T1-weighted MR image obtained
after intravenous administration of gadolinium shows heterogeneous contrast enhancement and exten-
sion through the diaphragm. Note that a large portion of the mass is intrahepatic and thus a tumor of
liver origin would have to be considered. Diaphragm invasion was documented at surgery.
Table 3
MR Imaging Signal Intensity Characteristics of 18 LFTP
in two malignant LFTP (40%) (Fig 20). Mass coronal or sagittal images (Figs 18, 19). Patterns
effect on the diaphragm was documented in 12 of signal intensity are summarized in Table 3. Six
(67%) cases (eight benign, four malignant) on lesions (five benign, one malignant) were imaged
after the administration of intravenous gadolin-
ium, and all exhibited heterogeneous contrast
enhancement (Figs 18, 20).
RG f Volume 23 ● Number 3 Rosado-de-Christenson et al 773
RadioGraphics
Angiography.—Ten LFTP were studied with Pathogenesis
angiography (seven benign, three malignant). All Different theories about the cell of origin for
lesions (100%) exhibited exuberant tumor vascu- LFTP have been proposed and probably account
larity (Fig 12c). for the varied nomenclature used to designate this
rare neoplasm. Localized mesothelioma, fibrous
Ultrasonography.—Nine benign LFTP were mesothelioma, benign fibrous mesothelioma, lo-
imaged with US. Six masses (67%) were hetero- calized fibrous mesothelioma, benign mesothe-
geneous and exhibited hypoechoic and hypere- lioma, benign localized mesothelioma, fibrosing
choic areas but no cysts or calcification. US visu- mesothelioma, and mesothelial fibroma represent
alization of the diaphragm was documented in nomenclature that alludes to early theories sup-
five cases, established the intrathoracic location of porting a mesothelial cell of origin (3,8 –12).
the lesion in all, and showed mass effect without Terms such as subpleural fibroma, submesothe-
invasion in three (Fig 17b). lial fibroma, and submesothelioma suggest a sub-
mesothelial precursor cell (3,10). Pleural fibroma,
Radiologic Reports.—A review of the radiologic benign pleural fibroma, pleural fibromyxoma,
reports generated at the time the lesions were solitary fibrous tumor of the pleura, and localized
originally imaged (n ⫽ 80) showed that the diag- fibrous tumor of the serosal cavities are additional
nosis of LFTP was not mentioned in the preop- terms that have been used to designate these le-
erative differential diagnosis for 43 (54%) cases sions (10,11,13). The currently accepted nomen-
(33 benign, 10 malignant). The specific diagnosis clature is localized fibrous tumor of the pleura,
of LFTP was included in the differential diagno- and a derivation from submesothelial mesenchy-
sis for 21 (26%) lesions (15 benign, six malig- mal cells with a fibroblastic differentiation is gen-
nant), and 16 (20%) lesions (14 benign, two ma- erally acknowledged (11).
lignant) were characterized as pleural in location. Lesions of similar histologic characteristics
have been reported in extrapleural thoracic loca-
Discussion tions, including the mediastinum, lung, pericar-
Primary pleural neoplasms are rare and represent dium, and heart (14 –16). Yousem and Flynn
less than 5% of all pleural neoplasia. The most (17) described three intrapulmonary localized
common primary pleural neoplasm is malignant fibrous tumors and suggested a common origin
mesothelioma, a highly aggressive malignancy for this subset of lesions from tissues in the inter-
that is associated with asbestos exposure. Malig- lobular septa. Localized fibrous tumors have also
nant mesothelioma is characterized by diffuse been reported in the abdomen (liver, peritoneum,
circumferential pleural involvement and a poor and retroperitoneum) and in the meninges, orbit,
prognosis (1). The first report of a primary pleu- thyroid, salivary gland, and the soft tissues includ-
ral neoplasm dates back to 1767 and is attributed ing the breast (15,16,18). It has been suggested
to Lieutaud (2). Wagner published the first mi- that the upper respiratory tract may be a preferred
croscopic description of a primary diffuse pleural extrathoracic location, with reports of localized
neoplasm in 1870 and proposed that the tumor fibrous tumors arising in the nose, paranasal si-
was derived from the endothelium of the pleural nuses, parapharyngeal tissues, nasopharynx and
lymphatics (3,4). In 1931, Klemperer and Rabin epiglottis (15,19).
classified primary pleural neoplasms as localized
and diffuse types and proposed that a subme- Clinical Characteristics
sothelial mesenchymal cell was the cell of origin LFTP affect male and female patients, with a
for the localized type (5). LFTP are rare neo- slight female predominance reported in some
plasms; only over 600 cases have been reported in studies (3). Women represented 55% of patients
the scientific literature (6). In a 1978 review of 60 with benign LFTP and 50% of patients with ma-
cases of LFTP, Okike and colleagues (7) reported lignant LFTP in our series. Affected patients
a prevalence of only 2.8 cases per 100,000 regis- display a wide age range (5– 87 years), with the
trations at their institution, the Mayo Clinic.
774 May-June 2003 RG f Volume 23 ● Number 3
RadioGraphics majority of cases reported in the 6th and 7th de- systemic effects typically occur with large tumors
cades of life and a reported mean age of 50 –57 and generally resolve after excision of the neo-
years (3,20,21). Although 83% of the patients in plasms but may recur with subsequent tumor re-
our series were older than 40 years, only 41% currences (12,30 –32). Non–islet cell tumor hypo-
were in the 6th and 7th decades of life. LFTP are glycemia has been described in association with
not thought to be related to exposure to cigarette both epithelial and mesenchymal neoplasms, in-
smoke, asbestos, or other environmental pollut- cluding LFTP (30,32–34). Many theories have
ants (22,23). However, there are at least two case been proposed to explain this phenomenon, in-
reports of patients with LFTP who were exposed cluding increased glucose consumption by the
to asbestos and one of a patient who developed a tumor; proliferation of insulin receptors; impaired
LFTP following thoracic irradiation for the treat- growth hormone counter-regulatory responses to
ment of a chest wall keloid (24,25). hypoglycemia; and tumor secretion of insulin-like
Up to 50% of patients with LFTP, particular- substances, such as insulin-like growth factor 2
ly those with small neoplasms, may be entirely (IGF-2) and a high molecular weight (“big”)
asymptomatic, and the tumor is discovered inci- IGF-2 (4,30,32,35–38). In some patients, hypo-
dentally because of radiographs obtained for glycemic coma may be the presenting manifesta-
other reasons (3,21,26). When signs and symp- tion of a LFTP (34). Hypertrophic osteoar-
toms (particularly digital clubbing and hypertro- thropathy may relate to hyaluronic acid produc-
phic osteoarthropathy) are present, they are usu- tion and its osteolytic effects and is reported in
ally associated with larger tumors and may occur 17%–35% of cases (4,8,37,39). Although hyper-
more frequently in association with malignant trophic osteoarthropathy has been described in
subtypes of LFTP (7,27). Interestingly, the aver- cases of lung carcinoma, malignant pleural me-
age size of LFTP that affected asymptomatic indi- sothelioma, and even empyema and other tho-
viduals in our series was 10.5 cm, whereas the racic infections, LFTP may produce it more com-
average size of those that occurred in symptom- monly (4). Interestingly, none of our patients pre-
atic patients was 16.6 cm. Studies of patients with sented with hypertrophic osteoarthropathy.
LFTP published from 1942 to 1972 reported that
72% of patients were symptomatic at presenta- Gross Features
tion, whereas studies published between 1973 LFTP are typically solitary lesions with rare oc-
and 1980 reported symptoms in only 54% of pa- currences of conglomerate or multifocal masses.
tients. This decrease in the prevalence of symp- Sixty-six percent to 70% of localized fibrous tu-
toms in patients with LFTP may relate to more mors arise from the visceral pleura, and nearly
widespread imaging of asymptomatic populations half are pedunculated, with the vascular supply to
and the resultant detection of a larger number of the tumor contained within the pedicle (21,40)
incidental tumors (3). (Figs 2, 3). Although a pedicle was present in
Reported symptoms are similar to those exhib- 49% of our cases, 55% of the lesions arose from
ited by our patients and include respiratory and the parietal pleural surface (Fig 1). This discrep-
thoracic complaints such as cough, dyspnea, he- ancy may relate to selection bias, because our
moptysis, chest pain, chest heaviness, and the cases were collected through consultation for sec-
sensation of a mass moving within the chest ond-opinion diagnosis or through case contribu-
(3,21) (Table 1). Abdominal pain has been re- tions by residents attending AFIP courses, and
ported in patients with supradiaphragmatic tu- only cases with imaging studies were included.
mors (28). Systemic complaints may occur and The AFIP series published by England and col-
include chills, sweats, weakness, and weight loss leagues (21), based on a larger number of cases
(3). Paraneoplastic syndromes such as hypoglyce- (n ⫽ 223) submitted to the Institute before 1988,
mia, digital clubbing, and hypertrophic osteoar- documented origin of LFTP from the visceral
thropathy are uncommon, but when they are as- pleura in 66% of cases. Adhesions to the adjacent
sociated with an intrathoracic mass they may sug- pleural surfaces and pericardium are common
gest the diagnosis of LFTP (4,29,30). These (3,21). LFTP are usually well-circumscribed
masses with lobular or smooth external surfaces
and a thin, glistening translucent serosa through
RG f Volume 23 ● Number 3 Rosado-de-Christenson et al 775
RadioGraphics
Figure 21. Benign LFTP. (a, b) High-power (original magnification, ⫻400) (a) and intermediate-power
(original magnification, ⫻200) (b) photomicrographs (hematoxylin-eosin [H-E] stain) demonstrate the bland,
low-grade appearance of spindle-shaped tumor cells (a) that are arranged in a haphazard or so-called pattern-
less pattern (b), the classic appearance of localized fibrous tumors. (c, d) High-power photomicrographs
(original magnification, ⫻400; H-E stain) of the same neoplasm demonstrate the variable cellularity character-
istic of localized fibrous tumors. A hypercellular area (c) is composed of abundant spindle-shaped cells with a
paucity of collagen fibers, whereas a hypocellular area (d) demonstrates abundant collagen (*) between tumor
cells.
which a network of prominent blood vessels may cell of mesenchymal differentiation as the cell of
be seen (Fig 1) (40). The cut surface is firm, origin. At histologic analysis, localized fibrous
pink-white or yellow-tan and often exhibits a tumors appear as low-grade neoplasms of variable
whorled or nodular pattern (Figs 3, 8c). Areas of cellularity. The tumor cells are ovoid to spindle-
necrosis, hemorrhage, or cystic degeneration may shaped with round to oval nuclei, an evenly dis-
be evident, particularly in large or malignant le- tributed fine chromatin, inconspicuous nucleoli,
sions (Figs 5, 12d, 16b, 17c). and bipolar faintly eosinophilic cytoplasm with
indistinct cell borders (Fig 21). Nuclear pleomor-
Histologic Characteristics phism is minimal and mitoses are usually rare or
England and colleagues (21) proposed that LFTP
originated from the submesothelial connective
tissues and suggested a primitive multipotential
776 May-June 2003 RG f Volume 23 ● Number 3
RadioGraphics
Figure 22. Benign LFTP. High-power photomicro- Figure 23. Benign LFTP. Low-power photomicro-
graph (original magnification, ⫻400; H-E stain) of a graph (original magnification, ⫻100; H-E stain) dem-
benign LFTP demonstrates the characteristic ropy col- onstrates the second most commonly encountered his-
lagen that occurs in the hypocellular areas of these tologic pattern (hemangiopericytoma-like) character-
lesions. ized by staghorn-like vessels.
RadioGraphics
Figure 26. Malignant LFTP in an asymptomatic 37-year-old man. (a) PA chest radiograph demonstrates a lobular
mass in the right cardiophrenic angle. Note poor visualization of the superolateral border of the lesion. (b) Unen-
hanced chest CT scan (mediastinal window) demonstrates a spherical heterogeneous right paracardiac mass that pro-
duces mass effect on the heart and perilesional atelectasis. Note the small ipsilateral right pleural effusion.
Figure 27. Benign LFTP in a 27-year-old woman with chest pain. Contrast-enhanced chest CT scans (medi-
astinal window) demonstrate a heterogeneously enhancing mass of the left middle hemithorax that produces
pressure erosion on an adjacent rib. The lesion forms obtuse and acute angles with the adjacent pleura and ex-
hibits a smoothly tapering margin (b). Note enhancing serpiginous linear structures within the lesion (a),
which likely represent vascular structures.
Mendelson and colleagues (46) state that the di- num, and pleural effusion were more common in
agnosis of LFTP should be considered when le- malignant LFTP. Calcification is described in 7%
sions abut the mediastinum or the paraspinal ar- of cases and is usually reported in large lesions in
eas. However, it should be noted that lesions of association with necrosis (11,18). Intralesional
identical histologic characteristics (ie, localized calcification was documented in 26% of our cases
fibrous tumors) may arise in the mediastinum and was characterized as punctate, linear, or
without any relationship to the pleura (14). Mass coarse (Table 2) (Figs 11b, 12b). Local invasion
effect on the adjacent lung and mediastinum is is rarely reported, and lymphadenopathy is not a
described as a typical finding (9,26) (Fig 26). In- feature of LFTP (52). Chest wall involvement
terestingly, although there was little difference in was seen at CT in 8% of our cases, manifesting as
size between the malignant and benign LFTP in sclerosis or pressure erosion on adjacent ribs, a
our series, atelectasis, mass effect on the mediasti- characteristic feature of chest wall and mediasti-
nal neoplasms of neurogenic origin that is rarely
reported in association with LFTP (Fig 27) (45).
RG f Volume 23 ● Number 3 Rosado-de-Christenson et al 779
RadioGraphics
Figure 28. Malignant LFTP in an asymptomatic HIV-positive 28-year-old woman. (a) Con-
trast-enhanced chest CT scan (mediastinal window) demonstrates an irregular soft-tissue mass in
the left middle hemithorax that exhibits heterogeneous enhancement and internal geographic areas
of low attenuation. (b) Photograph of the resected specimen demonstrates a centrally necrotic
mass. Note that complete excision of the LFTP required a pneumonectomy.
LFTP have been reported to exhibit intermedi- monly seen on unenhanced scans (Fig 13). Ho-
ate to high attenuation on unenhanced CT scans. mogeneity of attenuation may indirectly relate to
This appearance has been attributed to the high size (as small lesions may exhibit necrosis less fre-
physical density of collagen and the abundant quently) and was seen in lesions with average
capillary network within these lesions. However, sizes of 8.1 cm on unenhanced CT scans and 4.6
Francis and colleagues (53) studied nine cases of cm on enhanced scans. In addition, hemorrhage,
fibromatosis (a tumor that contains a dense col- necrosis, or cystic change were absent in seven of
lagenized matrix) and found no relationship be- eight lesions (88%) with homogeneous attenua-
tween CT attenuation and histologic content of tion on unenhanced scans and in all eight LFTP
the lesions. Enhancement may correlate with the with homogeneous attenuation on contrast-en-
vascular nature of these lesions and may result in hanced CT scans. Heterogeneous areas of low
higher attenuation than that of other soft tissues attenuation on unenhanced CT scans correlated
in the thorax (46). Lee and colleagues (50) stud- with gross presence of necrosis, hemorrhage, or
ied nine cases of LFTP with CT and demon- cystic change in 86% of benign LFTP.
strated significant enhancement in all. In addi- Heterogeneous attenuation of LFTP after con-
tion, they documented high attenuation (equal to trast material administration was typical (Figs
that of the surrounding muscles) on unenhanced 15–17) and correlated with gross descriptions of
CT scans in three of their cases (50). Enhance- hemorrhage, necrosis, or cystic change in 22 of 40
ment is typically heterogeneous (particularly in (55%) cases of benign LFTP with intrinsic areas
large lesions) with central areas of low attenua- of low attenuation (Figs 16, 17). Little has been
tion, which have been shown to correlate with written about the CT appearance of malignant
myxoid change, hemorrhage, necrosis, or cystic LFTP. They are described as indistinguishable
degeneration (11,37,46,50). Heterogeneity may from benign lesions: large masses of heteroge-
become more conspicuous after intravenous ad- neous attenuation and patchy enhancement (52).
ministration of contrast material because areas of Although there was no difference in lesion size
viable tumor exhibit contrast material enhance- when the benign and malignant LFTP in our se-
ment and those with necrosis do not. Although ries were compared, low-attenuation areas were
contrast material enhancement was common in seen in all malignant LFTP whether the scans
our series (62%) (Figs 10, 14), it was not demon- were obtained before or after the administration
strated in all cases. of intravenous contrast material and correlated
The majority of the lesions in our series exhib- with gross descriptions of hemorrhage, necrosis,
ited heterogeneous attenuation on CT scans be- or cystic change in 60% of unenhanced (Figs 12,
fore and after the administration of intravenous 26) and 54% of enhanced CT studies (Fig 28).
contrast material. Only benign LFTP exhibited
homogeneous attenuation, which was more com-
780 May-June 2003 RG f Volume 23 ● Number 3
RadioGraphics
Figure 29. Benign LFTP in a 64-year-old woman who presented with confusion and hypoglyce-
mia. (a) Sagittal T1-weighted MR image demonstrates a large soft-tissue mass of heterogeneous
intermediate signal intensity in the right hemithorax that produces mass effect on the hemidia-
phragm. Note flow void foci consistent with tumor vessels. (b) Sagittal T2-weighted MR image at
the same level demonstrates an overall heterogeneous increase in signal intensity in the lesion with
multiple hyperintense areas.
MR Imaging.—There are few reports of the MR diaphragm and chest wall (18). Sagittal and coro-
imaging features of LFTP (20,28,54,55). The nal images permit tumor localization within the
existing reports describe masses of predominant thorax and allow diaphragmatic evaluation. Flow
low or intermediate signal intensity on both T1- voids reflecting the vascular nature of some LFTP
and T2-weighted images and on proton density– have previously been reported (55) and were seen
weighted images, which is thought to relate to in 11% of our cases (Figs 19, 29). Intense hetero-
high content of fibrous collagenous tissue, hypo- geneous enhancement after administration of in-
cellularity, and relatively small numbers of mobile travenous gadolinium is typical (Figs 18, 20).
protons (8,11,54,55). However, high signal inten-
sity on T2-weighted images has also been re- Angiography.—Angiography is useful in deter-
ported and may relate to necrosis, cystic or myx- mining the vascular supply to the lesion, which
oid degeneration, prominent vascular structures, typically enters through the pedicle (26). Demon-
and hypercellular areas (20,55). In our study, stration of blood supply from the inferior phrenic,
predominantly heterogeneous lesions were seen intercostal, or internal mammary arteries may be
on both T1- and T2-weighted images (Fig 29). a helpful clue to the extrapulmonary origin of
Areas of low to intermediate signal intensity were large LFTP (8).
more commonly seen on T1-weighted images,
and areas of high signal intensity were demon- Ultrasonography.—US is helpful in the evalua-
strated more frequently on T2-weighted images tion of large masses in the inferior hemithorax
(Figs 18, 29). Low-signal-intensity septa were a through visualization of the diaphragm and estab-
common finding on T2-weighted MR images lishment of their intrathoracic location, particu-
(Fig 19). Most lesions increased in signal inten- larly those lesions located in the inferior hemitho-
sity on T2-weighted images, compared with T1- rax (4,26) (Fig 17b).
weighted images (Fig 29). MR imaging is supe-
rior to CT in assessment of tumor extent and is Patient Management
more sensitive in excluding local invasion of the
Diagnosis.—The diagnosis of LFTP may not be
suggested prospectively after imaging affected
patients. This is not surprising given the fact that
RG f Volume 23 ● Number 3 Rosado-de-Christenson et al 781
RadioGraphics classically described radiographic features (in- lesions can be successfully resected particularly if
complete visualization of the lesion borders, fis- they are pedunculated (49,60). Recurrent tumors
sural location) and cross-sectional imaging fea- may exhibit malignant histologic features and a
tures (obtuse angles, mobility, pedicle visualiza- more aggressive biologic and clinical behavior
tion) of focal pleural masses are usually not despite prior resection of histologically benign
present in these cases. The diagnosis of LFTP is lesions (14). Malignant tumors may metastasize,
typically made after surgical excision of the mass. and local recurrences are more common in cases
Confident preoperative diagnosis can be made of malignant lesions than in benign lesions (22).
with large-bore cutting needle biopsies, and be-
cause the needle can usually be introduced into Conclusions
the mass while avoiding aerated lung, the risk of Localized fibrous tumors are rare primary pleural
pneumothorax is minimal (56). Although fine- neoplasms that may grow to large sizes and typi-
needle aspiration may yield characteristic and di- cally affect symptomatic men and women over
agnostic morphologic features, cutting needle the age of 40 years. Small LFTP may be discov-
biopsy is probably preferable because of wider ered incidentally on chest radiographs of asymp-
tissue sampling (10,56). tomatic individuals. Although small lesions may
exhibit the characteristic imaging features of pleu-
Treatment.—The therapy for patients with ral masses, classic findings of focal pleural dis-
LFTP is complete surgical excision. Aggressive ease, such as the “incomplete border” sign, ob-
surgery is recommended given the possibility of tuse or right angles against the adjacent pleura,
recurrent disease (57). Vascular adhesions to ad- fissural location, or mobility within the pleural
jacent tissues may result in massive intraoperative space, are usually absent. The majority of LFTP
hemorrhage, which is adequately handled with occupy the inferior hemithorax, and those that
good exposure, prompt removal of the mass, and abut the ipsilateral hemidiaphragm may mimic
meticulous hemostasis (58). Cardillo and col- diaphragmatic elevation or eventration. The diag-
leagues (6) described 55 patients with LFTP who nosis should be considered in symptomatic adults
underwent resection via video-assisted thoraco- who present with solitary, large, lobular, hetero-
scopic surgery (VATS) and thoracotomy with no geneous intrathoracic masses in the absence of
operative deaths. Although VATS may be ad- local invasion, lymphadenopathy, or metastatic
equate for the resection of small tumors, large disease. Small LFTP without gross necrosis,
tumors usually require a thoracotomy (39). In hemorrhage, or cystic change may exhibit homo-
many cases, resection of the lesion with its pedicle geneous attenuation on unenhanced and less fre-
and a small patch of adjacent lung may be suffi- quently on contrast-enhanced chest CT scans.
cient (4). In some cases, complete excision may However, the majority of LFTP exhibit heteroge-
require lung resection (segmentectomy, lobec- neous attenuation on CT scans, characterized as
tomy, bilobectomy, pneumonectomy), partial intralesional geographic, focal or linear areas of
pleurectomy, or en bloc chest wall resection (Fig low attenuation that often correlate with hemor-
28b) (4,6,7,12). Long-term imaging follow-up is rhage, necrosis, or cystic changes. Calcification
recommended in all cases to exclude tumor recur- may occur in one-fourth of cases. Atelectasis of
rence or metastatic disease. Recurrent disease the adjacent lung and mass effect on the mediasti-
typically affects the ipsilateral pleura and may also num are common associated findings. MR imag-
affect the lung. Repeat resection of recurrent le- ing typically demonstrates intrathoracic lobular
sions is recommended (4,49). Adjuvant therapy is masses of heterogeneous signal intensity with
probably not helpful because of the low cellular both T1- and T2-weighted sequences. Internal
content and low mitotic rates characteristic of low-signal-intensity septa on T2-weighted images
LFTP (27,59). are common. Studies that allow direct visualiza-
tion of the diaphragm such as MR imaging and
Prognosis.—The prognosis for patients with US are useful in establishing the intrathoracic
LFTP is generally favorable. The majority of le- location of large lesions occupying the inferior
sions behave in a benign fashion (88%), but ap- hemithorax. Although there are no imaging fea-
proximately 12% of patients die of extensive in- tures that definitively distinguish benign from
trathoracic tumor growth or an unresectable re- malignant subtypes of LFTP, heterogeneity on
currence (3). Although tumor size does not cross-sectional images, mass effect, and pleural
directly correlate with prognosis, pedunculated effusion may be slightly more common in malig-
noninvasive tumors are less likely to recur when nant lesions.
complete excision of all neoplastic tissue is ac-
complished. In fact, complete excision is the best
prognostic indicator (3,7,8,21,25,60). Malignant
782 May-June 2003 RG f Volume 23 ● Number 3
RadioGraphics Acknowledgments: The authors wish to acknowl- 13. el-Naggar AK, Ro JY, Ayala AG, Ward R, Or-
edge the immense assistance of Linda C. Wilkins, BS, donez NG. Localized fibrous tumor of the serosal
who constructed the database used to collect and ana- cavities: immunohistochemical, electron-micro-
lyze the data for the manuscript. We also acknowledge scopic, and flow-cytometric DNA study. Am J
the input and contribution of Philip A. Templeton, Clin Pathol 1989; 92:561–565.
MD, during the early planning stages of the project. 14. Balassiano M, Reichert N, Rosenman Y, Hertcheg
Finally, we thank the countless residents who through- E, Lieberman Y, Yellin A. Localized fibrous me-
out the years have participated in the Radiologic Pa- sothelioma of the mediastinum devoid of pleural
thology Courses of the Department of Radiologic Pa- connections. Postgrad Med J 1989; 65:788 –790.
thology at the Armed Forces Institute of Pathology in 15. Ibrahim NB, Briggs JC, Corrin B. Double primary
Washington, DC. Their case contributions enriched localized fibrous tumours of the pleura and retro-
the content of the Institute’s archives and enhanced our peritoneum. Histopathology 1993; 22:282–284.
understanding of radiologic-pathologic correlation of 16. Vaswani K, Guttikonda S, Vitellas KM. Case 1.
thoracic diseases. Localized fibrous tumor of the liver. AJR Am J
Roentgenol 2000; 175:872,875– 876.
17. Yousem SA, Flynn SD. Intrapulmonary localized
References fibrous tumor: intraparenchymal so-called local-
1. Miller BH, Rosado-de-Christenson ML, Mason ized fibrous mesothelioma. Am J Clin Pathol 1988;
AC, Fleming MV, White CC, Krasna MJ. From 89:365–369.
the archives of the AFIP—Malignant pleural me- 18. Sandvliet RH, Heysteeg M, Paul MA. A large tho-
sothelioma: radiologic-pathologic correlation. Ra- racic mass in a 57-year-old patient: solitary fibrous
dioGraphics 1996; 16:613– 644. tumor of the pleura. Chest 2000; 117:897–900.
2. Kucuksu N, Thomas W, Ezdinli EZ. Chemo- 19. Safneck JR, Alguacil-Garcia A, Dort JC, Phillips
therapy of malignant diffuse mesothelioma. Can- SM. Solitary fibrous tumour: report of two new
cer 1976; 37:1265–1274. locations in the upper respiratory tract. J Laryngol
3. Briselli M, Mark EJ, Dickersin GR. Solitary fi- Otol 1993; 107:252–256.
brous tumors of the pleura: eight new cases and 20. Dynes MC, White EM, Fry WA, Ghahremani
review of 360 cases in the literature. Cancer 1981; GG. Imaging manifestations of pleural tumors.
47:2678 –2689. RadioGraphics 1992; 12:1191–1201.
4. Suter M, Gebhard S, Boumghar M, Peloponisios 21. England DM, Hochholzer L, McCarthy MJ. Lo-
N, Genton CY. Localized fibrous tumours of the calized benign and malignant fibrous tumors of
pleura: 15 new cases and review of the literature. the pleura: a clinicopathologic review of 223 cases.
Eur J Cardiothorac Surg 1998; 14:453– 459. Am J Surg Pathol 1989; 13:640 – 658.
5. Klemperer P, Rabin CB. Primary neoplasms of the 22. Harrison RI, McCaughan BC. Malignancy in a
pleura: a report of five cases. Arch Pathol 1931; massive localized fibrous tumour of pleura. Aust
11:385– 412. N Z J Surg 1992; 62:311–313.
6. Cardillo G, Facciolo F, Cavazzana AO, Capece G, 23. Scharifker D, Kaneko M. Localized fibrous “me-
Gasparri R, Martelli M. Localized (solitary) fi- sothelioma” of pleura (submesothelial fibroma): a
brous tumors of the pleura: an analysis of 55 pa- clinicopathologic study of 18 cases. Cancer 1979;
tients. Ann Thorac Surg 2000; 70:1808 –1812. 43:627– 635.
7. Okike N, Bernatz PE, Woolner LB. Localized me- 24. Metintas M, Gibbs AR, Harmanci E, et al. Malig-
sothelioma of the pleura: benign and malignant nant localized fibrous tumor of the pleura occur-
variants. J Thorac Cardiovasc Surg 1978; 75:363– ring in a person environmentally exposed to
372. tremolite asbestos. Respiration 1997; 64:236 –239.
8. Desser TS, Stark P. Pictorial essay: solitary fibrous 25. Bilbey JH, Muller NL, Miller RR, Nelems B. Lo-
tumor of the pleura. J Thorac Imaging 1998; 13: calized fibrous mesothelioma of pleura following
27–35. external ionizing radiation therapy. Chest 1988;
9. Desrumaux I, Baekelandt M, Verbeke W, 94:1291–1292.
Gryspeerdt S, Van Holsbeeck B, Lefere P. Local- 26. Dedrick CG, McLoud TC, Shepard JA, Shipley
ized benign fibrous mesothelioma mimicking a RT. Computed tomography of localized pleural
malignant tumor of the pleura. JBR-BTR 1998; mesothelioma. AJR Am J Roentgenol 1985; 144:
81:131–133. 275–280.
10. Dusenbery D, Grimes MM, Frable WJ. Fine- 27. Khan JH, Rahman SB, Clary-Macy C, et al. Giant
needle aspiration cytology of localized fibrous tu- solitary fibrous tumor of the pleura. Ann Thorac
mor of pleura. Diagn Cytopathol 1992; 8:444 – Surg 1998; 65:1461–1464.
450. 28. Tublin ME, Tessler FN, Rifkin MD. US case of
11. Ferretti GR, Chiles C, Choplin RH, Coulomb M. the day. Solitary fibrous tumor of the pleura
Localized benign fibrous tumors of the pleura. (SFTP). RadioGraphics 1998; 18:523–525.
AJR Am J Roentgenol 1997; 169:683– 686. 29. Chaugle H, Parchment C, Grotte GJ, Keenan
12. Pond F, Wilson A, McKelvie P. Localized fibrous DJ. Hypoglycaemia associated with a solitary fi-
tumour of the pleura: two case reviews. Aust N Z J brous tumour of the pleura. Eur J Cardiothorac
Surg 1997; 67:821– 824. Surg 1999; 15:84 – 86.
30. Cole FH Jr, Ellis RA, Goodman RC, Weber BC,
Courington DP. Benign fibrous pleural tumor
RG f Volume 23 ● Number 3 Rosado-de-Christenson et al 783
RadioGraphics with elevation of insulin-like growth factor and 46. Mendelson DS, Meary E, Buy JN, Pigeau I,
hypoglycemia. South Med J 1990; 83:690 – 694. Kirschner PA. Localized fibrous pleural mesothe-
31. Aufiero TX, McGary SA, Campbell DB, Phillips lioma: CT findings. Clin Imaging 1991; 15:105–
PP. Intrapulmonary benign fibrous tumor of the 108.
pleura. J Thorac Cardiovasc Surg 1995; 110:549 – 47. Bicer M, Yaldiz S, Gursoy S, Ulgan M. A case of
551. giant benign localized fibrous tumor of the pleura.
32. Mandal AK, Rozer MA, Salem FA, Oparah SS. Eur J Cardiothorac Surg 1998; 14:211–213.
Localized benign mesothelioma of the pleura asso- 48. Ulrik CS, Viskum K. Fibrous pleural tumour pro-
ciated with a hypoglycemic episode. Arch Intern ducing 171 litres of transudate. Eur Respir J 1998;
Med 1983; 143:1608 –1610. 12:1230 –1232.
33. Nelson R, Burman SO, Kiani R, Chertow BS, 49. de Perrot M, Kurt AM, Robert JH, Borisch B,
Shah J, Cantave I. Hypoglycemic coma associated Spiliopoulos A. Clinical behavior of solitary fi-
with benign pleural mesothelioma. J Thorac Car- brous tumors of the pleura. Ann Thorac Surg
diovasc Surg 1975; 69:306 –314. 1999; 67:1456 –1459.
34. Tirilomis T, Busch T, Sirbu H, Dalichau H. Giant 50. Lee KS, Im JG, Choe KO, Kim CJ, Lee BH. CT
localized fibrous mesothelioma: an unusual large findings in benign fibrous mesothelioma of the
intrathoracic tumor. Langenbecks Arch Surg pleura: pathologic correlation in nine patients.
2000; 385:482– 484. AJR Am J Roentgenol 1992; 158:983–986.
35. Kishi K, Homma S, Tanimura S, Matsushita H, 51. Spizarny DL, Gross BH, Shepard JA. CT findings
Nakata K. Hypoglycemia induced by secretion of in localized fibrous mesothelioma of the pleural
high molecular weight insulin-like growth factor-II fissure. J Comput Assist Tomogr 1986; 10:942–
from a malignant solitary fibrous tumor of the 944.
pleura. Intern Med 2001; 40:341–344. 52. Saifuddin A, Da Costa P, Chalmers AG, Carey
36. Shapiro ET, Bell GI, Polonsky KS, Rubenstein BM, Robertson RJ. Primary malignant localized
AH, Kew MC, Tager HS. Tumor hypoglycemia: fibrous tumours of the pleura: clinical, radiological
relationship to high molecular weight insulin-like and pathological features. Clin Radiol 1992; 45:
growth factor-II. J Clin Invest 1990; 85:1672– 13–17.
1679. 53. Francis IR, Dorovini-Zis K, Glazer GM, Lloyd
37. Rena O, Filosso PL, Papalia E, et al. Solitary fi- RV, Amendola MA, Martel W. The fibromatoses:
brous tumour of the pleura: surgical treatment. CT-pathologic correlation. AJR Am J Roentgenol
Eur J Cardiothorac Surg 2001; 19:185–189. 1986; 147:1063–1066.
38. Yokosaki Y, Kido M, Nagata N, et al. Hypoglyce- 54. Harris GN, Rozenshtein A, Schiff MJ. Benign fi-
mia associated with localized fibrous mesothe- brous mesothelioma of the pleura: MR imaging
lioma of the pleura. J UOEH 1995; 17:191–197. findings. AJR Am J Roentgenol 1995; 165:1143–
39. Robinson LA, Reilly RB. Localized pleural me- 1144.
sothelioma: the clinical spectrum. Chest 1994; 55. George JC. Benign fibrous mesothelioma of the
106:1611–1615. pleura: MR findings. AJR Am J Roentgenol 1993;
40. Dalton WT, Zolliker AS, McCaughey WT, 160:204 –205.
Jacques J, Kannerstein M. Localized primary tu- 56. Weynand B, Noel H, Goncette L, Noirhomme P,
mors of the pleura: an analysis of 40 cases. Cancer Collard P. Solitary fibrous tumor of the pleura: a
1979; 44:1465–1475. report of five cases diagnosed by transthoracic cut-
41. Moran CA, Suster S, Koss MN. The spectrum of ting needle biopsy. Chest 1997; 112:1424 –1428.
histologic growth patterns in benign and malig- 57. Utley JR, Parker JC Jr, Hahn RS, Bryant LR, Mo-
nant fibrous tumors of the pleura. Semin Diagn bin-Uddin K. Recurrent benign fibrous mesothe-
Pathol 1992; 9:169 –180. lioma of the pleura. J Thorac Cardiovasc Surg
42. Ellis R. Incomplete border sign of extrapleural 1973; 65:830 – 834.
masses. JAMA 1977; 237:2748. 58. McNicholas KW, Rose EA, Edie RN, Jaretzki A
43. Desai SR, Wilson AG. Pleura and pleural disor- 3rd. Resection of giant benign fibrous mesothe-
ders. In: Armstrong P, Wilson AG, Dee P, Hansell lioma of pleura. N Y State J Med 1980; 80:626 –
DM, eds. Imaging of diseases of the chest. 3rd ed. 629.
London, England: Mosby, 2000; 727–787. 59. McCormack PM, Nagasaki F, Hilaris BS, Martini
44. Berne AS, Heitzman ER. The roentgenologic N. Surgical treatment of pleural mesothelioma.
signs of pedunculated pleural tumors. Am J J Thorac Cardiovasc Surg 1982; 84:834 – 842.
Roentgenol Radium Ther Nucl Med 1962; 87: 60. Carter D, Otis CN. Three types of spindle cell
892– 895. tumors of the pleura: fibroma, sarcoma, and sarco-
45. Truong M, Munden RF, Kemp BL. Localized matoid mesothelioma. Am J Surg Pathol 1988;
fibrous tumor of the pleura. AJR Am J Roentgenol 12:747–753.
2000; 174:42.
This article meets the criteria for 1.0 credit hour in category 1 of the AMA Physician’s Recognition Award. To obtain
credit, see accompanying test at http://www.rsna.org/education/rg_cme.html.