Integrative Clinicoradiologic, Cytologic, and Histologic Diagnosis of Soft Tissue Tumors
Scott Ethan Kilpatrick
Click on any thumbnail images for an enlarged view of that slide
Patient is a 38-year-old man with a well-circumscribed mass involving the musculature of the right
anterior shoulder. A fine needle aspiration biopsy (FNAB) was performed followed by surgical
Patient is a 50-year-old woman with a posterior right thigh, soft tissue mass, localized between the adductor magnus and semimembranous muscles. Signal characteristics by MRI suggest a fatty mass. A simple excision was initially performed followed by wide local excision and radiation therapy.
Patient is a 36-year-old woman with a right medial distal thigh soft tissue mass. When the lesion is "bumped", she reports painful radiation to the groin, proximally, and the medial calf, distally. A FNAB was performed followed by simple excision.
Patient is a 13-year-old African-American boy with a 5 x 6 cm soft tissue mass in the right neck, first noticed 3 months ago. Physical examination showed no abnormalities in the oral cavity or oropharynx. A FNAB was performed. Following immediate interpretation, a MRI of the head and neck showed no masses in the brain, paranasal sinuses, or orbits. Surgical resection of the neck mass was performed, and the patient was enrolled on a pediatric therapeutic protocol.
Patient is a 58-year-old woman with a large, heterogeneous soft tissue mass within the left posterior-medial calf. A FNAB was performed followed by pre-operative chemotherapy, radiation therapy and radical excision.
Patient is a 65-year-old woman with a painful soft tissue mass immediately adjacent to the distal, medial left tibia. She had a history of "colon cancer" several years ago. A FNAB was performed followed by limb salvage resection of the distal tibia.
Patient is a 28-year-old man with a 10 cm, heterogeneous, soft tissue mass, replacing the anterior and lateral portions of the left deltoid muscle. A FNAB was performed. Subsequently, he received pre-operative chemotherapy and radiation therapy followed by radical resection of the anterior 2/3 of the deltoid muscle.
Patient is a 47-year-old man with a lobulated, anterior distal thigh mass, that reportedly has been present for approximately 3 years. A FNAB was performed, followed by radical resection.
The diagnosis of soft tissue tumors has always been difficult, in part, due to their absolute rarity
(< 1% of all malignant neoplasms), but also compounded by the realization that a malignant diagnosis
may lead to adverse consequences (i.e. amputation). Further apprehension may be related to the fact that
many sarcomas (and pseudosarcomas) affect children. According to Enzinger and Weiss, approximately 15%
of soft tissue sarcomas arise in patients < 15 years of age, while approximately 40% occur in older
adult patients (> 55 years). 
I. Biopsy Techniques
Traditionally, at most institutions, diagnostic tissue from soft tissue tumors has been (and for the
most part still is) obtained from open, incisional biopsies. Such samples usually provide the surgical
pathologist with more than enough tissue to analyze morphologic architecture and perform ancillary
studies, if deemed necessary. The disadvantages include a higher rate of intra- and post-operative
complications, the development of a scar, and the potential, especially if not carefully planned by an
experienced orthopedic oncologic surgeon, to affect subsequent surgical management. In the last decade,
greater emphasis has been placed on techniques that sample relatively smaller amounts of tissue (without
compromising quality), are more readily accepted by the patient, may be easily performed in outpatient
clinic settings, and are less likely to be associated with significant complications. Both core needle
biopsy (CNB) and fine needle aspiration biopsy (FNAB) meet the aforementioned demands and, in addition,
are relatively cost-effective. However, I believe, in many circumstances, FNAB is the superior
technique. The argument in favor of CNB over FNAB is generally related to the fact that CNB yields
actual "tissue" and subsequently allows assessment of architecture as well as providing a source for
ancillary studies (e.g. immunohistochemistry). Nevertheless, practitioners of FNAB readily acknowledge
that tissue samples are easily obtained by FNAB in the form of cell blocks. Additional FNAB passes may
be performed to obtain material for cytogenetic analysis, flow cytometry, and image analysis. Bennert
and Abdul-Karim analyzed 37 sarcomas by FNAB and reported successful subtyping in 30 (81%)
tumors. Concomitant CNBs were performed in all of these cases. In their series, the main advantage of
CNB over FNAB was that the former helped establish "the specific subtype of sarcoma when the fine needle
aspirate was reported as sarcoma, not otherwise specified."  However, among diagnostic FNABs,
CNB did not significantly contribute to overall patient management. I would agree with the authors that
an unsatisfactory FNAB, especially if clinically suspicious for sarcoma, should be further investigated
by repeat FNAB, CNB, or open biopsy. The obvious advantages of FNAB over CNB and open biopsy include
virtually no risk of significant complications and the technical ease of the procedure.
Additionally, the ability to render an immediate interpretation allows for planning of surgical
intervention and/or neoadjuvant therapy at the initial presenting clinic visit, avoiding a potentially
inconvenient second clinic visit.
Most cytopathologists define FNAB as the use of a 22-gauge or smaller needle; the risk of needle tract
seeding is exceedingly low and appears to be between 0.003 (3/100,000) and 0.009%(9/100,000). 
Major complications including death usually involve intra-abdominal and intrathoracic FNABs associated
with larger bore needles (>18-gauge).  A comparison of all of these biopsy
techniques is summarized in Table 1.
Table 1: Comparison of Fine Needle Aspiration Biopsy (FNAB), Core Needle Biopsy (CNB) and Open Biopsy
| Characteristics || FNAB ||CNB ||Open Biopsy|
|Technical Ease of the procedure ||Easily learned and applied ||Easily learned and applied ||Difficult; requires specialty training|
|Cytologic Smears ||Yes - recommend Diff-Quik for matrix material; Pap for nuclear detail ||No ||No|
|Histologic Tissue ||Yes (cell block) - usually most helpful for ancillary studies, not morphology ||Yes - helpful for both morphology and ancillary studies ||Yes - helpful for both morphology and ancillary studies|
|Availability of Ancillary Studies (e.g. immuno- histochemistry, cytogenetics, flow cytometry) ||Yes - requires additional FNAB passes ||Yes - may require additional CNB passes ||Yes - may require additional tissue|
|Speed of Interpretation ||Fast; often within minutes to hours ||Tissue requires fixation and processing; 1-2 days ||Tissue requires fixation and processing; 1-2 days|
|Risk of Complications ||Exceedingly low ||Low, but higher than FNAB ||Low, but significantly higher than either FNAB or CNB|
|Risk of Tumor Contamination of Biopsy Site ||Exceedingly low ||Low, but higher than FNAB ||Low, but significantly higher than either FNAB or CNB|
|Accuracy Rate for Histologic Subtyping ||30-75%, histologic classification unreliable in fatty tumors and adult pleomorphic sarcomas ||75-90% ||>95%|
Although the value of FNAB in distinguishing malignancy from benignancy has been established in
several studies and often approaches and exceeds 90%, its accuracy for establishing a specific histologic
subtype has been far less tested.  Layfield et al.  evaluated a total of 51
histologically-confirmed soft tissue sarcomas; an accurate histologic subtype was established by FNAB in
38 (74%) cases. FNAB from one example of well-differentiated liposarcoma was insufficient for diagnosis
and one case of rhabdomyosarcoma consisted of "fibrous tissue only."  Costa et
al.  retrospectively reviewed 45 sarcomas of which only approximately 21% were correctly
subtyped. The authors' conclusion was "FNA usually does not provide a specific diagnosis…" 
Among 110 soft tissue lesions (of which 23 were sarcomas), Hook et al.  documented specificity
and positive predictive values for the nonspecific diagnosis of sarcoma as 98% and 91%, respectively.
However, subclassification of sarcomas "was possible in only 30% of cases."  In our series,
61 (86%) of a total of 71 adequate samples were correctly recognized as sarcoma.  However,
an accurate histologic subtype was rendered in only 34 (48%) cases. However, as we will see momentarily,
failure to provide a specific histologic subtype does not always preclude initiation of appropriate
therapy. Excluding inadequate samples (three cases), we have observed only 1 false negative FNAB
specimen in our series. One case of hemangiopericytoma was initially misinterpreted as a giant cell
tumor of tendon sheath due to the presence of numerous multinucleated giant cells and its clinical
presentation in the popliteal fossa, adjacent to synovium. The lesion was excised and the patient has
suffered no adverse outcome as a consequence of the misinterpretation. In our experience with over 1200
bone and soft tissue aspirates, we have encountered no false positive cases (histologically-proven benign
tumor misinterpreted as malignant by FNAB). [9 ]Furthermore, we have documented no complications
or evidence of needle tract tumor seeding.
II. Understanding Sarcoma Therapy and its Relationship to Diagnosis
With the exception of pediatric small round cell tumors and osteosarcoma, therapy for most soft tissue
sarcomas, especially adult forms, is based predominantly on anatomic site and stage, the latter of which
incorporates histologic grade.  Although histologic subtyping may help determine the
histologic grade (e.g. Ewing sarcoma is definitionally high grade), it is not always necessary. For
example, at most institutions, the therapy for adult pleomorphic malignant fibrous histiocytoma (MFH) is
similar to that used for pleomorphic liposarcoma or pleomorphic leiomyosarcoma, as all represent high
grade sarcomas. Consequently, for the evaluation of FNAB in the therapeutic management of soft tissue
sarcomas, the percentage of cases in which FNAB is sufficient for the initiation of definitive therapy is
more important than its accuracy rate for histologic subtyping. In our recent series, among soft tissue
sarcomas, the diagnosis rendered by FNAB was sufficient for definitive therapy (not requiring additional
open incisional biopsy) in 73 patients (of 88 patients with adequate specimens, 83%). 
Histologic subtyping of Ewing sarcoma, skeletal osteosarcoma, and rhabdomyosarcoma is absolutely
essential as many patients are clinically eligible for histogenetic-specific protocols. Again, in our
series, among 18 patients clinically eligible for these protocols, an accurate diagnosis was rendered in
17 cases (94%).  Overall 12 pediatric patients were enrolled on a variety of
histogenetic-specific, pediatric protocols for neuroblastoma (POG), rhabdomyosarcoma (IRSG), osteosarcoma
(CCG/POG), and Ewing's sarcoma (POG).
III. Clinical and Radiologic Correlation
Correlation with clinical and, when available, radiologic features cannot be over-emphasized. Certain
tumors are virtually non-existent in children (e.g. liposarcoma) and others, for reasons not understood,
occur only rarely in certain ethnic groups (e.g. Ewing sarcoma is rare in African-Americans).
Radiographs, including computed tomography (CT) scans and magnetic resonance imaging (MRI), provide
information regarding location (e.g. visceral involvement, intramuscular, etc.), size, presence (or
absence) of bone involvement (e.g. primary bone tumor with soft tissue extension vs. primary soft tissue
tumor), homogeneity vs. heterogeneity, content (e.g. cystic vs. fatty), and, finally, relationships to
neurovascular structures (e.g. nerve sheath tumor). Obviously, combining clinical and radiographic
information helps the pathologist provide a more accurate and definitive diagnosis.
IV. Practical Approach for Diagnosis
Because of the enormous number of soft tissue tumors and tumor-like lesions, as well as, the often
subtle pathologic differences between prognostically different but related tumors, establishing a
practical approach to the diagnosis of soft tissue tumors is considerably more difficult than that of
bone tumors. Nevertheless, by combining clinicoradiologic data with pathology, one can make certain
generalizations regarding benignancy vs. malignancy. These features are summarized in Table 2.
Table 2: Clinicoradiologic, Histologic, and Cytologic Features of Benign vs. Malignant Soft Tissue Lesions
| Features ||Benign Characteristics ||Malignant Characteristics|
|Size and properties ||Small, often well-circumscribed ||Large, more often infiltrative|
|Anatomic location ||Superficial and/or cutaneous ||Deep and/or intramuscular|
|Tumor cellularity ||Slight to moderately cellular ||Moderately to markedly cellular|
|Cellular cohesion ||Tendency to form cohesive tissue ||Tendency toward discohesiveness|
|Nuclear features ||Uniform, "open" or vesicular chromatin ||More commonly pleomorphic and variable Coarse chromatin|
|Mitotic activity ||Less mitotically active (often < 1 mitoses/10 HPF) ||Often more mitotically active (often > 10 mitoses/10 HPF)|
|Necrosis ||Usually absent ||Often present|
Once one has decided that a given lesion is malignant, histologic subtyping should be attempted, as
certain sarcomas, such as rhabdomyosarcoma, osteosarcoma, and Ewing sarcoma, are treated with
histogenetic-specific protocols. Traditionally, the most common approach for histologic subtyping has
been based on histogenesis (e.g. fibrous/fibrohistiocytic, lipogenic, etc.). However, the use of such a
classification system depends upon the recognition of the histogenetic subtype and/or differential
diagnosis. To render an accurate diagnosis, one must at least consider the correct diagnosis. Indeed,
ancillary tests are only helpful if the appropriate studies are obtained. I think a better approach is
one that emphasizes the predominant cytomorphologic features of the sarcoma in question, as illustrated
in Table 3.
Table 3: Classification of Soft Tissue Sarcomas based on Predominant Cytomorphologic Features (based on the results of Kilpatrick and Geisinger )
| Small Round Cell Sarcomas ||Spindle Cell Sarcoma ||Myxoid Sarcomas|
|Ewing sarcoma ||Fibrosarcoma ||Myxofibrosarcoma|
|Rhabdomyosarcoma, childhood subtypes|| Leiomyosarcoma ||Myxoid liposarcoma|
|Neuroblastoma ||Synovial sarcoma* ||Myxoid chondrosarcoma|
|Mesenchymal chondrosarcoma ||Malignant peripheral nerve sheath tumor|| |
|Desmoplastic small round cell tumor ||Hemangiopericytoma|| |
| Epithelioid/Polygonal Cell Sarcomas ||Pleomorphic Sarcomas|
|Epithelioid sarcoma || Malignant Fibrous Histiocytoma|
|Epithelioid hemangio- endothelioma/angiosarcoma ||Pleomorphic liposarcoma|
|Epithelioid malignant schwannoma ||Pleomorphic leiomyosarcoma|
|Alveolar soft part sarcoma ||Pleomorphic rhabdomyosarcoma, adult type|
|Clear cell sarcoma (melanoma|
of soft parts)
|Synovial sarcoma * ||Angiosarcoma|
* Synovial sarcoma may be considered in either the spindle cell or epithelioid groups
This classification is useful both clinically and pathologically. For example, small round cell
sarcomas are far more commonly observed in children while pleomorphic and myxoid sarcomas are generally
observed in older adults (> 50 years of age). Spindle cell and epithelioid/polygonal cell sarcomas
are most commonly seen in young to middle-aged adults. As is expected, limitations exist regarding this
classification scheme. Certain sarcomas, such as embryonal rhabdomyosarcoma, may display a range of
cytomorphologic features, from typical small round cell forms to predominant spindle cell morphology and
even a myxoid stroma. In some examples, synovial sarcoma is best considered an epithelioid/polygonal
cell sarcoma. For these reasons, I do not favor abandoning the traditional
classification system, but rather I would propose using a combination of both the traditional and
cytomorphologic schemes, ensuring that all pertinent diagnoses are at least entertained. The
cytomorphologic system will also prove most advantageous when attempting the diagnosis of sarcomas by
fine needle aspiration biopsy (FNAB). In my experience, cell blocks, although often containing perfectly
adequate material for immunocytochemistry, generally lack sufficient tissue to confidently pursue a
histogenesis based purely on tissue architecture alone. Finally, one should always bear in mind that
non-mesenchymal tumors can involve soft tissue, presenting as a primary soft tissue mass (e.g. metastatic
carcinoma, malignant lymphoma, and malignant melanoma).
V. Histologic Grading of Sarcomas
Regardless of tumor type, histologic grading of malignant tumors is based predominantly on the
resemblance of the tumor cells to their "histologic counterparts".  Thus, histologic grade
essentially corresponds to the differentiation within an individual histologic subtype. Malignant tumors
composed of neoplastic cells closely resembling normal cells are considered well differentiated (low
grade); conversely, those tumors containing malignant cells least resembling normal cells are considered
poorly differentiated (high grade). From a practical point of view, histologic grading of malignant
tumors is only applicable in those that show significant histologic and/or cytologic variation (and
differentiation) from tumor to tumor. It may be argued that tumors showing little to no variation or
histologic differences (e.g. Ewing sarcoma) should not be graded, as such data provide no additional
The primary purpose of histologic grading is to separate malignant tumors associated with a good
prognosis ((low grade) from tumors with a poor prognosis (high grade). The value of any histologic
grading system is related not only to the ability to predict patient survival but also to identify
patients who may benefit from adjuvant therapy. Unfortunately, there is no universally-accepted system
for the histologic grading of soft tissue sarcomas. Arguably, the 2 most commonly utilized grading
systems are the National Cancer Institute (NCI) and the Fédération Nationale des Centres de Lutte Contre
le Cancer (FNCLCC), also known as the French Federation of Cancer Centers sarcoma group, systems.
The NCI system is summarized in Table 4. These authors propose a 3-scale grading system largely
incorporating histologic subtype and the percentage of tumor necrosis.  The grade 1 lesions
include the following sarcoma subtypes: well differentiated liposarcoma, myxoid liposarcoma,
dermatofibrosarcoma protuberans, leiomyosarcoma, malignant hemangiopericytoma, malignant peripheral nerve
sheath tumor (MPNST), and myxoid chondrosarcoma. Liposarcomas are classified as well differentiated or
myxoid only when other components (i.e. round cell or pleomorphic) do not exceed 15 to 20% of the tumor
volume. Leiomyosarcoma is classified as grade 1 "when they exhibited an orderly fasciculated pattern,
well differentiated cytologic features, absence of pleomorphism, absence of necrosis, and exceedingly low
mitotic activity". MPNSTs are also regarded as grade 1 when they closely resembled neurofibroma but show
significant mitotic activity and areas of high cellularity. Adult hemangiopericytomas lacking
significant necrosis with only rare mitotic figures are grade 1 lesions. Extraskeletal Ewing
sarcoma/PNET, extraskeletal osteosarcoma, rhabdomyosarcoma, pleomorphic liposarcoma, synovial sarcoma,
mesenchymal chondrosarcoma, and malignant triton tumor are automatically designated as grade 3 lesions
based solely on their histologic subtype.For sarcomas not qualifying as grade 1 or 3, the extent of
necrosis determines the grade. Sarcomas with minimal necrosis (up to 15%) are classified as grade 2;
those with moderate to marked necrosis (>15%) are grade 3. In 1990, Costa  modified this
grading system, and some sarcomas that were originally classified as grade 3 (e.g. synovial sarcoma,
pleomorphic liposarcoma) are currently regarded as either grade 2 or 3.
Table 4: Grading parameters of the NCI system: based on the results of Costa et al. 
|Histologic Parameter ||Grade|
|well-differentiated LPS, myxoid LPS, well-differentiated malignant HP (<1 mitotic figure per 10 HPF), well-differentiated fibrosarcoma and LMS (<6 mitotic figures per 10 HPF), MPNST (resemble neurofibroma, <6 mitotic figures per 10 HPF), myxoid CS (no mitotic activity) ||1|
|Any sarcoma not automatically designated as grade 3 with tumor necrosis <15% ||2|
|Any sarcoma with > 15% necrosis RMS (all subtypes), extraskeletal OS and ES/PNET, mesenchymal CS, ASPS, synovial sarcoma, pleomorphic LPS ||3|
LPS = Liposarcoma
HP = Hemangiopericytoma
LMS = Leiomyosarcoma
MPNST = Malignant Peripheral Nerve Sheath Tumor
CS = Chondrosarcoma
RMS = Rhabdomyosarcoma
ES/PNET = Ewing Sarcoma/Primitive Neuroectodermal Tumor
OS = Osteosarcoma
ASPS = Alveolar Soft Part Sarcoma
HPF = High Power Fields
The FNCLCC system was initially described by Trojani et al.  and later modified by Guillou
et al.  A summary of this grading system is provided in Table 5. This system uses three
factors to establish histologic grade - tumor differentiation, mitotic count, and volume of tumor
necrosis. For each of these parameters, a score of 0 to 3 is assigned and a total score obtained from
the addition of the individual scores. Tumor differentiation corresponds to specific histologic
subtypes. Well-differentiated liposarcoma, fibrosarcoma, malignant peripheral nerve sheath tumor,
leiomyosarcoma, and chondrosarcoma are assigned a score of 1. When histologic subtype is "certain", a
score of 2 is given. This group includes myxoid liposarcoma, conventional fibrosarcoma and malignant
peripheral nerve sheath tumor, well-differentiated malignant hemangiopericytoma, myxoid and
pleomorphic/storiform malignant fibrous histiocytoma (MFH), myxoid chondrosarcoma, and conventional
angiosarcoma. A score of 3 is applied to sarcomas of uncertain type, including poorly differentiated and
epithelioid MPNST, giant cell and inflammatory MFH, rhabdomyosarcoma, synovial sarcoma, poorly
differentiated leiomyosarcoma, round cell, pleomorphic, and dedifferentiated liposarcoma, Ewing sarcoma,
osteosarcoma, alveolar soft part sarcoma, epithelioid sarcoma, clear cell sarcoma, and mesenchymal
chondrosarcoma. Tumor necrosis is defined as none, £ 50%, and > 50% for scores of 0, 1, and 2,
respectively. Mitotic counts of 0-9, 10-19, and ³ 20 per 10 high power fields (HPF) yielded scores of 1,
2, and 3 respectively. A total score of 2 or 3 was classified as grade 1; 4 or 5 as grade 2; and 6, 7,
and 8, as grade 3.
Regarding my personal preferred grading system for the histologic examination of soft tissue sarcomas,
I like the FNCLCC system. Among my colleagues, the scoring system appears easily applied and highly
reproducible. As expected, it is not a perfect system. Epithelioid, clear cell, and alveolar soft part
sarcomas are definitionally high grade lesions and may, in fact, be "under-graded" by the FNCLCC system.
For this reason, I do not typically document a histologic grade in the pathology report for these
sarcomas, but, instead, I add the statement, "definitionally high grade." Until recently, I included
synovial sarcoma among this group; however, recent evidence from the French Federation of Cancer Centers
Sarcoma Group suggests that histologic grade (eg. 2 vs. 3) by the FNCLCC system is prognostically
significant in synovial sarcoma.  Also, I see little reason for the histologic grading of
small blue cell sarcomas.
Table 5: Scoring and grading parameters of the FNCLCC system: Modified From Guillou, et al. 
| || Histologic Parameter ||Score|
|I.|| Tumor Differentiation|
| || Sarcomas closely resemblingnormal adult tissue (e.g. well-differentiated LPS, fibrosarcoma, MPNST, LMS, chondrosarcoma)|| 1|
| || Sarcomas of certain histologic subtype (e.g. myxoid LPS, conventional fibrosarcoma and MPNST, well-differentiated malignant HP, myxoid and storiform/pleomorphic MFH, myxoid CS, conventional AS)|| 2|
| || Sarcomas of uncertain histologic subtype (e.g. poorly differentiated and epithelioid MPNST, giant cell and inflammatory MFH, RMS, SS, poorly differentiated LMS, round cell, pleomorphic and dedifferentiated LPS, ES/PNET, OS, ASPS, EPS, CCS, poorly differentiated/epithelioid AS, Mesenchymal CS|| 3|
|II.|| Tumor Necrosis|
| || None ||0|
| || ≤50% ||1|
| || >50% ||2|
|III. || Mitotic Count|
| || 0-9 mitoses per 10 HPF || 1|
| || 10-19 mitoses per 10 HPF || 2|
| || >20 mitoses per 10 HPF || 3|
| || Histologic Grade || Total Score|
| || Grade 1 || 2,3|
| || Grade 2 || 4,5|
| || Grade 3 || 6,7,8|
LPS = Liposarcoma
MPNST = Malignant Peripheral Sheath Tumor
LMS = Leiomyosarcoma
HP = Hemangiopericytoma
MFH = Malignant Fibrous Histiocytoma
CS = Chondrosarcoma
AS = Angiosarcoma
RMS = Rhabdomyosarcoma
ES/PNET = Ewing's Sarcoma/Primitive Neuroectodermal Tumor
OS = Osteosarcoma
EPS = Epithelioid Sarcoma
CCS = Clear Cell Sarcoma
ASPS = Alveolar Soft Part Sarcoma
SS = Synovial Sarcoma
HPF = High Power Fields
Although these grading systems are clearly applicable to open biopsy (histologic) specimens, their
usefulness for FNAB samples is limited, as features, such as mitotic activity and percentage of necrosis
are not easily evaluated in cytologic specimens. Nevertheless, one can make certain assumptions
regarding histologic subtype and grade. As the aforementioned grading systems illustrate, many sarcomas
subtypes are considered automatically high grade based solely on their histologic classification. For
example, Ewing sarcoma is a high grade tumor, regardless of mitotic rate or percent necrosis. Thus, for
many sarcomas, if one can establish a histologic subtype, the grade of the tumor may also be determined.
A summary of this approach, applicable to FNAB is in Table 6.
Table 6: Soft tissue sarcoma grading by FNAB (based on the results of Kilpatrick, et al. )
|Cytologic Grade || Subtype or Cytomorphologic Group|
|Low Grade (by definition)|| Well-differentiated liposarcoma |
Extraskeletal myxoid chondrosarcoma
|High Grade (by definition)|| Small round cell sarcomas |
Adult pleomorphic sarcomas
Round cell liposarcoma*
Clear cell sarcoma
Alveolar soft part sarcoma
|Potentially Low or High Grade || Leiomyosarcoma|
Gastrointestinal stromal tumor
Malignant granular cell tumor
MFH = malignant fibrous histiocytoma
MPNST = malignant peripheral nerve sheath tumor.
* Myxoid liposarcoma and round cell liposarcoma are generally considered to be low and high grade
spectrums of the same neoplasm, respectively.
For those sarcomas which may be low or high grade, I normally use a combination of nuclear atypia,
cellularity, and the presence (or absence) of tumor necrosis to designate a histologic grade (e.g. low
vs. high), if deemed clinically necessary. However, at our institution, this is not always required.
For myxofibrosarcoma, a designation of low, intermediate, or high grade is rendered based on tumor
cellularity and the amount of myxoid stroma. Generally, these parameters are inversely
VI. The Importance of Ancillary Studies
Recently, immunohistochemical and genetic studies have revealed many novel, relatively sensitive (but
not always specific) immunohistochemical markers, chromosomal translocations, and molecular mutations in
soft tissue tumors. These are summarized in Table 7. The detection of the chromosomal and molecular
aberrations have involved many established as well as relatively new techniques including
immunohistochemistry, conventional cytogenetic analysis, polymerase chain reaction (PCR), in situ
hybridization, and Southern and Northern blotting. For more detail regarding these techniques and their
application to soft tissue sarcomas, the reader is referred to additional articles more adequately
addressing these issues.
Of special interest is the frequency of morphologically,
immunohistochemically, and clinically distinct soft tissue sarcomas harboring translocations and fusion
genes involving the EWS gene (22q12). In addition to Ewing's sarcoma, abnormalities associated with the
EWS gene are observed in intra-abdominal desmoplastic small round cell tumor, myxoid chondrosarcoma, and
clear cell sarcoma.
Table 7: Histologic Soft Tissue Sarcoma Subtypes with Associated Immunohistochemical Markers and Molecular and Cytogenetic Abnormalities
| Histologic Subtype ||IHCmarkers ||Cytogenetic Abboldities ||Molecular Events ||Estimated Frequency of Events|
|Ewing sarcoma/Primitive neuroectodermal tumor ||CD99 ||t(11;22)(q24;q12)t(21;22)(q22;q12)t(7;22)(p22;q12)others ||FLI1-EWS fusionERG-EWS fusionETV1-EWS fusion ||>85%5-10%<5%1-5%|
|Alveolar Rhabdomyosarcoma ||MyogeninMyo-D1 ||t(2;13)(q35;q14)t(1;13)(p36;q14) ||PAX3-FKHR fusionPAX7-FKHR fusion ||80%10-20%|
|Embryonal Rhabdomyosarcoma ||MyogeninMyo-D1 ||+2q, +8, +20 || ||70-80%|
|Desmoplastic Small Round Cell Tumor ||CytokeratinDesminNSE ||t(11;22)(p13;q12) ||WT1-EWS fusion ||>80%|
|Clear cell Sarcoma(Melanoma of soft parts) ||S-100HMB 45 ||t(12;22)(q13;q12) ||ATF1-EWS fusion ||>80%|
|Myxoid Chondrosarcoma(Chordoid sarcoma) ||Vimentin ||t(9;22)(q21-31; q12) ||EWS-CHN(TEC) fusion ||>50%|
|Myxoid/Round Cell Liposarcoma ||Vimentin ||t(12;16)(q13;p11) ||CHOP-TLS fusion ||>80%|
|Synovial Sarcoma ||CytokeratinEMA CD99 ||t(X;18)(p11;q11) ||SYT-SSX1 or SSX2 fusions ||>90%|
|Extrarenal Rhabdoid Tumor ||Cytokeratin ||22q11 deletion || ||? >90%|
|Infantile Fibrosarcoma ||Vimentin ||t(12;15)(p13;q25)+8,+11, +17, +20 ||ETV6-NTRK3 fusion ||? >90%|
|Alveolar Soft Part Sarcoma ||Vimentin Desmin ||t(X;17)(p11;q25) ||ASPL-TFE3 fusion ||Not known|
|Epithelioid Hemangioendothelioma ||CD31, CD34FLI-1 ||t(1;3)(p36;q25) ||Not known ||Not known|
IHC markers = immunohistochemical markers which are positive in the majority of cases
NSE = neuron specific enolase
EMA = epithelial membrane antigen
SYN = synaptophysin.
VII. Words of Caution Regarding Ancillary Studies
Following intensive scientific investigation, the initial enthusiasm associated with the discovery of
specific immunohistochemical and molecular markers is virtually always replaced by a more realistic
perspective. For example, the immunohistochemical marker CD99, initially reported as specific for
Ewing's sarcoma, has now been convincingly documented in most cases of mesenchymal chondrosarcoma,
synovial sarcoma and a few examples of small cell osteosarcoma, rhabdomyosarcoma and desmoplastic small
round cell tumor.
Cytokeratin positivity is sometimes observed in
rhabdomyosarcoma.  Desmin, an immunohistochemical marker usually associated with
rhabdomyosarcoma, is often expressed by the blastemal portion of Wilm's tumor. 
Cytogenetic and molecular analyses have yielded similar surprisingly nonspecific results. The most
frequent Ewing's sarcoma translocation, t(11;22)(q24;q12) with the EWS-FLI1 fusion product, has been
reported in phenotypically classic examples of mesenchymal chondrosarcoma, embryonal and alveolar
rhabdomyosarcoma, polyphenotypic tumor of childhood, and neuroblastoma.
EWS-FLI1 and the EWS-ERG fusion transcripts have been described in cases of intra-abdominal desmoplastic
small round cell tumor.
For these reasons, immunohistochemical and
molecular studies should be considered ancillary not definitive tests. The "gold standard" for diagnosis
remains light microscopy. I do not want to imply that knowledge and application of these
ancillary tests are not helpful for the pathologist faced with a difficult soft tissue sarcoma, but such
data should only be interpreted with the complete knowledge of the pertinent clinical and microscopic
features. As summarized by Dehner, "…the gold standard for pathologic diagnosis is predicated on the
microscopic features of the tumor. Once the results of an ancillary method like immunohistochemistry
become disconnected from the clinical and pathologic findings, we find ourselves separated from the first
principles of diagnostic pathology that have served as one of the fundamental pillars in the practice of
Benign Mimickers of Malignancy
Nodular Fasciitis and Proliferative Fasciitis/Myositis
Nodular fasciitis is a relatively common, reactive, and frequently rapidly growing but localized
proliferation of fibroblasts and myofibroblasts, usually arising in young adults (20-40 years of age)
most often within the extremities. Although rare, when it occurs in children, localization to the head
and neck region is most commonly observed. Anatomically, nodular fasciitis may involve the dermis,
subcutaneous fat, skeletal muscle, and even intravascular spaces. Nodular fasciitis characteristically
forms a well-circumscribed lesion (< 2 cm) with minimally infiltrative margins. Histologically, the
lesion is usually comprised of proliferating but uniform-appearing fibroblasts arranged loosely (e.g.
tissue culture appearance) in short fascicles and faintly developed storiform patterns. The background
stroma is variably collagenous but may be significantly myxoid, especially in cases arising in children.
In places, the lesional tissue often appears to "tear apart" creating pseudocystic spaces. Mitotic
figures may be abundant but should never be atypical. Lymphocytes, plasma cells, and histiocytes are
variably scattered among the lesional cells but the presence of neutrophils tends to be an unusual
finding. In contrast to most benign soft tissue tumors, FNAB of nodular fasciitis typically yields
hypercellular smears composed of mostly discohesive fibroblasts/myofibroblasts. The latter range from
spindled to stellate forms with mostly round to ovoid, uniform nuclei surrounded by delicate to dense,
tapering cytoplasm. Most often, the nuclei are vesicular with an evenly distributed chromatin pattern
and prominent nucleoli. A myxoid granular background film is often observed in aspirates procured from
relatively "young", actively growing lesions. Background inflammatory cells, especially histiocytes and
lymphocytes, are frequently seen. Spontaneous regression following even incomplete excision or FNAB is
the rule rather than the exception.  In my practice, once the diagnosis of nodular fasciitis
is rendered on FNAB, I generally recommend close clinical follow-up and re-biopsy if the lesion continues
to grow or fails to resolve in at least 2 months.
Proliferative fasciitis and myositis differ from nodular fasciitis both clinically and
pathologically. Although all are considered benign, probably reactive lesions, proliferative
fasciitis/myositis generally arise in the extremities of middle aged to older patients, most often 40-60
years of age. In contrast to the rather nice circumscription of nodular fasciitis, proliferative
fasciitis and myositis are infiltrative, frequently stellate appearing lesions, involving the deep
subcutaneous fat and skeletal muscle, respectively. In addition to spindled cells and myxoid background,
the proliferative lesions display variable numbers of large ganglion-like cells with large round to
ovoid, vesicular nuclei and 1 or more prominent nucleoli. Mitoses may be abundant but, as in nodular
fasciitis, are not atypical.
Immunohistochemically, nodular and proliferative fasciitis exhibit characteristics of
fibroblast/myofibroblast differentiation. Consequently, the lesional cells may show variable expression
of actin and desmin. The distinction between a nodular fasciitis and a smooth
muscle tumor, namely leiomyosarcoma, should be based on morphologic and clinical features, not
Myositis ossificans is a typically localized, self-limited, ossifying process that often occurs
secondary to soft tissue trauma. In reality, the designation myositis ossificans encompasses a
heterogeneous group of lesions, involving skeletal muscle, skin and subcutaneous fat (osteoma cutis),
tendons, nail bed (subungual exostosis), and periosteum (bizarre parosteal osteochondromatous
proliferation and florid reactive periostitis). Classic myositis ossificans most commonly occurs in
young adults (20-30 years of age), beginning as a rapidly growing mass within intramuscular tissues of
the extremities, principally the quadriceps, gluteus, and brachialis muscles. Initially, myositis
ossificans produces a localized nonossified proliferation, but, over time, becomes more progressively
ossified and calcified. As a consequence, FNAB findings are variable, depending upon the stage of the
lesion. Most cytologic preparations obtained early in the evolution of the proliferation are
hypercellular and composed of mostly discohesive, reactive-appearing fibroblasts/ myofibroblasts (closely
resembling that observed in nodular or proliferative fasciitis) and minimal amounts of osteoid. With
progression, cytologic preparations become progressively less cellular and contain considerably greater
amounts of matrix material and calcified debris. Indeed, older examples are often so significantly
ossified that sampling by FNAB is not possible. The radiologic findings, especially on plain film
X-rays, are often characteristic and thus helpful in establishing a definitive diagnosis.
Malignant Mimicker of Benignancy
Well-Differentiated Liposarcoma (atypical lipoma)
Well differentiated liposarcoma (WDLS) is synonymous with the designation "atypical lipoma" referring to
an adult low grade sarcoma that may locally recur but, in the absence of dedifferentiation, will not
metastasize. In the past, the designation of "atypical lipoma" was generally reserved for
superficially-located lesions. However, several examples of such lesions have recently been described
having undergone dedifferentiation. Furthermore, cytogenetic analysis has revealed ring and giant marker
chromosomes in the majority of these atypical lipomatous tumors, regardless of anatomic origin (e.g.
subcutaneous vs. intramuscular).  For these reasons, it is probably best to diagnose all such
lesions as WDLS.
The classification of WDLS includes at least 3 subtypes: lipoma-like, sclerosing, and inflammatory
variants. For the purposes of this discussion, we will focus on the more common lipoma-like variant.
Histologically, WDLS may have areas indistinguishable from ordinary lipoma. Meticulous sampling is
sometimes necessary to establish (as well as exclude) the diagnosis, especially among deeply-seated fatty
tumors of the extremities or retroperitoneum. At low power, helpful clues include heterogeneity (both
size and shape) of the adipocytes, an increased amount of fibrous tissue (compared to conventional
lipomas), and the presence of atypical, enlarged and hyperchromatic nuclei. Multi-nucleated giant cells
are randomly distributed throughout the lesion but their absolute numbers tend to be variable. In my
experience, classic multi-vacuolated lipoblasts are less commonly observed and certainly not required for
The differential diagnosis of superficial and deeply-seated fatty tumors represents one of the greater
challenges in cytopathology. It encompasses the diagnoses of WDLS, lipoma, and inadvertently sampled
normal subcutaneous adipose tissue. Seemingly unremarkable adipose tissue may be observed in all of
these entities. As a result, FNAB is quite adequate for "ruling in" the diagnosis of WDLS but probably
inadequate for "ruling out" the diagnosis. The diagnosis of WDLS rests upon the cytologic finding of
atypical and enlarged hyperchromatic nuclei and/or multivacuolated lipoblasts. Even in the event of
benign cytologic findings, I would recommend that all deeply-seated fatty tumors (by imaging studies
and/or biopsy) of the extremities or retroperitoneum be completely excised and meticulously sampled to
fully exclude the possibility of WDLS.
A significant proportion of WDLS may undergo a process of "dedifferentiation," implying transformation
to a higher grade nonliposarcomatous sarcoma. By definition, the diagnosis of dedifferentiated
liposarcoma is justified by either the presence of WDLS juxtaposed to a nonliposarcomatous sarcoma or the
development of a high grade nonliposarcomatous sarcoma in the region of a previously excised WDLS. For this reason, we always instruct our residents to sample the adjacent soft tissue
immediately surrounding all adult sarcoma cases. Histologically, the high grade component most
often resembles a non-descript spindle cell or pleomorphic sarcoma but rarely may exhibit
leiomyosarcomatous, rhabdomyosarcomatous, osteosarcomatous and/or chondrosarcomatous components. Because
the diagnosis of dedifferentiated liposarcoma usually depends upon the finding of both low grade and high
grade components, it is usually not possible to render a specific diagnosis of dedifferentiated
liposarcoma on FNAB samples. In my experience, only the high grade component is generally sampled and
the diagnosis of pleomorphic sarcoma, not otherwise specified, is rendered.
Spindle Cell Tumors
Benign Nerve Sheath Tumors
Nerve sheath tumors, especially benign forms, are fairly common neoplasms occurring in virtually all
anatomic compartments and sites. Outside of neurofibromatosis, the vast majority of the deeply-seated
lesions are schwannomas (neurilemomas); neurofibromas generally present as solitary cutaneous masses and
are rarely observed within the deep soft tissues of the extremities, thorax, or abdomen. Most
pathologists are fairly familiar with the histology of neurofibroma and schwannoma; however, there are
some important points worth reviewing.
Neurofibromas may present as one of three forms: solitary and circumscribed (most common), diffuse,
and plexiform. For all practical purposes, the latter 2 subtypes are only seen in patients with
neurofibromatosis. Histologically, neurofibromas are unencapsulated and generally arise as fusiform
expansions of peripheral nerve, although this is often not evident in cutaneous tumors. Microscopically,
neurofibromas are typically comprised of haphazardly arranged Schwann cells accompanied by variable
amounts of wire-like strands of collagen, embedded in at least a partially myxoid stroma. Chronic
inflammatory cells and histiocytes are often randomly distributed throughout the tumor. More cellular
tumors appear more compact with less myxoid matrix, often exhibiting prominent whorls and storiform
patterns. As in schwannomas, degenerative features (ancient change) characterized by enlarged and
hyperchromatic Schwann cell nuclei (often having a "smudged" appearance) may be evident. Mitotic figures
are virtually absent; consequently, it is this criteria which is most helpful in establishing the
diagnosis of malignancy.
Schwannomas are well-circumscribed and encapsulated masses that usually arise as eccentric, sometimes
dumbbell-shaped lesions of peripheral nerve. As previously mentioned, they are only rarely cutaneous and
almost always solitary. Microscopically, the histologic hallmark of schwannomas is the alternating
Antoni A (cellular and compact spindle cell proliferation arranged in bundles and intersecting fascicles)
and Antoni B (hypocellular spindle cells proliferation arranged haphazardly within a loose matrix)
regions. Nuclear palisading and so-called Verocay bodies are often observed in the Antoni A areas, while
microcystic change and inflammatory cells are usually evident in the Antoni B zones. Another helpful
feature is the presence of hyalinized blood vessels. Atypia in the form of ancient change is often
present and may be widespread, but the degree of atypia exceeds the expected mitotic activity, which in
degenerative schwannomas is virtually absent. As a general rule, avoid the
diagnosis of malignancy in any spindle cell tumor with moderate to marked atypia but a correspondingly
low mitotic rate. An additional diagnostic pitfall is cellular schwannoma, a tumor comprised of
predominantly Antoni A. Mitotic activity may be quite high and focal necrosis is occasionally observed.
Helpful diagnostic clues include circumscription, the presence of hyalinized blood vessels, and strong
and diffuse S-100 protein positivity.
Malignant Peripheral Nerve Sheath Tumor
The diagnosis of malignant peripheral nerve sheath tumor (MPNST) usually requires at least one of the
following clinical settings: 1) origin from a peripheral nerve, usually a large nerve trunk, 2) origin
from a pre-existing benign nerve sheath tumor, usually a neurofibroma, and/or 3) occurrence in a patient
with neurofibromatosis. Rarely, the diagnosis may be rendered outside of these clinical settings, if
classic morphologic features are evident. Compared to other spindle cell tumors, the morphologic
diversity of MPNST is quite impressive. Most appear as obvious high grade sarcomas and manifest features
of fibrosarcoma. Surprisingly, classic nuclear palisading is often absent. Additionally, MPNST may be
very heterogeneous, containing elements of glandular differentiation, rhabdomyosarcoma, bone and/or
cartilaginous metaplasia. More anaplastic-appearing forms closely resemble pleomorphic malignant fibrous
Fine Needle Aspiration Biopsy of Nerve Sheath Tumors
In most cases, FNAB is extraordinarily useful for the differential diagnosis of benign nerve sheath
tumor vs. MPNST. Cytologic smears from schwannomas are often very cellular; however, the cells are
usually arranged as cohesive cellular tissue aggregates, comprised of the cytoplasmic processes of the
individual Schwann cells. Morphologically, individual tumor cells are spindled with wavy to serpentine
nuclei and may show significant atypia (ancient change). Mitotic activity is virtually never observed.
Unfortunately, nuclear palisading is a helpful but uncommon finding in cytologic
In contrast, MPNST are virtually always recognized as malignant. Smears are highly cellular and
comprised of mostly individually dispersed tumor cells. Nuclei tend to be quite large, atypical and
multinucleated forms are often present.
In difficult cases, the use of S-100 protein on core needle biopsies or cell block preparations is
invaluable. Strong and diffuse staining for S-100 protein is a hallmark of schwannoma or neurofibroma;
patchy and focal staining for S-100 protein is more typical of MPNST.  Furthermore, up to 50%
of cases of MPNST may be negative for S-100 protein. For these reasons, a spindle
cell lesion with "borderline" malignant features that is strongly and diffusely S-100 protein positive is
probably NOT malignant. Such a tumor is far more likely to represent a schwannoma than a MPNST.
Obviously anaplastic and malignant tumors that are strongly and diffusely S-100 protein positive are far
more likely to represent melanoma than MPNST.
Small Round Cell Sarcomas
Recent evidence suggests that Ewing sarcoma (ES) and primitive neuroectodermal tumor (PNET) represent
a spectrum of similar appearing bone and soft tissue sarcomas which usually share the same cytogenetic
abnormality, t(11;22)(q24;q12).  In the past, the distinction between these two entities
rested largely on the finding of a rather arbitrary number of rosettes, generally seen in PNET but less
obvious in ES. Most investigators now consider these tumors to form points along a continuous spectrum
of the same neoplasm, with classic mostly undifferentiated ES at one end and more neural-differentiated
PNET at the opposite end. 
Clinically, ES arises far more commonly as a primary bone tumor than an extraskeletal neoplasm. As a
concomitant soft tissue mass is virtually always present with the skeletal form, radiographs, including
CT scan and/or MRI, are essential for establishing extraskeletal origin. Within soft tissues,
extraskeletal ES most commonly arises, in decreasing order of frequency, from the trunk (chest wall and
paravertebral region), lower extremities, and retroperitoneum. Most patients are between 10 and 20 years
of age at diagnosis and, for reasons not completely understood, rarely afflicts
Regardless of anatomic site of origin, the histologic and cytomorphologic characteristics are
similar. At low power, ES cells may be arranged in diffuse "fields", variably-sized nests, and/or
fascicles separated by fibrovascular septa. The nuclei tend to be very uniform with hyperchromasia,
inconspicuous nucleoli, and scant rims of esoinophilic to clear cytoplasm. Rarely, ES may contain a
population of larger tumor cells with more irregular nuclear membranes and obvious nucleoli, the
so-called atypical large cell variant.  Bi- and multi-nucleated tumor cells are never
present. Variable numbers of rosettes may be observed, especially in the more differentiated PNET. Most
commonly, the rosettes resemble those of neuroblastoma, exhibiting an eosinophilic core of
neurofibrillary material surrounded by the tumor cells (Homer-Wright rosettes). Mitotic activity tends
to be variable and mitoses may be surprisingly sparse. Spontaneous tumor necrosis is frequently present,
sometimes leaving residual viable-appearing tumor cells forming collars around blood vessels. In both
open and needle biopsy specimens, the diagnosis of ES may be easily established, especially if sufficient
material is obtained for ancillary techniques. Cytologic smears are markedly cellular and composed of
mostly individually dispersed cells and scattered small cohesive cell clusters.  The tumor
cells are remarkably uniform, possessing a single round hyperchromatic nucleus, an extraordinarily high
nuclear to cytoplasmic ratio, and a thin rim of cytoplasm, sometimes containing small vacuoles indicative
of intracytoplasmic glycogen. Nucleoli are generally inconspicuous to absent. Rare forms of the
so-called atypical (large cell) ES may exhibit more uniformly enlarged nuclei, irregular nuclear
membranes, and prominent nucleoli. Background matrix material (e.g. cartilage or osteoid) is distinctly
absent. Rarely, tumor rosettes may be observed.
In general, ancillary studies are essential for a specific diagnosis. Immunocytochemically, most
cases of ES possess the mic-2 glycoprotein product demonstrated by positive cytoplasmic membranous
staining for CD99.  However, this finding is not specific for ES and may be observed in
synovial sarcoma, small cell osteosarcoma, and mesenchymal chondrosarcoma.
The ES gene
on chromosome 22 may be associated with several translocations including, in decreasing order of
frequency, t(11;22)(q24;q12), t(21;22)(q12;q12), and t(7;22)(q22;q12).  We have already
mentioned the fact that other phenotypically distinct tumors have also shown similar cytogenetic and
Rhabdomyosarcoma (RMS) is the most common soft tissue sarcoma of childhood. The most recent
International Classification of Rhabdomyosarcoma (ICR) recognizes 5 prognostically significant subtypes –
spindle cell and botryoid embryonal RMS (superior prognosis), embryonal RMS (intermediate prognosis), and
alveolar RMS and undifferentiated sarcoma (poor prognosis).
Recently, a diffuse
anaplastic variant has been added to the poor or unfavorable prognostic category.  Various
subtypes tend to arise in specific age groups and anatomic sites. The botryoid form of embryonal RMS
generally arises in the submucosa and therefore tends to occur in mucosa-lined hollow organs such as the
bladder, vagina, rectum, nasal cavity and nasopharynx. Most patients are < 10 years of age at
diagnosis. The spindle cell subtype of embryonal RMS is also more commonly observed in children < 10
years of age but most frequently arises in the paratesticular and head and neck regions. Conventional
embryonal RMS, the most common RMS subtype, is frequently seen in children < 10 years of age but
occasionally observed in adolescents and young adults. Common anatomic sites include the genitourinary
tract, head and neck, and retroperitoneum. In contrast to most other subtypes, alveolar RMS is more
frequently observed in adolescent patients in the extremities.
Conventional or classical embryonal RMS consists of mainly poorly differentiated tumor cells arranged
in diffuse aggregates and nests. Alternating hypo- and hypercellular regions are frequently present,
sometimes creating a "pulmonary edema" pattern. The background stroma ranges from myxoid to partly
hyalinized. Individual tumor cells are ovoid to slightly spindled with round to ovoid, hyperchromatic
nuclei, inconspicuous to prominent nuclei, and scant to abundant tapering cytoplasm. The latter may
display clearly identifiable cross striations. Large, polygonal-appearing rhabdomyoblasts are often
scattered among the more undifferentiated cells but tend to be more abundant and frequently observed in
postchemotherapy specimens. Importantly, conventional embryonal RMS lacks the features (or criteria)
necessary for the diagnosis of the various specific subtypes, such as botryoid or spindle cell.
In the past, the diagnosis of alveolar RMS was restricted to tumors exhibiting the classic
fibrovascular septa outlining cleft-like spaces, lined by a population of mostly round tumor cells.
Cytologically, the tumor cells are mostly uniform, round to slightly ovoid with round, hyperchromatic to
vesicular nuclei, one or more prominent nucleoli, and surrounded by a thin rim of eosinophilic
cytoplasm. Bi- and multi-nucleated tumor giant cells are usually present. Definite myogenesis and
so-called "strap" cells are not generally features of alveolar RMS. Less commonly, the neoplastic cells
may be vacuolated and have more abundant, clear cytoplasm, presumably due to the presence of glycogen.
In 1992, Tsokos and colleagues described a "solid" form of alveolar RMS, characterized as having the
cytologic features of alveolar RMS but appearing more compact, lacking the "alveolar"
architecture.  Importantly, there was no difference in survival between the classic
appearing alveolar and the more "solid" variant.  In addition to the solid nested pattern of
alveolar RMS, I have also encountered examples with tumor cells arranged in infiltrative, double-layered
to single file cords, resembling lobular carcinoma of the breast. It is not uncommon to encounter two or
more of these patterns within the same neoplasm. I should also point out that one may occasionally
observe examples of RMS with elements of both embryonal and alveolar differentiation. According to the
current IRC classification, such tumors should be considered alveolar RMS for therapeutic
For the diagnosis of botryoid RMS, the current ICR/IRSG requires the presence of an intact epithelium
overlying a condensed tumor layer of rhabdomyoblasts (cambium layer) in at least 1 microscopic
field.  The presence of a "grape-like' cluster of tumor cells visualized grossly and/or in
the intraoperative setting is not enough, by itself, to establish the diagnosis of botryoid embryonal
RMS. The diagnosis must be confirmed microscopically. Variable amounts of rhabdomyoblastic
differentiation may be evident, ranging from more undifferentiated tumor cells to diagnostic spindled to
stellate rhabdomyoblasts with definite cross striations. Additionally, the background stroma is
Spindle cell RMS is a relatively recently described entity, having been only clearly documented in the
literature since 1993.
By definition, spindle cell morphology must predominate; more
rounded rhabdomyoblasts or multinucleated tumor cells are rarely observed. More commonly, the spindle
cells exhibit ovoid nuclei, prominent nucleoli, and are arranged in whorls and storiform patterns;
however, occasional cases resemble leiomyosarcoma with tumors cells arranged in intersecting bundles or
fascicles and comprised of elongated spindled cells with cigar-shaped nuclei.
Anaplasia is defined as tumor cells containing enlarged (3 times the size of adjacent nuclei) and
lobated, hyperchromatic nuclei and the presence of atypical, multipolar mitotic figures. It is of
prognostic import only when diffusely present, that is "where anaplastic
cells aggregate in clusters or form a continuous sheet." 
Given the range of histologic subtypes in RMS, it is not surprising that there exists a range in
cytomorphology as well. In general, the degree of cytomorphologic variability tends to exceed that
observed in other small round cell sarcomas, especially in the embryonal subtype.  Regardless
of subtype, FNAB cytologic smears of RMS are highly cellular and comprised of mostly discohesive
neoplastic cells. More specifically, the embryonal subtype is composed of small to intermediate-sized
cells with round and polygonal to spindled cell contours.  Nuclei are round to ovoid,
hyperchromatic, and may have one or more prominent nucleoli. Scant to abundant amounts of dense to
stringy cytoplasm, rarely associated with cross striations, may be present, depending upon the degree of
differentiation. Well-differentiated rhabdomyoblasts are more frequently observed in post-chemotherapy
specimens. Myxoid stromal tumor fragments may occasionally be seen. Compared to embryonal RMS, the
tumor cells in the alveolar form are larger and more uniformly round to polygonal with a round nucleus
and a thin rim of dense cytoplasm.  Indeed, the uniformity of the tumor cells in alveolar RMS
is a characteristic cytologic feature and more likely to cause confusion with other small round cell
neoplasms, such as lymphoma or Ewing's sarcoma. A helpful diagnostic feature of alveolar RMS is the
presence of bi- and multi-nucleated tumor giant cells. Although uncommon, both embryonal and alveolar
RMS may exhibit a frothy background (tigroid appearance) resembling that seen in aspirates of
Despite the distinct histologic (and seemingly distinct cytologic ) differences between the embryonal
and alveolar forms of RMS, absolute distinction is usually but not always possible in small biopsy
specimens. In this situation, determination of DNA ploidy may be helpful, as most cases of embryonal RMS
are hyperdiploid while alveolar RMS is frequently tetraploid. Cytogenetic analysis may also be useful as
the majority of cases of alveolar RMS have a specific translocation, either the t(2;13)(q35;q14) with the
PAX3/FKHR fusion or the t(1;13)(q36;q14) with the PAX7/FKHR fusion.  We should point out
that recent IRS protocols for low risk and low stage patients call for the addition of an alkylating
agent (cyclophosphamide) to the usual regimen (vincristine and actinomycin-D) if the RMS subtype is
unfavorable (e.g. alveolar). However, adolescent patients with metastatic disease at presentation (Group
IV) are treated similarly regardless of subtype. For these reasons, close communication of the
pathologist with the pediatrician is recommended in difficult cases to avoid unnecessary tests and
additional biopsies. In some circumstances, a diagnosis of simply "rhabdomyosarcoma" will suffice for
initial treatment purposes. Immunocytochemical positivity for desmin, actin, myo-D1, and/or myogenin
help support the diagnosis of RMS. Occasional tumor cells may express cytokeratins, especially the
Desmoplastic Small Round Cell Tumor
Desmoplastic small round cell tumor (DSRCT) is a relatively recently described entity, generally
occurring within the abdomen and afflicting adolescents and young adults, usually between 15 and 35
years.  The characteristic histologic feature of DSRCT is the presence of sharply
circumscribed nests and clusters of small round tumor cells within a densely fibrotic to fibromyxoid
stroma. The neoplastic cells are usually uniform, small, round to ovoid with high nuclear cytoplasmic
ratios, hyperchromasia, and inconspicuous nucleoli. Cytoplasm is typically scant and eosinophilic but
may be abundant and clear to vacuolated. Uncommon histologic features include cells with spindle-like
morphology, signet ring features, rhabdoid phenotype, insular growth patterns, and tumor
rosettes.  Although data is limited, most FNAB samples are variably cellular, probably
dependent on the amount of desmoplastic stroma within the tumor. The neoplastic cells are mostly small,
uniformly round to ovoid, and dispersed individually and in small aggregates. Nuclei exhibit a finely
granular chromatin pattern and inconspicuous to small nucleoli. As expected, the tumor cells share a
close resemblance to ES. The presence of variably cellular collagenous stromal fragments, a feature not
observed in ES, is a useful finding, supporting the diagnosis of DSRCT. Immunocytochemically, DSRCT
typically displays multipotential differentiation, including but not limited to expression of WT-1,
cytokeratins, neuroendocrine markers, and desmin. Due to the later, it may sometimes be difficult to
distinguish DSRCT from RMS. Determination of DNA ploidy by image analysis may be helpful. The vast
majority of RMS are aneuploid while most cases of DSRCT are diploid. Cytogenetic analysis in up to 75%
of cases reveals a characteristic translocation involving the ES gene, t(11;22)(p13;q12) with the
formation of the WT1/EWS fusion gene product. Because of its highly aggressive behavior despite
initially favorable responses to chemotherapy, DSRCT should be separated from other small round cell
tumors of childhood.
Adult Pleomorphic Sarcomas
The term adult pleomorphic sarcoma encompasses a diverse group of sarcomas, all of which share a
histologic pattern of fascicles and nests of markedly pleomorphic, spindled to round cells and
anaplastic, multi-nucleated tumor giant cells. In the past, such tumors were often "lumped" into the
rather vague category of "pleomorphic malignant fibrous histiocytoma" (MFH). We now know that this
histologic pattern is relatively non-specific.  In fact, through meticulous attention to
morphologic detail and judicious use of immunohistochemistry, the majority of pleomorphic sarcomas may be
subclassified as pleomorphic liposarcoma, leiomyosarcoma , rhabdomyosarcoma, dedifferentiated
liposarcoma, and myxofibrosarcoma. Less commonly, the pleomorphic MFH pattern may be observed in
angiosarcoma, MPNST, and extraskeletal osteosarcoma. From a strict therapeutic perspective, the
separation of these sarcoma subtypes is generally not necessary, as all are high grade sarcomas.
However, included in the differential of pleomorphic sarcoma, are metastatic carcinoma, melanoma, and
anaplastic lymphoma, all of which require therapeutic options and exhibit prognostic features distinctly
different from pleomorphic sarcoma. A careful review of the clinical and radiographic findings will
uncover most such cases. I should also point out that recent evidence suggests that pleomorphic sarcomas
with myogenic differentiation (eg. leiomyosarcoma or rhabdomyosarcoma) appear to behave more
aggressively, but, at present, histogenetic-specific therapy is not available. 
As expected, the cytologic features of the pleomorphic sarcoma subtypes are also quite similar.
Histologic subtyping by FNAB is often impossible but usually not problematic, as long as the tumor can be
recognized as a pleomorphic sarcoma. Cytologic smears are usually moderately cellular and comprised of
small clusters and solitary, markedly pleomorphic, spindled to epithelioid tumor cells. Bi- and
multi-nucleated tumor giant cells are virtually always evident. In my experience, cell blocks are most
useful for ancillary studies, helping to exclude non-mesenchymal lesions, if clinically indicated.
Epithelioid/Polygonal Cell Sarcomas
The category of epithelioid/polygonal cell sarcomas encompasses a group of rather diverse entities,
all of which share the presence of an epithelioid to polygonal shaped cell usually with abundant
cytoplasm and a mostly round to slightly ovoid nucleus. It is also worth mentioning that some neoplasms
included here, especially synovial sarcoma, may also be placed within the differential diagnosis of
spindle cell tumors. Although the reader may assume that this category of sarcoma shares little in their
biologic characteristics, common features exist among many of these neoplasms that deserve special
consideration. Despite very distinct morphologic and immunohistochemical properties, epithelioid
sarcoma, synovial sarcoma, alveolar soft part sarcoma, and clear cell carcinoma, exhibit similar
epidemiologic and biologic characteristics. All tend to afflict adolescents and young adults, arising
mostly within the extremities. Although 5-year survival rates may range from 60 to 75%, continued
declines in 10 and 15-year survival rates are classically observed. It is not unusual for an afflicted
patient to succumb to metastatic disease twenty years after initial presentation. Another interesting
feature shared by these tumors is their relatively high frequency of regional lymph node involvement, a
finding uncommonly observed among most soft tissue sarcomas. Consequently, these sarcomas are considered
high grade lesions and, at most institutions, are treated with adjuvant radiation therapy and/or
Synovial sarcoma (SS) is a soft tissue malignancy of uncertain histogenesis. Based on its morphologic
and immunohistochemical properties, it may be more accurately characterized as a "carcinosarcoma" of soft
tissues. The original designation of "synovial sarcoma" was partly due to the fact that these tumors
frequently arose in soft tissues near, but rarely within, large joints and, histologically, had an
epithelial component that was likened to synovium. However, it should be remembered that normal synovium
lacks immunoreactivity to keratins.
SS typically arises in adolescence and young adults, with a peak incidence between 20 and 40 years of
age. The lower extremity is more commonly afflicted than the upper extremity and proximal more often
than distal. Radiologically, SS are almost always deeply-seated, presenting as deceptively
well-circumscribed but nonspecific soft tissue masses, often containing stippled calcifications.
On the basis of morphologic features, SS may be subtyped into four distinct categories: monophasic
fibrous, monophasic epithelial, biphasic, and poorly differentiated.  The latter
categorization likely represents a poorly differentiated form of monophasic fibrous. The monophasic
fibrous subtype, according to most recent series, represents the most common variant, followed in
decreasing order of frequency by biphasic, poorly differentiated, and monophasic epithelial
subtypes. Arguably, many, if not most, examples of monophasic epithelial subtype are biphasic tumors with
an overwhelmingly predominant epithelial component. Regardless of subtype, SS often exhibit
intratumoral calcifications and may undergo cystic degeneration.
Monophasic fibrous SS is characterized by a relatively uniform population of monotonous slightly
spindled tumor cells arranged in short fascicular patterns associated with variable amounts of stromal
collagen deposition. Alternating hypercellular and hypocellular, loose-appearing areas are frequently
observed. Less commonly, the latter may be abundantly mucinous, so extensive in some cases as to
confound the diagnosis. A hemangiopericytoma-like vascular pattern is also frequently seen in the more
cellular regions. Cytologically, individual tumor cells appear slightly spindled with a round to ovoid,
hyperchromatic nucleus and inconspicuous nucleoli. Nuclear to cytoplasmic ratios are generally quite
high with scant amounts of slightly tapering cytoplasm. Surprisingly, mitotic activity may be quite low,
ranging from <1 mitosis to >10 mitoses/10 high power fields. Poorly differentiated SS are also
generally comprised of a mostly uniform population of rounded to polygonal and/or slightly spindled cells
with round to ovoid nuclei. However, a vesicular or coarsely granular chromatin pattern and prominent
nucleoli are characteristic. Additionally, a high mitotic rate, often exceeding 20 mitoses/10 high power
fields, is frequently observed.  These tumors usually lack obvious epithelial or glandular
differentiation. Infrequently, poorly differentiated SS may be composed of small round cells with more
uniformly hyperchromatic nuclei, resembling a Ewing sarcoma. The epithelial component of biphasic and
monophasic epithelial subtypes is most commonly glandular, resembling a moderate to well differentiated
adenocarcinoma. Less common patterns include papillary differentiation and solid nests of poorly
differentiated glandular cells.
Cytologic preparations from SS are generally highly cellular and composed of a mostly discohesive
tumor cell population accompanied by occasional tumor cell aggregates.  In most cases,
individual tumor cells are remarkably uniform with high nuclear to cytoplasmic ratios, round to ovoid
hyperchromatic nuclei, inconspicuous nucleoli, and absent to scant, slightly tapering cytoplasm. As
expected, the epithelial tumor cells typically exhibit a mostly round, vesicular nucleus and occasional
prominent nucleoli. Unfortunately, in our experience, the vast majority of SS, including the biphasic
variant, lack such cells on cytologic smears.  This is usually not problematic, as the
diagnosis is easily rendered when material is available for ancillary studies.
Both monophasic and epithelial components usually express keratins and epithelial membrane antigen.
Epithelial membrane antigen appears to represent a more sensitive marker than cytokeratin, especially for
the poorly differentiated subtype of SS.  Cytoplasmic CD99 reactivity has been reported in
greater than 50% of tumors.  A potential source of confusion with the differential diagnosis
of MPNST is the fact that up to approximately one-third of SS may show at least focal positivity for
S-100 protein. Because MPNST may occasionally exhibit cytokeratin and/or epithelial membrane antigen
positivity, some have advocated the use of cytokeratin subsets for distinguishing monophasic fibrous SS
from MPNST. Utilizing cytokeratins 7 and 19, Smith et al. evaluated twenty-nine cases of monophasic
fibrous SS.  Twenty-three cases expressed both cytokeratin subsets whereas only two cases
appeared negative for both cytokeratins. Conversely, among 22 MPNST, two cases expressed cytokeratin 7,
one case stained with cytokeratin 19, but no examples expressed both cytokeratins. 
In a majority of cases, conventional cytogenetic analysis reveals a balanced relatively specific
translocation t(X;18)(p11.2;q11.2) with the fusion product SYT-SSX.  This finding is
observed in both biphasic and monophasic subtypes. Furthermore, we now know that two related but
distinct X-chromosomal genes may be rearranged in this translocation, producing distinct fusion products,
SYT-SSX1 and SYT-SSX2. Kawai et al.  documented an association between SS subtype and the
SYT-SSX subtype. In their series, biphasic tumors exclusively harbored the SYT-SSX1 rearrangement while
monophasic fibrous types were predominantly (but not exclusively) found to have the SYT-SSX2
rearrangement. However, other investigators have not shown such a relationship.  We should
also point out that fine needle aspirates have proven successful in detecting the t(X;18) and its fusion
product, SYT-SSX. 
Estimated 5-, 10-, and 15-year survival rates for SS range from 50 to 75%, 20 to 50%, and 10-45%,
respectively. Risk factors for disease progression have included older age of patient (>25 years),
large tumor size (>5 cm), poorly differentiated subtype (high nuclear grade), extensive tumor
necrosis (>50%), presence of rhabdoid morphology, presence of bone and/or neurovascular invasion, and
high tumor stage. Bergh and colleagues retrospectively reviewed their experience with 121 patients with
SS, dividing them into two prognostic categories: low and high risk.  The low risk group
included patients aged <25 years, tumor size <5 cm, and the absence of poorly differentiated
histology. Overall disease free survival among patients in the latter group was estimated at 88%.
Conversely, the high risk category, defined as patient ages > 25 years, tumor size >5
cm, and the presence of poorly differentiated histology, had an overall disease free survival of
approximately 18%. 
SS containing the SYT-SSX1-type fusion appear to be associated with worse prognosis than those
harboring the SYT-SSX2 subtype.
Inagaki et al. documented a statistically significant
relationship between the SYT-SSX1 fusion and high Ki-67 expression and a high mitotic rate, suggesting
that the SYT-SSX fusion subtype is associated with the tumor cell proliferative activity. 
At most institutions, SS is considered an intermediate to high grade sarcoma requiring not only
surgical resection, but adjuvant therapy, including chemotherapy and/or radiation therapy. Adequate
primary surgery appears to represent the most significant factor for local control and the prevention of
Like SS, the histogenesis of epithelioid sarcoma (EPS) is uncertain. The morphologic and
immunohistochemical profile of EPS suggests that it may represent a "carcinoma" of soft parts. EPS
typically arises in adolescent and young adults, with a usual age range of 10-40 years. It predominantly
occurs in the distal extremities and has a propensity, unlike most sarcomas, to involve the upper
extremity, especially the hands and forearm.  Consequently, EPS often presents as a
painless, soft tissue mass, superficially located within the dermis and subcutis, especially along the
fascial planes and tendon sheaths. Radiologically, EPS are generally solitary but often multinodular
masses that, similar to SS, frequently harbor calcifications, in approximately 20-30% of
EPS typically grow as multinodular proliferations comprised of epithelioid to polygonal-appearing
tumor cells with round to ovoid vesicular nuclei, inconspicuous to prominent nucleoli, and abundant
Nuclear pleomorphism generally ranges from slight to moderate
with occasional bi-nucleated forms. Intracytoplasmic vacuoles may occasionally be observed, mimicking
the appearance of epithelioid hemangioendothelioma. When abundant necrosis is present, the multinodular
growth pattern exhibits a "granulomatous" appearance, reminiscent of rheumatoid nodule or granuloma
annulare. These necrotic zones are comprised of acellular debris and/or abundant hyalinized collagen.
Less commonly, focal areas of myxoid change may be observed. Rarely, multinucleated giant cells may be
haphazardly scattered among the cellular regions. Such cells may result in an erroneous interpretation
of giant cell tumor of tender sheath. Although virtually all cases exhibit predominant epithelioid
morphology, focal areas displaying spindle cell patterns and even storiform growth often occur.
Dystrophic calcifications with or without osseous metaplasia are seen in up to one-fifth of
cases.  Surprisingly, mitotic activity is quite variable, ranging from <1 to >20
mitotic figures/10 high power fields. Recently, Guillou and others have described a "proximal-type" of
EPS displaying marked cytologic atypia and rhabdoid features histologically and arising within more
proximal anatomic sites, including the pelvis and perineal regions, buttocks and hip. 
Limited follow-up data suggest that this unusual variant of EPS may behave more aggressively.
Cytologic smears generally range from moderate to highly cellular with mostly individually dispersed
tumor cells ranging from epithelioid and polygonal to slightly spindled. Nuclei are generally large,
round to ovoid, with a slightly granular to vesicular chromatin pattern, and occasional prominent
nucleoli. The cytoplasm is dense, often abundant, and may be slightly tapering. 
Intracytoplasmic vacuoles, presumably degenerative, appear at least focally, present in a majority of
Immunohistochemically, the vast majority of EPS express cytokeratin and epithelial membrane antigen.
Likewise, vimentin is positive in >95% of cases. Up to 50% of tumors, at least focally, express
As metastatic carcinomas are almost always negative for this marker, the use of
CD34 may, in difficult cases, help distinguish these two entities.  Unlike epithelioid
hemangioendothelioma, CD31 is uniformly negative in EPS. Although S-100 protein has also typically
produced negative results, HMB45 has been rarely reported as positive.  Cytogenetic analysis
has revealed no reproducible chromosomal abnormalities characteristic of EPS.
Local recurrence and metastatic rates in EPS are quite high ranging from 38-77% and 40-50%,
As in SS, prolonged follow-up is mandatory in these patients.
Tumor-related death may be observed up to 20 years following initial presentation. A more aggressive
clinical course has been associated with tumor size (>5 cm), proximal location, and histologic
features such as necrosis, mitotic activity, and the presence of vascular invasion. Evans and Baer have
emphasized the initial treatment has having a strong relationship with local recurrence but not affecting
ultimate clinical outcome.  Indeed, regional lymph node metastasis appears to be strongly
correlated with tumor size.
Adult Myxoid Sarcomas
Although many tumors may develop myxoid changes, we will focus only on those soft tissue sarcomas that
are, by definition, defined by and/or predominantly contain a myxoid matrix. For the purposes of our
discussion, this includes, in decreasing order of frequency, myxofibrosarcoma, myxoid/round cell
liposarcoma, and myxoid chondrosarcoma.
Myxofibrosarcoma encompasses a group of low to high grade lesions that in the past were often referred
to as myxoid MFH. It is one of the most common sarcomas of the extremities in elderly people. In
contrast to most soft tissue sarcomas, localization in the dermis and subcutaneous fat is more commonly
observed than within deeper intramuscular locations.
At one end of the
spectrum, the low grade lesions are characteristically multi-lobular, hypocellular and myxoid, and
contain a mostly spindle cell population; at the opposite end are high grade, solid, pleomorphic tumors
with minimal (usually only focal) myxoid stroma. As expected, compared to low grade lesions,
intermediate grade tumors show more cellularity but retain a prominent myxoid matrix. Common to all
grades is the presence of randomly-distributed, thin-walled, curvilinear blood vessels. So-called
pseudolipoblasts, containing 1 or more mucin-filled cytoplasmic vacuoles, may be present, mimicking a
liposarcoma. Local recurrence occurs in up to 50% of cases, regardless of grade. Metastases are
generally observed only in intermediate and high grade tumors.
Recent morphologic and cytogenetic evidence suggests that myxoid and round cell liposarcoma occupy a
spectrum of low and high grade sarcoma, respectively.  Clinically, most patients are middle
aged to older adults and the deep soft tissues of the extremities (especially the thigh) are most
commonly affected. In its pure form, myxoid liposarcoma is characterized by a population of mostly
uniform, hyperchromatic, ovoid to slightly spindled cells within a prominent myxoid stroma and
accompanied by arborizing blood vessels. Signet-ring cell lipoblasts are often observed but
multi-vacuolated lipoblasts tend to be uncommon.  Significant nuclear pleomorphism is not a
feature of "pure" myxoid liposarcoma and should suggest the possibility of myxofibrosarcoma or round cell
differentiation. The latter is characterized by larger cells with uniformly-shaped, round cells with a
vesicular chromatin pattern and conspicuous nucleoli. Occasionally, the round cell nuclei may appear
more epithelioid and show significant nuclear pleomorphism, suggesting the diagnosis of carcinoma.
Mixtures of myxoid and round cell liposarcoma are common and have prognostic significance. In our
series, unfavorable prognostic indicators included age >45 years, round cell differentiation >25%,
and the presence of spontaneous tumor necrosis. 
Extraskeletal Myxoid Chondrosarcoma
Extraskeletal myxoid chondrosarcoma (EMC) is the rarest of this group of tumors. Most occur in the
deep soft tissues of the extremities and afflict middle to older adults. Morphologically, EMC is
abundantly myxoid and multilobular, with individual lobules frequently transversed by fibrous tissue
septa. Most are hypovascular, at least compared to the other myxoid sarcomas. The tumor cells are
classically arranged in anastomosing strands, rings, and nests.  Cytologically, individual
tumor cells are remarkably uniform, round to ovoid. Likewise, nuclei are also round to ovoid with
inconspicuous to prominent nucleoli surrounded by scant amounts of eosinophilic to vacuolated cytoplasm.
A rhabdoid appearance may rarely be observed. Despite the name, well-developed hyaline cartilage is
virtually never seen. Although classically considered a low grade sarcoma, recent evidence with
long-term follow-up suggests local recurrence and metastatic rates approaching 50%.  Adverse
prognostic indicators include older age of the patient, large tumor size, and localization in the
Fine Needle Aspiration Biopsy
Cytologically, myxofibrosarcoma, myxoid liposarcoma, and EMC share the presence of a myxoid stroma.
However, the character of the stroma, degree of cellular atypia, and the arrangement of the tumor cells
allows reliable separation of these neoplasms in the majority of cases. In myxofibrosarcoma, the myxoid
stroma is manifested by a diffuse granular background, covering virtually the entire surface area of the
smear. Among the myxoid sarcomas, the degree of cytologic atypia and nuclear pleomorphism in
myxofibrosarcoma exceeds that typically observed in myxoid liposarcoma and chondrosarcoma. 
Conversely, the latter neoplasms more commonly exhibit distinct myxoid stromal fragments containing a
uniform cell population, usually with minimal to no nuclear pleomorphism. Additionally, the tumor cells
in myxoid chondrosarcoma are usually arranged in cords and anastomosing strands and may even appear
lodged in lacunae, albeit true hyaline cartilage differentiation is rarely observed. 
Soft Tissue Sarcoma Template
| DIAGNOSIS ||Soft tissue mass, anatomic location, procedure:|
|GROSS DESCRIPTION|| |
|Specimen Fixation: ||Fresh or in Formalin|
|Specimen Received: ||Limb-Salvage Resection, Amputation, Intra-Abdominal/Thoracic Resection, Retroperitoneal Resection, and Tumor Debulking|
|Anatomic Depth/Site: ||Dermal, subcutis, intramuscular, etc./Toe, Foot, Leg, Arm, etc.|
|Tumor Dimensions: ||3 Dimensions (cm)|
|Estimation of Tumor Necrosis: ||__ %, usually appear chalky yellow-white; include cystic area volume separately|
|Growth Pattern: ||Infiltrative vs. Well-Circumscribed; Encapsulated vs. Nonencapsulated|
|Surgical Margins: ||Positive vs. Negative; Anatomic site (e.g. anterior, posterior, etc) and distance from margin (cm) if applicable|
|Other Unusual Features: ||Bone involvement, extension into viscera, etc.|
|Block Summary:|| |
|LIGHT MICROSCOPY|| |
|Tumor Histologic Type:|| |
|Tumor Microscopic Size (Largest Dimension):|| |
|Tumor Grade (FNCLCC System): Grade|| |
| Tumor Differentiation Score:|| |
| Tumor Necrosis Score:|| |
| Mitotic Count Score:|| |
| Total Score:|| |
| Percent Necrosis: ____%, (Pretherapy/Posttherapy)|| |
|Lymph Nodes:|| |
|Surgical Margins:|| |
- Enzinger FM, Weiss SW. General Considerations In: Soft Tissue Tumors. 3rd ed. St. Louis, MO: Mosby; 1995:1-16.
- Bennert KW, Abdul-Karim FW. Fine needle aspiration cytology vs. needle core biopsy of soft tissue tumors: a comparison. Acta Cytol 1994;38:381-384.
- The Papanicolaou Society. Guidelines of the Papanicolaou Society of Cytopathology for fine-needle aspiration procedure and reporting. Mod Pathol 1997;10:739-47.
- Kilpatrick SE, Geisinger KR. Soft tissue sarcomas: the usefulness and limitations of fine-needle aspiration biopsy. Am J Clin Pathol 1998;110:50-68.
- Layfield LJ, Anders KH, Glasgow BJ, et al. Fine-needle aspiration of primary soft-tissue lesions. Arch Pathol Lab Med 1986;110:420-424.
- Costa MJ, Campman SC, Davis RL, et al. Fine-needle aspiration cytology of sarcoma: retrospective review of diagnostic utility and specificity. Diagn Cytopathol 1996;15:23-32.
- Hook SY, Jorda M, Mailhot S, et al. Fine needle aspiration cytology of superficial soft tissue lesions: a study of 110 cases (abstract). Acta Cytol 1998;42:1289.
- Kilpatrick SE, Ward WG, Cappellari JO, et al. Fine needle aspiration biopsy of soft tissue sarcomas: a cytomorphologic analysis with emphasis on histologic subtyping, grading, and therapeutic significance. Am J Clin Pathol 1999;112:179-188.
- Kilpatrick SE, Cappellari JO, Bos GD, et al. Is fine needle aspiration biopsy a practical alternative to open biopsy for the primary diagnosis of sarcoma? Experience with 140 patients. Am J Clin Pathol 2001;115:59-68.
- Pollock RE, Karnell LH, Menck HR, et al. The National Cancer data base report on soft tissue sarcoma. Cancer 1996;78:2247-2257.
- Unni KK, Dahlin DC: Grading of bone tumors. Semin in Diagn Pathol 1984;1:165-172.
- Costa J, Wesley RA, Glatstein E, et al. The grading of soft tissue sarcomas: results of a clinicohistopathologic correlation in a series of 163 cases. Cancer 1984;53:530-541.
- Costa J. The grading of soft tissue sarcomas, in Fletcher CDM, McKee PH (eds). Pathobiology of soft tissue tumors. Edinburgh, UK, Churchill Livingston, 1990, pp221-238.
- Trojani M, Contesso G, Coindre JM, et al. Soft tissue sarcomas of adults: study of pathological prognostic variables and definition of a histopathologic grading system. Int J Cancer 1984;33:37-42.
- Guillou L, Coindre JM, Bonichon F, et al. Comparative study of the National Cancer Institute and French Federation of Cancer Centers sarcoma group grading systems in a population of 410 adult patients with soft tissue sarcomas. J Clin Oncol 1997;15:350-362.
- Coindre JM, Terrier P, Guillou L, et al. Predictive value of grade for metastasis development in the main histologic types of adult soft tissuesarcomas: a study of 1240 patients from the French Federation of Cancer Centers Sarcoma Group. Cancer 2001; 91:1914-1926
- Mentzel T, Calonje E, Wadden C, et al. Myxofibrosarcoma: clinicopathologic analysis of 75 cases with emphasis on the low-grade variant. Am J Surg Pathol 1996;20:391-405.
- Ladanyi M. The emerging molecular genetics of sarcoma translocation. Diagn Mol Pathol 1995;4:162-173.
- Fletcher JA. Cytogenetics and molecular biology of soft tissue tumors. In: Weiss SW, Brooks JSJ, eds. Soft Tissue Tumors. Monographs in Pathology; No. 38. Philadelphia, PA: William & Wilkins, 1996;37-64.
- Stevenson AJ, Chatten J, Bertoni F, Miettinen M. CD99 (p30/32 MIC2) neuroectodermal/Ewing's sarcoma antigen as an immunohistochemical marker: review of more than 600 tumors and the literature experience. Appl Immunohistochem 1994;2:231-240.
- Granter SR, Renshaw AA, Fletcher CD, et al. CD99 reactivity in mesenchymal chondrosarcoma. Hum Pathol 1996;27:1273-1276.
- Dei Tos AP, Wadden C, Calonje E, et al. Immunohistochemical demonstration of glycoprotein p30/32 MIC2 (CD99) in synovial sarcoma: a potential cause of diagnostic confusion. Appl Immunohistochem 1995;3:168-173.
- Dehner LP. On trial: a malignant small cell tumor in a child: four wrongs do not make a right. Am J Clin Pathol 1998;109:662-668.
- Miettinen M, Rapola J. Immunohistochemical spectrum of rhabdomyosarcoma and rhabdomyosarcoma-like tumors: expression of cytokeratin and the 68-kD neurofilament protein. Am J Surg Pathol 1989;13:120-132.
- Folpe AL, Patterson K, Gown AM. Antibodies to desmin identify the blastemal component of nephroblastoma. Mod Pathol 1997;10:895-900.
- Thorner P, Squire J, Chilton-McNeill S, et al. Is the EWS/FLI-1 fusion transcript specific for Ewing's sarcoma and peripheral primitive neuroectodermal tumor? A report of four cases showing this transcript in a wider range of tumor types. Am J Pathol 1996;148-1125-1138.
- Sorenson PHB, Shimada H, Liu XF, et al. Biphenotypic sarcomas with myogenic and neural differentiation express the Ewing's sarcoma EWS/FLI1 fusion gene. Cancer Res 1995;55:1385-1392.
- Burchill SA, Wheeldon J, Cullinane C, et al. EWS-FLI-1 fusion transcripts identified in patients with typical neuroblastoma. Eur J Cancer 1997;33:239-243.
- Sainati L, Scapinello A, Montaldi A, et al. A mesenchymal chondrosarcoma of a child with the reciprocal translocation (11;22)(q24;q12). Cancer Genet Cytogenet 1993;71:144-147.
- Ordi J, de Alava E, Torne A, et al. Intraabdominal desmoplastic small round cell tumor with the EWS/ERG fusion transcript. Am J Surg Pathol 1998;22:1026-1032.
- Katz Rl, Quezado M, Senderowicz AM, et al. An intra-abdominal small round cell neoplasm with features of primitive neuroectodermal and desmoplastic round cell tumor and a EWS/FLI-1 fusion transcript. Hum Pathol 1997;28:502-509.
- Stanley MW, Skoog L, Tani EM, Horwitz CA. Nodular fasciitis: spontaneous resolution following diagnosis by fine-needle aspiration. Diagn Cytopathol 1993;9:322-324.
- Fletcher CDM, Akerman M, Cin PD, et al. Correlation between clinicopathological features and karyotype in lipomatous tumors: a report of 178 cases from the Chromosomes and Morphology (CHAMP) Collaborative Study Group. Am J Pathol 1996;148:623-630.
- McGee RS, Ward WG, Kilpatrick SE. Malignant peripheral nerve sheath tumor: a fine needle aspiration biopsy study. Diagn Cytopathol 1997;17:298-305.
- Weiss SW, Langloss JM, Enzinger FM. Value of S-100 protein in the diagnosis of soft tissue tumors with particular reference to benign and malignant Schwann cell tumors. Lab Invest 1983;49:299-308.
- Dehner LP. Primitive neuroectodermal tumor and Ewing's sarcoma. Am J Surg Pathol 1993;17:1
- Kissane JM, Askin FB, Foulkes M, et al. Ewing's sarcoma of bone: clinicopathologic aspects of 303 cases from the intergroup Ewing's sarcoma study. Hum Pathol 1993;14:773-779.
- Nascimento AG, Cooper KL, Unni KK, et al. A clinicopathologic study of 20 cases of large-cell (atypical) Ewing's sarcoma of bone. Am J Surg Pathol 1980;4:29-36.
- Renshaw AA, Perez-Atayde AR, Fletcher JA, Granter SR. Cytology of typical and atypical Ewing's sarcoma / PNET. Am J Clin Pathol 1996;106:620-624.
- Perlman EJ, Dickman PS, Askin FB, et al. Ewing's sarcoma – routine diagnostic utilization of MIC2 analysis: a Pediatric Oncology Group/Children's Cancer Group Intergroup Study. Hum Pathol 1994;25:304-307.
- Newton WA Jr., Gehan EA, Webber BL, et al. Classification of rhabdomyosarcomas and related sarcomas, pathologic aspects and proposal for a new classification -- an Intergroup Rhabdomyosarcoma Study. Cancer 1995;76:1073-1085.
- Qualman SJ, Coffin CM, Newton WA, et al. Intergroup Rhabdomyosarcoma Study: update for pathologists. Ped Develop Pathol 1998;1:550-561.
- Tsokos M, Webber BL, Parham DM, et al. Rhabdomyosarcoma: a new classification scheme related to prognosis. Arch Pathol Lab Med 1992;116:847-855.
- Ahktar M, Ali MA, Bakry M, et al. Fine needle aspiration biopsy diagnosis of rhabdomyosarcoma: cytology, histologic, and ultrastructural correlations. Diagn Cytopathol 1992;8:465-474.
- Ordonez NG. Desmoplastic small round cell tumor: a histopathologic study of 39 cases with emphasis on unusual histological patterns. Am J Surg Pathol 1998;22:1303-1313.
- Fletcher CDM. Pleomorphic malignant fibrous histiocytoma: fact or fiction? A critical appraisal based on 159 tumors diagnosed as pleomorphic sarcoma. Am J Surg Pathol 1992;16:213-228.
- Fletcher CD, Gustafson P, Rydholm A, et al. Clinicopathologic re-evaluation of 100 malignant fibrous histiocytomas: prognostic relevance of subclassification. J Clin Oncol 2001;19:3045-3050.
- Bergh P, Meis-Kindblom JM, Gherlinzoni F, et al. Synovial sarcoma: identification of low and high risk groups. Cancer 1999;85:2596-2607.
- Folpe AL, Schmidt RA, Chapman D, Gown AM. Poorly differentiated synovial sarcoma: immunohistochemical distinction from primitive neuroectodermal tumors and high-grade malignant peripheral nerve sheath tumors. Am J Surg Pathol 1998;22:673-682.
- Kilpatrick SE, Teot LA, Stanley MW, et al. Fine-needle aspiration biopsy of synovial sarcoma: a cytomorphologic analysis of primary, recurrent, and metastatic tumors. Am J Clin Pathol 1996;106:769-775.
- Smith TA, Machen SK, Fisher C, Goldblum JR. Usefulness of cytokeratin subsets for distinguishing monophasic synovial sarcoma from malignant peripheral nerve sheath tumor. Am J Clin Pathol 1999;112:641-648.
- Kawai A, Woodruff J, Healey JH, et al. SYT-SSX gene fusion as a determinant of morphology and prognosis in synovial sarcoma. N Engl J Med 1998;338:153-160.
- Inagaki H, Nagasaka T, Otsuka T, et al. Association of SYT-SSX fusion types with proliferative activity and prognosis in synovial sarcoma. Mod Pathol 2000;13:482-488.
- Nilsson G, Wang M, Wejde J, et al. Reverse transcriptase polymerase chain reaction on fine needle aspirates for rapid detection of translocations in synovial sarcoma. Acta Cytol 1998;42:1317-1324.
- Chase DR, Enzinger FM. Epithelioid sarcoma: diagnosis, prognostic indicators, and treatment. Am J Surg Pathol 1985;241-263.
- Evans HL, Baer SC. Epithelioid sarcoma: a clinicopathologic and prognostic study of 26 cases. Semin Diagn Pathol 1993;10:286-291.
- Guillou L, Wadden C, Coindre JM, et al. "Proximal-type" epithelioid sarcoma, a distinctive aggressive neoplasm showing rhabdoid features. Am J Surg Pathol 1997;21:130-146.
- Arber DA, Kandalaft PL, Mehta P, Battifora H. Vimentin-negative epithelioid sarcoma: the value of an immunohistochemical panel that includes CD34. Am J Surg Pathol 1993;17:302-307.
- Kilpatrick SE, Ward WG. Myxofibrosarcoma of soft tissues: A cytomorphologic analysis of a series. Diagn Cytopathol 1999;20:6-9.
- Kilpatrick SE, Doyon J, Choong PFM, Sim FH, Nascimento AG. The clinicopathologic spectrum of myxoid and round cell liposarcoma: a study of 95 cases. Cancer 1996;77:1450-8.
- Meis-Kindblom JM, Bergh P, Gunterberg B, Kindblom LG. Extraskeletal myxoid chondrosarcoma: a reappraisal of its morphologic spectrum and prognostic factors based on 117 cases. Am J Surg Pathol 1999;23:636-650.
- Kilpatrick SE, Ward WG, Bos GD. The value of fine needle aspiration biopsy in the differential diagnosis of adult myxoid sarcoma. Cancer (Cancer Cytopathol) 2000;90:167-177.