The diagnosis of a soft tissue lesion requires a modicum of clinical information and adequate, well-processed tissue. At a minimum, the pathologist should be apprised of patient age, tumor location, and its growth characteristics. In some cases, the results of imaging studies, particularly magnetic resonance imaging (MRI), enhance one’s understanding of the clinical extent of the lesion and its relationship to normal structures (see Chapter 3 ).
Although age rarely, if ever, suggests a particular diagnosis, it is important to know whether the patient is a child. In general, there is little overlap between soft tissue tumors occurring in children and those in adults. Therefore, this critical piece of information essentially presents the pathologist with two groups of tumors from which a differential diagnosis can be constructed. For example, undifferentiated pleomorphic sarcoma (formerly known as malignant fibrous histiocytoma, or MFH) is essentially unheard of during childhood, so one should consider other diagnoses for a pleomorphic tumor in a child. On the other hand, neuroblastoma virtually never occurs after childhood, and such diagnoses should always be made cautiously in adults.
Location also often provides direction in the differential diagnosis. Sarcomas, for the most part, develop as deeply located masses and infrequently present as superficial lesions. Exceptions do occur, however, and include lesions such as dermatofibrosarcoma protuberans (DFSP), epithelioid sarcoma, Kaposi sarcoma, and angiosarcoma ( Box 5.1 ). It is also useful to recall that when carcinomas or melanomas metastasize to soft tissue, they do so usually as small, superficial nodules rather than as large, deeply situated masses. The most common carcinomas that present as soft tissue metastases are pulmonary and renal carcinomas, the former usually appearing as a subcutaneous mass on the chest wall and the latter as a soft tissue mass in almost any location.
Angiomatoid fibrous histiocytoma
Plexiform fibrohistiocytic tumor
CD34-positive fibroblastic tumor
Unfortunately, there is a great overlap in the clinical presentation of benign and malignant soft tissue masses, so this information may be least helpful to the pathologist. Most soft tissue sarcomas of the extremities are detected by the patient as a slowly growing mass that has been present for 6 months or more at the time of diagnosis. The duration of benign lesions may be similar, although such lesions are generally described as static or slowly growing. An exception is the rapid development of some cases of nodular fasciitis and related lesions. These superficial, pseudosarcomatous, benign lesions may develop rapidly over 1 to 3 weeks, and we have even encountered some that evolved in a few days, a pattern of growth that seldom, if ever, is encountered with a sarcoma. Therefore, an astute general surgeon can sometimes suggest the diagnosis of fasciitis for a rapidly evolving superficial lesion of the extremity.
In the past, the choice of biopsy technique for soft tissue masses was dictated by the size and location of the lesion (see Chapter 2 ). An incisional biopsy was considered the gold standard for large, deeply situated masses and provided ample material for diagnosis and ancillary studies. Its principal disadvantages included spillage of tumor into adjacent compartments as a result of poor hemostasis or faulty biopsy placement, complications of wound infection, and the usual requirement for hospitalization of the patient. Excisional biopsy, although more expedient and providing the entire lesion for examination, was performed on only small, superficial lesions amenable to complete resection. The current reliance on minimally invasive techniques to procure tissue has significantly changed the biopsy paradigm in the direction of core-needle biopsy (CNB). Use of CNB increased from less than 10% to almost 80% of tumors during the early 1990s and now is performed in our hospitals on essentially all deep soft tissue masses. Consequently, the amount of material available to type and grade sarcomas has significantly decreased, and this trend is likely to continue unabated because of the emphasis on less costly outpatient care. Therefore, it is important to be aware of the limitations and pitfalls of CNB and to keep in mind a few basic principles.
First, the pathologist should be aware of the expectation of the clinician. In some cases the goal of a CNB may be simply to establish that a soft tissue mass is a mesenchymal neoplasm as opposed to a lymphoma or metastatic lesion, a distinction that can usually be made in the majority of cases either morphologically or with judicious use of immunohistochemistry. If definitive surgery will be performed following the needle biopsy, then the most important priority is to determine whether the lesion is a sarcoma. If, however, the intention is to provide preoperative (neoadjuvant) radiotherapy or chemotherapy, every attempt should be made not only to make the diagnosis of sarcoma, but also to specifically classify and grade the lesion. However, as discussed in Chapter 1 , it is not always possible to grade a sarcoma reliably on the basis of a CNB. In particular, it is difficult to discriminate a grade 2 from a grade 3 lesion. Pathologists may find that the best assessment that they can give is the designation low grade or high grade, recognizing that high grade will encompass both grade 2 and 3 lesions. Information is inevitably lost when collapsing a three-tiered system into a two-tiered one; however, a two-tiered system still performs reasonably well and is consistent with therapeutic considerations.
In grading CNBs, one can usually accept the presence of high-grade areas in several needle cores as diagnostic of a high-grade sarcoma, because of the improbability that additional material will result in downgrading the lesion. At the same time, one should also be unwilling to accept a lesion as low grade if the number of core biopsies is small, if the lesion has not been adequately sampled, or if imaging studies suggest features of a high-grade sarcoma (i.e., necrosis). CNBs containing necrosis usually imply a high-grade sarcoma, again because of the improbability that limited material captures a solitary or limited focus of necrosis. However, pathologists must be certain that necrosis is of the coagulative and not the hyaline type and is not reflective of prior therapy or surgical intervention. The corollary to the latter portion of this statement is that as soon as radiation or chemotherapy has been administered, grading becomes unreliable because of alterations in nuclear features, mitotic activity, cellularity, and interstitial hyalinization. Therapy also induces necrosis, although it is not possible to discriminate spontaneous from therapy-induced necrosis. Most important, interpreting CNBs necessitates a close dialogue with clinicians to resolve any inconsistencies between clinical and pathologic diagnoses.
Frozen Section Diagnosis
In the past, frozen section examinations were performed with the expectation that definitive surgery would be accomplished during the same intraoperative procedure, but this trend has changed. Frozen sections are now obtained primarily to assure the surgeon that representative, viable tissue has been obtained that is adequate for a permanent section diagnosis or to evaluate margins. The former may be accomplished by freezing a portion of the biopsy material or sometimes, as in the case of a needle biopsy, by performing a touch preparation. The presence of malignant cells in a nonnecrotic background on a touch preparation usually ensures that the specimen is adequate. A background of reactive or necrotic cells suggests that a biopsy has been performed on the pseudocapsule, or that the specimen is largely necrotic, requiring additional material depending on the clinical impression.
Evaluation of Resection Specimens
Because the preference is to perform limb-sparing surgery for sarcomas of all types, if possible, the number of major amputation specimens received in the surgical pathology laboratory has greatly decreased. Most extremity sarcomas are removed with wide local excision, usually combined with preoperative or postoperative radiotherapy. Ideally, such specimens are received fresh and unfixed so that tissue for ancillary studies can be obtained. As with many other surgical specimens, the margins should be marked with permanent ink and blotted dry before the dissection of the specimen. After the incision, the gross characteristics of the tumor should be noted. If malignancy is suspected, a careful assessment of the tumor as to its surroundings is mandatory. This includes the location of the lesion (e.g., subcutis, muscle), its size, its relation to vital structures (e.g., bone, neurovascular bundle), and the relative amount of necrosis, if it can be judged grossly. Size is important for providing an accurate T descriptor for the surgeon if the lesion is a sarcoma. Lesions less than 5 cm are classified as T1, whereas those larger than 5 cm are classified as T2. An assessment of the degree of necrosis is important for untreated sarcomas because this parameter is used in most grading systems, including the French (FNCLCC) system. The extent of necrosis in lesions treated with preoperative irradiation or chemotherapy is also important to note because it helps the clinician assess the efficacy of therapy, although it does not carry the same implication as necrosis in an untreated lesion. Often, the gross appearance of the tumor is deceptive. Sarcomas may appear well circumscribed, and benign tumors occasionally infiltrate. Use of the term encapsulation can be misleading and can invite inadequate excision by shelling out, or enucleating, the tumor. In reality, sarcomas lack a true capsule and instead are surrounded by a compressed zone of normal tissue, known as a pseudocapsule .
There are no rigid guidelines for sampling soft tissue tumors; to some extent, sampling is dictated by the specific case. In the case of a known benign lesion, a few representative sections suffice (or the entire lesion if it is small). With a sarcoma, the questions to be answered are different. For example, it may be less important to submit numerous sections for a high-grade sarcoma than for a low-grade lesion in which the sampling is being driven by the need to rule out the presence of high-grade areas. We generally obtain one section for each centimeter of tumor diameter, with no more than about 10 sections if the lesion appears more or less uniform. Representative sections of the margins or sections designed to show impingement of vital structures are also obtained. We select blocks for margins judiciously, depending on the gross appearance of the lesion. Lesions several centimeters away from a margin seldom have positive margins microscopically, so extensive margin sampling in these situations is less critical than with excisions containing grossly close margins. One notable exception is epithelioid sarcoma, a lesion that may be grossly deceptive in its clinical extent. Digital images can be useful for providing visual data as to the orientation of the specimen and sampling sites.
Most specimens are handled adequately as described previously. In diagnostically challenging cases, it is useful to have frozen tissue in reserve if ancillary studies are required. Even though the ability to perform molecular tests on formalin-fixed material has significantly diminished the need to have fresh tissue on every case, freezing tissue is an excellent practice, particularly because frozen tissue is often a requirement for patient entry into national protocols (e.g., childhood rhabdomyosarcoma). Approximately 1 cm 3 is sufficient, although less is acceptable from small biopsy specimens. The tissue should be cut into 0.2-cm fragments, stored at −70° C, and shipped on dry ice, if necessary.
The first and most important step in reaching a correct diagnosis is careful scrutiny of conventionally stained sections at low-power magnification. Useful microscopic features that can be identified at this point include the size and depth of the lesion, its relation to overlying skin and underlying fascia, and the nature of the borders (e.g., pushing, infiltrative).
The most important decision to be made initially is whether the lesion is reactive or neoplastic. Reactive lesions occur in either superficial or deep soft tissue but tend to occur more frequently in superficial tissue ( Box 5.2 ). A number of histologic features suggest a reactive process. First, many reactive lesions display a zonation. For example, in the case of fascial forms of nodular fasciitis and ischemic fasciitis, there is a cuff of proliferating fibroblasts that surround a central hypocellular zone of fibrinoid change. Myositis ossificans also displays zonation, manifested by centrifugal maturation of fibroblastic to osteoblastic mesenchyme. Cells comprising reactive lesions often have the appearance of tissue culture fibroblasts, with large vesicular nuclei, prominent nucleoli, and striking cytoplasmic basophilia, reflecting the presence of abundant, rough endoplasmic reticulum. Although mitotic figures may be numerous, important negative observations include the absence of atypical mitotic figures or nuclear atypia, as one would expect in a sarcoma.
Intravascular and cranial fasciitis
Ischemic fasciitis (atypical decubital fibroplasia)
Proliferative fasciitis and myositis
Intravascular papillary endothelial hyperplasia
Myositis and panniculitis ossificans
Fibrodysplasia ossificans progressiva
Fibroosseous pseudotumor of the digits
Localized massive lymphedema
When satisfied that a reactive lesion can be excluded, the pathologist proceeds with analysis of the neoplasm. At low power, the architectural pattern, appearance of the cells, and characteristics of the stroma should be noted. These characteristics lend themselves to the development of a number of differential diagnostic categories, as follows:
Fasciculated spindle cell tumors. These lesions comprise a large group of tumors (e.g., adult-type fibrosarcoma, malignant peripheral nerve sheath tumor, synovial sarcoma) characterized by spindled cells arranged in long fascicles ( Box 5.3 ). Cellular schwannoma and fibromatosis must be distinguished from the others because they are nonmetastasizing tumors. Unlike the others, fibromatosis is typically a lesion of low cellularity and nuclear grade. Cellular schwannoma, unlike the others, is characterized by diffuse, intense S-100 protein immunoreactivity.
Fibromatosis (desmoid tumor)
Fibrosarcomatous dermatofibrosarcoma protuberans
Spindle cell rhabdomyosarcoma
Synovial sarcoma (monophasic fibrous type)
Malignant peripheral nerve sheath tumor
Low-grade fibromyxoid sarcoma
Myxoid lesions ( Box 5.4 ). Although almost any soft tissue tumor may appear myxoid from time to time, a number of lesions display myxoid features consistently. In adults the differential diagnosis of myxoid tumors includes myxoma, myxofibrosarcoma, myxoid liposarcoma, and extraskeletal myxoid chondrosarcoma. Analysis of the vascular pattern, degree of nuclear atypia, and organization of the cells aid in this distinction. For example, an intricate vasculature is a feature of both myxoid liposarcoma and myxofibrosarcoma, but it is not a feature of extraskeletal myxoid chondrosarcoma or myxoma.
Myxoma, cutaneous, intramuscular
Myxoinflammatory fibroblastic sarcoma
Nerve sheath myxoma
Spindle cell lipoma
Low-grade fibromyxoid sarcoma
Ossifying fibromyxoid tumor
Extraskeletal myxoid chondrosarcoma
Dermatofibrosarcoma protuberans, myxoid
Botryoid embryonal rhabdomyosarcoma
Myxoid leiomyosarcoma (rare)
Epithelioid tumors ( Box 5.5 ). For the differential diagnosis of epithelioid tumors, it is important to rule out metastatic carcinoma, melanoma, and even large-cell lymphomas before assuming that the mass is an epithelioid soft tissue tumor. Immunohistochemistry (IHC) plays a decidedly pivotal role in this regard (see Chapter 6 ).
Alveolar soft part sarcoma
Gastrointestinal stromal tumor
Epithelioid malignant peripheral nerve sheath tumor
Malignant rhabdoid tumor
Perivascular epithelioid cell tumor (PEComa)
Sclerosing epithelioid fibrosarcoma
Synovial sarcoma (biphasic and predominantly monophasic epithelial)
Round cell tumors ( Box 5.6 ). As with epithelioid lesions, the differential diagnosis of round cell lesions can be broad; it presupposes excluding non–soft tissue lesions that may mimic a round cell sarcoma (e.g., lymphoma, carcinoma) and is greatly facilitated by the use of IHC and increasingly, molecular genetic testing. One should also bear in mind that the round cell tumor is not synonymous with “round cell sarcoma” because benign lesions (e.g., glomus tumor, giant cell-poor forms of tenosynovial giant cell tumor) also enter the differential diagnosis. In general, patient age helps narrow the possibilities. In children these lesions include neuroblastoma, alveolar rhabdomyosarcoma, Ewing sarcoma, and the rare desmoplastic small round cell tumor.