Trauma is frequently implicated in the development of sarcomas. Many of these reports are anecdotal, however, and the integrity of the injured part was not clearly established before the injury. Consequently, trauma often seems to be an event that merely calls attention to the underlying neoplasm. Rare soft tissue sarcomas have been reported arising in scar tissue following surgical procedures or thermal or acid burns, at fracture sites, and in the vicinity of plastic or metal implants, usually after a latent period of several years. Kirkpatrick et al. studied the histologic features in capsules surrounding the implantation site of a variety of biomaterials. Interestingly, they noted a spectrum of changes ranging from focal proliferative lesions through preneoplastic proliferations to incipient sarcomas and suggested a model of multistage tumorigenesis similar to the adenoma-carcinoma sequence.

Environmental carcinogens have been related to the development of sarcomas, but their role is largely unexplored, and only a few substances have been identified as playing a role in the induction of sarcomas in humans. A variety of animal models now exist to induce sarcomas, allowing a better understanding of their pathogenesis.

Phenoxyacetic acid herbicides, chlorophenols, and their contaminants, such as 2,3,7,8-tetrachlorodibenzo-para-dioxin (dioxin), have been linked to sarcomagenesis. A series of case-control studies from Sweden from 1979 to 1990 reported an up to six-fold increased risk of soft tissue sarcoma associated with exposure to phenoxyacetic acids or chlorophenols in individuals exposed to these herbicides in agricultural or forestry work. Similar reports of an increased risk of sarcoma associated with these herbicides came from Italy, Great Britain, and New Zealand. Although a study by Leiss and Savitz linked the use of phenoxyacetic acid lawn pesticides with soft tissue sarcomas in children, other studies with more detailed exposure histories did not confirm this association. These inconsistencies may be caused in part by the predominant phenoxyacetic herbicide used in different locations. In the United States, 2,4-dichlorophenoxyacetic acid is the primary phenoxyacetic herbicide used, whereas in Sweden the main herbicides contain 2,4,5-trichlorophenoxyacetic acid and 2-methyl-4-chlorophenoxyacetic acid, both of which are more likely contaminated with dioxin. ^, High levels of dioxin exposure from accidental environmental contamination near Seveso, Italy, from an explosion at a chemical factory was followed by a three-fold increased risk of soft tissue sarcomas among individuals living near the factory. Similarly, Collins et al. found a significantly higher risk of soft tissue sarcomas in trichlorophenol workers in Midland, Michigan, who were exposed to 2,3,7,8-tetrachlorodibenzo- p -dioxin. In addition, the possibility of an increased incidence of sarcomas was claimed for some of the 2 million soldiers stationed in Vietnam between 1965 and 1970 who were exposed to Agent Orange, a defoliant that contained dioxin as a contaminant. ^, However, in several case-control and proportional mortality studies, no excess risk of soft tissue sarcoma was reported among those Vietnam veterans who were directly involved with the spraying of Agent Orange.

Vinyl chloride exposure is clearly associated with the development of hepatic angiosarcoma. ^, There are also rare reports of extrahepatic angiosarcoma associated with this agent.

Radiation exposure is related to the development of sarcomas, but considering the frequency of radiotherapy, radiation-induced soft tissue sarcomas are actually quite uncommon. The incidence of postradiation sarcoma is difficult to estimate, but reports generally range from 0.03% to 0.80%. ^, Much of the data regarding the incidence of postradiation sarcomas are derived from large cohorts of breast cancer patients treated with postoperative radiation therapy. ^, To qualify as a postradiation sarcoma, there must be documentation that the sarcoma developed in the irradiated field, a histologic confirmation of the diagnosis, a period of latency of at least 3 years between irradiation and the appearance of a tumor, and documentation that the region bearing the tumor was normal before the administration of the radiation. ^, Almost all postradiation sarcomas occur in adults, and women develop these tumors more frequently, an observation that might reflect the common use of radiation for the treatment of breast and gynecologic malignancies.

Postradiation sarcomas do not display the wide range of appearances associated with sporadic non–radiation-induced tumors. The most common postradiation soft tissue sarcoma is undifferentiated pleomorphic sarcoma, which accounts for almost 70% of cases. Unfortunately, most postradiation sarcomas are high-grade lesions and are detected at a relatively higher stage than their sporadic counterparts. Therefore, the survival rate associated with these lesions is quite poor.

The prognosis of postradiation sarcomas is most closely related to the anatomic site, which in turn probably reflects resectability. Patients with radiation-induced sarcomas of the extremities have the best survival (approximately 30% at 5 years), whereas those with lesions arising in the vertebral column, pelvis, and shoulder girdle generally have survival rates of less than 5% at 5 years. ^,

The total dose of radiation seems to influence the incidence of postradiation sarcoma; most are reported to occur at doses of 5000 cGy or more. Mutations of the TP53 gene have been implicated in the pathogenesis of these tumors. Extravasated Thorotrast (thorium dioxide), although no longer used for diagnostic or therapeutic purposes, has induced soft tissue sarcomas, particularly angiosarcomas, at the site of injection. ^,

The role of oncogenic viruses in the pathogenesis of soft tissue sarcomas is still poorly understood, although there is convincing evidence that the human herpesvirus 8 (HHV8) is the causative agent of Kaposi sarcoma. ^, In addition, a large body of literature supports the role of the Epstein–Barr virus (EBV) in the pathogenesis of smooth muscle tumors in patients with immunodeficiency syndromes or following therapeutic immunosuppression in the transplant setting. ^, Aside from these settings, there is no conclusive evidence that human-transmissible viral agents constitute a major risk factor in the development of soft tissue sarcomas.

As mentioned previously, immunodeficiency and therapeutic immunosuppression are associated with the development of soft tissue sarcomas, particularly smooth muscle tumors and Kaposi sarcoma. In addition, acquired regional immunodeficiency, or loss of regional immune surveillance, may play a central role in the development of the relatively rare angiosarcomas that arise in the setting of chronic lymphedema, ^, secondary to radical mastectomy ^, or congenital or infectious conditions.

A number of genetic diseases are associated with the development of soft tissue tumors, and the list will undoubtedly lengthen as we continue to further understand the molecular underpinnings of mesenchymal neoplasia. Neurofibromatosis types 1 and 2 and familial adenomatous polyposis (FAP, Gardner syndrome) are classic examples of genetic diseases associated with soft tissue tumors. Familial cancer syndromes associated with soft tissue sarcomas are more fully described in Chapter 4 .

Grading of soft tissue sarcomas was first proposed in 1939 by Broders. He suggested that fibrosarcomas could be divided into several subtypes (fibrous, fibrocellular, and cellular), and that those that were highly cellular should be considered grade 4 regardless of the level of mitotic activity. These principles persist in current grading systems, in that certain parameters (e.g., mitotic activity) should be evaluated in sarcomas, some histologic subtypes a priori dictate a grade, and the level of differentiation must be factored into the assignment of a grade. Over the ensuing decades, numerous studies reaffirmed the importance of grading and emphasized the primacy of necrosis and mitotic activity in assessing a grade. Some studies have further proposed the use of Ki-67 immunoreactivity or MIB-1 score/index to accurately assess mitotic activity.

The first large-scale effort to grade and stage sarcomas occurred in 1977 when Russell et al., using a database of 1000 cases and the tumor-node-metastasis (TNM) staging system, showed that incorporating grade into the staging system improved prediction of outcome. Most important, in the absence of metastatic disease, grade essentially defined the clinical stage . This study is most often cited as providing the first reliable grading system in the United States, yet paradoxically, it did not provide objective criteria for grading. Rather, the grade was determined by a panel of experts based on their years of experience. The paper’s real contribution to grading was the implied concept that certain histologic types of sarcomas were inherently low grade and others were high grade, a premise of many subsequent grading systems.

Following that seminal publication, many grading systems were published internationally, including one from the U.S. National Cancer Institute (NCI). Although differing in emphasis, most relied on mitotic activity and necrosis in deriving a grade, and some proposed that sarcoma-specific parameters should be used. The number of grades varies among the systems, ranging from two to four. Three-grade systems seem best suited for predicting patterns for survival and a likely response to therapy ( Fig. 1.1 ). Four-grade systems usually show little difference between the two lowermost grades; two-grade systems, which distinguish between only low-grade and high-grade sarcomas, are more readily related to the type of surgical therapy but make it difficult to deal with sarcomas that lie between these two extremes.

Grading system for soft tissue sarcomas based on three grades of malignancy.

From Myhre Jensen O, Kaae S, Madsen EH, et al. Histopathological grading in soft tissue tumours: relation to survival in 261 surgically treated patients. Acta Pathol Microbiol Immunol Scand. 1983;91A:145.

The French system published by Trojani et al. in 1984 was developed by the French Federation of Cancer Centers Sarcoma Group (FNCLCC), based on an analysis of 155 adult patients with soft tissue sarcomas. On the basis of a multivariate analysis of histologic features, a combination of cellular differentiation, mitotic index, and extent of tumor necrosis was determined to be the most useful parameters for sarcoma grading. This system assigns a score to each parameter and adds the scores together for a combined grade ( Table 1.2 ). This study concluded that histologic grade was the most important factor for predicting survival rates; tumor depth (superficial vs. deep) was another important prognostic parameter. The reproducibility of this system was tested by 15 pathologists; an agreement was reached in 81% of the cases for tumor necrosis, 74% for tumor differentiation, 73% for mitotic index, and 75% for overall tumor grade, although the agreement as to histologic type was only 61%. Importantly, this reproducibility study was performed in the early 1980s before more sophisticated immunohistochemical and molecular genetic techniques, which would have improved the agreement on histologic type.

Although the French system relies on a balanced evaluation of parameters (differentiation score, mitotic index, extent of necrosis), its principal weakness lies in the assignment of the differentiation score. The differentiation score is defined as the extent to which a tumor resembles adult mesenchymal tissue (score 1), the extent to which the histologic type is known (score 2), or the observation that the tumor is undifferentiated (score 3). Although a listing of the differentiation scores for the common tumors has been reported ( Table 1.3 ), the rationale for some of these scores is not clear. It must also be remembered that this system has been derived from resected specimens unmodified by treatment, a situation that is not analogous to our current practices, which are heavily weighted toward grading on biopsies or on resection specimens following preoperative radiation or chemotherapy.

Despite these issues, the French system is the most widely used grading system for sarcomas throughout the world. In a study of soft tissue pathologists from 30 countries, more preferred the French system (37.3%) over the NCI (24%), Broders’ criteria (12%), Markhede’s system (1.3%), and other systems (15.3%), probably because it is more precisely defined and therefore reproducible. It also performed superior to the NCI system in a large dataset comparison. Guillou et al. compared the two systems in adult patients with nonmetastatic soft tissue sarcomas. By a univariate analysis, both systems were of prognostic value for predicting metastasis and overall survival. By a multivariate analysis, tumor size of 10 cm or more, deep location, and high tumor grade, regardless of the system used, were found to be independent prognostic factors for predicting metastases. Interestingly, there were grade discrepancies using these two grading systems in 34.6% of cases. Use of the FNCLCC system resulted in an increased number of grade 3 tumors, a reduced number of grade 2 tumors, and a better correlation with overall and metastasis-free survival compared with results from the NCI system. It is recognized that tumor grade is the most important prognostic factor for adult soft tissue sarcomas and is a critical component of the AJCC eighth edition staging system. In a study of 1240 patients with localized soft tissue sarcomas, 5-year metastasis-free survival rates were 91%, 71%, and 43% for grades 1, 2, and 3, respectively.

Despite the widespread use of some form of grading system in the diagnosis and management of sarcomas, experts agree that no grading system performs well on every type of sarcoma. ^{, ,} There are several reasons for this. In the most obvious situation, there are sarcomas in which the histologic subtype essentially defines behavior, and therefore grade becomes redundant. This is best illustrated by a well-differentiated liposarcoma (atypical lipomatous tumor), an inherently low-grade, nonmetastasizing lesion, and the majority of round cell sarcomas (e.g., alveolar rhabdomyosarcoma), which are inherently high grade.

Also problematic are the rare sarcomas that are considered difficult, if not impossible, to grade. Epithelioid sarcoma, clear cell sarcoma, and alveolar soft part sarcoma are the most frequently cited examples of ungradable sarcomas, yet it is difficult to find a cogent explanation for this long-standing bias in the literature. It is possible that our grading systems fail to capture the correct histologic information in grading these rare sarcomas, or compared to other sarcomas, nonhistologic factors may be much more influential in predicting outcome than histologic factors. Clearly, however, there is a substantial risk of distant metastasis in the long term, whereas in the short term (5 years), the interval for which traditional grading systems are most accurate, the risk may be low. Therefore, the assignment of grade to these tumors does not guarantee biologic equivalency to other sarcomas of comparable grade.

In a number of sarcomas, clinical features play a larger role in determining a prognosis. Cutaneous angiosarcomas are usually ungraded because multifocality and size are more predictive of outcome; paradoxically, angiosarcomas of deep soft tissue are probably amenable to grading. The difficulty of grading synovial sarcomas by histologic features has been noted in many studies, leading Bergh et al. to stratify synovial sarcoma into low- and high-risk groups using a combination of age, size, and presence or absence of poorly differentiated areas. Extraskeletal myxoid chondrosarcoma, long considered a low-grade lesion histologically, has late metastasis in approximately 40% of cases. By stratifying lesions by age, distal versus proximal location, and grade, Meis-Kindblom et al. were able to predict clinical outcome.

Leiomyosarcomas present another interesting paradigm. Various studies present conflicting views as to the predictive power of grading these tumors, and there is some evidence that, as a group, myogenic tumors have a worse prognosis when matched for other variables. The reasons for this are not clear. Even among sarcomas that have traditionally been graded, we have begun to recognize the limitations of grading. For example, malignant peripheral nerve sheath tumors have customarily been graded, but the FNCLCC has indicated that an assigned grade does not predict metastasis.

Strictly speaking, these models are not grading systems because they incorporate histologic, clinical, and demographic variables. Nonetheless, their use suggests we may gradually move in the direction of sarcoma-specific analyses, which may be used in conjunction with or, in some cases, instead of grade. The advantage of such an approach is that it allows the most appropriate criteria to be used for each sarcoma type, theoretically to improve the ability to prognosticate. The disadvantage of this approach is that it presupposes an inordinate amount of clinical information for each sarcoma type, a challenge considering the rarity of some subtypes of these tumors. Moreover, the more specific these systems become, the more complicated they also become.

Another means of integrating clinical and pathologic data in a manner that accounts for sarcoma subtype is the use of nomograms. This method collates multiple clinical and histologic parameters in a given patient and compares the data against a large population of patients with similar parameters whose outcome is known. A nomogram for a 12-year sarcoma-specific mortality has been devised by the Memorial Sloan-Kettering Cancer Center (MSKCC). ^, Other sarcoma-specific nomograms can be found on the MSKCC website ( https://www.mskcc.org/nomograms/sarcoma ).

Despite these limitations, grading remains one of the most powerful and inexpensive ways of assessing the prognosis in a sarcoma and is currently regarded as a major independent predictor of metastasis in the major histologic types of adult soft tissue sarcomas. Consequently, a grade should be provided by the pathologist, whenever possible, although it should not be considered a substitute for an accurate histologic diagnosis. Grading, as with diagnosing soft tissue sarcomas, requires representative, well-fixed, well-stained histologic material that should be obtained before neoadjuvant therapy, because this process alters many of the features necessary for accurate grading. Thick or heavily stained sections are misleading because they may suggest less cellular differentiation than is actually present. Selection of the tissue sample and the length of fixation may also influence the degree of necrosis and the mitotic index. Necrosis may be prominent in tumors for which a biopsy has been previously performed or that have been irradiated or embolized and therefore cannot be accurately assessed in these situations. Grading is usually based on the least differentiated area of a tumor, unless it comprises a very minor component of the overall tumor.