Sample Chapter

June 2007

Section 3 Breast, Skin and Soft Tissue

Soft tissue sarcoma (STS) is a collective term for an unusual and diverse group of malignancies that arise from cells of the embryonic mesoderm. Although tissues derived from the mesoderm contain approximately 75% of the cells in the human body, sarcomas represent only 1% of adult tumors and 15% of pediatric tumors. Sarcomas may occur anywhere in the body and comprise more than 50 distinct histologic subtypes. Approximately 43% of STSs occur in the extremities, 15% in the retroperitoneum, 10% in the trunk, 19% in the viscera, and 13% in other locations.¹ In addition, some sarcomas occur in the GI tract. It has been estimated that in 2007, there will be approximately 9,220 cases of sarcoma in the United States, with 3,560 deaths.² STS-related mortality has been quite constant over the years, indicating that relatively little progress has been made in the treatment of most sarcomas.

The etiology of STS is unclear and somewhat controversial. Genetic factors, irradiation, chemical exposure, and lymphedema have all been shown to have a strong correlation with the evolution of these lesions.

Chromosomal abnormalities associated with STS may be due to translocations, point mutations, or deletions resulting in balanced or unbalanced karyotypes. Mutations in regulatory genes, especially p53 and RB1, are a common finding.^3,4 In particular, persons with Li-Fraumeni syndrome, familial retinoblastoma, Werner syndrome, Gardner syndrome, tuberous sclerosis, basal cell nevus syndrome, or neurofibromatosis have an increased incidence of STS (and of other tumors as well).¹

Ionizing radiation has long been recognized as a factor predisposing to the development of osteosarcoma and STS. It may exert this effect at doses as low as 8 Gy. Sarcomas arising from radiation exposure often do not become clinically apparent until long after the inciting exposure; consequently, they are frequently associated with definitive radiation treatments that lead to a long expected survival, such as may be provided in the setting of breast cancer, cervical cancer, lymphoma, or childhood malignancies. The most common of the sarcomas linked with radiation exposure are osteosarcoma (accounting for 35% of radiation-related STS) and malignant fibrous histiocytomas (accounting for 22%).⁵

Chemical carcinogenesis has been linked to a variety of sarcomas. The thorium dioxide-containing contrast agent Thorotrast, vinyl chloride, and arsenic all have been conclusively shown to play a role in the development of hepatic angiosarcomas.^6–8 Phenoxyacetic acids, chlorophenols, and dioxins, which have been used in industry and agriculture, may contribute to the genesis of STS, but the precise causal relations are unclear and remain controversial.^9–11

Chronic lymphedema occurring after lymphadenectomy or secondary to filarial infection is a well-documented predisposing factor for the development of lymphangiosarcomas.^12,13

The relation between trauma and sarcoma development continues to be highly controversial. Perhaps the best evidence for a causal connection between the two is that parturition and extremity injuries have a predilection for giving rise to desmoid tumors.¹⁴

STS may be broadly categorized on the basis of the tissue type from which the tumor is believed to have originated [see Table 1]. Subtypes within a particular histologic category may be defined by means of histochemistry, flow cytometry, electron microscopy, tissue culture, and cytogenetic analysis. This more detailed classification can be useful in determining which therapy is to be employed for a given tumor, but it is not part of the staging system set forth by the American Joint Commission on Cancer (AJCC) [see Tables 2 and 3], which is the one most widely employed at present. The best indicator of a tumor's biologic aggressiveness and metastatic potential is its grade, regardless of its histologic type. The histologic grade is defined by the tumor's cellularity, nuclear atypia, degree of necrosis, and mitotic activity. Determining the grade of a particular specimen on the basis of these characteristics can be difficult; as prospective reviews have shown, expert pathologists may differ by as much as 40% in their assessments of specimens.^15,16

The AJCC staging system integrates tumor grade, tumor size, depth of tissue invasion, degree of nodal involvement, and presence or absence of metastases. It should be kept in mind that time and anatomic location, in concert with tumor size and grade, help determine the likely outcome and the type of recurrence that might be expected. Different staging variables tend to be more strongly related to some outcomes than to others. Tumor grade is related to early recurrence, tumor size is related to late recurrence, and anatomic location is related to the site or sites at which metastases are found.¹⁷

Clinical Evaluation and Investigative Studies

---

Figure 1. CT scan: retroperitoneal liposarcoma

STS typically presents as an asymptomatic mass. Such masses are frequently painless and large, and they often are brought to patients' attention by a history of recent trauma to the area [see Figure 1]. Approximately 38% are larger than 10 cm at presentation, with the remaining 62% about equally divided between those that are 5 cm or smaller and those that are larger than 5 cm but not larger than 10 cm.¹

Figure 2. Malignant fibrous histiocytoma of the thigh

The optimal modality for imaging a sarcoma remains a matter of debate. Although it is accepted that for extremity lesions, magnetic resonance imaging provides more information than computed tomography, a 1997 study by the Radiology Diagnostic Oncology Group did not find MRI to possess any significant advantages over CT in this setting [see Figure 2].¹⁸ Positron emission tomography (PET), either alone or in combination with CT (PET/CT), does not currently have a clearly defined role in the workup and diagnosis of STS, but it does appear to be useful for identifying the sites of distant metastases and, in particular, for determining the response to adjuvant therapy in patients with gastrointestinal stromal tumors (GISTs).¹⁹

Distant sites of metastases vary in accordance with the tumor histology and the site of the primary tumor. Extremity sarcomas most commonly metastasize to the lung, whereas abdominal and retroperitoneal sarcomas more frequently metastasize to the liver, with the lung being a secondary site of metastatic implantation.^20,21 In addition, there are a few sarcomas that tend to metastasize to regional lymph nodes [see Table 4].²² These differing patterns of metastatic spread must be carefully considered during the initial evaluation of sarcoma patients.

A mass that is larger than 5 cm, is growing, or has persisted for longer than 4 weeks constitutes an indication for biopsy. In choosing the method, location, and orientation of the biopsy, it is critical to consider possible future surgical interventions. For masses larger than 5 cm, either an incisional biopsy or a core-needle biopsy (CNB) is appropriate. Incisional biopsies increase the possibility of tumor spread through the tissue planes of the incision. For this reason, many practitioners now prefer CNB for evaluation of soft tissue masses. Several studies have shown CNB to have very high sensitivity and specificity with respect to both malignant and benign soft tissue masses.^23–25 For most masses smaller than 5 cm, an excisional biopsy with a clear surgical margin is indicated. Fine-needle aspiration is not a useful modality in the workup of soft tissue masses, because it does not provide a large enough sample to permit determination of histologic architecture and tumor grade.²³

Surgery is the foundation of the treatment of STS. Amputation was once considered the only option for cure, but this view changed with the 1982 publication of the landmark trial carried out by Rosenberg and colleagues from the National Cancer Institute (NCI). In this trial, 43 patients with extremity sarcomas were randomly assigned to undergo either amputation of the affected limb (N = 16) or limb-sparing surgery (LSS) plus radiation therapy (N = 27); all patients received adjuvant chemotherapy. The 5-year local recurrence rate was slightly higher in the LSS-radiation group, though the difference was not statistically significant. The 5-year survival was nearly identical in the two groups: 70% in the LSS-radiation group and 71% in the amputation group.²⁶ In the years since the NCI trial, LSS has become the standard of care. Currently, fewer than 10% of all patients with surgically treated extremity sarcomas resections undergo amputation, and local recurrence rates after LSS are in the range of 5% to 15%.^27,28 Amputation now is usually reserved for patients with extremity sarcomas that involve major vessels, nerves, or bones. Although en bloc resection with reconstruction is also possible for these types of sarcomas, long-term function is often poor, and the associated morbidity tends to be high. In these cases, amputation is the better option for accomplishing definitive surgical treatment and allowing a rapid transition to rehabilitation.²⁹

Adequate surgical resection involves excising a margin of normal tissue around the tumor, along with any areas through which biopsies have previously been performed. Compartmental resection or resection of entire muscle groups provides no additional benefit over wide local excision. The ideal extent of a wide local excision has not yet been defined; however, it is usually considered that a margin of at least 1 to 2 cm, when possible, should be the goal. Resection of STSs with negative margins is associated with a 12% to 31% recurrence rate.^30,31 In contrast, simple enucleation along the pseudocapsule is associated with a 33% to 67% recurrence rate and should therefore be avoided.³² Frozen-section analysis may be useful when an area is thought to be a close margin.

Regional lymphadenectomy is not usually indicated in the treatment of sarcomas, because only 2.6% of sarcomas metastasize to lymph nodes.³³ One exception to this general rule is a case in which the tumor is in proximity to a lymph node basin. Another is a case in which the tumor is one of the several specific subtypes that more commonly metastasize to the lymphatic system [see Table 4]. These subtypes include rhabdomyosarcoma, epithelioid sarcoma, clear cell sarcoma, synovial sarcoma, and vascular sarcoma (i.e., angiosarcoma and lymphangiosarcoma). As an alternative to regional lymphadenectomy for these specific sarcoma subtypes, sentinel lymph node biopsy (SLNB) may be considered [see 3:6 Lymphatic Mapping and Sentinel Lymph Node Biopsy].²²

Radiation Therapy

As noted [see Surgical Therapy, above], adjuvant radiation therapy has dramatically changed the surgical treatment of sarcomas, providing local control of the tumor in cases where LSS is appropriate. Three distinct types of radiotherapy are currently in use: brachytherapy, postoperative external beam radiotherapy, and preoperative external beam radiotherapy.

To date, only three prospective, randomized trials examining the utility of radiotherapy combined with surgery have been completed. The first was a trial of adjuvant brachytherapy in STS patients that was performed at Memorial Sloan-Kettering Cancer Center (MSKCC). In this study, 164 patients were randomly assigned during operation to receive either adjuvant brachytherapy or no further therapy after resection of an STS in an extremity or the superficial trunk.³⁰ Brachytherapy consisted of 42 to 45 Gy delivered over a period of 4 to 6 days. The median follow-up period was 76 months. The 5-year actuarial local control rate was 82% for the brachytherapy group and 69% for the no-brachytherapy group. In the high-grade lesion subset, this difference in local control rates was maintained (89% for the brachytherapy group and 66% for the no-brachytherapy group) and was statistically significant; however, in the low-grade lesion subset, there was no significant difference between the two groups. Despite the overall improvement in local control with brachytherapy, there was no significant difference between the brachytherapy group and the no-brachytherapy group with respect to 5-year disease-specific survival (84% and 81%, respectively). Similar results were noted when tumor grade subsets were analyzed. The investigators concluded that brachytherapy afforded superior local control but ultimately did not affect disease-specific survival.

The second study was a prospective, randomized trial from the NCI that examined adjuvant external beam radiotherapy. A total of 91 patients with high-grade STS were randomly assigned to receive either radiotherapy (N = 44) or no radiotherapy (N = 47) after complete surgical excision of the sarcoma; all patients received adjuvant chemotherapy.³¹ In the radiotherapy group, there were no local recurrences, whereas in the no-radiotherapy group, there were nine. There was no difference between the two groups in terms of overall survival. An additional 50 patients with low-grade sarcomas were treated according to the same protocol. In the radiotherapy group (N = 26), there was one local recurrence, whereas in the no-radiotherapy group (N = 24), there were eight. As with the patients who had high-grade STS, there was no difference in overall survival between the two groups. This NCI trial, along with the MSKCC brachytherapy trial, demonstrated that adjuvant radiotherapy effectively controlled local recurrence in patients with high-grade STS. In addition, it demonstrated that external beam radiotherapy was advantageous in preventing local recurrences. Unfortunately, neither study identified a means of prolonging overall survival.

The third and most recent study was a prospective, randomized trial carried out by the NCI of Canada Clinical Trials Group/Canadian Sarcoma Group, which compared preoperative and postoperative radiotherapy in patients with extremity STS.³⁴ A total of 190 patients were randomly assigned to receive either preoperative radiotherapy (N = 94) or postoperative radiotherapy (N = 96). This trial was closed before completion of the planned accrual because a preliminary analysis by the data monitoring committee demonstrated a significant difference in wound complications between the two groups. At the time of closure, the median follow-up period was 3 years. The wound complication rate in the preoperative radiotherapy group was 35%, compared with 17% in the postoperative radiotherapy group. At the time of the analysis, there was no difference in local control rate between the two groups (93% in both), but there was a significant difference in overall survival that slightly favored the preoperative radiotherapy group. The wound complication rate was in line with data from previous preoperative radiotherapy studies and was not unexpected; however, the significance of the unanticipated improvement in overall survival at this early time point remains unclear.^35,36

Currently, there is no consensus on which form of radiation therapy is best for treatment of STS. Several retrospective studies suggest that small (< 5 cm) STSs that are resected with negative margins may be treated with surgical therapy alone.^37,38

Chemotherapy

Postoperative chemotherapy for the treatment of STS has been studied in numerous prospective, randomized trials, but the small sample sizes and the various differences among the trials have made it difficult to determine how efficacious such therapy is. In an attempt to overcome the shortcomings of these small studies, the Sarcoma Meta-analysis Collaboration performed a meta-analysis of 14 doxorubicin-based adjuvant trials that included a total of 1,568 patients.³⁹ The median follow-up across all of the trials was 9.4 years. Hazard ratios were calculated for local recurrence-free interval, distant recurrence-free interval, recurrence-free survival, and overall survival, which when analyzed together allowed assessment of the absolute effects of treatment. A statistically significant improvement in 10-year disease-free survival (from 45% to 55%) was reported in the chemotherapy group. An improvement in 10-year local disease-free survival (from 75% to 81%) was also noted in this group. A trend towards improved overall 10-year survival (from 50% to 54%) was observed in the chemotherapy group but did not reach statistical significance; however, there was a statistically significant improvement in 10-year overall survival (7% absolute benefit) within a subset of the chemotherapy group comprising 886 extremity sarcomas. Several trials that made use of various other chemotherapeutic agents (e.g., epirubicin and ifosfamide) documented modest survival benefits that echoed the results of this very well performed meta-analysis. At present, given the uncertainty regarding its efficacy, postoperative adjuvant chemotherapy for treatment of STS is probably best employed in the context of appropriate clinical trials.

Preoperative adjuvant chemotherapy has several theoretical advantages over postoperative therapy in this setting. It allows delivery of agents through the native vasculature, permits assessment of the effectiveness of a particular treatment by means of pathologic analysis, may facilitate treatment of micrometastases, and may downstage tumors or make them more amenable to surgical treatment. Certainly, the application of this modality to the treatment of osteosarcoma and Ewing sarcoma in the pediatric population has met with great success. As a consequence of this success, several groups have performed studies aimed at determining the effects of preoperative chemotherapy in adult STS patients.

A retrospective analysis of 46 patients with extremity STS at M. D. Anderson Cancer Center examined a preoperative chemotherapy regimen consisting of doxorubicin, cyclophosphamide, and dacarbazine.⁴⁰ The overall tumor response rate was 40%. In patients who exhibited a complete, partial, or minor response, there was a statistically significant improvement in both disease-free survival and overall survival. Subsequently, a prospective trial was completed at MSKCC in which 29 patients with large (> 10 cm) high-grade stage IIIB extremity sarcomas underwent preoperative chemotherapy (involving a regimen similar to that employed in the M. D. Anderson study) and were compared with historic control subjects.⁴¹ Unlike the M. D. Anderson trial, the MSKCC trial did not find preoperative chemotherapy to confer any survival advantage over postoperative chemotherapy or no chemotherapy after resection. Given the lack of sufficient evidence for any survival benefit, preoperative chemotherapy may reasonably be considered in attempting to preserve limb function, but otherwise, its use should be limited to clinical trials.

Targeted Therapeutics

Perhaps the most interesting and exciting advances in the treatment of STS are those currently being achieved in targeted therapeutics. In particular, the characterization and targeting of the tyrosine kinase receptor c-kit has generated a new approach to treating GISTs. Phase III trials comparing standard and high-dose imatinib mesylate in the treatment of these lesions have been completed in the United States and Europe.^42,43 In these two trials, the progression-free response rate ranged from 43% to 53%, and the estimated 2-year survival ranged from 69% to 78%. The use of imatinib mesylate in adjuvant and neoadjuvant settings is still evolving. Data from ACOSOG-Z9000, which has finished its patient accrual, and from several other trials, which were still accruing patients as of April 2007 (ACOSOG-Z9001, EORTC 62024, SSGXVIII, RTOG-S0132, MDACC AD03-0023), should provide some valuable clarification in this regard.⁴⁴ The initial results achieved with imatinib mesylate and its therapeutic analogues have given rise to hopes that other similar molecules will be discovered that can be directed at tumor-specific targets for effective treatment of these largely drug-resistant tumors.

Hyperthermic Isolated Limb Perfusion

Hyperthermic isolated limb perfusion is a technique that is currently reserved for patients in whom LSS is not a possibility. An isolated circuit is created by cannulating the inflow arterial system and the outflow venous system of a limb and applying a proximal tourniquet. This isolated circuit is usually monitored by means of radiolabeled albumin to identify any leakage into the systemic circulation. A heart-lung bypass machine maintains mild hyperthermia (39° to 40° C) and oxygenation and circulates chemotherapeutic agents in the isolated limb.

The largest trials of this technique to date were conducted in the Netherlands by Eggermont and colleagues, who used tumor necrosis factor (TNF) and melphalan.^45,46 A total of 246 patients with extremity sarcomas who were not candidates for LSS underwent one or two sessions of isolated limb perfusion. After an interval of 2 to 4 months, 76% of these patients were able to undergo a resection with a negative margin, and 71% underwent successful LSS. This approach is very well developed in Europe, but it is still in the early investigational phase in the United States, where TNF is not currently available.

Patients with extremity STSs experience local recurrences at a rate of 8% to 20% despite appropriate primary resection, and those with resected retroperitoneal sarcomas experience local recurrences at a rate of 38% to 50%.^47–49 These patients are often candidates for re-resection of the locally recurring sarcoma. A retrospective review from the Brigham and Women's Hospital found that 67% of patients with recurrent STSs were able to undergo salvage surgery with excellent long-term survival.⁴⁸ Patients who have not received radiation therapy are candidates for preoperative or postoperative treatment. Those who have already received radiation therapy may be considered for brachytherapy, which has a reported 5-year actuarial control rate of 68% and a reported morbidity of 12.5%.⁵⁰ However, given that low-grade sarcomas do not show a significant response to brachytherapy, this modality should be reserved for treatment of high-grade sarcomas.^30,51 In some instances, such as when the tumor burden is disseminated or limb function cannot be preserved, salvage surgery may necessitate amputation.

Pulmonary metastases are present in approximately 20% of patients with trunk or extremity sarcomas.⁵² If the patient is medically fit, the primary tumor is controlled, no extrathoracic disease is present, and the metastases are resectable, pulmonary metastasectomy may be attempted.⁵³ In a retrospective study performed at MSKCC, 148 of 716 patients with extremity sarcomas experienced a recurrence.⁵² In 135 (91%) of the 148, the lung was the only site of recurrence. Of these 135 patients, 78 (58%) were judged to be candidates for operative management, and ultimately, 65 (48%) were able to undergo a complete pulmonary metastasectomy. The 3-year survival in this group was 23%, and the median survival time was 19 months. The 3-year survival in the group that did not undergo thoracotomy was 0%. The overall 3-year survival for all patients with pulmonary metastases was 11%. Large studies of resection of pulmonary metastases from STS have yielded 3-year survival rates ranging from 23% to 54%.^54–61

Unresectable

Despite adequate resection of STS and the use of adjuvant therapy, distant metastases may develop in as many as 50% of cases.¹ As noted (see above), the tumor's site of origin and histology determine the site or sites at which metastatic disease will occur. Unfortunately, for the vast majority of patients, the only available treatment option is systemic chemotherapy. There has been considerable debate regarding the best chemotherapy regimen for treating metastatic sarcoma, with most of the controversy centering on whether it is better to employ a single agent or a combination of agents. Doxorubicin has been the backbone of most chemotherapeutic regimens, and it has proved to be as effective as many single-agent or combination chemotherapy regimens with respect to recurrence rates and survival.^62–65 Some combination regimens have yielded higher response rates than doxorubicin alone, but with increased toxicity and without any improvement in overall survival.^66,67