434 Radiation Oncology Radiation therapy (RT) uses ionizing radiation (IR), which damages DNA to exert the end effect of cell death. IR is a type of radiation that contains enough energy to ionize or eject orbital electrons from an atom and may take the form of particles or electromagnetic waves. This chapter examines the roles of RT in the treatment of different sarcomas. In localized soft tissue sarcomas, the backbone of curative treatment is surgery with wide negative margins. Preoperative, intraoperative, or postoperative RT can maximize local control, though omission of radiation may be considered in select cases, particularly for small (<5 cm), low-grade, superficial tumors. In Ewing sarcoma, RT may be used to treat localized disease when wide margins cannot be achieved with surgery or if surgery would impair function, though the risk of radiation-associated malignancy, especially in pediatric patients, must be considered. In chordoma and chondrosarcoma, RT may have a role when wide local resection cannot be achieved. The chapter also discusses re-irradiation and stereotactic body RT for metastases from sarcomas. cell death, DNA damage, ionizing radiation, radiation therapy, re-irradiation, soft tissue sarcomas, stereotactic body radiation therapy Cell Death, DNA Damage, Radiation, Ionizing, Radiosurgery, Radiotherapy, Re-Irradiation, Sarcoma INTRODUCTION In 1921, James Ewing described his eponymous tumor as a diffuse endothelioma of bone. In contrast to osteosarcoma, he reported that Ewing sarcoma appeared in adolescents, favored long and flat bones, involved most of the bone shaft while sparing the ends, showed no bone production, and responded exquisitely to radiation.1 The following year, in his 1922 presentation of the Mütter lecture, An Analysis of Radiation Therapy in Cancer, he noted that “basal cell carcinomas,” “endothelioma of lymph nodes, endothelioma of bone and myeloma,” and other rapidly growing tumors were all “peculiarly susceptible to radiation.”2 Under the microscope, he observed that irradiated tumor cells developed “swelling, hyperchromatism, vacuolar degeneration, and solution of fragmentation of nuclei.” A century later, Ewing’s remarks on the relationship between radiation and cancer cells remain relevant. Radiation therapy (RT) uses ionizing radiation (IR), which damages DNA to exert the end effect of cell death that Ewing described. IR is a type of radiation that contains enough energy to ionize or eject orbital electrons from an atom and may take the form of particles or electromagnetic waves. Ionizing electromagnetic waves include x-rays, which are produced when high-energy electrons hit a target atom such as tungsten or gold, and gamma rays, which are produced by nuclear decay of radioactive isotopes such as iodine-125, iodine-131, and iridium-192. Ionizing particles include electrons, protons, and alpha particles. The susceptibility of cancer cells to IR relies mainly on DNA damage, which results from direct action on DNA or occurs through indirect effects. In direct action, IR directly ionizes the DNA molecule, causing structural damage to DNA bases or the sugar–phosphate backbone that causes single-stranded DNA breaks or double-stranded DNA breaks, which are especially lethal to cells. In indirect action, radiation interacts with water or other organic molecules, which results in free radicals such as hydroxyl and alkoxy molecules, which then subsequently react with and ionize DNA. In most normal cells, DNA damage activates the DNA damage response that can lead to repair of DNA damage and survival after exposure to IR. However, in cancer cells, mutations in genes that regulate the DNA damage response or that drive DNA synthesis and mitosis even in the presence of unrepaired DNA double-strand breaks result in cell death. Broadly speaking, IR can selectively kill cancer cells over normal tissue, though the therapeutic ratio varies for different cancer cells and normal tissues. In addition to the biologic basis (i.e., cancer-specific mutations) for a therapeutic ratio for IR, RT also achieves a therapeutic ratio from physically delivering more radiation to the tumor than to normal tissue. RT can be delivered either from a radiation source outside the patient, referred to as external beam RT, or from a source adjacent to or inside the patient, in the form of brachytherapy or radionuclide therapy. External beam RT is commonly delivered with a medical linear accelerator (LINAC), which produces a beam of x-ray photons or electrons directed at the tumor within the patient. The x-ray photons are deeply penetrating and are used most frequently to treat sarcomas. The x-rays enter the patient and deposit their maximum energy at a specific depth, which depends on the energy of the x-rays, before exiting out of the patient. Electrons are superficially penetrating, with the depth of penetration proportional to their energy, and may be delivered en face to the patient to treat cancers up to approximately 4 cm in depth. External beam RT can also be delivered using a beam of protons, which like x-rays can deeply penetrate tissue to a specific depth based on energy, but then rapidly deposit all of the remaining energy without exiting out of the patient. Protons may be particularly useful when a sarcoma is adjacent to a critical normal tissue, such as a chondrosarcoma in the skull base. Since the inception of LINAC-based external beam RT in the 1950s, refinements in treatment planning have improved coverage, conformity, and homogeneity of radiation dose to the tumor. 44Three-dimensional conformal RT (3D-CRT), first used in the 1980s, starts with creating a 3D model of the patient via CT or MRI. Tumor target volumes and normal tissues are then delineated slice by slice, and radiation beams are manually optimized to achieve the prescribed tumor dose to the tumor target volumes while minimizing dose to normal tissues. In the 1990s, treatment planning advanced with intensity-modulated RT (IMRT), in which computerized optimization of multiple radiation beams as well as nonuniform intensity within each beam provide even more conformal and homogeneous coverage, especially for tumors with complex shapes. Modern 3D-CRT or IMRT is often image guided with x-ray or low-dose CT imaging in the treatment room prior to each fraction, to optimize patient position before treatment and thereby improve accuracy. Improved accuracy of radiation delivery also results from advances in patient immobilization, so that the patient is in the same treatment position each day and in approaches to minimize tumor motion, such as respiratory gating strategies. Another important consideration of RT is fractionation and total dose. Radiation dose is prescribed in gray (Gy) and the unit of absorbed IR in joules per kilogram. Conventional fractionation is typically 1.8 to 2 Gy/day to a total dose of about 50 to 70 Gy in the curative setting. Stereotactic radiosurgery (SRS), which was developed in the 1950s, is used for brain tumors, and it describes a single highly conformal fraction of approximately 18 to 24 Gy. Over the past two decades, this technology has been applied to other sites and termed stereotactic body RT (SBRT) or stereotactic ablative RT (SABR), in which one to five large fractions are delivered with highly accurate stereotactic guidance to target extracranial tumors, such as pulmonary or spine metastases from sarcomas. In contrast to external beam RT, brachytherapy places a sealed radioactive isotope such as iridium-192 directly within or adjacent to the tumor for a specified period of time. In this setting, the radiation acts over a short distance because the dose falls off with the square of the distance from the radiation source(s). Brachytherapy is typically used following surgery for soft tissue sarcomas (STSs) when catheters can be placed in the tumor bed to facilitate delivery of the radioactive isotope. In radionuclide therapy, an unsealed radioactive isotope is ingested by or injected into the patient. Examples include iodine-131, which accumulates in thyroid cells and is used to treat hyperthyroidism and thyroid cancer; yttrium-90–labeled rituximab in non-Hodgkin lymphoma; and transarterial radioembolization with yttrium-90 in liver metastases. Intraoperative RT (IORT), which is a single fraction delivered at the time of surgery, can use either external beam RT with electrons or brachytherapy with radioactive isotopes. IORT has been used to treat extremity and retroperitoneal sarcomas. An advantage of IORT is direct target visualization and physical manipulation of normal tissues, such as the intestine, away from the target. SOFT TISSUE SARCOMA The cornerstone of curative treatment for localized STS is surgery with negative margins. Historically, local control rates following local excision alone were 10% to 50%. Thus, radical resection or amputation was often favored over local excision, which could improve local control rates to >80% at the cost of limb function. Adjuvant RT enhances local control after limb-sparing surgery and is typically employed for large (>5 cm), intermediate, or high-grade sarcomas located deep to the fascia. RT may be delivered preoperatively, intraoperatively, or postoperatively, and the timing of RT with respect to surgery can be tailored to individual patients. Preoperative RT generally allows for smaller radiation field size, lower total radiation dose, and improved functional outcomes, however, with higher rates of acute wound complications. Thus, preoperative RT may be favored for young patients, who would likely recover well from an acute wound complication and who, if cured of their tumor, would be expected to live for many years and would benefit from the anticipated reduction in late radiation toxicity with preoperative RT. In contrast, postoperative RT may be preferable for elderly patients with multiple comorbid conditions, who may be at higher risk for significant morbidity from a wound complication and for whom late radiation effects may be a lesser concern. Current guidelines from the National Comprehensive Cancer Network recommend 50 Gy in the preoperative setting, with postoperative boost reserved for positive margins only.3 Fifty gray is also utilized in the postoperative setting; however, a tumor bed boost of at least 10 Gy is given for negative margins, 16 to 18 Gy for microscopically positive margins, and 20 to 26 Gy for gross residual disease. Ultimately, the decision of preoperative versus postoperative RT should be made with the patient in a multidisciplinary evaluation involving the surgeon and radiation oncologist, taking into account clinical features such as size and grade of the sarcoma, type and extent of surgery required, patient comorbid conditions, and the potential consequences of a wound complication and late effects of radiation. 45Postoperative RT In the 1970s, external beam RT was advocated as a means to improve local control following local excision of STSs. In 1982, the National Cancer Institute (NCI) published a prospective randomized trial of 43 patients with high-grade STS, comparing amputation to wide local excision with adjuvant RT.4 Both groups received adjuvant chemotherapy with doxorubicin, cyclophosphamide, and methotrexate. RT dose was 50 Gy to the area at risk for microscopic spread and an additional 10 to 20 Gy tumor bed boost. On multivariate analysis, positive margins were associated with increased local recurrence. Local recurrence was observed in 4/27 patients who underwent limb-sparing surgery and in 0/16 patients who underwent amputation. No difference in 5-year disease-free survival (71% vs. 78%) or overall survival (83% vs. 88%) was seen between the limb-sparing versus amputation groups. This study concluded that the local excision with postoperative RT was an acceptable alternative to amputation. Following the 1982 NCI study, the benefit of postoperative RT after limb-sparing surgery was directly tested in another NCI study published in 1998.5 This trial randomized 141 patients, 50 patients with low-grade STS and 91 patients with high-grade STS, to receive postoperative RT of 45 Gy to a wide field and an additional 18 Gy tumor bed boost. Patients with high-grade STS also received postoperative chemotherapy with doxorubicin and cyclophosphamide, either alone or concurrently with radiation. At a median 9.6 years of follow-up, in the high-grade STS group, local recurrence was observed in 9/47 patients who received postoperative chemotherapy alone and 0/44 patients who received postoperative chemotherapy and RT (p = .003). In the low-grade STS group, local recurrence was observed in 8/24 patients who received no adjuvant therapy and 1/26 patients who received postoperative RT (p = .016). No difference in distant metastases or overall survival was seen with the addition of RT. A concurrent quality-of-life study demonstrated that patients who received RT had transient worsening of muscle strength and edema in the first year after surgery as well as decreased joint motion up to 3 years after surgery, though no significant differences in global quality of life or performance of activities of daily living were seen. Long-term outcomes from this randomized trial were assessed after a median follow-up of 17.9 years.6 Twenty-year overall survival rate was similar for both groups: 64% for the limb-sparing surgery group and 71% for the limb-sparing surgery plus RT group. Although there was a trend to more edema and functional limb deficits in patients treated with RT, this study established that the addition of postoperative RT to limb-sparing surgery improves local control with acceptable treatment-related morbidity. Retrospective studies have attempted to define a subgroup of surgical patients in which RT improves survival. A 1998 to 2005 Surveillance, Epidemiology, and End Results (SEER) database analysis suggested benefit was confined to high-grade tumors.7 This analysis included 6,960 patients with extremity STS who underwent limb-sparing surgery; 47% of patients received RT, of whom 13.5% received preoperative RT. In low-grade tumors, no difference was seen in overall survival at 3 years. In high-grade tumors, patients who received radiation had improved 3-year overall survival versus patients who did not receive radiation (73% vs. 63%, p < .001). A retrospective matched-pair analysis from the National Cancer Database (NCDB) of patients with high-grade extremity sarcoma treated from 2004 to 2013 also concluded that patients treated with limb-sparing surgery and postoperative or preoperative RT had improved overall survival (hazard ratio [HR]: approximately 0.7) compared to patients treated with limb-sparing surgery alone.8 Preoperative RT A comparison of preoperative versus postoperative RT was addressed in the SR2 trial by the NCI of Canada Clinical Trials Group.9 This trial randomized 190 patients with extremity STS to 50 Gy of preoperative RT or postoperative RT. Patients in the preoperative group with tumor cells at the resection margin also received a 16 to 20 Gy boost. All patients in the postoperative group received a 16 to 20 Gy boost. The primary endpoint was wound complications. Major wound complications included secondary operation or invasive procedure for wound repair, deep wound packing, and readmission for wound care. The rate of major wound complications was higher in the preoperative group compared to the postoperative group (35% vs. 17%, p = .01), which suggests that preoperative RT does not cause wound complications, but instead exacerbates the underlying risk by approximately twofold. The most common site of wound complications was the upper leg followed by the lower leg. Interestingly, wound complications in the upper extremity were rare: 0/11 patients in the postoperative group and 1/10 patients in the preoperative group. In addition to preoperative RT and anatomic site, maximum tumor size >10 cm was also associated with wound complications on logistic regression. No difference in local control (approximately 90%) or overall survival was observed between the groups. An analysis 46of late effects found that a greater proportion of patients who received postoperative RT had grade 2 or greater fibrosis compared to the preoperative arm (48% vs. 32%, p = .07), which was likely due to larger radiation field size and larger radiation dose.10 Joint stiffness and edema were also more frequent in the postoperative arm, though the difference was not statistically significant. RTOG 0630 was a multi-institutional Phase 2 clinical trial of reduced target volume preoperative image-guided RT for extremity STS.11 Local control for the 79 eligible patients was 93%, and all five local treatment failures were in the radiation field. Of the 57 patients assessed for late toxicities at 2 years, 10.5% experienced at least one grade ≥2 toxicity, as compared with 37% of patients from the preoperative arm of the SR2 trial by the NCI of Canada. A meta-analysis of 1,098 patients also addressed the question of preoperative versus postoperative RT.12 This analysis found that a higher proportion of patients who underwent preoperative RT had an initial tumor size >10 cm. The risk of local recurrence was lower in patients who had received preoperative radiation (odds ratio: approximately 0.66). The average survival rate was 76% for patients who received preoperative RT and 67% in patients who received postoperative RT. This meta-analysis concluded that the delay in surgery inherent in preoperative radiation did not appear to adversely affect survival and may even be associated with decreased local recurrence. The feasibility of preoperative RT with chemotherapy was evaluated in a single-arm, multi-institutional Phase 2 study in RTOG 9514.13 In this study, 66 patients with high-grade STS ≥8 cm received neoadjuvant chemotherapy (three cycles of modified mesna, doxorubicin, ifosfamide, and dacarbazine), split course 44 Gy RT interdigitated between chemotherapy cycles 1 and 2 and between cycles 2 and 3, as well as three cycles of postoperative chemotherapy. Also, 61/66 patients underwent surgery and 58 had R0 resections including five amputations, while three had R1 resections. The 3-year local–regional failure rate was 17.6% (including amputation as failure), disease-free survival rate was 57%, and overall survival rate was 75%. Intraoperative RT IORT may also be used as a boost in addition to preoperative or postoperative RT. IORT is typically delivered using an electron beam from a LINAC or with high dose rate (HDR) brachytherapy. Current National Comprehensive Cancer Network (NCCN) guidelines recommend 10 to 12.5 Gy for microscopic disease and 15 Gy for gross residual disease following preoperative RT, and 10 to 16 Gy boost prior to postoperative RT. The decision to administer IORT is made following multidisciplinary evaluation by the surgeon and radiation oncologist, and may be considered when surgery is likely to leave microscopic residual or gross disease that cannot be adequately treated with conventional external beam RT owing to nearby dose-limiting normal tissue. IORT may be especially useful for certain anatomic sites such as the retroperitoneum, where the radiation dose tolerance of adjacent bowel may limit the amount of radiation that can be delivered safely to a large field. In IORT, bowel or other dose-limiting normal tissues may be manually displaced or shielded during radiation delivery. Data on IORT with external beam RT are mostly limited to small series. A tumor bed boost with IORT appears to provide similar local control as boosting with external beam RT, with 5-year local control rates of 82% to 97% in IORT series and 93% to 93% in non-IORT series.14 One prospective randomized trial has examined the use of IORT. In this trial, 35 patients with resected retroperitoneal sarcomas were randomized to 20 Gy IORT with 35 to 40 Gy postoperative radiation or to 50 to 55 Gy postoperative radiation. Six of 15 patients who received IORT developed locoregional recurrences, while 16 of 20 patients who received external beam RT alone developed locoregional recurrences. The IORT group also had lower rates of radiation-related enteritis but, however, had higher rates of peripheral neuropathy. Median overall survival was similar between the two groups.15 Brachytherapy Brachytherapy can be used as a boost in conjunction with preoperative RT. In this case, brachytherapy can be given during or after surgery. However, brachytherapy alone can also be used as adjuvant therapy after surgery (i.e., no external beam RT). A single-institution randomized trial of 164 patients with completely resected sarcomas compared limb-sparing surgery alone to surgery plus adjuvant brachytherapy.16 Brachytherapy technique consisted of percutaneous catheters placed intraoperatively in the tumor bed in a single plane. Catheters were loaded with low dose rate (LDR) iridium-192, which was designed to deliver 42 to 45 Gy over the course of 4 to 6 days. Five-year local control was 82% in the patients who received brachytherapy and 69% in the patients who received no further therapy (p = .04). This benefit was limited to patients with high-grade sarcomas. This local control rate is lower than that reported in studies with external beam RT, which may reflect the challenge 47in using a single plane of brachytherapy catheters to cover the volume at risk for large sarcomas.9 No difference was observed in distant metastases or disease-specific survival. Initially, patients who received brachytherapy had a significantly higher risk of wound complications. As an interim analysis demonstrated a relationship between early catheter loading and wound complications, the trial was amended such that catheters were loaded at least 6 days after surgery. After this adjustment, no significant difference in wound complications was observed between the patients who received brachytherapy and those who received no further treatment. In LDR brachytherapy, the patient is hospitalized while an implanted radioactive source delivers radiation at a rate of 0.4 to 2.0 Gy/hour over several days. Healthcare staff may be at risk of radiation exposure during this period. In contrast, HDR brachytherapy delivers radiation at a rate of >12 Gy/hour, which allows for treatment delivery in minutes, reducing the need for hospitalization and virtually eliminating radiation exposure to healthcare personnel. A comparison of LDR versus HDR brachytherapy in 37 patients with STS showed no significant difference in 2-year local control rates or wound complications.17 Definitive RT Surgery is generally the backbone of definitive treatment of sarcoma. However, in patients with unacceptable medical comorbid conditions or unresectable tumors, definitive RT alone may be considered. Massachusetts General Hospital has published a series of 112 patients with STS who received a median dose of 64 Gy for gross disease, either following biopsy or attempted radical surgery.18 Twenty-one percent of patients received chemotherapy. Overall, 5-year local control rate and overall survival rate were 45% and 35%, respectively. On multivariate analysis, local control was related to the tumor size, total dose, and stage, but not chemotherapy, grade, age, or location. Tumor sizes of <5, 5 to 10, and >10 cm had 5-year local control rates of 51%, 45%, and 9%, respectively, while radiation doses of <63 and ≥63 Gy had 5-year local control rates of 22% and 60%, respectively. Major complications, the most common of which was wound healing disease requiring major surgery, were seen in 14% of patients and were positively correlated with increased radiation dose. The addition of concurrent chemotherapy to definitive RT as a radiosensitizer has also been examined in a series from Germany.19 In this retrospective analysis, 11 patients with macroscopic sarcoma received a median of 60 Gy RT, mostly conventionally fractionated, with a single agent ifosfamide at 1.0 or 1.5 g/m2 during the first and fifth weeks of radiation. Overall, the 2- and 5-year local control rate was 70%; both the patients with local failures had tumors >10 cm. The results from this small series appear to compare favorably to the larger series from Massachusetts General Hospital, where patients were mostly treated with RT alone. Omitting RT In addition to the trials of limb- or function-sparing surgery with adjuvant RT described in the preoperative and postoperative RT sections, several series have also sought to characterize the risk of local recurrence following function-sparing surgery alone, in order to define a cohort of favorable-risk patients who may not need adjuvant RT. To that end, a retrospective series published in 1999 reported outcomes of 74 patients who received function-sparing surgery alone for trunk or extremity sarcomas.20 The median tumor size was 4 cm and the majority were low grade. Ninety-two percent of patients had negative margins, 7% had microscopically positive margins, and one patient had grossly positive margins. Seven patients received adjuvant doxorubicin-based chemotherapy, and none received adjuvant radiation. The 10-year actuarial local control rate and survival rate were 93% and 73%, respectively. Margin status was the most significant predictor of local control, with a local control rate of 100% when the margins were at least 1 cm wide. However, in this patient cohort, tumor size, grade, site, and depth were not significantly associated with local control. This study demonstrated that function-sparing surgery alone with wide negative margins can achieve local control in highly selected patients with STSs. A recently published nomogram estimating risk of 3- and 5-year local recurrence after function-sparing surgery may also help inform patient and provider decisions on adjuvant RT.21 The authors analyzed 684 patients who underwent limb-sparing surgery alone without adjuvant chemotherapy or RT from 1982 to 2006 at Memorial Sloan Kettering Cancer Center. The 3- and 5-year actuarial risks of local recurrence were 11% and 13%, respectively. On multivariate analysis, age ≥50, size >5 cm, close or positive margins, high grade, and histologic types other than well-differentiated liposarcoma and atypical lipomatous tumor were all associated with increased risk of local recurrence; patient sex and depth were not significant. The decision of whether to treat a patient with STS with surgery alone or with surgery and RT should be made after the patient is evaluated by a surgeon and a radiation oncologist with expertise in the 48care of STSs and after discussion at a multidisciplinary tumor board.22 Remarkably, the French Sarcoma Group has found that referral to a multidisciplinary tumor board with expertise in sarcomas prior to initial treatment increased the rate of local relapse-free survival and distant relapse-free survival on multivariate analysis.23 At our multidisciplinary tumor board, we consider treatment of STSs with surgery alone for small (<5 cm), low-grade, superficial tumors, particularly in an anatomic location where salvage surgery and RT for a future local recurrence would not lead to unacceptable morbidity. Special Considerations Myxoid Liposarcoma While most STSs do not shrink during a course of RT, myxoid liposarcomas, which frequently harbor a FUS–DDIT3 translocation t(12;16)(q13;p11), are exquisitely radiosensitive. In a retrospective analysis of 691 patients with extremity STS who received surgery and RT, patients with myxoid liposarcoma compared to other subtypes had improved local recurrence-free survival, improved metastasis-free survival, and improved overall survival.24 The radiosensitivity of myxoid liposarcomas has been documented by quantifying the tumor size before and after RT. A study of 16 patients with myxoid liposarcomas who underwent preoperative radiation found mean pre- and postradiation tumor volumes of 415 and 199 cm3, respectively, for a proportional reduction of 59% following radiation.25 In contrast, a control group of patients with malignant fibrous histiocytoma (now referred to as undifferentiated pleomorphic sarcoma or pleomorphic fibrosarcoma) had mean pre- and postradiation volumes of 264 and 273 cm3, respectively, for a proportional increase in 7%. Lymph Node Metastases Most STS histologic types rarely metastasize to lymph nodes (<1% risk), and elective radiation of lymph nodes is not recommended. However, synovial sarcoma, clear cell sarcoma, angiosarcoma, rhabdomyosarcoma, and epithelioid sarcoma histologic types are associated with 10% to 20% rates of lymph node metastases.26 Clinically suspicious lymph nodes may be included in the preoperative radiation volume. Clinically or pathologically involved lymph node stations may also be included in the postoperative radiation volume, especially in the setting of limited lymph node dissection or extracapsular extension. However, the potential benefit of extending the radiation field to cover draining lymph nodes must be considered within the context of potential increased toxicity. While PET/CT27 and sentinel node biopsies28 have been assessed for their utility in staging sarcoma patients, their role for determining lymph node metastases remains uncertain. Angiosarcoma of the Scalp Surgery can be utilized for angiosarcoma of the scalp. However, local failure rates after surgery alone are high. In retrospective series, use of RT and multimodality therapy has been associated with improved local control.29 The optimal radiation dose and field size are not well established. Our practice is to manage most patients with scalp angiosarcoma without surgery. We administer total scalp irradiation with bilateral neck lymph node irradiation using a shrinking field technique with concurrent paclitaxel chemotherapy. Total dose to the primary angiosarcoma is typically 70 to 74 Gy. BONE SARCOMA Ewing Sarcoma Following standard induction chemotherapy for Ewing sarcoma, localized disease is treated with surgery and/or RT. As in STSs, RT may be preoperative, postoperative, or definitive. Preoperative RT may be considered when close or positive margins are expected. Postoperative RT is indicated for close margins, poor histologic response, or tumor spill. Definitive RT may be offered when function-preserving surgery is not an option, such as tumors of the acetabulum or pelvis. Chemotherapy is routinely administered during RT. Surgery and radiation have not been compared head-to-head in prospective, randomized studies. A retrospective analysis of local control in pelvic Ewing sarcoma from INT-0091 found no significant differences in local control between surgery and RT.30 Another retrospective analysis of patients treated on CESS 81, CESS 86, and EICESS 92 trials suggested that postoperative RT improved local control compared to surgery alone for intralesional resections and poor histologic response. In this analysis, definitive RT had higher local failure rates than surgery, though the authors note that these RT-only patients represented a negatively selected subgroup, with comparatively more paraspinal and pelvic tumors and fewer limb tumors than the surgical patients.31 49For metastatic Ewing sarcoma, RT may have a role along with surgery for local control of primary tumor and for patients with limited pulmonary, soft tissue, or bone metastases. For pulmonary metastases, whole lung irradiation may be considered even after patients have had resection of limited pulmonary metastases or have shown an excellent response to chemotherapy. Retrospective analysis of cooperative group trials have found improved event-free survival in patients who received low-dose whole lung irradiation.32 Use of RT in pediatric patients with Ewing sarcoma is balanced against the risk of second malignancy from irradiation of normal tissue. The most common second malignancy is osteosarcoma. Use of IMRT can reduce the amount of normal tissue exposed to high-dose radiation (45–54 Gy), and proton therapy can also reduce the amount of normal tissue exposed to low-dose radiation (<20 Gy). Osteosarcoma Osteosarcoma is routinely treated with induction chemotherapy, surgical resection, and adjuvant chemotherapy. However, RT can be utilized to treat osteosarcoma, but high radiation doses are needed to impact local control. Adjuvant RT can be utilized for close or positive surgical margins and inoperable tumors. A retrospective study of 41 patients who received RT for osteosarcoma suggested that local control was improved for doses ≥55 Gy.33 Chordoma and Chondrosarcoma Chordoma, a malignant bone neoplasm arising from notochord remnants, and chondrosarcoma, a malignant bone neoplasm with production of chondroid matrix, both tend to recur locally, particularly when arising in the mobile spine, sacrum, skull base, or pelvis. En bloc resection is recommended, and a systematic review has found decreased local recurrence and mortality following wide or marginal margins en bloc.34 However, resection may be limited by patient comorbid conditions, surgical access, and proximity to critical structures. Definitive or adjuvant radiation may, therefore, be considered in the setting of absent or incomplete resection. Historically, conventional radiation doses (<60 Gy) of adjuvant RT resulted in poor local control. The same systematic review suggested that radiation doses of >60 to 65 Gy may improve local control. However, as with surgery, proximity and radiation dose constraints of critical structures, especially the brainstem and spinal cord, may limit the radiation plan. Particle RT with protons or carbon ions has been utilized in this setting, especially for the base of skull chordomas and chondrosarcomas, to achieve favorable dosimetry to spare critical structures. RE-IRRADIATION Radiation-associated sarcomas arise within a previously irradiated field. Undifferentiated sarcoma, angiosarcoma, or osteosarcoma, or less commonly, other sarcoma subtypes are also seen in patients who have received radiation, particularly among survivors of childhood cancers. Radiation-associated angiosarcomas are recognized as a complication following breast or chest wall radiation for breast cancer. The mainstay of treatment for sarcomas arising within a previously irradiated field, either for locally recurrent or radiation-associated sarcomas, is wide local excision with or without systemic therapy. When wide margins cannot be achieved, adjuvant or definitive re-irradiation may be considered. The potential benefit of re-irradiation must be balanced by the potential toxicity because of dose tolerance of adjacent normal tissues such as brainstem, spinal cord, brachial plexus, and other major nerves, blood vessels, and bowel. Potential complications of re-irradiation are similar to those of upfront radiation, but the risk and severity increase; of utmost concern are impaired wound healing or nonhealing wounds, tissue necrosis, fistulas, bowel perforation, vascular blowout, and neuropathy. Retrospective series and single-arm prospective trials have examined re-irradiation with either external beam RT or brachytherapy. A retrospective analysis of 10 patients with locally recurrent limb sarcoma who received electron or photon external beam re-irradiation reported adequate local control, with a median survival of 14 months after re-irradiation. Radionecrosis was seen in a single patient who had received a total dose of 145 Gy.35 Another retrospective analysis of patients with locally recurrent sarcoma who received re-irradiation with perioperative brachytherapy reported adequate local control with a 5-year local recurrence-free rate of 52% and an overall survival rate of 33%. Complication rates were deemed acceptable, with wound breakdown or osteonecrosis or neuropathy occurring in 5/26 patients.36 Re-irradiation with proton therapy has also been utilized and may provide favorable dosimetry in respecting adjacent normal tissue constraints. 50Hyperfractionated, accelerated RT with smaller fraction sizes delivered two to three times per day rather than once daily can also be used for re-irradiation. A retrospective series of 14 patients who developed angiosarcoma following whole breast radiation for breast cancer used photon and/or electron external beam re-irradiation with adequate local control and limited toxicity.37 The re-irradiation regimen used was 1 Gy three times per day to total doses of 45, 60, or 75 Gy for low-risk subclinical, high-risk subclinical, and gross disease. All patients received mastectomy for angiosarcoma either before or after RT. The 5- and 10-year overall survival rates were 79% and 63%, respectively, and overall toxicity appeared to be acceptable, with three patients developing lymphedema, one patient developing recurrent benign pleural effusion, and four patients developing self-limited rib fractures. SBRT FOR METASTASES FROM SARCOMAS For several decades, surgical metastasectomy has been used for palliative and local control of sarcoma metastases. Some retrospective studies have suggested a longer disease-free interval or even survival benefit in carefully selected patients, typically with a limited number of metastatic sites and/or indolent disease, where resection of all metastatic sites is feasible.38 In patients unfit for surgery, SBRT has also been used for the local control of metastatic disease to the lungs and spine. Several institutions have published experiences with pulmonary and spine SBRT for sarcoma, with a local control rate >85% and minimal toxicity. A retrospective series from the University of Rochester examined 52 patients with pulmonary metastases from STS (median four lesions; range 1–16 lesions); 31% had surgical metastasectomy, while 29% received SBRT.39 No grade ≥3 toxicities were observed with SBRT. Also, 7/74 pulmonary metastases treated with SBRT progressed, with a 3-year local lesion control rate of 82% with SBRT. The median and mean survival for patients who received SBRT was 2.1 years. Another prospective observational study from Milan of 28 patients who received SBRT for pulmonary metastases reported a 5-year local control rate of 96%, also without grade ≥3 toxicity.40 Five-year overall survival for this cohort was 60.5%. A retrospective analysis from Memorial Sloan Kettering Cancer Center of 88 patients who received SBRT for spinal metastases from STS reported a 12-month local failure-free survival rate of 85.9%.41,42 Also, 83.3% of local failures failed distantly. The 12-month overall survival rate was 62.1%, and the median overall survival was 18.9 months. No grade >3 toxicities were observed; 1% had acute grade 3 toxicity and 4.5% had chronic grade 3 toxicity. SUMMARY RT uses IR to kill sarcoma cells, primarily through DNA damage. RT is typically delivered via external beam RT or brachytherapy. Advances including image-guided RT, IMRT, and particle therapy with protons or carbon ions can improve the therapeutic ratio by maximizing the tumor dose while minimizing the dose to adjacent normal tissues. In localized STS, the backbone of curative treatment is surgery with wide negative margins. Preoperative, intraoperative, or postoperative RT can maximize local control, though omission of radiation may be considered in select cases, particularly for small (<5 cm), low-grade, superficial tumors. Compared to postoperative RT, preoperative RT typically uses a lower total radiation dose and a smaller radiation field size, which causes less late effects. However, the rates of wound complications increase approximately twofold. IORT may also be utilized as a tumor bed boost when normal tissue is displaced out of the radiation field. Brachytherapy is typically used in the setting of an intraoperative or postoperative boost to the tumor bed or as adjuvant postoperative RT. Myxoid liposarcoma is especially radiosensitive, and a significant reduction in tumor volume can occur following RT. Scalp angiosarcoma has a propensity for local, cervical nodal, and distant recurrence. RT to the total scalp with bilateral neck node irradiation has been advocated to decrease the risk of recurrence. In Ewing sarcoma, RT may be used to treat localized disease when wide margins cannot be achieved with surgery or if surgery would impair function, though the risk of radiation-associated malignancy, especially in pediatric patients, must be considered. In Ewing sarcoma for limited pulmonary metastases, whole lung irradiation can also be considered. In chordoma and chondrosarcoma, RT may also have a role when wide local resection cannot be achieved. Particle therapy with protons has also been utilized, especially at the skull base, to adequately cover the tumor while minimizing the dose to critical structures such as brainstem and spinal cord. Re-irradiation for radiation-associated sarcomas or locally recurrent sarcomas provides local control; however, feasibility depends on dose tolerance of 51adjacent normal tissues. Complications of re-irradiation include impaired wound healing, soft tissue necrosis or osteonecrosis, fistulas, and neuropathy. Hyperfractionated RT and/or particle therapy with protons have been used to limit normal tissue toxicity for re-irradiation. SBRT may be used for local control of metastatic sites in appropriately selected patients with few metastases and/or indolent disease. Local control rates are typically >85% with minimal grade >3 toxicity.
Radiation Oncology
Jessica W. Lee and David G. Kirsch
